diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,111337 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 15901,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.560257448818305,
+      "learning_rate": 2.092050209205021e-08,
+      "loss": 1.3878,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.05368035791786,
+      "learning_rate": 4.184100418410042e-08,
+      "loss": 1.4129,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.118477403459769,
+      "learning_rate": 6.276150627615063e-08,
+      "loss": 1.4392,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.818196701992624,
+      "learning_rate": 8.368200836820084e-08,
+      "loss": 1.4417,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.52241286190301,
+      "learning_rate": 1.0460251046025104e-07,
+      "loss": 1.4274,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.166198039046842,
+      "learning_rate": 1.2552301255230126e-07,
+      "loss": 1.4287,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.152000828367362,
+      "learning_rate": 1.4644351464435148e-07,
+      "loss": 1.3883,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.282759445642426,
+      "learning_rate": 1.6736401673640168e-07,
+      "loss": 1.5214,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.187048651757747,
+      "learning_rate": 1.882845188284519e-07,
+      "loss": 1.3853,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.755813758937292,
+      "learning_rate": 2.092050209205021e-07,
+      "loss": 1.3802,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.471968662690486,
+      "learning_rate": 2.3012552301255234e-07,
+      "loss": 1.3958,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.023481948686502,
+      "learning_rate": 2.5104602510460253e-07,
+      "loss": 1.3625,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.407879675934606,
+      "learning_rate": 2.7196652719665275e-07,
+      "loss": 1.3987,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.002417532604325,
+      "learning_rate": 2.9288702928870297e-07,
+      "loss": 1.3685,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.046397935114733,
+      "learning_rate": 3.1380753138075313e-07,
+      "loss": 1.3761,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.74413617531794,
+      "learning_rate": 3.3472803347280335e-07,
+      "loss": 1.3515,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.668487707646594,
+      "learning_rate": 3.5564853556485363e-07,
+      "loss": 1.3325,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.55703442979226,
+      "learning_rate": 3.765690376569038e-07,
+      "loss": 1.3312,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.86102637540386,
+      "learning_rate": 3.97489539748954e-07,
+      "loss": 1.3139,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.885150017781749,
+      "learning_rate": 4.184100418410042e-07,
+      "loss": 1.313,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.508917024905127,
+      "learning_rate": 4.393305439330544e-07,
+      "loss": 1.182,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.155384192157996,
+      "learning_rate": 4.6025104602510467e-07,
+      "loss": 1.1837,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.159131473054676,
+      "learning_rate": 4.811715481171549e-07,
+      "loss": 1.2192,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.591628396174414,
+      "learning_rate": 5.020920502092051e-07,
+      "loss": 1.2123,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.2459320704076555,
+      "learning_rate": 5.230125523012552e-07,
+      "loss": 1.1723,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.686533379285876,
+      "learning_rate": 5.439330543933055e-07,
+      "loss": 1.2776,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.744258441539658,
+      "learning_rate": 5.648535564853557e-07,
+      "loss": 1.1493,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.774128304322188,
+      "learning_rate": 5.857740585774059e-07,
+      "loss": 1.1565,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7154225724871135,
+      "learning_rate": 6.066945606694561e-07,
+      "loss": 1.0396,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5116914798097727,
+      "learning_rate": 6.276150627615063e-07,
+      "loss": 1.0286,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3769413257117877,
+      "learning_rate": 6.485355648535565e-07,
+      "loss": 1.0221,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.715927901242416,
+      "learning_rate": 6.694560669456067e-07,
+      "loss": 0.7152,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0664296690038766,
+      "learning_rate": 6.90376569037657e-07,
+      "loss": 1.0258,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8471000703188283,
+      "learning_rate": 7.112970711297073e-07,
+      "loss": 1.0315,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.04394481220455,
+      "learning_rate": 7.322175732217573e-07,
+      "loss": 1.0299,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9290763986647264,
+      "learning_rate": 7.531380753138076e-07,
+      "loss": 1.0159,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.560701757054639,
+      "learning_rate": 7.740585774058578e-07,
+      "loss": 0.9823,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4094589147970704,
+      "learning_rate": 7.94979079497908e-07,
+      "loss": 0.9859,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.304616304731331,
+      "learning_rate": 8.158995815899583e-07,
+      "loss": 0.9283,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.336308342883063,
+      "learning_rate": 8.368200836820084e-07,
+      "loss": 0.9183,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6741125301434066,
+      "learning_rate": 8.577405857740586e-07,
+      "loss": 0.9356,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5818907495124943,
+      "learning_rate": 8.786610878661088e-07,
+      "loss": 0.909,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5428484367199093,
+      "learning_rate": 8.995815899581591e-07,
+      "loss": 0.9047,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.563139490584407,
+      "learning_rate": 9.205020920502093e-07,
+      "loss": 0.9273,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.559972040338691,
+      "learning_rate": 9.414225941422594e-07,
+      "loss": 0.9096,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.176763259124502,
+      "learning_rate": 9.623430962343098e-07,
+      "loss": 0.8882,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.147504876876301,
+      "learning_rate": 9.8326359832636e-07,
+      "loss": 0.8603,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9553524513603393,
+      "learning_rate": 1.0041841004184101e-06,
+      "loss": 0.8538,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7258382720756227,
+      "learning_rate": 1.0251046025104603e-06,
+      "loss": 0.8622,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.227266937437163,
+      "learning_rate": 1.0460251046025104e-06,
+      "loss": 0.9289,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6795622373939423,
+      "learning_rate": 1.0669456066945608e-06,
+      "loss": 0.8651,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.254037734651701,
+      "learning_rate": 1.087866108786611e-06,
+      "loss": 0.8048,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7583196392527818,
+      "learning_rate": 1.1087866108786612e-06,
+      "loss": 0.8187,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6957034862495282,
+      "learning_rate": 1.1297071129707113e-06,
+      "loss": 0.8168,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6473460490856577,
+      "learning_rate": 1.1506276150627615e-06,
+      "loss": 0.7906,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.614593057786042,
+      "learning_rate": 1.1715481171548119e-06,
+      "loss": 0.8337,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.572763056170728,
+      "learning_rate": 1.192468619246862e-06,
+      "loss": 0.8324,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0357680203943334,
+      "learning_rate": 1.2133891213389122e-06,
+      "loss": 0.8232,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.814797864213227,
+      "learning_rate": 1.2343096234309624e-06,
+      "loss": 0.8144,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6438167910451045,
+      "learning_rate": 1.2552301255230125e-06,
+      "loss": 0.7745,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8748382016688478,
+      "learning_rate": 1.276150627615063e-06,
+      "loss": 0.8187,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.783334835407333,
+      "learning_rate": 1.297071129707113e-06,
+      "loss": 0.7347,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6367872868851776,
+      "learning_rate": 1.3179916317991635e-06,
+      "loss": 0.7966,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7926974408303888,
+      "learning_rate": 1.3389121338912134e-06,
+      "loss": 0.7534,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6606628925876652,
+      "learning_rate": 1.3598326359832636e-06,
+      "loss": 0.7751,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3262653732005614,
+      "learning_rate": 1.380753138075314e-06,
+      "loss": 0.7957,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9727336258636625,
+      "learning_rate": 1.4016736401673641e-06,
+      "loss": 0.7589,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5456704266811179,
+      "learning_rate": 1.4225941422594145e-06,
+      "loss": 0.7719,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.696764818970671,
+      "learning_rate": 1.4435146443514645e-06,
+      "loss": 0.7103,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8835448556085705,
+      "learning_rate": 1.4644351464435146e-06,
+      "loss": 0.7208,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5943338456870952,
+      "learning_rate": 1.485355648535565e-06,
+      "loss": 0.7296,
+      "step": 71
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.142273603062082,
+      "learning_rate": 1.5062761506276152e-06,
+      "loss": 0.7343,
+      "step": 72
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9454596449518564,
+      "learning_rate": 1.5271966527196656e-06,
+      "loss": 0.7321,
+      "step": 73
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.384812557384574,
+      "learning_rate": 1.5481171548117155e-06,
+      "loss": 0.76,
+      "step": 74
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7705882044117423,
+      "learning_rate": 1.5690376569037657e-06,
+      "loss": 0.7091,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6371238362918907,
+      "learning_rate": 1.589958158995816e-06,
+      "loss": 0.709,
+      "step": 76
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.931138171116971,
+      "learning_rate": 1.6108786610878662e-06,
+      "loss": 0.7118,
+      "step": 77
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.790129601746314,
+      "learning_rate": 1.6317991631799166e-06,
+      "loss": 0.7066,
+      "step": 78
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.561249405813863,
+      "learning_rate": 1.6527196652719666e-06,
+      "loss": 0.6805,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.136737674376449,
+      "learning_rate": 1.6736401673640167e-06,
+      "loss": 0.6977,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7710100366758756,
+      "learning_rate": 1.694560669456067e-06,
+      "loss": 0.722,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5677734826636633,
+      "learning_rate": 1.7154811715481173e-06,
+      "loss": 0.678,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.876098097240444,
+      "learning_rate": 1.7364016736401676e-06,
+      "loss": 0.6647,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.079061521700847,
+      "learning_rate": 1.7573221757322176e-06,
+      "loss": 0.6866,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7513118734948716,
+      "learning_rate": 1.7782426778242678e-06,
+      "loss": 0.6748,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5300678728195052,
+      "learning_rate": 1.7991631799163181e-06,
+      "loss": 0.6544,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5018200347221655,
+      "learning_rate": 1.8200836820083683e-06,
+      "loss": 0.6691,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6743180612769364,
+      "learning_rate": 1.8410041841004187e-06,
+      "loss": 0.6736,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.8169151334734304,
+      "learning_rate": 1.8619246861924686e-06,
+      "loss": 0.5151,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.103169025054942,
+      "learning_rate": 1.8828451882845188e-06,
+      "loss": 0.6861,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7336268979816554,
+      "learning_rate": 1.9037656903765692e-06,
+      "loss": 0.6725,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9197089606721394,
+      "learning_rate": 1.9246861924686196e-06,
+      "loss": 0.6709,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6507475729497776,
+      "learning_rate": 1.9456066945606697e-06,
+      "loss": 0.6621,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5919123304273,
+      "learning_rate": 1.96652719665272e-06,
+      "loss": 0.6644,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6086503140886639,
+      "learning_rate": 1.98744769874477e-06,
+      "loss": 0.6779,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6852890600303034,
+      "learning_rate": 2.0083682008368202e-06,
+      "loss": 0.6649,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.709169433281033,
+      "learning_rate": 2.0292887029288704e-06,
+      "loss": 0.6657,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.255857063906131,
+      "learning_rate": 2.0502092050209206e-06,
+      "loss": 0.6827,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2727597545863,
+      "learning_rate": 2.071129707112971e-06,
+      "loss": 0.6371,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7926032772173868,
+      "learning_rate": 2.092050209205021e-06,
+      "loss": 0.6583,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.099978743369382,
+      "learning_rate": 2.112970711297071e-06,
+      "loss": 0.6969,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0217607651610785,
+      "learning_rate": 2.1338912133891217e-06,
+      "loss": 0.6691,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8662523120478176,
+      "learning_rate": 2.154811715481172e-06,
+      "loss": 0.6395,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4402038425145702,
+      "learning_rate": 2.175732217573222e-06,
+      "loss": 0.6115,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7596873929734194,
+      "learning_rate": 2.196652719665272e-06,
+      "loss": 0.6733,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6899417778676438,
+      "learning_rate": 2.2175732217573223e-06,
+      "loss": 0.6507,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.920649295828148,
+      "learning_rate": 2.2384937238493725e-06,
+      "loss": 0.6683,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7786419616325029,
+      "learning_rate": 2.2594142259414227e-06,
+      "loss": 0.6756,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.516926447022803,
+      "learning_rate": 2.2803347280334732e-06,
+      "loss": 0.6141,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8608800321743426,
+      "learning_rate": 2.301255230125523e-06,
+      "loss": 0.64,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.547921974374885,
+      "learning_rate": 2.322175732217573e-06,
+      "loss": 0.6765,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5918671185078295,
+      "learning_rate": 2.3430962343096237e-06,
+      "loss": 0.6028,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.542704232862084,
+      "learning_rate": 2.364016736401674e-06,
+      "loss": 0.6367,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9542813058768365,
+      "learning_rate": 2.384937238493724e-06,
+      "loss": 0.6052,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1428940037238218,
+      "learning_rate": 2.4058577405857742e-06,
+      "loss": 0.6141,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5871575608949053,
+      "learning_rate": 2.4267782426778244e-06,
+      "loss": 0.6421,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6380030671795447,
+      "learning_rate": 2.4476987447698746e-06,
+      "loss": 0.6206,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.768553551606372,
+      "learning_rate": 2.4686192468619247e-06,
+      "loss": 0.6136,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6324952719566632,
+      "learning_rate": 2.4895397489539753e-06,
+      "loss": 0.6165,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6563571805125223,
+      "learning_rate": 2.510460251046025e-06,
+      "loss": 0.6085,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7670514673682962,
+      "learning_rate": 2.5313807531380757e-06,
+      "loss": 0.6067,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.996143798308078,
+      "learning_rate": 2.552301255230126e-06,
+      "loss": 0.6103,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.245567993212247,
+      "learning_rate": 2.5732217573221756e-06,
+      "loss": 0.6164,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7354389724223707,
+      "learning_rate": 2.594142259414226e-06,
+      "loss": 0.6175,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7223077760804373,
+      "learning_rate": 2.6150627615062763e-06,
+      "loss": 0.6069,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7790619431990171,
+      "learning_rate": 2.635983263598327e-06,
+      "loss": 0.6347,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9358395244420905,
+      "learning_rate": 2.6569037656903767e-06,
+      "loss": 0.5972,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.876919263803925,
+      "learning_rate": 2.677824267782427e-06,
+      "loss": 0.6134,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8284656253275764,
+      "learning_rate": 2.6987447698744774e-06,
+      "loss": 0.5963,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0927052573703575,
+      "learning_rate": 2.719665271966527e-06,
+      "loss": 0.6374,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.043065973554129,
+      "learning_rate": 2.7405857740585778e-06,
+      "loss": 0.6321,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8387676107474462,
+      "learning_rate": 2.761506276150628e-06,
+      "loss": 0.5928,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6784610566868574,
+      "learning_rate": 2.7824267782426777e-06,
+      "loss": 0.6031,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7524319607012582,
+      "learning_rate": 2.8033472803347283e-06,
+      "loss": 0.5967,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6078432497800754,
+      "learning_rate": 2.8242677824267784e-06,
+      "loss": 0.5957,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5260696980523862,
+      "learning_rate": 2.845188284518829e-06,
+      "loss": 0.6297,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9377468490909344,
+      "learning_rate": 2.8661087866108788e-06,
+      "loss": 0.5889,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.5307926184565455,
+      "learning_rate": 2.887029288702929e-06,
+      "loss": 0.6061,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9059167405758104,
+      "learning_rate": 2.9079497907949795e-06,
+      "loss": 0.6077,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9100292543552058,
+      "learning_rate": 2.9288702928870293e-06,
+      "loss": 0.5915,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.732542946589111,
+      "learning_rate": 2.94979079497908e-06,
+      "loss": 0.5694,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5538540167703245,
+      "learning_rate": 2.97071129707113e-06,
+      "loss": 0.6137,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3343185435662486,
+      "learning_rate": 2.9916317991631798e-06,
+      "loss": 0.623,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0435238439705445,
+      "learning_rate": 3.0125523012552303e-06,
+      "loss": 0.5857,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9264488409367186,
+      "learning_rate": 3.0334728033472805e-06,
+      "loss": 0.5854,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8358898581063288,
+      "learning_rate": 3.054393305439331e-06,
+      "loss": 0.573,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0776269822206817,
+      "learning_rate": 3.075313807531381e-06,
+      "loss": 0.6185,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.758152422938119,
+      "learning_rate": 3.096234309623431e-06,
+      "loss": 0.5661,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2387002347198237,
+      "learning_rate": 3.1171548117154816e-06,
+      "loss": 0.6174,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8649137544286947,
+      "learning_rate": 3.1380753138075313e-06,
+      "loss": 0.5836,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1614961662618675,
+      "learning_rate": 3.158995815899582e-06,
+      "loss": 0.6087,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.963025160385293,
+      "learning_rate": 3.179916317991632e-06,
+      "loss": 0.6015,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.127885928484131,
+      "learning_rate": 3.200836820083682e-06,
+      "loss": 0.5827,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8926972991118225,
+      "learning_rate": 3.2217573221757324e-06,
+      "loss": 0.5639,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9778972234978238,
+      "learning_rate": 3.2426778242677826e-06,
+      "loss": 0.5472,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.742073166608714,
+      "learning_rate": 3.263598326359833e-06,
+      "loss": 0.62,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9624844320438002,
+      "learning_rate": 3.284518828451883e-06,
+      "loss": 0.576,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0381244644824084,
+      "learning_rate": 3.305439330543933e-06,
+      "loss": 0.5507,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2545084375913989,
+      "learning_rate": 3.3263598326359837e-06,
+      "loss": 0.4356,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1873318854202206,
+      "learning_rate": 3.3472803347280334e-06,
+      "loss": 0.5827,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2282400425869424,
+      "learning_rate": 3.368200836820084e-06,
+      "loss": 0.6281,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.897405087693351,
+      "learning_rate": 3.389121338912134e-06,
+      "loss": 0.5442,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.088965814031687,
+      "learning_rate": 3.410041841004184e-06,
+      "loss": 0.5998,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.026569481356135,
+      "learning_rate": 3.4309623430962345e-06,
+      "loss": 0.5726,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.103211691829716,
+      "learning_rate": 3.4518828451882847e-06,
+      "loss": 0.4828,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9849847813201476,
+      "learning_rate": 3.4728033472803353e-06,
+      "loss": 0.5415,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1630055011267317,
+      "learning_rate": 3.493723849372385e-06,
+      "loss": 0.5518,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4194171052375233,
+      "learning_rate": 3.514644351464435e-06,
+      "loss": 0.5851,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8093937261849506,
+      "learning_rate": 3.5355648535564858e-06,
+      "loss": 0.5697,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9064649752754856,
+      "learning_rate": 3.5564853556485355e-06,
+      "loss": 0.6195,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.846338624409405,
+      "learning_rate": 3.577405857740586e-06,
+      "loss": 0.576,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.829929989511185,
+      "learning_rate": 3.5983263598326363e-06,
+      "loss": 0.5492,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9034714933492618,
+      "learning_rate": 3.619246861924686e-06,
+      "loss": 0.5418,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.962239415108419,
+      "learning_rate": 3.6401673640167366e-06,
+      "loss": 0.5805,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6926327780212296,
+      "learning_rate": 3.6610878661087868e-06,
+      "loss": 0.5272,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8187430316592739,
+      "learning_rate": 3.6820083682008374e-06,
+      "loss": 0.4574,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.24986259368454,
+      "learning_rate": 3.702928870292887e-06,
+      "loss": 0.5984,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2422838567432923,
+      "learning_rate": 3.7238493723849373e-06,
+      "loss": 0.5598,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8202869130922215,
+      "learning_rate": 3.744769874476988e-06,
+      "loss": 0.5432,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2684918363837006,
+      "learning_rate": 3.7656903765690376e-06,
+      "loss": 0.5954,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.840220183982628,
+      "learning_rate": 3.786610878661088e-06,
+      "loss": 0.5471,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7375381450379523,
+      "learning_rate": 3.8075313807531384e-06,
+      "loss": 0.5398,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5833188190590233,
+      "learning_rate": 3.8284518828451885e-06,
+      "loss": 0.5593,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4203088015902696,
+      "learning_rate": 3.849372384937239e-06,
+      "loss": 0.5341,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4037462352952135,
+      "learning_rate": 3.870292887029289e-06,
+      "loss": 0.5241,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9018114196493612,
+      "learning_rate": 3.8912133891213395e-06,
+      "loss": 0.5541,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8346681758313153,
+      "learning_rate": 3.912133891213389e-06,
+      "loss": 0.5884,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.6060256359553735,
+      "learning_rate": 3.93305439330544e-06,
+      "loss": 0.5501,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1610451908072212,
+      "learning_rate": 3.9539748953974895e-06,
+      "loss": 0.5964,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.0619098539338365,
+      "learning_rate": 3.97489539748954e-06,
+      "loss": 0.5723,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.487926344923484,
+      "learning_rate": 3.995815899581591e-06,
+      "loss": 0.5874,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8608181899237632,
+      "learning_rate": 4.0167364016736405e-06,
+      "loss": 0.5233,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4239428611178764,
+      "learning_rate": 4.037656903765691e-06,
+      "loss": 0.5519,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5520657816429004,
+      "learning_rate": 4.058577405857741e-06,
+      "loss": 0.6036,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.956756479512331,
+      "learning_rate": 4.0794979079497905e-06,
+      "loss": 0.5487,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.179677188583239,
+      "learning_rate": 4.100418410041841e-06,
+      "loss": 0.5362,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.58054581350178,
+      "learning_rate": 4.121338912133892e-06,
+      "loss": 0.5526,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0882473223940354,
+      "learning_rate": 4.142259414225942e-06,
+      "loss": 0.5691,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.225018522453562,
+      "learning_rate": 4.163179916317992e-06,
+      "loss": 0.5446,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2548977860892365,
+      "learning_rate": 4.184100418410042e-06,
+      "loss": 0.5493,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0015897504392632,
+      "learning_rate": 4.205020920502092e-06,
+      "loss": 0.5727,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8549759738478968,
+      "learning_rate": 4.225941422594142e-06,
+      "loss": 0.4758,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7218688114187135,
+      "learning_rate": 4.246861924686193e-06,
+      "loss": 0.5437,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4956100779292383,
+      "learning_rate": 4.267782426778243e-06,
+      "loss": 0.5583,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3497272517836394,
+      "learning_rate": 4.288702928870293e-06,
+      "loss": 0.561,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1347763158373505,
+      "learning_rate": 4.309623430962344e-06,
+      "loss": 0.5512,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.017505861291735,
+      "learning_rate": 4.330543933054393e-06,
+      "loss": 0.545,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.944054774884085,
+      "learning_rate": 4.351464435146444e-06,
+      "loss": 0.6029,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.05288008370838,
+      "learning_rate": 4.372384937238494e-06,
+      "loss": 0.5386,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.965382181850134,
+      "learning_rate": 4.393305439330544e-06,
+      "loss": 0.5262,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5857770698218623,
+      "learning_rate": 4.414225941422595e-06,
+      "loss": 0.5269,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.834596608275258,
+      "learning_rate": 4.435146443514645e-06,
+      "loss": 0.5422,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.057630453079115,
+      "learning_rate": 4.456066945606695e-06,
+      "loss": 0.5334,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6220490371524603,
+      "learning_rate": 4.476987447698745e-06,
+      "loss": 0.5374,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.286435524178221,
+      "learning_rate": 4.4979079497907956e-06,
+      "loss": 0.5227,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.227202842739513,
+      "learning_rate": 4.518828451882845e-06,
+      "loss": 0.5751,
+      "step": 216
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.829600359946089,
+      "learning_rate": 4.539748953974896e-06,
+      "loss": 0.5109,
+      "step": 217
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1186628117992368,
+      "learning_rate": 4.5606694560669465e-06,
+      "loss": 0.5155,
+      "step": 218
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.757460139156587,
+      "learning_rate": 4.581589958158996e-06,
+      "loss": 0.5528,
+      "step": 219
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.913175574549681,
+      "learning_rate": 4.602510460251046e-06,
+      "loss": 0.5263,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8302369226695772,
+      "learning_rate": 4.6234309623430966e-06,
+      "loss": 0.548,
+      "step": 221
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.472263108281696,
+      "learning_rate": 4.644351464435146e-06,
+      "loss": 0.5217,
+      "step": 222
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.901098226008417,
+      "learning_rate": 4.665271966527197e-06,
+      "loss": 0.5094,
+      "step": 223
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.23530258377056,
+      "learning_rate": 4.6861924686192475e-06,
+      "loss": 0.4972,
+      "step": 224
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.901061367814879,
+      "learning_rate": 4.707112970711297e-06,
+      "loss": 0.5148,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5767890434567686,
+      "learning_rate": 4.728033472803348e-06,
+      "loss": 0.5601,
+      "step": 226
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.314143929513127,
+      "learning_rate": 4.7489539748953976e-06,
+      "loss": 0.5433,
+      "step": 227
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1529240549180026,
+      "learning_rate": 4.769874476987448e-06,
+      "loss": 0.5143,
+      "step": 228
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1515149718204225,
+      "learning_rate": 4.790794979079498e-06,
+      "loss": 0.5434,
+      "step": 229
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.20706496494061,
+      "learning_rate": 4.8117154811715485e-06,
+      "loss": 0.4952,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.154685233114359,
+      "learning_rate": 4.832635983263599e-06,
+      "loss": 0.5263,
+      "step": 231
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.338673730530193,
+      "learning_rate": 4.853556485355649e-06,
+      "loss": 0.5515,
+      "step": 232
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.332063111503053,
+      "learning_rate": 4.874476987447699e-06,
+      "loss": 0.5559,
+      "step": 233
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7427624508109318,
+      "learning_rate": 4.895397489539749e-06,
+      "loss": 0.5355,
+      "step": 234
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2863543898558905,
+      "learning_rate": 4.9163179916318e-06,
+      "loss": 0.5394,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.183148941750467,
+      "learning_rate": 4.9372384937238495e-06,
+      "loss": 0.493,
+      "step": 236
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8999604880172074,
+      "learning_rate": 4.9581589958159e-06,
+      "loss": 0.5232,
+      "step": 237
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.024650317117212,
+      "learning_rate": 4.979079497907951e-06,
+      "loss": 0.5463,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.519833654324174,
+      "learning_rate": 5e-06,
+      "loss": 0.5053,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1145062860634294,
+      "learning_rate": 5.02092050209205e-06,
+      "loss": 0.5421,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8230791631428602,
+      "learning_rate": 5.041841004184101e-06,
+      "loss": 0.5175,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2292071552421118,
+      "learning_rate": 5.062761506276151e-06,
+      "loss": 0.5188,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.060899197962938,
+      "learning_rate": 5.083682008368201e-06,
+      "loss": 0.5064,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.287968133304338,
+      "learning_rate": 5.104602510460252e-06,
+      "loss": 0.5456,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8504400266281082,
+      "learning_rate": 5.125523012552301e-06,
+      "loss": 0.4325,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1361266792185174,
+      "learning_rate": 5.146443514644351e-06,
+      "loss": 0.5422,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4283629300625877,
+      "learning_rate": 5.167364016736403e-06,
+      "loss": 0.5613,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.004467872375992,
+      "learning_rate": 5.188284518828452e-06,
+      "loss": 0.5277,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1897735043913804,
+      "learning_rate": 5.209205020920503e-06,
+      "loss": 0.5529,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0391670505960278,
+      "learning_rate": 5.230125523012553e-06,
+      "loss": 0.5311,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0097009295986283,
+      "learning_rate": 5.251046025104602e-06,
+      "loss": 0.5372,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1829047992119124,
+      "learning_rate": 5.271966527196654e-06,
+      "loss": 0.5075,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8215535134615874,
+      "learning_rate": 5.292887029288704e-06,
+      "loss": 0.5389,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.054601898394612,
+      "learning_rate": 5.313807531380753e-06,
+      "loss": 0.549,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.329629461954194,
+      "learning_rate": 5.334728033472804e-06,
+      "loss": 0.5311,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8914673884674622,
+      "learning_rate": 5.355648535564854e-06,
+      "loss": 0.52,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0859952944448994,
+      "learning_rate": 5.376569037656904e-06,
+      "loss": 0.4608,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.503935817868153,
+      "learning_rate": 5.397489539748955e-06,
+      "loss": 0.5126,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6643267775576076,
+      "learning_rate": 5.418410041841005e-06,
+      "loss": 0.5213,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.65610581106084,
+      "learning_rate": 5.439330543933054e-06,
+      "loss": 0.5778,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5025442905122772,
+      "learning_rate": 5.460251046025105e-06,
+      "loss": 0.5002,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.520819133078781,
+      "learning_rate": 5.4811715481171555e-06,
+      "loss": 0.5489,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.272951313335712,
+      "learning_rate": 5.502092050209205e-06,
+      "loss": 0.563,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.413045100161984,
+      "learning_rate": 5.523012552301256e-06,
+      "loss": 0.5402,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8929401036698851,
+      "learning_rate": 5.543933054393306e-06,
+      "loss": 0.4515,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.325230474399367,
+      "learning_rate": 5.564853556485355e-06,
+      "loss": 0.55,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.049356261368777,
+      "learning_rate": 5.585774058577407e-06,
+      "loss": 0.5259,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.621157778314054,
+      "learning_rate": 5.6066945606694565e-06,
+      "loss": 0.5395,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 96.63248774085258,
+      "learning_rate": 5.627615062761507e-06,
+      "loss": 0.5568,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.91911523630248,
+      "learning_rate": 5.648535564853557e-06,
+      "loss": 0.5073,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.378666811200259,
+      "learning_rate": 5.669456066945607e-06,
+      "loss": 0.5402,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.561323053565177,
+      "learning_rate": 5.690376569037658e-06,
+      "loss": 0.5315,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6279745240004773,
+      "learning_rate": 5.711297071129708e-06,
+      "loss": 0.4996,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.221717607952609,
+      "learning_rate": 5.7322175732217575e-06,
+      "loss": 0.5017,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.705689661615929,
+      "learning_rate": 5.753138075313808e-06,
+      "loss": 0.5159,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9104386727427634,
+      "learning_rate": 5.774058577405858e-06,
+      "loss": 0.5142,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8517039944287434,
+      "learning_rate": 5.7949790794979084e-06,
+      "loss": 0.5703,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7616299815648704,
+      "learning_rate": 5.815899581589959e-06,
+      "loss": 0.5456,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6609993986366334,
+      "learning_rate": 5.836820083682009e-06,
+      "loss": 0.5327,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6076800181703126,
+      "learning_rate": 5.8577405857740585e-06,
+      "loss": 0.505,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3287901339772996,
+      "learning_rate": 5.878661087866109e-06,
+      "loss": 0.4926,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9176150332150999,
+      "learning_rate": 5.89958158995816e-06,
+      "loss": 0.4637,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.836208971568334,
+      "learning_rate": 5.92050209205021e-06,
+      "loss": 0.539,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.56516241457902,
+      "learning_rate": 5.94142259414226e-06,
+      "loss": 0.5162,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.234022868978037,
+      "learning_rate": 5.96234309623431e-06,
+      "loss": 0.5201,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.193157674838078,
+      "learning_rate": 5.9832635983263595e-06,
+      "loss": 0.5133,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6235735117222805,
+      "learning_rate": 6.004184100418411e-06,
+      "loss": 0.5325,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0685788179195264,
+      "learning_rate": 6.025104602510461e-06,
+      "loss": 0.5088,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8676759066530564,
+      "learning_rate": 6.046025104602511e-06,
+      "loss": 0.5303,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0962004596269437,
+      "learning_rate": 6.066945606694561e-06,
+      "loss": 0.4912,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3877027702814257,
+      "learning_rate": 6.087866108786611e-06,
+      "loss": 0.5036,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7071208206826958,
+      "learning_rate": 6.108786610878662e-06,
+      "loss": 0.4957,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2847240073491752,
+      "learning_rate": 6.129707112970712e-06,
+      "loss": 0.5106,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3774041487743665,
+      "learning_rate": 6.150627615062762e-06,
+      "loss": 0.4879,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6449473691246475,
+      "learning_rate": 6.171548117154812e-06,
+      "loss": 0.5333,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4353294447704243,
+      "learning_rate": 6.192468619246862e-06,
+      "loss": 0.4978,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.020778042253329,
+      "learning_rate": 6.213389121338913e-06,
+      "loss": 0.5214,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0942726866993144,
+      "learning_rate": 6.234309623430963e-06,
+      "loss": 0.5359,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6716584551967406,
+      "learning_rate": 6.255230125523013e-06,
+      "loss": 0.4982,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0646153430845535,
+      "learning_rate": 6.276150627615063e-06,
+      "loss": 0.5266,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.94715980004556,
+      "learning_rate": 6.297071129707113e-06,
+      "loss": 0.5234,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1027040732914546,
+      "learning_rate": 6.317991631799164e-06,
+      "loss": 0.4561,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.206545821119485,
+      "learning_rate": 6.3389121338912145e-06,
+      "loss": 0.4935,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8473207773595686,
+      "learning_rate": 6.359832635983264e-06,
+      "loss": 0.5527,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.023532167096835,
+      "learning_rate": 6.380753138075314e-06,
+      "loss": 0.4862,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.292990098601755,
+      "learning_rate": 6.401673640167364e-06,
+      "loss": 0.4951,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8132000573019575,
+      "learning_rate": 6.422594142259415e-06,
+      "loss": 0.5869,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.973838233332685,
+      "learning_rate": 6.443514644351465e-06,
+      "loss": 0.4958,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.539826559036644,
+      "learning_rate": 6.4644351464435155e-06,
+      "loss": 0.5604,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0504923999789053,
+      "learning_rate": 6.485355648535565e-06,
+      "loss": 0.4794,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0248026253469127,
+      "learning_rate": 6.506276150627615e-06,
+      "loss": 0.4841,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.060222864652489,
+      "learning_rate": 6.527196652719666e-06,
+      "loss": 0.5296,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5536836715360005,
+      "learning_rate": 6.548117154811716e-06,
+      "loss": 0.5435,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8216575901049348,
+      "learning_rate": 6.569037656903766e-06,
+      "loss": 0.4525,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2872488197895846,
+      "learning_rate": 6.5899581589958165e-06,
+      "loss": 0.4808,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0154227553774837,
+      "learning_rate": 6.610878661087866e-06,
+      "loss": 0.5075,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.689464889095058,
+      "learning_rate": 6.631799163179918e-06,
+      "loss": 0.5294,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.511514014884,
+      "learning_rate": 6.652719665271967e-06,
+      "loss": 0.5001,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.462426397042369,
+      "learning_rate": 6.673640167364017e-06,
+      "loss": 0.5201,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.309178481355211,
+      "learning_rate": 6.694560669456067e-06,
+      "loss": 0.4914,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.00046890267073,
+      "learning_rate": 6.7154811715481175e-06,
+      "loss": 0.4846,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.771069502433317,
+      "learning_rate": 6.736401673640168e-06,
+      "loss": 0.5383,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7459182369145854,
+      "learning_rate": 6.757322175732219e-06,
+      "loss": 0.5196,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0963134515292685,
+      "learning_rate": 6.778242677824268e-06,
+      "loss": 0.4883,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 21.49385232659443,
+      "learning_rate": 6.799163179916318e-06,
+      "loss": 0.522,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6575515829395526,
+      "learning_rate": 6.820083682008368e-06,
+      "loss": 0.4965,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7250280279432313,
+      "learning_rate": 6.841004184100419e-06,
+      "loss": 0.5186,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0326723726262537,
+      "learning_rate": 6.861924686192469e-06,
+      "loss": 0.5194,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.874319558287596,
+      "learning_rate": 6.88284518828452e-06,
+      "loss": 0.4986,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.5858321447557,
+      "learning_rate": 6.903765690376569e-06,
+      "loss": 0.4747,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.85784679462047,
+      "learning_rate": 6.924686192468619e-06,
+      "loss": 0.5098,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.824972859212857,
+      "learning_rate": 6.9456066945606706e-06,
+      "loss": 0.5029,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1117225820256342,
+      "learning_rate": 6.96652719665272e-06,
+      "loss": 0.5309,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.061982984691296,
+      "learning_rate": 6.98744769874477e-06,
+      "loss": 0.5131,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.536437490932801,
+      "learning_rate": 7.008368200836821e-06,
+      "loss": 0.5196,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.309975716765149,
+      "learning_rate": 7.02928870292887e-06,
+      "loss": 0.4742,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.570204434848717,
+      "learning_rate": 7.050209205020922e-06,
+      "loss": 0.5131,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.707895248879186,
+      "learning_rate": 7.0711297071129716e-06,
+      "loss": 0.5506,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.673949592438475,
+      "learning_rate": 7.092050209205021e-06,
+      "loss": 0.4867,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7316392716115172,
+      "learning_rate": 7.112970711297071e-06,
+      "loss": 0.4904,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3951559468838584,
+      "learning_rate": 7.133891213389122e-06,
+      "loss": 0.4647,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.448340965348166,
+      "learning_rate": 7.154811715481172e-06,
+      "loss": 0.4623,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6649219852341175,
+      "learning_rate": 7.175732217573223e-06,
+      "loss": 0.5209,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.585114417633767,
+      "learning_rate": 7.1966527196652726e-06,
+      "loss": 0.5184,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.452898799199118,
+      "learning_rate": 7.217573221757322e-06,
+      "loss": 0.4877,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.833547011782983,
+      "learning_rate": 7.238493723849372e-06,
+      "loss": 0.478,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7953763531070988,
+      "learning_rate": 7.2594142259414235e-06,
+      "loss": 0.5042,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.734182011955151,
+      "learning_rate": 7.280334728033473e-06,
+      "loss": 0.4758,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9624647174911558,
+      "learning_rate": 7.301255230125524e-06,
+      "loss": 0.5058,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3363783927689292,
+      "learning_rate": 7.3221757322175736e-06,
+      "loss": 0.479,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9336687093258327,
+      "learning_rate": 7.343096234309623e-06,
+      "loss": 0.5096,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.266888841938558,
+      "learning_rate": 7.364016736401675e-06,
+      "loss": 0.5157,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5073325128843336,
+      "learning_rate": 7.3849372384937245e-06,
+      "loss": 0.4959,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8126613086801358,
+      "learning_rate": 7.405857740585774e-06,
+      "loss": 0.5085,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.60687001755244,
+      "learning_rate": 7.426778242677825e-06,
+      "loss": 0.5306,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.341686498326082,
+      "learning_rate": 7.4476987447698746e-06,
+      "loss": 0.4989,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0904637210259118,
+      "learning_rate": 7.468619246861926e-06,
+      "loss": 0.4616,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.133117015555545,
+      "learning_rate": 7.489539748953976e-06,
+      "loss": 0.5255,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.006984132530126,
+      "learning_rate": 7.5104602510460255e-06,
+      "loss": 0.5218,
+      "step": 359
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3938720688264263,
+      "learning_rate": 7.531380753138075e-06,
+      "loss": 0.5064,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.870676793551916,
+      "learning_rate": 7.552301255230127e-06,
+      "loss": 0.4847,
+      "step": 361
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8390125025598012,
+      "learning_rate": 7.573221757322176e-06,
+      "loss": 0.4413,
+      "step": 362
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.327764398915044,
+      "learning_rate": 7.594142259414227e-06,
+      "loss": 0.5116,
+      "step": 363
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1647685751562404,
+      "learning_rate": 7.615062761506277e-06,
+      "loss": 0.5016,
+      "step": 364
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3640020871193026,
+      "learning_rate": 7.635983263598326e-06,
+      "loss": 0.5247,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.700581385258537,
+      "learning_rate": 7.656903765690377e-06,
+      "loss": 0.4815,
+      "step": 366
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.767697256841907,
+      "learning_rate": 7.677824267782428e-06,
+      "loss": 0.4991,
+      "step": 367
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4927873233606883,
+      "learning_rate": 7.698744769874478e-06,
+      "loss": 0.5097,
+      "step": 368
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.828020437858698,
+      "learning_rate": 7.719665271966527e-06,
+      "loss": 0.4657,
+      "step": 369
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6332343549170316,
+      "learning_rate": 7.740585774058578e-06,
+      "loss": 0.502,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4241080414330596,
+      "learning_rate": 7.761506276150628e-06,
+      "loss": 0.4871,
+      "step": 371
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4052948087097104,
+      "learning_rate": 7.782426778242679e-06,
+      "loss": 0.5261,
+      "step": 372
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4405844871804074,
+      "learning_rate": 7.80334728033473e-06,
+      "loss": 0.4534,
+      "step": 373
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.648179101276863,
+      "learning_rate": 7.824267782426778e-06,
+      "loss": 0.4664,
+      "step": 374
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7911410791896376,
+      "learning_rate": 7.845188284518829e-06,
+      "loss": 0.5241,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8902105884147422,
+      "learning_rate": 7.86610878661088e-06,
+      "loss": 0.4666,
+      "step": 376
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0296579259691927,
+      "learning_rate": 7.88702928870293e-06,
+      "loss": 0.4474,
+      "step": 377
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.222181680520652,
+      "learning_rate": 7.907949790794979e-06,
+      "loss": 0.4888,
+      "step": 378
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0631162291398097,
+      "learning_rate": 7.92887029288703e-06,
+      "loss": 0.4995,
+      "step": 379
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.209846140694516,
+      "learning_rate": 7.94979079497908e-06,
+      "loss": 0.537,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.463680463708773,
+      "learning_rate": 7.97071129707113e-06,
+      "loss": 0.5255,
+      "step": 381
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.891988986891982,
+      "learning_rate": 7.991631799163181e-06,
+      "loss": 0.4869,
+      "step": 382
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.867591904677801,
+      "learning_rate": 8.01255230125523e-06,
+      "loss": 0.4612,
+      "step": 383
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.430308944036147,
+      "learning_rate": 8.033472803347281e-06,
+      "loss": 0.5192,
+      "step": 384
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.747116918822297,
+      "learning_rate": 8.054393305439332e-06,
+      "loss": 0.5017,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.085915787920984,
+      "learning_rate": 8.075313807531382e-06,
+      "loss": 0.473,
+      "step": 386
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3728873767492407,
+      "learning_rate": 8.096234309623433e-06,
+      "loss": 0.4877,
+      "step": 387
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5845719704457917,
+      "learning_rate": 8.117154811715482e-06,
+      "loss": 0.5671,
+      "step": 388
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9173579598015733,
+      "learning_rate": 8.138075313807532e-06,
+      "loss": 0.4929,
+      "step": 389
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.883561647904503,
+      "learning_rate": 8.158995815899581e-06,
+      "loss": 0.4941,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.940398886539253,
+      "learning_rate": 8.179916317991633e-06,
+      "loss": 0.4837,
+      "step": 391
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.961712747262717,
+      "learning_rate": 8.200836820083682e-06,
+      "loss": 0.4857,
+      "step": 392
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7196777020309737,
+      "learning_rate": 8.221757322175733e-06,
+      "loss": 0.5571,
+      "step": 393
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3298620477337977,
+      "learning_rate": 8.242677824267783e-06,
+      "loss": 0.5027,
+      "step": 394
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6061366433259003,
+      "learning_rate": 8.263598326359832e-06,
+      "loss": 0.4978,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0405769078662233,
+      "learning_rate": 8.284518828451885e-06,
+      "loss": 0.4555,
+      "step": 396
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8937000549595946,
+      "learning_rate": 8.305439330543934e-06,
+      "loss": 0.4902,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.972537027473449,
+      "learning_rate": 8.326359832635984e-06,
+      "loss": 0.4724,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.187539954145826,
+      "learning_rate": 8.347280334728035e-06,
+      "loss": 0.4782,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5411039017670323,
+      "learning_rate": 8.368200836820084e-06,
+      "loss": 0.5113,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5085262051947974,
+      "learning_rate": 8.389121338912136e-06,
+      "loss": 0.4803,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.848215888898236,
+      "learning_rate": 8.410041841004185e-06,
+      "loss": 0.4969,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.679526385946745,
+      "learning_rate": 8.430962343096235e-06,
+      "loss": 0.469,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1049472628050294,
+      "learning_rate": 8.451882845188284e-06,
+      "loss": 0.4728,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.378765128456705,
+      "learning_rate": 8.472803347280335e-06,
+      "loss": 0.4759,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8512740769910905,
+      "learning_rate": 8.493723849372385e-06,
+      "loss": 0.5282,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.277942641095943,
+      "learning_rate": 8.514644351464436e-06,
+      "loss": 0.4712,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.443730718018829,
+      "learning_rate": 8.535564853556487e-06,
+      "loss": 0.4821,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.321172925160859,
+      "learning_rate": 8.556485355648536e-06,
+      "loss": 0.4636,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8545919242223279,
+      "learning_rate": 8.577405857740586e-06,
+      "loss": 0.4965,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4878827595054522,
+      "learning_rate": 8.598326359832637e-06,
+      "loss": 0.4793,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.266346876657598,
+      "learning_rate": 8.619246861924687e-06,
+      "loss": 0.4922,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5792097367050015,
+      "learning_rate": 8.640167364016738e-06,
+      "loss": 0.5004,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.009757320195426,
+      "learning_rate": 8.661087866108787e-06,
+      "loss": 0.5171,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2238853932080604,
+      "learning_rate": 8.682008368200837e-06,
+      "loss": 0.5391,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2712477846853827,
+      "learning_rate": 8.702928870292888e-06,
+      "loss": 0.5026,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1988046779112986,
+      "learning_rate": 8.723849372384939e-06,
+      "loss": 0.4858,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.406534027788669,
+      "learning_rate": 8.744769874476987e-06,
+      "loss": 0.4656,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.163961435074915,
+      "learning_rate": 8.765690376569038e-06,
+      "loss": 0.5427,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3597195024144497,
+      "learning_rate": 8.786610878661089e-06,
+      "loss": 0.4897,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.168965457878025,
+      "learning_rate": 8.80753138075314e-06,
+      "loss": 0.479,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5988091819064887,
+      "learning_rate": 8.82845188284519e-06,
+      "loss": 0.5255,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7920088032537924,
+      "learning_rate": 8.849372384937239e-06,
+      "loss": 0.4933,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.983311565053605,
+      "learning_rate": 8.87029288702929e-06,
+      "loss": 0.4882,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6590108562139867,
+      "learning_rate": 8.89121338912134e-06,
+      "loss": 0.5219,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1317040286033073,
+      "learning_rate": 8.91213389121339e-06,
+      "loss": 0.4815,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.274968257780614,
+      "learning_rate": 8.933054393305441e-06,
+      "loss": 0.4822,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.686968625635076,
+      "learning_rate": 8.95397489539749e-06,
+      "loss": 0.4893,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5596408404374777,
+      "learning_rate": 8.97489539748954e-06,
+      "loss": 0.4605,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 17.2740122622564,
+      "learning_rate": 8.995815899581591e-06,
+      "loss": 0.5216,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.921884062267339,
+      "learning_rate": 9.016736401673642e-06,
+      "loss": 0.4743,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4626916917319113,
+      "learning_rate": 9.03765690376569e-06,
+      "loss": 0.4867,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.073554583318843,
+      "learning_rate": 9.058577405857741e-06,
+      "loss": 0.4803,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6827488402593347,
+      "learning_rate": 9.079497907949792e-06,
+      "loss": 0.4818,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.575804476795184,
+      "learning_rate": 9.10041841004184e-06,
+      "loss": 0.4579,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9555600587994975,
+      "learning_rate": 9.121338912133893e-06,
+      "loss": 0.4641,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9832609347832768,
+      "learning_rate": 9.142259414225942e-06,
+      "loss": 0.4641,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5867908607748533,
+      "learning_rate": 9.163179916317992e-06,
+      "loss": 0.4515,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.206710679839185,
+      "learning_rate": 9.184100418410043e-06,
+      "loss": 0.5006,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.877142684214976,
+      "learning_rate": 9.205020920502092e-06,
+      "loss": 0.4721,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8918168834279614,
+      "learning_rate": 9.225941422594144e-06,
+      "loss": 0.4877,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2351645066296113,
+      "learning_rate": 9.246861924686193e-06,
+      "loss": 0.4819,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 24.542472415486223,
+      "learning_rate": 9.267782426778244e-06,
+      "loss": 0.5727,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9593789970001019,
+      "learning_rate": 9.288702928870293e-06,
+      "loss": 0.4963,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7693233306816372,
+      "learning_rate": 9.309623430962343e-06,
+      "loss": 0.5029,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3267296066825596,
+      "learning_rate": 9.330543933054394e-06,
+      "loss": 0.476,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9083774240017217,
+      "learning_rate": 9.351464435146444e-06,
+      "loss": 0.5121,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.886945427880566,
+      "learning_rate": 9.372384937238495e-06,
+      "loss": 0.4735,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8086180482800336,
+      "learning_rate": 9.393305439330544e-06,
+      "loss": 0.4796,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5469885536000434,
+      "learning_rate": 9.414225941422594e-06,
+      "loss": 0.5053,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7621867233967463,
+      "learning_rate": 9.435146443514645e-06,
+      "loss": 0.4892,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3418689487785005,
+      "learning_rate": 9.456066945606696e-06,
+      "loss": 0.4976,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7565291667069471,
+      "learning_rate": 9.476987447698746e-06,
+      "loss": 0.4702,
+      "step": 453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7055592954521366,
+      "learning_rate": 9.497907949790795e-06,
+      "loss": 0.4975,
+      "step": 454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.125122571133811,
+      "learning_rate": 9.518828451882846e-06,
+      "loss": 0.4854,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.916873860259975,
+      "learning_rate": 9.539748953974896e-06,
+      "loss": 0.4663,
+      "step": 456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0397319347096645,
+      "learning_rate": 9.560669456066947e-06,
+      "loss": 0.4953,
+      "step": 457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7578682358617486,
+      "learning_rate": 9.581589958158996e-06,
+      "loss": 0.501,
+      "step": 458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.93713540725994,
+      "learning_rate": 9.602510460251046e-06,
+      "loss": 0.5161,
+      "step": 459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.481061522409949,
+      "learning_rate": 9.623430962343097e-06,
+      "loss": 0.4816,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1183860924267104,
+      "learning_rate": 9.644351464435148e-06,
+      "loss": 0.488,
+      "step": 461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5309859152257093,
+      "learning_rate": 9.665271966527198e-06,
+      "loss": 0.4803,
+      "step": 462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6623926328087257,
+      "learning_rate": 9.686192468619247e-06,
+      "loss": 0.4625,
+      "step": 463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.226731031134517,
+      "learning_rate": 9.707112970711298e-06,
+      "loss": 0.4777,
+      "step": 464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9974017132717794,
+      "learning_rate": 9.728033472803348e-06,
+      "loss": 0.4999,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9782826327473388,
+      "learning_rate": 9.748953974895399e-06,
+      "loss": 0.4746,
+      "step": 466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.280861114366562,
+      "learning_rate": 9.76987447698745e-06,
+      "loss": 0.4949,
+      "step": 467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.561532321958314,
+      "learning_rate": 9.790794979079498e-06,
+      "loss": 0.5523,
+      "step": 468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6482834591286095,
+      "learning_rate": 9.811715481171549e-06,
+      "loss": 0.4785,
+      "step": 469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.788555807619979,
+      "learning_rate": 9.8326359832636e-06,
+      "loss": 0.4918,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7239375482508028,
+      "learning_rate": 9.85355648535565e-06,
+      "loss": 0.4521,
+      "step": 471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.075375377188315,
+      "learning_rate": 9.874476987447699e-06,
+      "loss": 0.4852,
+      "step": 472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5660880924744864,
+      "learning_rate": 9.89539748953975e-06,
+      "loss": 0.4766,
+      "step": 473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0997494916527482,
+      "learning_rate": 9.9163179916318e-06,
+      "loss": 0.4651,
+      "step": 474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.197086697835969,
+      "learning_rate": 9.937238493723849e-06,
+      "loss": 0.5007,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.509517393333771,
+      "learning_rate": 9.958158995815901e-06,
+      "loss": 0.4993,
+      "step": 476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7187968677337175,
+      "learning_rate": 9.97907949790795e-06,
+      "loss": 0.4827,
+      "step": 477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.282998909013323,
+      "learning_rate": 1e-05,
+      "loss": 0.5512,
+      "step": 478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7084881852837643,
+      "learning_rate": 9.99999989627056e-06,
+      "loss": 0.4992,
+      "step": 479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8868923475095896,
+      "learning_rate": 9.999999585082243e-06,
+      "loss": 0.5299,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6316744119836013,
+      "learning_rate": 9.999999066435062e-06,
+      "loss": 0.5353,
+      "step": 481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.980509745621231,
+      "learning_rate": 9.99999834032904e-06,
+      "loss": 0.4819,
+      "step": 482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7208825151127956,
+      "learning_rate": 9.999997406764204e-06,
+      "loss": 0.4674,
+      "step": 483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.930274641225573,
+      "learning_rate": 9.999996265740595e-06,
+      "loss": 0.474,
+      "step": 484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.46637256655527,
+      "learning_rate": 9.99999491725826e-06,
+      "loss": 0.4848,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.962705189780874,
+      "learning_rate": 9.999993361317255e-06,
+      "loss": 0.4832,
+      "step": 486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9013581764155414,
+      "learning_rate": 9.999991597917645e-06,
+      "loss": 0.4701,
+      "step": 487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.392350399151913,
+      "learning_rate": 9.999989627059501e-06,
+      "loss": 0.4972,
+      "step": 488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7267714061114867,
+      "learning_rate": 9.999987448742909e-06,
+      "loss": 0.4847,
+      "step": 489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.632008755771487,
+      "learning_rate": 9.999985062967955e-06,
+      "loss": 0.4746,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6553610090016668,
+      "learning_rate": 9.99998246973474e-06,
+      "loss": 0.4669,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8526190594946903,
+      "learning_rate": 9.999979669043371e-06,
+      "loss": 0.4777,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9881684812188742,
+      "learning_rate": 9.999976660893964e-06,
+      "loss": 0.496,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.192218782535668,
+      "learning_rate": 9.999973445286645e-06,
+      "loss": 0.5122,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8663884967123074,
+      "learning_rate": 9.999970022221547e-06,
+      "loss": 0.4729,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7233098792878956,
+      "learning_rate": 9.999966391698814e-06,
+      "loss": 0.5244,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1689089902666683,
+      "learning_rate": 9.999962553718592e-06,
+      "loss": 0.5363,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2179232638230273,
+      "learning_rate": 9.999958508281042e-06,
+      "loss": 0.4806,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.830937218899892,
+      "learning_rate": 9.999954255386336e-06,
+      "loss": 0.5072,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5490979182484326,
+      "learning_rate": 9.999949795034643e-06,
+      "loss": 0.4835,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6043327569945123,
+      "learning_rate": 9.999945127226153e-06,
+      "loss": 0.473,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0698345439739447,
+      "learning_rate": 9.999940251961062e-06,
+      "loss": 0.4923,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.404757938812203,
+      "learning_rate": 9.999935169239564e-06,
+      "loss": 0.4903,
+      "step": 503
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.681536235612271,
+      "learning_rate": 9.999929879061879e-06,
+      "loss": 0.4953,
+      "step": 504
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.832575757357449,
+      "learning_rate": 9.999924381428221e-06,
+      "loss": 0.4759,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6505871660183593,
+      "learning_rate": 9.99991867633882e-06,
+      "loss": 0.4857,
+      "step": 506
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2055095409109557,
+      "learning_rate": 9.999912763793912e-06,
+      "loss": 0.5042,
+      "step": 507
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.707154179903497,
+      "learning_rate": 9.999906643793741e-06,
+      "loss": 0.4825,
+      "step": 508
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.032646923161125,
+      "learning_rate": 9.999900316338566e-06,
+      "loss": 0.4628,
+      "step": 509
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3106848272052103,
+      "learning_rate": 9.999893781428643e-06,
+      "loss": 0.4963,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1321540785077566,
+      "learning_rate": 9.999887039064248e-06,
+      "loss": 0.4897,
+      "step": 511
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9729172626842642,
+      "learning_rate": 9.999880089245659e-06,
+      "loss": 0.5195,
+      "step": 512
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9921180065923847,
+      "learning_rate": 9.999872931973163e-06,
+      "loss": 0.4993,
+      "step": 513
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4222433489211916,
+      "learning_rate": 9.999865567247058e-06,
+      "loss": 0.4661,
+      "step": 514
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4729050680514097,
+      "learning_rate": 9.999857995067652e-06,
+      "loss": 0.4864,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9891660124937793,
+      "learning_rate": 9.999850215435255e-06,
+      "loss": 0.4939,
+      "step": 516
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6145310260308054,
+      "learning_rate": 9.999842228350191e-06,
+      "loss": 0.5025,
+      "step": 517
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6840151848583955,
+      "learning_rate": 9.999834033812795e-06,
+      "loss": 0.4761,
+      "step": 518
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9479719370056958,
+      "learning_rate": 9.999825631823404e-06,
+      "loss": 0.445,
+      "step": 519
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9412623736620975,
+      "learning_rate": 9.999817022382365e-06,
+      "loss": 0.424,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.256004528716746,
+      "learning_rate": 9.999808205490038e-06,
+      "loss": 0.525,
+      "step": 521
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0329283510561034,
+      "learning_rate": 9.999799181146787e-06,
+      "loss": 0.4958,
+      "step": 522
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.29614027965801,
+      "learning_rate": 9.999789949352988e-06,
+      "loss": 0.4886,
+      "step": 523
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8802866180224957,
+      "learning_rate": 9.999780510109023e-06,
+      "loss": 0.4997,
+      "step": 524
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.958253765695981,
+      "learning_rate": 9.999770863415286e-06,
+      "loss": 0.5277,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2046301239602006,
+      "learning_rate": 9.999761009272174e-06,
+      "loss": 0.4725,
+      "step": 526
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5465710295017514,
+      "learning_rate": 9.999750947680096e-06,
+      "loss": 0.4969,
+      "step": 527
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.031761355371023,
+      "learning_rate": 9.999740678639471e-06,
+      "loss": 0.5108,
+      "step": 528
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0828099016664163,
+      "learning_rate": 9.999730202150726e-06,
+      "loss": 0.4825,
+      "step": 529
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9273603673172606,
+      "learning_rate": 9.999719518214293e-06,
+      "loss": 0.4704,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5018699726099425,
+      "learning_rate": 9.999708626830617e-06,
+      "loss": 0.4905,
+      "step": 531
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.504510016209695,
+      "learning_rate": 9.999697528000151e-06,
+      "loss": 0.5068,
+      "step": 532
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.626186179716801,
+      "learning_rate": 9.999686221723353e-06,
+      "loss": 0.4726,
+      "step": 533
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9030201544718344,
+      "learning_rate": 9.999674708000692e-06,
+      "loss": 0.4925,
+      "step": 534
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.297824608304791,
+      "learning_rate": 9.99966298683265e-06,
+      "loss": 0.5128,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4063063427872162,
+      "learning_rate": 9.999651058219708e-06,
+      "loss": 0.4323,
+      "step": 536
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9257458540615224,
+      "learning_rate": 9.999638922162363e-06,
+      "loss": 0.463,
+      "step": 537
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9692464760787463,
+      "learning_rate": 9.99962657866112e-06,
+      "loss": 0.4524,
+      "step": 538
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5905931244414884,
+      "learning_rate": 9.999614027716488e-06,
+      "loss": 0.5167,
+      "step": 539
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.410373295565918,
+      "learning_rate": 9.999601269328994e-06,
+      "loss": 0.4753,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5532275741843966,
+      "learning_rate": 9.99958830349916e-06,
+      "loss": 0.463,
+      "step": 541
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8810677571179277,
+      "learning_rate": 9.999575130227526e-06,
+      "loss": 0.4694,
+      "step": 542
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9790666546297664,
+      "learning_rate": 9.999561749514642e-06,
+      "loss": 0.4966,
+      "step": 543
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9908036371688174,
+      "learning_rate": 9.999548161361058e-06,
+      "loss": 0.457,
+      "step": 544
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.997095918538681,
+      "learning_rate": 9.999534365767342e-06,
+      "loss": 0.5354,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.08301974810039,
+      "learning_rate": 9.999520362734065e-06,
+      "loss": 0.4981,
+      "step": 546
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.231020304541261,
+      "learning_rate": 9.999506152261809e-06,
+      "loss": 0.4632,
+      "step": 547
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.551668641095965,
+      "learning_rate": 9.999491734351162e-06,
+      "loss": 0.4733,
+      "step": 548
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.180686411917383,
+      "learning_rate": 9.999477109002722e-06,
+      "loss": 0.4842,
+      "step": 549
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.72808462322627,
+      "learning_rate": 9.999462276217096e-06,
+      "loss": 0.4543,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2164396078924913,
+      "learning_rate": 9.999447235994902e-06,
+      "loss": 0.4276,
+      "step": 551
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0792152173641463,
+      "learning_rate": 9.999431988336762e-06,
+      "loss": 0.4595,
+      "step": 552
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4026681979538793,
+      "learning_rate": 9.999416533243309e-06,
+      "loss": 0.465,
+      "step": 553
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.006484644283491,
+      "learning_rate": 9.999400870715182e-06,
+      "loss": 0.4442,
+      "step": 554
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6070388377684557,
+      "learning_rate": 9.999385000753034e-06,
+      "loss": 0.4676,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.040094846791098,
+      "learning_rate": 9.999368923357525e-06,
+      "loss": 0.4711,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0506551892026024,
+      "learning_rate": 9.999352638529316e-06,
+      "loss": 0.4602,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.770924960955531,
+      "learning_rate": 9.999336146269088e-06,
+      "loss": 0.4597,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.444707020009515,
+      "learning_rate": 9.999319446577523e-06,
+      "loss": 0.4944,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0794848971406736,
+      "learning_rate": 9.999302539455314e-06,
+      "loss": 0.4328,
+      "step": 560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.968220333607118,
+      "learning_rate": 9.999285424903163e-06,
+      "loss": 0.4923,
+      "step": 561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2070225116189857,
+      "learning_rate": 9.99926810292178e-06,
+      "loss": 0.4732,
+      "step": 562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9672062941744224,
+      "learning_rate": 9.999250573511883e-06,
+      "loss": 0.4457,
+      "step": 563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7397800191770814,
+      "learning_rate": 9.999232836674202e-06,
+      "loss": 0.42,
+      "step": 564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.955989009342422,
+      "learning_rate": 9.99921489240947e-06,
+      "loss": 0.4446,
+      "step": 565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9205981841495485,
+      "learning_rate": 9.999196740718432e-06,
+      "loss": 0.438,
+      "step": 566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.522660114624852,
+      "learning_rate": 9.999178381601842e-06,
+      "loss": 0.4845,
+      "step": 567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4620507763502344,
+      "learning_rate": 9.999159815060462e-06,
+      "loss": 0.4804,
+      "step": 568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.337410846560435,
+      "learning_rate": 9.99914104109506e-06,
+      "loss": 0.4626,
+      "step": 569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8850082114866145,
+      "learning_rate": 9.999122059706418e-06,
+      "loss": 0.4671,
+      "step": 570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8278181600093955,
+      "learning_rate": 9.999102870895323e-06,
+      "loss": 0.4578,
+      "step": 571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.489333267750027,
+      "learning_rate": 9.99908347466257e-06,
+      "loss": 0.4449,
+      "step": 572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.618328682479948,
+      "learning_rate": 9.999063871008963e-06,
+      "loss": 0.4658,
+      "step": 573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9802982929547728,
+      "learning_rate": 9.999044059935319e-06,
+      "loss": 0.4391,
+      "step": 574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8110250724553485,
+      "learning_rate": 9.999024041442455e-06,
+      "loss": 0.4685,
+      "step": 575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2616464759803057,
+      "learning_rate": 9.999003815531206e-06,
+      "loss": 0.457,
+      "step": 576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.187140130650438,
+      "learning_rate": 9.998983382202408e-06,
+      "loss": 0.4411,
+      "step": 577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3819418136366752,
+      "learning_rate": 9.998962741456912e-06,
+      "loss": 0.449,
+      "step": 578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7313485182956,
+      "learning_rate": 9.998941893295572e-06,
+      "loss": 0.4712,
+      "step": 579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.063108026944435,
+      "learning_rate": 9.998920837719254e-06,
+      "loss": 0.46,
+      "step": 580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.088599257715193,
+      "learning_rate": 9.998899574728832e-06,
+      "loss": 0.4623,
+      "step": 581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.208467206475676,
+      "learning_rate": 9.998878104325186e-06,
+      "loss": 0.5056,
+      "step": 582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.566290177886027,
+      "learning_rate": 9.998856426509208e-06,
+      "loss": 0.4744,
+      "step": 583
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.133215283066703,
+      "learning_rate": 9.998834541281798e-06,
+      "loss": 0.4992,
+      "step": 584
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.530779373872894,
+      "learning_rate": 9.998812448643866e-06,
+      "loss": 0.4671,
+      "step": 585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6625767048203386,
+      "learning_rate": 9.998790148596326e-06,
+      "loss": 0.4315,
+      "step": 586
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8671590408093837,
+      "learning_rate": 9.998767641140103e-06,
+      "loss": 0.4998,
+      "step": 587
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0928109248997626,
+      "learning_rate": 9.998744926276132e-06,
+      "loss": 0.4703,
+      "step": 588
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.2506176259906834,
+      "learning_rate": 9.998722004005356e-06,
+      "loss": 0.4782,
+      "step": 589
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9063269538020393,
+      "learning_rate": 9.998698874328725e-06,
+      "loss": 0.4609,
+      "step": 590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.027797630615375,
+      "learning_rate": 9.998675537247199e-06,
+      "loss": 0.4696,
+      "step": 591
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0533268864376435,
+      "learning_rate": 9.998651992761746e-06,
+      "loss": 0.4694,
+      "step": 592
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4804998394103683,
+      "learning_rate": 9.998628240873342e-06,
+      "loss": 0.4419,
+      "step": 593
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.795088503320507,
+      "learning_rate": 9.998604281582975e-06,
+      "loss": 0.4994,
+      "step": 594
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9220824194919697,
+      "learning_rate": 9.998580114891638e-06,
+      "loss": 0.4541,
+      "step": 595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8091027763660537,
+      "learning_rate": 9.998555740800335e-06,
+      "loss": 0.4528,
+      "step": 596
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.594109389576467,
+      "learning_rate": 9.998531159310074e-06,
+      "loss": 0.462,
+      "step": 597
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0457223813804037,
+      "learning_rate": 9.998506370421876e-06,
+      "loss": 0.4449,
+      "step": 598
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.364388799628461,
+      "learning_rate": 9.998481374136773e-06,
+      "loss": 0.4438,
+      "step": 599
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.225475904704627,
+      "learning_rate": 9.998456170455796e-06,
+      "loss": 0.4612,
+      "step": 600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.540161650441191,
+      "learning_rate": 9.998430759379999e-06,
+      "loss": 0.458,
+      "step": 601
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.602362536891887,
+      "learning_rate": 9.998405140910427e-06,
+      "loss": 0.4778,
+      "step": 602
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5625773310302615,
+      "learning_rate": 9.99837931504815e-06,
+      "loss": 0.4648,
+      "step": 603
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3451024621950423,
+      "learning_rate": 9.998353281794235e-06,
+      "loss": 0.4826,
+      "step": 604
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.599579513154246,
+      "learning_rate": 9.998327041149766e-06,
+      "loss": 0.4693,
+      "step": 605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9164031153911307,
+      "learning_rate": 9.998300593115831e-06,
+      "loss": 0.4559,
+      "step": 606
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9919266447646287,
+      "learning_rate": 9.998273937693525e-06,
+      "loss": 0.4275,
+      "step": 607
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.858787089836137,
+      "learning_rate": 9.998247074883956e-06,
+      "loss": 0.4338,
+      "step": 608
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.952562833224354,
+      "learning_rate": 9.998220004688238e-06,
+      "loss": 0.5,
+      "step": 609
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9821066259376316,
+      "learning_rate": 9.998192727107492e-06,
+      "loss": 0.4838,
+      "step": 610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8465287076621513,
+      "learning_rate": 9.998165242142855e-06,
+      "loss": 0.4511,
+      "step": 611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8975261047111254,
+      "learning_rate": 9.998137549795462e-06,
+      "loss": 0.4217,
+      "step": 612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.664265661419895,
+      "learning_rate": 9.998109650066465e-06,
+      "loss": 0.4556,
+      "step": 613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.024994077208901,
+      "learning_rate": 9.998081542957021e-06,
+      "loss": 0.4598,
+      "step": 614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.332872265825148,
+      "learning_rate": 9.998053228468296e-06,
+      "loss": 0.4686,
+      "step": 615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6914635597004524,
+      "learning_rate": 9.998024706601467e-06,
+      "loss": 0.4297,
+      "step": 616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9640858067208848,
+      "learning_rate": 9.997995977357712e-06,
+      "loss": 0.4164,
+      "step": 617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.731815997672232,
+      "learning_rate": 9.99796704073823e-06,
+      "loss": 0.4402,
+      "step": 618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4062689009583236,
+      "learning_rate": 9.997937896744216e-06,
+      "loss": 0.4367,
+      "step": 619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8611460221301015,
+      "learning_rate": 9.997908545376883e-06,
+      "loss": 0.4446,
+      "step": 620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9255996259506856,
+      "learning_rate": 9.997878986637446e-06,
+      "loss": 0.4364,
+      "step": 621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.832861131931658,
+      "learning_rate": 9.997849220527132e-06,
+      "loss": 0.461,
+      "step": 622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2911431679187215,
+      "learning_rate": 9.997819247047177e-06,
+      "loss": 0.439,
+      "step": 623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1394718060547646,
+      "learning_rate": 9.997789066198824e-06,
+      "loss": 0.459,
+      "step": 624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.768603478672137,
+      "learning_rate": 9.997758677983327e-06,
+      "loss": 0.5346,
+      "step": 625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9595764953684403,
+      "learning_rate": 9.997728082401946e-06,
+      "loss": 0.4677,
+      "step": 626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.689702247590821,
+      "learning_rate": 9.997697279455947e-06,
+      "loss": 0.4479,
+      "step": 627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8691749958551407,
+      "learning_rate": 9.997666269146612e-06,
+      "loss": 0.4684,
+      "step": 628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.588557160746207,
+      "learning_rate": 9.997635051475227e-06,
+      "loss": 0.4452,
+      "step": 629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.332900061870853,
+      "learning_rate": 9.997603626443088e-06,
+      "loss": 0.4617,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7491579573049527,
+      "learning_rate": 9.997571994051497e-06,
+      "loss": 0.4861,
+      "step": 631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.600134242711322,
+      "learning_rate": 9.997540154301766e-06,
+      "loss": 0.4686,
+      "step": 632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2919323119550588,
+      "learning_rate": 9.99750810719522e-06,
+      "loss": 0.4816,
+      "step": 633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3933176511759915,
+      "learning_rate": 9.997475852733183e-06,
+      "loss": 0.4721,
+      "step": 634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7704070087461385,
+      "learning_rate": 9.997443390916999e-06,
+      "loss": 0.426,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6750412033533468,
+      "learning_rate": 9.99741072174801e-06,
+      "loss": 0.4595,
+      "step": 636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7392006563073865,
+      "learning_rate": 9.997377845227577e-06,
+      "loss": 0.4473,
+      "step": 637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.982753584421421,
+      "learning_rate": 9.997344761357057e-06,
+      "loss": 0.4834,
+      "step": 638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.542893874456173,
+      "learning_rate": 9.997311470137828e-06,
+      "loss": 0.4372,
+      "step": 639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3783093034128138,
+      "learning_rate": 9.99727797157127e-06,
+      "loss": 0.4505,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.863024260620193,
+      "learning_rate": 9.997244265658774e-06,
+      "loss": 0.4273,
+      "step": 641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.010156445925229,
+      "learning_rate": 9.997210352401736e-06,
+      "loss": 0.4213,
+      "step": 642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9513288428258124,
+      "learning_rate": 9.997176231801565e-06,
+      "loss": 0.4659,
+      "step": 643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5284652329277972,
+      "learning_rate": 9.997141903859675e-06,
+      "loss": 0.5051,
+      "step": 644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.767869147714192,
+      "learning_rate": 9.997107368577492e-06,
+      "loss": 0.4745,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1508694750132054,
+      "learning_rate": 9.99707262595645e-06,
+      "loss": 0.4139,
+      "step": 646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.11583835038122,
+      "learning_rate": 9.997037675997987e-06,
+      "loss": 0.4698,
+      "step": 647
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.515650691332356,
+      "learning_rate": 9.997002518703556e-06,
+      "loss": 0.455,
+      "step": 648
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9175224977998657,
+      "learning_rate": 9.996967154074615e-06,
+      "loss": 0.4673,
+      "step": 649
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9242834891154732,
+      "learning_rate": 9.99693158211263e-06,
+      "loss": 0.4762,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9173214362306026,
+      "learning_rate": 9.99689580281908e-06,
+      "loss": 0.4902,
+      "step": 651
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2670351609030837,
+      "learning_rate": 9.996859816195446e-06,
+      "loss": 0.4644,
+      "step": 652
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8798348386074981,
+      "learning_rate": 9.996823622243224e-06,
+      "loss": 0.4319,
+      "step": 653
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8049037930911456,
+      "learning_rate": 9.996787220963915e-06,
+      "loss": 0.4488,
+      "step": 654
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.365755338439417,
+      "learning_rate": 9.996750612359026e-06,
+      "loss": 0.4799,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.477361241354372,
+      "learning_rate": 9.99671379643008e-06,
+      "loss": 0.4389,
+      "step": 656
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7752975651097125,
+      "learning_rate": 9.996676773178604e-06,
+      "loss": 0.4557,
+      "step": 657
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6547541148359164,
+      "learning_rate": 9.996639542606135e-06,
+      "loss": 0.4828,
+      "step": 658
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2922398431103956,
+      "learning_rate": 9.996602104714215e-06,
+      "loss": 0.4689,
+      "step": 659
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6653628345815956,
+      "learning_rate": 9.996564459504398e-06,
+      "loss": 0.4443,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5935263181365933,
+      "learning_rate": 9.996526606978248e-06,
+      "loss": 0.4537,
+      "step": 661
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.212712442498005,
+      "learning_rate": 9.996488547137335e-06,
+      "loss": 0.4419,
+      "step": 662
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9478331010182452,
+      "learning_rate": 9.996450279983236e-06,
+      "loss": 0.4565,
+      "step": 663
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5278859392506945,
+      "learning_rate": 9.99641180551754e-06,
+      "loss": 0.4433,
+      "step": 664
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.027332182723469,
+      "learning_rate": 9.996373123741843e-06,
+      "loss": 0.4834,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.073061441534379,
+      "learning_rate": 9.996334234657751e-06,
+      "loss": 0.4703,
+      "step": 666
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1932471783137433,
+      "learning_rate": 9.996295138266877e-06,
+      "loss": 0.4501,
+      "step": 667
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.318954452112392,
+      "learning_rate": 9.996255834570844e-06,
+      "loss": 0.4785,
+      "step": 668
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.867713605240658,
+      "learning_rate": 9.996216323571283e-06,
+      "loss": 0.446,
+      "step": 669
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7452091963528424,
+      "learning_rate": 9.996176605269832e-06,
+      "loss": 0.5069,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.681731236826786,
+      "learning_rate": 9.996136679668138e-06,
+      "loss": 0.4774,
+      "step": 671
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5679787343301403,
+      "learning_rate": 9.99609654676786e-06,
+      "loss": 0.4378,
+      "step": 672
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.285230500434291,
+      "learning_rate": 9.996056206570662e-06,
+      "loss": 0.4563,
+      "step": 673
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.549352866928325,
+      "learning_rate": 9.996015659078218e-06,
+      "loss": 0.4421,
+      "step": 674
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6743044639961289,
+      "learning_rate": 9.995974904292211e-06,
+      "loss": 0.4519,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.944103245358746,
+      "learning_rate": 9.995933942214331e-06,
+      "loss": 0.439,
+      "step": 676
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.178578626812516,
+      "learning_rate": 9.995892772846276e-06,
+      "loss": 0.4657,
+      "step": 677
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8834378335862514,
+      "learning_rate": 9.995851396189759e-06,
+      "loss": 0.4766,
+      "step": 678
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3943764794326166,
+      "learning_rate": 9.995809812246493e-06,
+      "loss": 0.4434,
+      "step": 679
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0256693185920414,
+      "learning_rate": 9.995768021018204e-06,
+      "loss": 0.4597,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.066097959989994,
+      "learning_rate": 9.995726022506627e-06,
+      "loss": 0.4364,
+      "step": 681
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.765308871003166,
+      "learning_rate": 9.995683816713503e-06,
+      "loss": 0.4723,
+      "step": 682
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.847147381689037,
+      "learning_rate": 9.995641403640585e-06,
+      "loss": 0.4761,
+      "step": 683
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4559454985726097,
+      "learning_rate": 9.995598783289631e-06,
+      "loss": 0.4727,
+      "step": 684
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.113195011765131,
+      "learning_rate": 9.99555595566241e-06,
+      "loss": 0.436,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1883784100307166,
+      "learning_rate": 9.9955129207607e-06,
+      "loss": 0.4867,
+      "step": 686
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.553270239468023,
+      "learning_rate": 9.995469678586286e-06,
+      "loss": 0.5049,
+      "step": 687
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5839827658726273,
+      "learning_rate": 9.995426229140963e-06,
+      "loss": 0.4729,
+      "step": 688
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6757771629648346,
+      "learning_rate": 9.995382572426531e-06,
+      "loss": 0.425,
+      "step": 689
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8114969729388422,
+      "learning_rate": 9.995338708444804e-06,
+      "loss": 0.4355,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.459675253701771,
+      "learning_rate": 9.995294637197602e-06,
+      "loss": 0.4854,
+      "step": 691
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7511074507439903,
+      "learning_rate": 9.995250358686753e-06,
+      "loss": 0.4154,
+      "step": 692
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.316640599823098,
+      "learning_rate": 9.995205872914094e-06,
+      "loss": 0.4188,
+      "step": 693
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.359760985100658,
+      "learning_rate": 9.995161179881469e-06,
+      "loss": 0.442,
+      "step": 694
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.788610007610766,
+      "learning_rate": 9.995116279590735e-06,
+      "loss": 0.4302,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0605553313415297,
+      "learning_rate": 9.995071172043755e-06,
+      "loss": 0.4065,
+      "step": 696
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.709277814057277,
+      "learning_rate": 9.9950258572424e-06,
+      "loss": 0.4429,
+      "step": 697
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8173910399442992,
+      "learning_rate": 9.994980335188549e-06,
+      "loss": 0.4684,
+      "step": 698
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8807990544036073,
+      "learning_rate": 9.994934605884093e-06,
+      "loss": 0.4423,
+      "step": 699
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.07708957760124,
+      "learning_rate": 9.994888669330927e-06,
+      "loss": 0.4473,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5759907952586523,
+      "learning_rate": 9.99484252553096e-06,
+      "loss": 0.4489,
+      "step": 701
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.401807864783039,
+      "learning_rate": 9.9947961744861e-06,
+      "loss": 0.458,
+      "step": 702
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8853034430839495,
+      "learning_rate": 9.99474961619828e-06,
+      "loss": 0.4362,
+      "step": 703
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4146421279508816,
+      "learning_rate": 9.994702850669426e-06,
+      "loss": 0.4408,
+      "step": 704
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6001923652749257,
+      "learning_rate": 9.994655877901479e-06,
+      "loss": 0.4403,
+      "step": 705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7221897854324726,
+      "learning_rate": 9.994608697896386e-06,
+      "loss": 0.4425,
+      "step": 706
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9029907396395576,
+      "learning_rate": 9.994561310656107e-06,
+      "loss": 0.4155,
+      "step": 707
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.219955758206292,
+      "learning_rate": 9.99451371618261e-06,
+      "loss": 0.4613,
+      "step": 708
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9205200134365303,
+      "learning_rate": 9.994465914477866e-06,
+      "loss": 0.4491,
+      "step": 709
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3594627345880355,
+      "learning_rate": 9.994417905543863e-06,
+      "loss": 0.4623,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7961548354447747,
+      "learning_rate": 9.994369689382586e-06,
+      "loss": 0.4737,
+      "step": 711
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.153950418069649,
+      "learning_rate": 9.994321265996043e-06,
+      "loss": 0.4557,
+      "step": 712
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7803043259836775,
+      "learning_rate": 9.994272635386238e-06,
+      "loss": 0.4391,
+      "step": 713
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.105188709639193,
+      "learning_rate": 9.99422379755519e-06,
+      "loss": 0.4386,
+      "step": 714
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4916189281607455,
+      "learning_rate": 9.994174752504926e-06,
+      "loss": 0.4487,
+      "step": 715
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8734568379953702,
+      "learning_rate": 9.994125500237482e-06,
+      "loss": 0.4182,
+      "step": 716
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.291847542999621,
+      "learning_rate": 9.994076040754898e-06,
+      "loss": 0.4724,
+      "step": 717
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.647137404633726,
+      "learning_rate": 9.994026374059232e-06,
+      "loss": 0.4919,
+      "step": 718
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3424631174226507,
+      "learning_rate": 9.99397650015254e-06,
+      "loss": 0.4566,
+      "step": 719
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.928470910158565,
+      "learning_rate": 9.993926419036893e-06,
+      "loss": 0.4845,
+      "step": 720
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3142700080597045,
+      "learning_rate": 9.993876130714367e-06,
+      "loss": 0.4495,
+      "step": 721
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.046292210969857,
+      "learning_rate": 9.993825635187052e-06,
+      "loss": 0.4376,
+      "step": 722
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7066143872941533,
+      "learning_rate": 9.99377493245704e-06,
+      "loss": 0.442,
+      "step": 723
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0008424541582586,
+      "learning_rate": 9.993724022526436e-06,
+      "loss": 0.4551,
+      "step": 724
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7770143116553527,
+      "learning_rate": 9.993672905397353e-06,
+      "loss": 0.4866,
+      "step": 725
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3870371251726765,
+      "learning_rate": 9.993621581071913e-06,
+      "loss": 0.4225,
+      "step": 726
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.239640924878757,
+      "learning_rate": 9.993570049552242e-06,
+      "loss": 0.4549,
+      "step": 727
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.186189260968605,
+      "learning_rate": 9.993518310840481e-06,
+      "loss": 0.4115,
+      "step": 728
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8149784852264021,
+      "learning_rate": 9.993466364938776e-06,
+      "loss": 0.4411,
+      "step": 729
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4569245728822007,
+      "learning_rate": 9.993414211849281e-06,
+      "loss": 0.428,
+      "step": 730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.769733089295754,
+      "learning_rate": 9.993361851574162e-06,
+      "loss": 0.431,
+      "step": 731
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3420693054026698,
+      "learning_rate": 9.99330928411559e-06,
+      "loss": 0.4612,
+      "step": 732
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2476509824120647,
+      "learning_rate": 9.993256509475746e-06,
+      "loss": 0.4488,
+      "step": 733
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.840328309573542,
+      "learning_rate": 9.993203527656822e-06,
+      "loss": 0.437,
+      "step": 734
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.792612558054085,
+      "learning_rate": 9.993150338661014e-06,
+      "loss": 0.4652,
+      "step": 735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.49519143632042,
+      "learning_rate": 9.99309694249053e-06,
+      "loss": 0.4302,
+      "step": 736
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.260931105451376,
+      "learning_rate": 9.993043339147584e-06,
+      "loss": 0.4485,
+      "step": 737
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9443299972263728,
+      "learning_rate": 9.992989528634403e-06,
+      "loss": 0.4193,
+      "step": 738
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8855198794919493,
+      "learning_rate": 9.992935510953216e-06,
+      "loss": 0.4948,
+      "step": 739
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.852162438384942,
+      "learning_rate": 9.992881286106268e-06,
+      "loss": 0.4686,
+      "step": 740
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6678759582183533,
+      "learning_rate": 9.992826854095806e-06,
+      "loss": 0.4802,
+      "step": 741
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4981249056088424,
+      "learning_rate": 9.992772214924089e-06,
+      "loss": 0.4366,
+      "step": 742
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9979541013190596,
+      "learning_rate": 9.992717368593385e-06,
+      "loss": 0.4504,
+      "step": 743
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0834487058102034,
+      "learning_rate": 9.99266231510597e-06,
+      "loss": 0.4484,
+      "step": 744
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8185227599352154,
+      "learning_rate": 9.992607054464128e-06,
+      "loss": 0.4698,
+      "step": 745
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7366592197539074,
+      "learning_rate": 9.99255158667015e-06,
+      "loss": 0.4665,
+      "step": 746
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8722742353505182,
+      "learning_rate": 9.99249591172634e-06,
+      "loss": 0.4252,
+      "step": 747
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.049576533626131,
+      "learning_rate": 9.992440029635007e-06,
+      "loss": 0.4645,
+      "step": 748
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5672077229256496,
+      "learning_rate": 9.99238394039847e-06,
+      "loss": 0.4056,
+      "step": 749
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5925498825462083,
+      "learning_rate": 9.992327644019053e-06,
+      "loss": 0.434,
+      "step": 750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8303015713255473,
+      "learning_rate": 9.992271140499096e-06,
+      "loss": 0.4563,
+      "step": 751
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4236142884002383,
+      "learning_rate": 9.992214429840944e-06,
+      "loss": 0.459,
+      "step": 752
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.752914195951192,
+      "learning_rate": 9.992157512046947e-06,
+      "loss": 0.4489,
+      "step": 753
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.115485119650311,
+      "learning_rate": 9.992100387119468e-06,
+      "loss": 0.4452,
+      "step": 754
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8854156175407226,
+      "learning_rate": 9.992043055060876e-06,
+      "loss": 0.4402,
+      "step": 755
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7833261170781056,
+      "learning_rate": 9.991985515873552e-06,
+      "loss": 0.4629,
+      "step": 756
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9153728045052216,
+      "learning_rate": 9.991927769559882e-06,
+      "loss": 0.4416,
+      "step": 757
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.783200603276375,
+      "learning_rate": 9.991869816122262e-06,
+      "loss": 0.4675,
+      "step": 758
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4475071637171941,
+      "learning_rate": 9.991811655563096e-06,
+      "loss": 0.4638,
+      "step": 759
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8403988649058063,
+      "learning_rate": 9.991753287884797e-06,
+      "loss": 0.4626,
+      "step": 760
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.055641677306497,
+      "learning_rate": 9.99169471308979e-06,
+      "loss": 0.4386,
+      "step": 761
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.063193682721805,
+      "learning_rate": 9.991635931180504e-06,
+      "loss": 0.4363,
+      "step": 762
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7133765269253347,
+      "learning_rate": 9.991576942159374e-06,
+      "loss": 0.4595,
+      "step": 763
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9417508822029572,
+      "learning_rate": 9.991517746028851e-06,
+      "loss": 0.5405,
+      "step": 764
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.548362709307829,
+      "learning_rate": 9.991458342791394e-06,
+      "loss": 0.4753,
+      "step": 765
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7744718817859155,
+      "learning_rate": 9.991398732449461e-06,
+      "loss": 0.4488,
+      "step": 766
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.506629010419364,
+      "learning_rate": 9.991338915005531e-06,
+      "loss": 0.4439,
+      "step": 767
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.746150729627741,
+      "learning_rate": 9.991278890462083e-06,
+      "loss": 0.4498,
+      "step": 768
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1926730472137432,
+      "learning_rate": 9.991218658821609e-06,
+      "loss": 0.4984,
+      "step": 769
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.846841983617649,
+      "learning_rate": 9.991158220086606e-06,
+      "loss": 0.4453,
+      "step": 770
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.907955257438044,
+      "learning_rate": 9.991097574259583e-06,
+      "loss": 0.446,
+      "step": 771
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5396188166622897,
+      "learning_rate": 9.991036721343058e-06,
+      "loss": 0.4506,
+      "step": 772
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9119711306037535,
+      "learning_rate": 9.990975661339554e-06,
+      "loss": 0.4549,
+      "step": 773
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1540435742691395,
+      "learning_rate": 9.990914394251605e-06,
+      "loss": 0.434,
+      "step": 774
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2126268391679975,
+      "learning_rate": 9.990852920081753e-06,
+      "loss": 0.4532,
+      "step": 775
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0377088083931834,
+      "learning_rate": 9.990791238832547e-06,
+      "loss": 0.4291,
+      "step": 776
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8288013685619777,
+      "learning_rate": 9.99072935050655e-06,
+      "loss": 0.4408,
+      "step": 777
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.781560694504767,
+      "learning_rate": 9.990667255106326e-06,
+      "loss": 0.4358,
+      "step": 778
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9714217622609607,
+      "learning_rate": 9.990604952634452e-06,
+      "loss": 0.4844,
+      "step": 779
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6237259170251106,
+      "learning_rate": 9.990542443093518e-06,
+      "loss": 0.4287,
+      "step": 780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0796091992235115,
+      "learning_rate": 9.990479726486111e-06,
+      "loss": 0.4779,
+      "step": 781
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5816331887509902,
+      "learning_rate": 9.990416802814838e-06,
+      "loss": 0.4233,
+      "step": 782
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.198578168807564,
+      "learning_rate": 9.990353672082307e-06,
+      "loss": 0.4779,
+      "step": 783
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.31231238158569,
+      "learning_rate": 9.990290334291137e-06,
+      "loss": 0.4262,
+      "step": 784
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7891061349189314,
+      "learning_rate": 9.990226789443959e-06,
+      "loss": 0.4686,
+      "step": 785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9574795554447961,
+      "learning_rate": 9.990163037543407e-06,
+      "loss": 0.451,
+      "step": 786
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9366174622724797,
+      "learning_rate": 9.990099078592128e-06,
+      "loss": 0.4428,
+      "step": 787
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.047931041875674,
+      "learning_rate": 9.990034912592774e-06,
+      "loss": 0.4289,
+      "step": 788
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2617071894422747,
+      "learning_rate": 9.98997053954801e-06,
+      "loss": 0.514,
+      "step": 789
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6326084843093283,
+      "learning_rate": 9.989905959460502e-06,
+      "loss": 0.4517,
+      "step": 790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8286255469673063,
+      "learning_rate": 9.989841172332932e-06,
+      "loss": 0.4562,
+      "step": 791
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.398080419481011,
+      "learning_rate": 9.989776178167991e-06,
+      "loss": 0.422,
+      "step": 792
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.451760998367041,
+      "learning_rate": 9.989710976968375e-06,
+      "loss": 0.4161,
+      "step": 793
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8979467286199225,
+      "learning_rate": 9.989645568736786e-06,
+      "loss": 0.4751,
+      "step": 794
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3096835223998924,
+      "learning_rate": 9.98957995347594e-06,
+      "loss": 0.4297,
+      "step": 795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.255040445356041,
+      "learning_rate": 9.98951413118856e-06,
+      "loss": 0.4463,
+      "step": 796
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3613276402916803,
+      "learning_rate": 9.989448101877375e-06,
+      "loss": 0.466,
+      "step": 797
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3231044170074253,
+      "learning_rate": 9.989381865545128e-06,
+      "loss": 0.4539,
+      "step": 798
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7764446748922416,
+      "learning_rate": 9.989315422194562e-06,
+      "loss": 0.4217,
+      "step": 799
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.944426116231744,
+      "learning_rate": 9.98924877182844e-06,
+      "loss": 0.426,
+      "step": 800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.413016157089324,
+      "learning_rate": 9.989181914449523e-06,
+      "loss": 0.4499,
+      "step": 801
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8796354863110096,
+      "learning_rate": 9.989114850060586e-06,
+      "loss": 0.4537,
+      "step": 802
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.30681700893175,
+      "learning_rate": 9.989047578664414e-06,
+      "loss": 0.4339,
+      "step": 803
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8686538396109308,
+      "learning_rate": 9.988980100263796e-06,
+      "loss": 0.468,
+      "step": 804
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3910489222265263,
+      "learning_rate": 9.988912414861531e-06,
+      "loss": 0.4618,
+      "step": 805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0476596917988434,
+      "learning_rate": 9.98884452246043e-06,
+      "loss": 0.4335,
+      "step": 806
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9450486515086292,
+      "learning_rate": 9.988776423063309e-06,
+      "loss": 0.4389,
+      "step": 807
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.002990635676293,
+      "learning_rate": 9.988708116672991e-06,
+      "loss": 0.4735,
+      "step": 808
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.92066119516322,
+      "learning_rate": 9.988639603292315e-06,
+      "loss": 0.3943,
+      "step": 809
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7646101488911214,
+      "learning_rate": 9.988570882924118e-06,
+      "loss": 0.4199,
+      "step": 810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6045767909297195,
+      "learning_rate": 9.988501955571257e-06,
+      "loss": 0.4504,
+      "step": 811
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7229917539663575,
+      "learning_rate": 9.988432821236588e-06,
+      "loss": 0.473,
+      "step": 812
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.726949425061288,
+      "learning_rate": 9.98836347992298e-06,
+      "loss": 0.4415,
+      "step": 813
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.338904531000487,
+      "learning_rate": 9.988293931633312e-06,
+      "loss": 0.4535,
+      "step": 814
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.609909544928862,
+      "learning_rate": 9.988224176370468e-06,
+      "loss": 0.4924,
+      "step": 815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.219390158335007,
+      "learning_rate": 9.988154214137345e-06,
+      "loss": 0.4382,
+      "step": 816
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.407382476139027,
+      "learning_rate": 9.98808404493684e-06,
+      "loss": 0.4437,
+      "step": 817
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7609613674124398,
+      "learning_rate": 9.988013668771872e-06,
+      "loss": 0.4226,
+      "step": 818
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8812049522958381,
+      "learning_rate": 9.987943085645355e-06,
+      "loss": 0.5146,
+      "step": 819
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9054287777952734,
+      "learning_rate": 9.98787229556022e-06,
+      "loss": 0.4275,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3132004263071244,
+      "learning_rate": 9.987801298519404e-06,
+      "loss": 0.4744,
+      "step": 821
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.808889943407105,
+      "learning_rate": 9.987730094525854e-06,
+      "loss": 0.4171,
+      "step": 822
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.472432442961767,
+      "learning_rate": 9.987658683582522e-06,
+      "loss": 0.4471,
+      "step": 823
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.818475863719265,
+      "learning_rate": 9.98758706569237e-06,
+      "loss": 0.4252,
+      "step": 824
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.308317465855414,
+      "learning_rate": 9.987515240858375e-06,
+      "loss": 0.4322,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9523186706404483,
+      "learning_rate": 9.987443209083513e-06,
+      "loss": 0.4434,
+      "step": 826
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.841784401817416,
+      "learning_rate": 9.987370970370773e-06,
+      "loss": 0.4176,
+      "step": 827
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.95961024478376,
+      "learning_rate": 9.987298524723153e-06,
+      "loss": 0.4287,
+      "step": 828
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4942654397408663,
+      "learning_rate": 9.98722587214366e-06,
+      "loss": 0.4205,
+      "step": 829
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.350426410910016,
+      "learning_rate": 9.987153012635305e-06,
+      "loss": 0.4199,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.12802650657785,
+      "learning_rate": 9.987079946201114e-06,
+      "loss": 0.4457,
+      "step": 831
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8872701395081841,
+      "learning_rate": 9.987006672844119e-06,
+      "loss": 0.453,
+      "step": 832
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 18.343687856141752,
+      "learning_rate": 9.98693319256736e-06,
+      "loss": 0.3863,
+      "step": 833
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1537149097587385,
+      "learning_rate": 9.986859505373882e-06,
+      "loss": 0.4922,
+      "step": 834
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8026116562799057,
+      "learning_rate": 9.986785611266749e-06,
+      "loss": 0.4246,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8120638282725512,
+      "learning_rate": 9.986711510249021e-06,
+      "loss": 0.4441,
+      "step": 836
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0593819798683164,
+      "learning_rate": 9.986637202323777e-06,
+      "loss": 0.3827,
+      "step": 837
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4497291118023137,
+      "learning_rate": 9.986562687494096e-06,
+      "loss": 0.4348,
+      "step": 838
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.510626177549304,
+      "learning_rate": 9.986487965763073e-06,
+      "loss": 0.4411,
+      "step": 839
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9839747214140684,
+      "learning_rate": 9.986413037133806e-06,
+      "loss": 0.4852,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.030492457333062,
+      "learning_rate": 9.986337901609407e-06,
+      "loss": 0.4245,
+      "step": 841
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9773631136619981,
+      "learning_rate": 9.986262559192992e-06,
+      "loss": 0.4883,
+      "step": 842
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7266203884732154,
+      "learning_rate": 9.986187009887685e-06,
+      "loss": 0.4491,
+      "step": 843
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7682562026424797,
+      "learning_rate": 9.986111253696625e-06,
+      "loss": 0.4583,
+      "step": 844
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.088482057876728,
+      "learning_rate": 9.98603529062295e-06,
+      "loss": 0.4289,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.249553220163194,
+      "learning_rate": 9.985959120669816e-06,
+      "loss": 0.4257,
+      "step": 846
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1260195024993584,
+      "learning_rate": 9.985882743840383e-06,
+      "loss": 0.412,
+      "step": 847
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6891161433123363,
+      "learning_rate": 9.985806160137816e-06,
+      "loss": 0.4203,
+      "step": 848
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8324367024281996,
+      "learning_rate": 9.985729369565299e-06,
+      "loss": 0.4353,
+      "step": 849
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.344083468436015,
+      "learning_rate": 9.985652372126013e-06,
+      "loss": 0.4629,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7443403482437772,
+      "learning_rate": 9.985575167823154e-06,
+      "loss": 0.4648,
+      "step": 851
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2278136312942904,
+      "learning_rate": 9.985497756659927e-06,
+      "loss": 0.4308,
+      "step": 852
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.054966974169479,
+      "learning_rate": 9.985420138639543e-06,
+      "loss": 0.4502,
+      "step": 853
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7331177037284515,
+      "learning_rate": 9.985342313765223e-06,
+      "loss": 0.4313,
+      "step": 854
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8738889044291418,
+      "learning_rate": 9.985264282040195e-06,
+      "loss": 0.4138,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.393667567747854,
+      "learning_rate": 9.985186043467697e-06,
+      "loss": 0.431,
+      "step": 856
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.335268424978182,
+      "learning_rate": 9.985107598050973e-06,
+      "loss": 0.4578,
+      "step": 857
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.31628435357004,
+      "learning_rate": 9.985028945793282e-06,
+      "loss": 0.4262,
+      "step": 858
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6258180779200866,
+      "learning_rate": 9.984950086697886e-06,
+      "loss": 0.4547,
+      "step": 859
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.423719826873862,
+      "learning_rate": 9.984871020768056e-06,
+      "loss": 0.4663,
+      "step": 860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8648510078726412,
+      "learning_rate": 9.984791748007074e-06,
+      "loss": 0.4288,
+      "step": 861
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9127701198918514,
+      "learning_rate": 9.984712268418228e-06,
+      "loss": 0.4648,
+      "step": 862
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.992699198148132,
+      "learning_rate": 9.984632582004817e-06,
+      "loss": 0.4029,
+      "step": 863
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4562612420210286,
+      "learning_rate": 9.984552688770145e-06,
+      "loss": 0.4631,
+      "step": 864
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.197226952886568,
+      "learning_rate": 9.984472588717528e-06,
+      "loss": 0.4876,
+      "step": 865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.444824121345561,
+      "learning_rate": 9.984392281850293e-06,
+      "loss": 0.4299,
+      "step": 866
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1793522233262013,
+      "learning_rate": 9.984311768171766e-06,
+      "loss": 0.4564,
+      "step": 867
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.162687932448794,
+      "learning_rate": 9.984231047685292e-06,
+      "loss": 0.4551,
+      "step": 868
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.997032886043869,
+      "learning_rate": 9.984150120394219e-06,
+      "loss": 0.4321,
+      "step": 869
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7485735587581321,
+      "learning_rate": 9.984068986301902e-06,
+      "loss": 0.4042,
+      "step": 870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.026706024737377,
+      "learning_rate": 9.983987645411712e-06,
+      "loss": 0.4477,
+      "step": 871
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.288813397372106,
+      "learning_rate": 9.983906097727023e-06,
+      "loss": 0.4386,
+      "step": 872
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6025153100994587,
+      "learning_rate": 9.983824343251216e-06,
+      "loss": 0.4004,
+      "step": 873
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.700468949812591,
+      "learning_rate": 9.983742381987685e-06,
+      "loss": 0.4663,
+      "step": 874
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6293938910155337,
+      "learning_rate": 9.983660213939832e-06,
+      "loss": 0.4464,
+      "step": 875
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.527513601607223,
+      "learning_rate": 9.983577839111062e-06,
+      "loss": 0.4658,
+      "step": 876
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9417977450619628,
+      "learning_rate": 9.983495257504799e-06,
+      "loss": 0.4022,
+      "step": 877
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.195997308949818,
+      "learning_rate": 9.983412469124462e-06,
+      "loss": 0.4557,
+      "step": 878
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3393453275261773,
+      "learning_rate": 9.983329473973494e-06,
+      "loss": 0.5276,
+      "step": 879
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1995698085811766,
+      "learning_rate": 9.983246272055333e-06,
+      "loss": 0.4331,
+      "step": 880
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8356954093128823,
+      "learning_rate": 9.983162863373433e-06,
+      "loss": 0.4788,
+      "step": 881
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.297221437287468,
+      "learning_rate": 9.983079247931255e-06,
+      "loss": 0.4292,
+      "step": 882
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.57184333886103,
+      "learning_rate": 9.982995425732267e-06,
+      "loss": 0.4302,
+      "step": 883
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 26.614849061559507,
+      "learning_rate": 9.982911396779949e-06,
+      "loss": 0.4599,
+      "step": 884
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.812496817272605,
+      "learning_rate": 9.982827161077787e-06,
+      "loss": 0.436,
+      "step": 885
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.850573999627721,
+      "learning_rate": 9.982742718629275e-06,
+      "loss": 0.4297,
+      "step": 886
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1018236009745332,
+      "learning_rate": 9.982658069437916e-06,
+      "loss": 0.409,
+      "step": 887
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.761668117737086,
+      "learning_rate": 9.982573213507225e-06,
+      "loss": 0.4194,
+      "step": 888
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1370168746072733,
+      "learning_rate": 9.982488150840722e-06,
+      "loss": 0.4014,
+      "step": 889
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9043261773199873,
+      "learning_rate": 9.982402881441933e-06,
+      "loss": 0.3951,
+      "step": 890
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6126271997581667,
+      "learning_rate": 9.9823174053144e-06,
+      "loss": 0.4129,
+      "step": 891
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9991985625527473,
+      "learning_rate": 9.982231722461669e-06,
+      "loss": 0.4392,
+      "step": 892
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.043952618343807,
+      "learning_rate": 9.982145832887294e-06,
+      "loss": 0.5015,
+      "step": 893
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.939395607295308,
+      "learning_rate": 9.98205973659484e-06,
+      "loss": 0.4329,
+      "step": 894
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6488668660224204,
+      "learning_rate": 9.981973433587876e-06,
+      "loss": 0.426,
+      "step": 895
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5073712420631002,
+      "learning_rate": 9.981886923869988e-06,
+      "loss": 0.3905,
+      "step": 896
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8621661400270306,
+      "learning_rate": 9.981800207444762e-06,
+      "loss": 0.4081,
+      "step": 897
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1146298803252797,
+      "learning_rate": 9.981713284315796e-06,
+      "loss": 0.4397,
+      "step": 898
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2228346828692689,
+      "learning_rate": 9.981626154486696e-06,
+      "loss": 0.5177,
+      "step": 899
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4933527790971017,
+      "learning_rate": 9.981538817961082e-06,
+      "loss": 0.4235,
+      "step": 900
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8595059330336282,
+      "learning_rate": 9.981451274742572e-06,
+      "loss": 0.4814,
+      "step": 901
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7674160092007893,
+      "learning_rate": 9.9813635248348e-06,
+      "loss": 0.4082,
+      "step": 902
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.783556821691551,
+      "learning_rate": 9.98127556824141e-06,
+      "loss": 0.4332,
+      "step": 903
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6950369875321318,
+      "learning_rate": 9.981187404966047e-06,
+      "loss": 0.426,
+      "step": 904
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4850559204818143,
+      "learning_rate": 9.98109903501237e-06,
+      "loss": 0.4538,
+      "step": 905
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6481703676970185,
+      "learning_rate": 9.981010458384048e-06,
+      "loss": 0.4888,
+      "step": 906
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.853290704029964,
+      "learning_rate": 9.980921675084755e-06,
+      "loss": 0.448,
+      "step": 907
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5318247712162347,
+      "learning_rate": 9.980832685118173e-06,
+      "loss": 0.4558,
+      "step": 908
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.614367986623324,
+      "learning_rate": 9.980743488487999e-06,
+      "loss": 0.4599,
+      "step": 909
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.1062925944638415,
+      "learning_rate": 9.980654085197928e-06,
+      "loss": 0.4321,
+      "step": 910
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7061385243800808,
+      "learning_rate": 9.980564475251673e-06,
+      "loss": 0.4315,
+      "step": 911
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.262690557515717,
+      "learning_rate": 9.980474658652953e-06,
+      "loss": 0.4286,
+      "step": 912
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.435930689025621,
+      "learning_rate": 9.98038463540549e-06,
+      "loss": 0.426,
+      "step": 913
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8404883744451876,
+      "learning_rate": 9.980294405513024e-06,
+      "loss": 0.4525,
+      "step": 914
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6347995801011797,
+      "learning_rate": 9.980203968979298e-06,
+      "loss": 0.4543,
+      "step": 915
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9011771571903395,
+      "learning_rate": 9.980113325808062e-06,
+      "loss": 0.4417,
+      "step": 916
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.625756947977176,
+      "learning_rate": 9.980022476003081e-06,
+      "loss": 0.4301,
+      "step": 917
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4732928132745444,
+      "learning_rate": 9.97993141956812e-06,
+      "loss": 0.4579,
+      "step": 918
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1499894763709215,
+      "learning_rate": 9.979840156506959e-06,
+      "loss": 0.4579,
+      "step": 919
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8541051688817138,
+      "learning_rate": 9.979748686823386e-06,
+      "loss": 0.4396,
+      "step": 920
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4980284504653731,
+      "learning_rate": 9.979657010521194e-06,
+      "loss": 0.4325,
+      "step": 921
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.016309885443545,
+      "learning_rate": 9.979565127604186e-06,
+      "loss": 0.4334,
+      "step": 922
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6505891850016337,
+      "learning_rate": 9.97947303807618e-06,
+      "loss": 0.4258,
+      "step": 923
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7367054042675574,
+      "learning_rate": 9.97938074194099e-06,
+      "loss": 0.5969,
+      "step": 924
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0760989197594975,
+      "learning_rate": 9.97928823920245e-06,
+      "loss": 0.4242,
+      "step": 925
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.921672511976158,
+      "learning_rate": 9.979195529864397e-06,
+      "loss": 0.4731,
+      "step": 926
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.157068536120661,
+      "learning_rate": 9.979102613930676e-06,
+      "loss": 0.469,
+      "step": 927
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.628234387008205,
+      "learning_rate": 9.979009491405145e-06,
+      "loss": 0.4292,
+      "step": 928
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6049554572463292,
+      "learning_rate": 9.978916162291666e-06,
+      "loss": 0.4482,
+      "step": 929
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.037473235289058,
+      "learning_rate": 9.978822626594112e-06,
+      "loss": 0.4495,
+      "step": 930
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.896781965707907,
+      "learning_rate": 9.978728884316363e-06,
+      "loss": 0.4556,
+      "step": 931
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8908137079365919,
+      "learning_rate": 9.97863493546231e-06,
+      "loss": 0.4368,
+      "step": 932
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5531223267122365,
+      "learning_rate": 9.97854078003585e-06,
+      "loss": 0.429,
+      "step": 933
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.149434927746202,
+      "learning_rate": 9.97844641804089e-06,
+      "loss": 0.4543,
+      "step": 934
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4435713127937873,
+      "learning_rate": 9.978351849481347e-06,
+      "loss": 0.4593,
+      "step": 935
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.406142026912929,
+      "learning_rate": 9.978257074361142e-06,
+      "loss": 0.4534,
+      "step": 936
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.09372384551579,
+      "learning_rate": 9.97816209268421e-06,
+      "loss": 0.4175,
+      "step": 937
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0932196752914467,
+      "learning_rate": 9.978066904454489e-06,
+      "loss": 0.4157,
+      "step": 938
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9765605268754234,
+      "learning_rate": 9.97797150967593e-06,
+      "loss": 0.4274,
+      "step": 939
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.193357439145338,
+      "learning_rate": 9.977875908352493e-06,
+      "loss": 0.4479,
+      "step": 940
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1720783248148368,
+      "learning_rate": 9.977780100488142e-06,
+      "loss": 0.4759,
+      "step": 941
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.947354107383864,
+      "learning_rate": 9.977684086086853e-06,
+      "loss": 0.4512,
+      "step": 942
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8346989509604015,
+      "learning_rate": 9.977587865152609e-06,
+      "loss": 0.4239,
+      "step": 943
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7483032110461862,
+      "learning_rate": 9.977491437689403e-06,
+      "loss": 0.4594,
+      "step": 944
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0645624753941836,
+      "learning_rate": 9.977394803701238e-06,
+      "loss": 0.4333,
+      "step": 945
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7286216204365414,
+      "learning_rate": 9.97729796319212e-06,
+      "loss": 0.4436,
+      "step": 946
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.362368679846375,
+      "learning_rate": 9.97720091616607e-06,
+      "loss": 0.4532,
+      "step": 947
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3733509548490734,
+      "learning_rate": 9.977103662627112e-06,
+      "loss": 0.4085,
+      "step": 948
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.704821619867478,
+      "learning_rate": 9.977006202579284e-06,
+      "loss": 0.4379,
+      "step": 949
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.910476077056555,
+      "learning_rate": 9.976908536026628e-06,
+      "loss": 0.4538,
+      "step": 950
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.498882365552425,
+      "learning_rate": 9.976810662973198e-06,
+      "loss": 0.4737,
+      "step": 951
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8904724739380527,
+      "learning_rate": 9.976712583423053e-06,
+      "loss": 0.4469,
+      "step": 952
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.05992217119996,
+      "learning_rate": 9.976614297380264e-06,
+      "loss": 0.4107,
+      "step": 953
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9441060804384573,
+      "learning_rate": 9.976515804848907e-06,
+      "loss": 0.4377,
+      "step": 954
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1253849536855154,
+      "learning_rate": 9.97641710583307e-06,
+      "loss": 0.4028,
+      "step": 955
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6826962564161625,
+      "learning_rate": 9.97631820033685e-06,
+      "loss": 0.4501,
+      "step": 956
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9882186886883442,
+      "learning_rate": 9.976219088364347e-06,
+      "loss": 0.4102,
+      "step": 957
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3046200112231214,
+      "learning_rate": 9.976119769919677e-06,
+      "loss": 0.4329,
+      "step": 958
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.461627477595595,
+      "learning_rate": 9.976020245006957e-06,
+      "loss": 0.4451,
+      "step": 959
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8276651557840978,
+      "learning_rate": 9.97592051363032e-06,
+      "loss": 0.5161,
+      "step": 960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5219751753702964,
+      "learning_rate": 9.975820575793902e-06,
+      "loss": 0.4213,
+      "step": 961
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6107616287879676,
+      "learning_rate": 9.975720431501851e-06,
+      "loss": 0.427,
+      "step": 962
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.973225956565384,
+      "learning_rate": 9.975620080758321e-06,
+      "loss": 0.431,
+      "step": 963
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6719909709611906,
+      "learning_rate": 9.975519523567477e-06,
+      "loss": 0.4253,
+      "step": 964
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.450332923063095,
+      "learning_rate": 9.97541875993349e-06,
+      "loss": 0.4618,
+      "step": 965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2104242610244804,
+      "learning_rate": 9.97531778986054e-06,
+      "loss": 0.4643,
+      "step": 966
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.771495474982805,
+      "learning_rate": 9.975216613352818e-06,
+      "loss": 0.44,
+      "step": 967
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.79462285334802,
+      "learning_rate": 9.975115230414524e-06,
+      "loss": 0.4188,
+      "step": 968
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.115747433755278,
+      "learning_rate": 9.97501364104986e-06,
+      "loss": 0.4444,
+      "step": 969
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6015233518485714,
+      "learning_rate": 9.974911845263045e-06,
+      "loss": 0.4033,
+      "step": 970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4811623886152963,
+      "learning_rate": 9.974809843058299e-06,
+      "loss": 0.4037,
+      "step": 971
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.290643090711748,
+      "learning_rate": 9.974707634439858e-06,
+      "loss": 0.4017,
+      "step": 972
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8200089712297691,
+      "learning_rate": 9.974605219411962e-06,
+      "loss": 0.4622,
+      "step": 973
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3624872532897834,
+      "learning_rate": 9.974502597978858e-06,
+      "loss": 0.4331,
+      "step": 974
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.27210189356096,
+      "learning_rate": 9.974399770144807e-06,
+      "loss": 0.4147,
+      "step": 975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4383478828633045,
+      "learning_rate": 9.974296735914072e-06,
+      "loss": 0.4111,
+      "step": 976
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.056648793814065,
+      "learning_rate": 9.974193495290931e-06,
+      "loss": 0.44,
+      "step": 977
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.879102672129695,
+      "learning_rate": 9.974090048279666e-06,
+      "loss": 0.4472,
+      "step": 978
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8466662539655996,
+      "learning_rate": 9.973986394884571e-06,
+      "loss": 0.4116,
+      "step": 979
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.037496370016901,
+      "learning_rate": 9.973882535109944e-06,
+      "loss": 0.4803,
+      "step": 980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9611846114870142,
+      "learning_rate": 9.973778468960099e-06,
+      "loss": 0.4179,
+      "step": 981
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0710228622316924,
+      "learning_rate": 9.973674196439349e-06,
+      "loss": 0.4056,
+      "step": 982
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8868523869531129,
+      "learning_rate": 9.973569717552022e-06,
+      "loss": 0.4231,
+      "step": 983
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5643636961505556,
+      "learning_rate": 9.973465032302455e-06,
+      "loss": 0.4403,
+      "step": 984
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8418721685888957,
+      "learning_rate": 9.973360140694987e-06,
+      "loss": 0.4365,
+      "step": 985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.678896806142568,
+      "learning_rate": 9.973255042733976e-06,
+      "loss": 0.3718,
+      "step": 986
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2221814726497073,
+      "learning_rate": 9.973149738423779e-06,
+      "loss": 0.405,
+      "step": 987
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2032963475640974,
+      "learning_rate": 9.973044227768765e-06,
+      "loss": 0.4615,
+      "step": 988
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6657263537472482,
+      "learning_rate": 9.972938510773313e-06,
+      "loss": 0.4407,
+      "step": 989
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7025416284652275,
+      "learning_rate": 9.972832587441811e-06,
+      "loss": 0.4401,
+      "step": 990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0383827337350766,
+      "learning_rate": 9.972726457778651e-06,
+      "loss": 0.4169,
+      "step": 991
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.36485014621816,
+      "learning_rate": 9.972620121788238e-06,
+      "loss": 0.4141,
+      "step": 992
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4310161874528062,
+      "learning_rate": 9.972513579474982e-06,
+      "loss": 0.4076,
+      "step": 993
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.305654451233364,
+      "learning_rate": 9.97240683084331e-06,
+      "loss": 0.431,
+      "step": 994
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1923634812750334,
+      "learning_rate": 9.972299875897641e-06,
+      "loss": 0.3946,
+      "step": 995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7901803929292885,
+      "learning_rate": 9.972192714642422e-06,
+      "loss": 0.4345,
+      "step": 996
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.465272511542452,
+      "learning_rate": 9.972085347082094e-06,
+      "loss": 0.4202,
+      "step": 997
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5262679006939772,
+      "learning_rate": 9.971977773221115e-06,
+      "loss": 0.4095,
+      "step": 998
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4850430263095706,
+      "learning_rate": 9.971869993063947e-06,
+      "loss": 0.4032,
+      "step": 999
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4255973979231296,
+      "learning_rate": 9.97176200661506e-06,
+      "loss": 0.4027,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9043882780359147,
+      "learning_rate": 9.971653813878937e-06,
+      "loss": 0.4484,
+      "step": 1001
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1930700474658518,
+      "learning_rate": 9.971545414860067e-06,
+      "loss": 0.4517,
+      "step": 1002
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1065434574616355,
+      "learning_rate": 9.971436809562948e-06,
+      "loss": 0.4328,
+      "step": 1003
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7849230866270132,
+      "learning_rate": 9.971327997992085e-06,
+      "loss": 0.4287,
+      "step": 1004
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.079922648594418,
+      "learning_rate": 9.971218980151993e-06,
+      "loss": 0.4515,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.026995958465551,
+      "learning_rate": 9.971109756047197e-06,
+      "loss": 0.4135,
+      "step": 1006
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3688983241560737,
+      "learning_rate": 9.971000325682225e-06,
+      "loss": 0.4248,
+      "step": 1007
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4013386275445587,
+      "learning_rate": 9.970890689061622e-06,
+      "loss": 0.4466,
+      "step": 1008
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4509037141099472,
+      "learning_rate": 9.970780846189934e-06,
+      "loss": 0.4171,
+      "step": 1009
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.894729520330879,
+      "learning_rate": 9.970670797071719e-06,
+      "loss": 0.4144,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.0500302607849745,
+      "learning_rate": 9.970560541711547e-06,
+      "loss": 0.4362,
+      "step": 1011
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8937281240678299,
+      "learning_rate": 9.970450080113988e-06,
+      "loss": 0.4361,
+      "step": 1012
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.432129687251775,
+      "learning_rate": 9.970339412283624e-06,
+      "loss": 0.4471,
+      "step": 1013
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7322069181436954,
+      "learning_rate": 9.970228538225051e-06,
+      "loss": 0.4365,
+      "step": 1014
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5036771104025033,
+      "learning_rate": 9.97011745794287e-06,
+      "loss": 0.4034,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4627252751655098,
+      "learning_rate": 9.970006171441684e-06,
+      "loss": 0.4404,
+      "step": 1016
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5231895430952793,
+      "learning_rate": 9.969894678726118e-06,
+      "loss": 0.3948,
+      "step": 1017
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5009359351294047,
+      "learning_rate": 9.969782979800791e-06,
+      "loss": 0.428,
+      "step": 1018
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6515350137500342,
+      "learning_rate": 9.969671074670341e-06,
+      "loss": 0.4681,
+      "step": 1019
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1735423046277487,
+      "learning_rate": 9.969558963339414e-06,
+      "loss": 0.4073,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6301082636370194,
+      "learning_rate": 9.969446645812656e-06,
+      "loss": 0.4251,
+      "step": 1021
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.353792866562552,
+      "learning_rate": 9.969334122094732e-06,
+      "loss": 0.4101,
+      "step": 1022
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4371540032356265,
+      "learning_rate": 9.969221392190307e-06,
+      "loss": 0.497,
+      "step": 1023
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8390282407206398,
+      "learning_rate": 9.96910845610406e-06,
+      "loss": 0.4329,
+      "step": 1024
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9311063342477017,
+      "learning_rate": 9.968995313840678e-06,
+      "loss": 0.4064,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.249935201981566,
+      "learning_rate": 9.968881965404855e-06,
+      "loss": 0.3997,
+      "step": 1026
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5217798343803648,
+      "learning_rate": 9.968768410801292e-06,
+      "loss": 0.4276,
+      "step": 1027
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3442816268055013,
+      "learning_rate": 9.968654650034702e-06,
+      "loss": 0.4113,
+      "step": 1028
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8157652085876863,
+      "learning_rate": 9.968540683109806e-06,
+      "loss": 0.4587,
+      "step": 1029
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5606612589384359,
+      "learning_rate": 9.968426510031333e-06,
+      "loss": 0.397,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1053549705834635,
+      "learning_rate": 9.968312130804019e-06,
+      "loss": 0.395,
+      "step": 1031
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1756119353760828,
+      "learning_rate": 9.968197545432608e-06,
+      "loss": 0.4166,
+      "step": 1032
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0672867495576734,
+      "learning_rate": 9.968082753921857e-06,
+      "loss": 0.4351,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.497903565945721,
+      "learning_rate": 9.967967756276528e-06,
+      "loss": 0.4164,
+      "step": 1034
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6953064491487724,
+      "learning_rate": 9.967852552501394e-06,
+      "loss": 0.453,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.212831139459771,
+      "learning_rate": 9.967737142601233e-06,
+      "loss": 0.4428,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.885711416467117,
+      "learning_rate": 9.967621526580836e-06,
+      "loss": 0.442,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.419841719673243,
+      "learning_rate": 9.967505704444996e-06,
+      "loss": 0.4371,
+      "step": 1038
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1092049370758943,
+      "learning_rate": 9.967389676198522e-06,
+      "loss": 0.4586,
+      "step": 1039
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5785517336382928,
+      "learning_rate": 9.967273441846227e-06,
+      "loss": 0.4596,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1897340497462054,
+      "learning_rate": 9.967157001392933e-06,
+      "loss": 0.4071,
+      "step": 1041
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7182912408548263,
+      "learning_rate": 9.967040354843473e-06,
+      "loss": 0.4325,
+      "step": 1042
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9268781515312112,
+      "learning_rate": 9.966923502202688e-06,
+      "loss": 0.4316,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8581954392547293,
+      "learning_rate": 9.966806443475423e-06,
+      "loss": 0.4576,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2620286989720024,
+      "learning_rate": 9.966689178666537e-06,
+      "loss": 0.4151,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9879301501050124,
+      "learning_rate": 9.966571707780894e-06,
+      "loss": 0.3877,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5951202653168173,
+      "learning_rate": 9.96645403082337e-06,
+      "loss": 0.4672,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.892616323634653,
+      "learning_rate": 9.966336147798848e-06,
+      "loss": 0.436,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5057467579778137,
+      "learning_rate": 9.966218058712218e-06,
+      "loss": 0.4144,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5211997236715327,
+      "learning_rate": 9.966099763568377e-06,
+      "loss": 0.3761,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.036891383244653,
+      "learning_rate": 9.96598126237224e-06,
+      "loss": 0.4391,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3484219984213954,
+      "learning_rate": 9.965862555128717e-06,
+      "loss": 0.3985,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5805476026012735,
+      "learning_rate": 9.965743641842737e-06,
+      "loss": 0.4209,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.843236713409933,
+      "learning_rate": 9.965624522519233e-06,
+      "loss": 0.4373,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0949125824287727,
+      "learning_rate": 9.965505197163148e-06,
+      "loss": 0.4292,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.22939719849107,
+      "learning_rate": 9.965385665779432e-06,
+      "loss": 0.4244,
+      "step": 1056
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6796170096341314,
+      "learning_rate": 9.965265928373045e-06,
+      "loss": 0.3717,
+      "step": 1057
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8208949409810051,
+      "learning_rate": 9.965145984948956e-06,
+      "loss": 0.4295,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6048724862682364,
+      "learning_rate": 9.96502583551214e-06,
+      "loss": 0.4128,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.182544860851816,
+      "learning_rate": 9.964905480067585e-06,
+      "loss": 0.4421,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8770526281876805,
+      "learning_rate": 9.964784918620284e-06,
+      "loss": 0.4144,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0502433368725024,
+      "learning_rate": 9.964664151175235e-06,
+      "loss": 0.4149,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6852493174725927,
+      "learning_rate": 9.964543177737453e-06,
+      "loss": 0.4279,
+      "step": 1063
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3863314076720705,
+      "learning_rate": 9.964421998311957e-06,
+      "loss": 0.4309,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9774735470449674,
+      "learning_rate": 9.964300612903775e-06,
+      "loss": 0.3857,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7865301839072318,
+      "learning_rate": 9.964179021517943e-06,
+      "loss": 0.4374,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9105237189090487,
+      "learning_rate": 9.964057224159505e-06,
+      "loss": 0.434,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0308961812961612,
+      "learning_rate": 9.963935220833516e-06,
+      "loss": 0.5032,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.567470420085035,
+      "learning_rate": 9.963813011545039e-06,
+      "loss": 0.4449,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9238137775653517,
+      "learning_rate": 9.963690596299142e-06,
+      "loss": 0.4169,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.224238352628365,
+      "learning_rate": 9.963567975100906e-06,
+      "loss": 0.4024,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8992630344406685,
+      "learning_rate": 9.963445147955417e-06,
+      "loss": 0.4147,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.146992246234404,
+      "learning_rate": 9.963322114867775e-06,
+      "loss": 0.4716,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.068810859692916,
+      "learning_rate": 9.963198875843082e-06,
+      "loss": 0.4104,
+      "step": 1074
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.121934838520355,
+      "learning_rate": 9.963075430886451e-06,
+      "loss": 0.4189,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.8409386562456875,
+      "learning_rate": 9.962951780003005e-06,
+      "loss": 0.4545,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.44691412556006,
+      "learning_rate": 9.962827923197875e-06,
+      "loss": 0.4159,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8674712042238808,
+      "learning_rate": 9.9627038604762e-06,
+      "loss": 0.4232,
+      "step": 1078
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7475315454587137,
+      "learning_rate": 9.962579591843126e-06,
+      "loss": 0.4344,
+      "step": 1079
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9700184410379884,
+      "learning_rate": 9.962455117303813e-06,
+      "loss": 0.385,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.734268009587939,
+      "learning_rate": 9.96233043686342e-06,
+      "loss": 0.4488,
+      "step": 1081
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3792889161990405,
+      "learning_rate": 9.962205550527124e-06,
+      "loss": 0.4301,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.816078388293769,
+      "learning_rate": 9.962080458300105e-06,
+      "loss": 0.4468,
+      "step": 1083
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.786762985149165,
+      "learning_rate": 9.961955160187555e-06,
+      "loss": 0.4022,
+      "step": 1084
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6825212039123876,
+      "learning_rate": 9.961829656194672e-06,
+      "loss": 0.4142,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4717471390365557,
+      "learning_rate": 9.961703946326664e-06,
+      "loss": 0.4195,
+      "step": 1086
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7048649803787788,
+      "learning_rate": 9.961578030588746e-06,
+      "loss": 0.3955,
+      "step": 1087
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1645420543212817,
+      "learning_rate": 9.961451908986142e-06,
+      "loss": 0.4688,
+      "step": 1088
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.104394571412978,
+      "learning_rate": 9.961325581524086e-06,
+      "loss": 0.4157,
+      "step": 1089
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7553422306302857,
+      "learning_rate": 9.961199048207819e-06,
+      "loss": 0.4036,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3138483785122332,
+      "learning_rate": 9.961072309042592e-06,
+      "loss": 0.5112,
+      "step": 1091
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5564848942173555,
+      "learning_rate": 9.960945364033662e-06,
+      "loss": 0.4448,
+      "step": 1092
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9577214807056948,
+      "learning_rate": 9.9608182131863e-06,
+      "loss": 0.4981,
+      "step": 1093
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.209846204917101,
+      "learning_rate": 9.960690856505774e-06,
+      "loss": 0.4315,
+      "step": 1094
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3829740417530734,
+      "learning_rate": 9.960563293997377e-06,
+      "loss": 0.4202,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6975487269010596,
+      "learning_rate": 9.960435525666397e-06,
+      "loss": 0.4224,
+      "step": 1096
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1979383948438187,
+      "learning_rate": 9.960307551518135e-06,
+      "loss": 0.4143,
+      "step": 1097
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6130717949337823,
+      "learning_rate": 9.960179371557905e-06,
+      "loss": 0.4302,
+      "step": 1098
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.193657055310098,
+      "learning_rate": 9.960050985791021e-06,
+      "loss": 0.4207,
+      "step": 1099
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8108961622421267,
+      "learning_rate": 9.959922394222811e-06,
+      "loss": 0.4081,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8118247307835853,
+      "learning_rate": 9.959793596858614e-06,
+      "loss": 0.4129,
+      "step": 1101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.982502282567653,
+      "learning_rate": 9.959664593703769e-06,
+      "loss": 0.4397,
+      "step": 1102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6872548116434403,
+      "learning_rate": 9.95953538476363e-06,
+      "loss": 0.4436,
+      "step": 1103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8915972803886736,
+      "learning_rate": 9.959405970043558e-06,
+      "loss": 0.4035,
+      "step": 1104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.005443454230022,
+      "learning_rate": 9.959276349548926e-06,
+      "loss": 0.3938,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1111697114362995,
+      "learning_rate": 9.959146523285108e-06,
+      "loss": 0.4682,
+      "step": 1106
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.906603648905359,
+      "learning_rate": 9.959016491257491e-06,
+      "loss": 0.4771,
+      "step": 1107
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8274519497356905,
+      "learning_rate": 9.958886253471474e-06,
+      "loss": 0.4155,
+      "step": 1108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3098968910760096,
+      "learning_rate": 9.958755809932457e-06,
+      "loss": 0.4178,
+      "step": 1109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9589235520258326,
+      "learning_rate": 9.958625160645855e-06,
+      "loss": 0.4624,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.923709855815237,
+      "learning_rate": 9.958494305617087e-06,
+      "loss": 0.3995,
+      "step": 1111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.103399335480662,
+      "learning_rate": 9.958363244851584e-06,
+      "loss": 0.4158,
+      "step": 1112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.800277095827046,
+      "learning_rate": 9.95823197835478e-06,
+      "loss": 0.4377,
+      "step": 1113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8306837780654002,
+      "learning_rate": 9.958100506132127e-06,
+      "loss": 0.4303,
+      "step": 1114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.907673716085669,
+      "learning_rate": 9.957968828189076e-06,
+      "loss": 0.4096,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.076372980055298,
+      "learning_rate": 9.957836944531091e-06,
+      "loss": 0.4238,
+      "step": 1116
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5765367603892482,
+      "learning_rate": 9.957704855163648e-06,
+      "loss": 0.4335,
+      "step": 1117
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.268987081114445,
+      "learning_rate": 9.957572560092223e-06,
+      "loss": 0.424,
+      "step": 1118
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5252904351103718,
+      "learning_rate": 9.957440059322308e-06,
+      "loss": 0.4293,
+      "step": 1119
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5133172714908736,
+      "learning_rate": 9.957307352859397e-06,
+      "loss": 0.4297,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.894375399108354,
+      "learning_rate": 9.957174440709e-06,
+      "loss": 0.4009,
+      "step": 1121
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.095415619491392,
+      "learning_rate": 9.957041322876632e-06,
+      "loss": 0.4252,
+      "step": 1122
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1931537983196945,
+      "learning_rate": 9.956907999367815e-06,
+      "loss": 0.4436,
+      "step": 1123
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1922907199920667,
+      "learning_rate": 9.956774470188079e-06,
+      "loss": 0.427,
+      "step": 1124
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1401379075208133,
+      "learning_rate": 9.956640735342966e-06,
+      "loss": 0.4364,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.34329941398725,
+      "learning_rate": 9.956506794838025e-06,
+      "loss": 0.4483,
+      "step": 1126
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9979221997681402,
+      "learning_rate": 9.956372648678814e-06,
+      "loss": 0.4196,
+      "step": 1127
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8312909577519454,
+      "learning_rate": 9.956238296870898e-06,
+      "loss": 0.4273,
+      "step": 1128
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.130057516734946,
+      "learning_rate": 9.956103739419852e-06,
+      "loss": 0.3799,
+      "step": 1129
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.3755194095153955,
+      "learning_rate": 9.955968976331258e-06,
+      "loss": 0.4219,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2437667514805772,
+      "learning_rate": 9.95583400761071e-06,
+      "loss": 0.4259,
+      "step": 1131
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1191427593579175,
+      "learning_rate": 9.955698833263805e-06,
+      "loss": 0.4436,
+      "step": 1132
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3600928176122453,
+      "learning_rate": 9.955563453296154e-06,
+      "loss": 0.4326,
+      "step": 1133
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.174961780325701,
+      "learning_rate": 9.955427867713372e-06,
+      "loss": 0.4094,
+      "step": 1134
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.103957737047937,
+      "learning_rate": 9.955292076521088e-06,
+      "loss": 0.4169,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.719030182825631,
+      "learning_rate": 9.955156079724932e-06,
+      "loss": 0.4353,
+      "step": 1136
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.67841001777247,
+      "learning_rate": 9.955019877330549e-06,
+      "loss": 0.3822,
+      "step": 1137
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6936061193769136,
+      "learning_rate": 9.95488346934359e-06,
+      "loss": 0.4211,
+      "step": 1138
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8270868845291233,
+      "learning_rate": 9.954746855769717e-06,
+      "loss": 0.4136,
+      "step": 1139
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.942249183739099,
+      "learning_rate": 9.954610036614595e-06,
+      "loss": 0.4024,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.315704927234245,
+      "learning_rate": 9.954473011883903e-06,
+      "loss": 0.4155,
+      "step": 1141
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.843726684819333,
+      "learning_rate": 9.954335781583326e-06,
+      "loss": 0.4286,
+      "step": 1142
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.640572597785077,
+      "learning_rate": 9.954198345718556e-06,
+      "loss": 0.4089,
+      "step": 1143
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0486714901349705,
+      "learning_rate": 9.954060704295298e-06,
+      "loss": 0.3875,
+      "step": 1144
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6723275491556393,
+      "learning_rate": 9.953922857319261e-06,
+      "loss": 0.4276,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.724395943504821,
+      "learning_rate": 9.953784804796167e-06,
+      "loss": 0.401,
+      "step": 1146
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2957490366865714,
+      "learning_rate": 9.95364654673174e-06,
+      "loss": 0.4555,
+      "step": 1147
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.170708342495129,
+      "learning_rate": 9.953508083131722e-06,
+      "loss": 0.4553,
+      "step": 1148
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8269200120497504,
+      "learning_rate": 9.953369414001853e-06,
+      "loss": 0.4604,
+      "step": 1149
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.901983238895689,
+      "learning_rate": 9.95323053934789e-06,
+      "loss": 0.4097,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4885513730789306,
+      "learning_rate": 9.953091459175595e-06,
+      "loss": 0.426,
+      "step": 1151
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.654939040249722,
+      "learning_rate": 9.952952173490735e-06,
+      "loss": 0.3987,
+      "step": 1152
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.620085677907411,
+      "learning_rate": 9.952812682299093e-06,
+      "loss": 0.4093,
+      "step": 1153
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.807227735743264,
+      "learning_rate": 9.952672985606457e-06,
+      "loss": 0.3993,
+      "step": 1154
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.077883064036308,
+      "learning_rate": 9.95253308341862e-06,
+      "loss": 0.4111,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7373548437655704,
+      "learning_rate": 9.952392975741389e-06,
+      "loss": 0.3972,
+      "step": 1156
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4599863742025554,
+      "learning_rate": 9.95225266258058e-06,
+      "loss": 0.4592,
+      "step": 1157
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8306714239615562,
+      "learning_rate": 9.952112143942008e-06,
+      "loss": 0.4112,
+      "step": 1158
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4581276301566324,
+      "learning_rate": 9.951971419831509e-06,
+      "loss": 0.4061,
+      "step": 1159
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.24926755423996,
+      "learning_rate": 9.951830490254918e-06,
+      "loss": 0.4054,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8121588491986924,
+      "learning_rate": 9.951689355218088e-06,
+      "loss": 0.4246,
+      "step": 1161
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8822317861508915,
+      "learning_rate": 9.951548014726868e-06,
+      "loss": 0.4248,
+      "step": 1162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.457734983831385,
+      "learning_rate": 9.951406468787128e-06,
+      "loss": 0.4024,
+      "step": 1163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6538322895061792,
+      "learning_rate": 9.95126471740474e-06,
+      "loss": 0.4229,
+      "step": 1164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.180062949779296,
+      "learning_rate": 9.951122760585582e-06,
+      "loss": 0.3761,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.985139784360479,
+      "learning_rate": 9.950980598335548e-06,
+      "loss": 0.4176,
+      "step": 1166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.098600871684249,
+      "learning_rate": 9.950838230660535e-06,
+      "loss": 0.4318,
+      "step": 1167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4070008307892183,
+      "learning_rate": 9.95069565756645e-06,
+      "loss": 0.4293,
+      "step": 1168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2088169739092502,
+      "learning_rate": 9.950552879059208e-06,
+      "loss": 0.3926,
+      "step": 1169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.45918126709504,
+      "learning_rate": 9.950409895144732e-06,
+      "loss": 0.4406,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6032233842578774,
+      "learning_rate": 9.95026670582896e-06,
+      "loss": 0.3942,
+      "step": 1171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1697066862160463,
+      "learning_rate": 9.950123311117828e-06,
+      "loss": 0.4011,
+      "step": 1172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.169470340729822,
+      "learning_rate": 9.949979711017288e-06,
+      "loss": 0.4097,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9054993124774207,
+      "learning_rate": 9.949835905533298e-06,
+      "loss": 0.4122,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.157218184394324,
+      "learning_rate": 9.949691894671824e-06,
+      "loss": 0.4312,
+      "step": 1175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.020711252280423,
+      "learning_rate": 9.949547678438842e-06,
+      "loss": 0.4044,
+      "step": 1176
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.324584809973775,
+      "learning_rate": 9.949403256840334e-06,
+      "loss": 0.3903,
+      "step": 1177
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0822794369792312,
+      "learning_rate": 9.949258629882295e-06,
+      "loss": 0.4084,
+      "step": 1178
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.833722949501575,
+      "learning_rate": 9.949113797570724e-06,
+      "loss": 0.3842,
+      "step": 1179
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.186114115992921,
+      "learning_rate": 9.948968759911633e-06,
+      "loss": 0.4584,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.030559565827121,
+      "learning_rate": 9.948823516911034e-06,
+      "loss": 0.4066,
+      "step": 1181
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8127397266979124,
+      "learning_rate": 9.948678068574959e-06,
+      "loss": 0.4168,
+      "step": 1182
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9327679928843606,
+      "learning_rate": 9.948532414909442e-06,
+      "loss": 0.4453,
+      "step": 1183
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5909469409027934,
+      "learning_rate": 9.948386555920525e-06,
+      "loss": 0.4246,
+      "step": 1184
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.654456156371072,
+      "learning_rate": 9.94824049161426e-06,
+      "loss": 0.3741,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.50992240356473,
+      "learning_rate": 9.94809422199671e-06,
+      "loss": 0.432,
+      "step": 1186
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8768086792182466,
+      "learning_rate": 9.947947747073939e-06,
+      "loss": 0.3945,
+      "step": 1187
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.157471389093322,
+      "learning_rate": 9.947801066852029e-06,
+      "loss": 0.4125,
+      "step": 1188
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.194982551871048,
+      "learning_rate": 9.947654181337063e-06,
+      "loss": 0.4154,
+      "step": 1189
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6524354804752717,
+      "learning_rate": 9.947507090535139e-06,
+      "loss": 0.4158,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4113212327423694,
+      "learning_rate": 9.947359794452356e-06,
+      "loss": 0.4155,
+      "step": 1191
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1654870426499024,
+      "learning_rate": 9.94721229309483e-06,
+      "loss": 0.4089,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3239226693757096,
+      "learning_rate": 9.947064586468677e-06,
+      "loss": 0.4077,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.045717960251605,
+      "learning_rate": 9.946916674580028e-06,
+      "loss": 0.4048,
+      "step": 1194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.94778507161363,
+      "learning_rate": 9.946768557435019e-06,
+      "loss": 0.3821,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.863795468418417,
+      "learning_rate": 9.946620235039797e-06,
+      "loss": 0.4062,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1985856913958752,
+      "learning_rate": 9.946471707400514e-06,
+      "loss": 0.4419,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0466579467587525,
+      "learning_rate": 9.946322974523336e-06,
+      "loss": 0.4051,
+      "step": 1198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9695428330223819,
+      "learning_rate": 9.94617403641443e-06,
+      "loss": 0.4178,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3333009429864418,
+      "learning_rate": 9.946024893079977e-06,
+      "loss": 0.4116,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.266405677726087,
+      "learning_rate": 9.945875544526168e-06,
+      "loss": 0.3841,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.924999768845305,
+      "learning_rate": 9.945725990759197e-06,
+      "loss": 0.4005,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.431551423993694,
+      "learning_rate": 9.94557623178527e-06,
+      "loss": 0.3906,
+      "step": 1203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.394086168640765,
+      "learning_rate": 9.945426267610603e-06,
+      "loss": 0.4211,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.124702489165753,
+      "learning_rate": 9.945276098241413e-06,
+      "loss": 0.4062,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0348838838892327,
+      "learning_rate": 9.945125723683934e-06,
+      "loss": 0.4045,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.079576986386569,
+      "learning_rate": 9.944975143944407e-06,
+      "loss": 0.404,
+      "step": 1207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.741809985860796,
+      "learning_rate": 9.944824359029078e-06,
+      "loss": 0.4253,
+      "step": 1208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7501764025367916,
+      "learning_rate": 9.944673368944202e-06,
+      "loss": 0.397,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.159395742750831,
+      "learning_rate": 9.944522173696047e-06,
+      "loss": 0.4116,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.466016323510857,
+      "learning_rate": 9.944370773290883e-06,
+      "loss": 0.3907,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.971238187842781,
+      "learning_rate": 9.944219167734994e-06,
+      "loss": 0.4272,
+      "step": 1212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0298059740919654,
+      "learning_rate": 9.94406735703467e-06,
+      "loss": 0.406,
+      "step": 1213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4026259661066875,
+      "learning_rate": 9.943915341196209e-06,
+      "loss": 0.3856,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.624219757935532,
+      "learning_rate": 9.94376312022592e-06,
+      "loss": 0.405,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4718592693709014,
+      "learning_rate": 9.943610694130117e-06,
+      "loss": 0.408,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.6270700810101255,
+      "learning_rate": 9.943458062915126e-06,
+      "loss": 0.4193,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9171228652969508,
+      "learning_rate": 9.94330522658728e-06,
+      "loss": 0.3958,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.02687721984281,
+      "learning_rate": 9.94315218515292e-06,
+      "loss": 0.3846,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.270919265580942,
+      "learning_rate": 9.942998938618394e-06,
+      "loss": 0.4047,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6915723864732783,
+      "learning_rate": 9.942845486990064e-06,
+      "loss": 0.5804,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5053069801244665,
+      "learning_rate": 9.942691830274293e-06,
+      "loss": 0.3884,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.195396899547578,
+      "learning_rate": 9.942537968477461e-06,
+      "loss": 0.3948,
+      "step": 1223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8535971708343282,
+      "learning_rate": 9.94238390160595e-06,
+      "loss": 0.4287,
+      "step": 1224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.925765481633316,
+      "learning_rate": 9.942229629666152e-06,
+      "loss": 0.3974,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.254763746431971,
+      "learning_rate": 9.942075152664467e-06,
+      "loss": 0.3776,
+      "step": 1226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6502737838639416,
+      "learning_rate": 9.941920470607306e-06,
+      "loss": 0.4086,
+      "step": 1227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8380067080178115,
+      "learning_rate": 9.941765583501088e-06,
+      "loss": 0.4707,
+      "step": 1228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.995217067426309,
+      "learning_rate": 9.941610491352238e-06,
+      "loss": 0.405,
+      "step": 1229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9206367129209652,
+      "learning_rate": 9.94145519416719e-06,
+      "loss": 0.3878,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0937950878177984,
+      "learning_rate": 9.94129969195239e-06,
+      "loss": 0.3835,
+      "step": 1231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.681959798676605,
+      "learning_rate": 9.94114398471429e-06,
+      "loss": 0.3898,
+      "step": 1232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.38673455622357,
+      "learning_rate": 9.94098807245935e-06,
+      "loss": 0.4006,
+      "step": 1233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7452893009158057,
+      "learning_rate": 9.940831955194036e-06,
+      "loss": 0.3908,
+      "step": 1234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.893389830279717,
+      "learning_rate": 9.94067563292483e-06,
+      "loss": 0.4397,
+      "step": 1235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4763339006276053,
+      "learning_rate": 9.940519105658217e-06,
+      "loss": 0.4231,
+      "step": 1236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9206715546328543,
+      "learning_rate": 9.94036237340069e-06,
+      "loss": 0.4188,
+      "step": 1237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.226804987608132,
+      "learning_rate": 9.940205436158753e-06,
+      "loss": 0.4057,
+      "step": 1238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2201140495236245,
+      "learning_rate": 9.940048293938918e-06,
+      "loss": 0.398,
+      "step": 1239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 23.169631949745703,
+      "learning_rate": 9.939890946747703e-06,
+      "loss": 0.4126,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8901675996120868,
+      "learning_rate": 9.93973339459164e-06,
+      "loss": 0.4422,
+      "step": 1241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0173839554868,
+      "learning_rate": 9.939575637477266e-06,
+      "loss": 0.3825,
+      "step": 1242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3312057219372173,
+      "learning_rate": 9.939417675411123e-06,
+      "loss": 0.3937,
+      "step": 1243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.41994531044071,
+      "learning_rate": 9.939259508399767e-06,
+      "loss": 0.3645,
+      "step": 1244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.130732966256115,
+      "learning_rate": 9.939101136449763e-06,
+      "loss": 0.4355,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9407138873887051,
+      "learning_rate": 9.938942559567677e-06,
+      "loss": 0.5359,
+      "step": 1246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1873362989230545,
+      "learning_rate": 9.938783777760095e-06,
+      "loss": 0.3971,
+      "step": 1247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6647545068782758,
+      "learning_rate": 9.938624791033599e-06,
+      "loss": 0.4553,
+      "step": 1248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.726292817792635,
+      "learning_rate": 9.93846559939479e-06,
+      "loss": 0.4015,
+      "step": 1249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6179117598872832,
+      "learning_rate": 9.938306202850272e-06,
+      "loss": 0.4852,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.38919296381999,
+      "learning_rate": 9.938146601406657e-06,
+      "loss": 0.4089,
+      "step": 1251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.947093361026362,
+      "learning_rate": 9.937986795070568e-06,
+      "loss": 0.4459,
+      "step": 1252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2415261426460065,
+      "learning_rate": 9.937826783848636e-06,
+      "loss": 0.4175,
+      "step": 1253
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0143115187190905,
+      "learning_rate": 9.9376665677475e-06,
+      "loss": 0.3941,
+      "step": 1254
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.884912048352357,
+      "learning_rate": 9.93750614677381e-06,
+      "loss": 0.4685,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.615011997154566,
+      "learning_rate": 9.93734552093422e-06,
+      "loss": 0.3747,
+      "step": 1256
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4576862985160397,
+      "learning_rate": 9.937184690235393e-06,
+      "loss": 0.3972,
+      "step": 1257
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.301713426062173,
+      "learning_rate": 9.937023654684004e-06,
+      "loss": 0.3979,
+      "step": 1258
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3974970802635305,
+      "learning_rate": 9.936862414286734e-06,
+      "loss": 0.4505,
+      "step": 1259
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.901122120553632,
+      "learning_rate": 9.936700969050275e-06,
+      "loss": 0.4301,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5712528297749246,
+      "learning_rate": 9.936539318981323e-06,
+      "loss": 0.4034,
+      "step": 1261
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5061825489424003,
+      "learning_rate": 9.936377464086586e-06,
+      "loss": 0.4692,
+      "step": 1262
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.062917052282411,
+      "learning_rate": 9.936215404372783e-06,
+      "loss": 0.4135,
+      "step": 1263
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.024460585702295,
+      "learning_rate": 9.936053139846631e-06,
+      "loss": 0.5626,
+      "step": 1264
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.647596296601089,
+      "learning_rate": 9.93589067051487e-06,
+      "loss": 0.3908,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8476435586334785,
+      "learning_rate": 9.935727996384237e-06,
+      "loss": 0.4074,
+      "step": 1266
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.1113709153265035,
+      "learning_rate": 9.93556511746148e-06,
+      "loss": 0.401,
+      "step": 1267
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0001077011543593,
+      "learning_rate": 9.935402033753364e-06,
+      "loss": 0.3933,
+      "step": 1268
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.4100065995475255,
+      "learning_rate": 9.935238745266648e-06,
+      "loss": 0.408,
+      "step": 1269
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9698265874125425,
+      "learning_rate": 9.935075252008113e-06,
+      "loss": 0.4248,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6037970756036097,
+      "learning_rate": 9.934911553984539e-06,
+      "loss": 0.3888,
+      "step": 1271
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7803594986696203,
+      "learning_rate": 9.934747651202718e-06,
+      "loss": 0.484,
+      "step": 1272
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9670363109487332,
+      "learning_rate": 9.934583543669454e-06,
+      "loss": 0.4171,
+      "step": 1273
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.392762481259209,
+      "learning_rate": 9.934419231391554e-06,
+      "loss": 0.3777,
+      "step": 1274
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6645278071234206,
+      "learning_rate": 9.934254714375834e-06,
+      "loss": 0.4242,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2565647136530593,
+      "learning_rate": 9.934089992629122e-06,
+      "loss": 0.3958,
+      "step": 1276
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5859524170815797,
+      "learning_rate": 9.933925066158254e-06,
+      "loss": 0.4164,
+      "step": 1277
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3127556176510717,
+      "learning_rate": 9.933759934970069e-06,
+      "loss": 0.4036,
+      "step": 1278
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8300530787503981,
+      "learning_rate": 9.93359459907142e-06,
+      "loss": 0.3761,
+      "step": 1279
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.703908562364397,
+      "learning_rate": 9.933429058469171e-06,
+      "loss": 0.3914,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.140590584560028,
+      "learning_rate": 9.933263313170187e-06,
+      "loss": 0.4299,
+      "step": 1281
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9763097183198512,
+      "learning_rate": 9.933097363181346e-06,
+      "loss": 0.4018,
+      "step": 1282
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0209916346738606,
+      "learning_rate": 9.932931208509533e-06,
+      "loss": 0.3795,
+      "step": 1283
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6783799119102611,
+      "learning_rate": 9.93276484916164e-06,
+      "loss": 0.4336,
+      "step": 1284
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7725227031042743,
+      "learning_rate": 9.932598285144575e-06,
+      "loss": 0.3956,
+      "step": 1285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7385715317313024,
+      "learning_rate": 9.932431516465244e-06,
+      "loss": 0.4128,
+      "step": 1286
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.744366053658705,
+      "learning_rate": 9.932264543130568e-06,
+      "loss": 0.4132,
+      "step": 1287
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.683341016651291,
+      "learning_rate": 9.932097365147477e-06,
+      "loss": 0.3865,
+      "step": 1288
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.334883256421227,
+      "learning_rate": 9.931929982522906e-06,
+      "loss": 0.4004,
+      "step": 1289
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0309873328446724,
+      "learning_rate": 9.931762395263798e-06,
+      "loss": 0.3918,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 76.55529773301637,
+      "learning_rate": 9.93159460337711e-06,
+      "loss": 0.4079,
+      "step": 1291
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.8661975404978115,
+      "learning_rate": 9.931426606869802e-06,
+      "loss": 0.4158,
+      "step": 1292
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.954796512963682,
+      "learning_rate": 9.931258405748846e-06,
+      "loss": 0.3868,
+      "step": 1293
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3411782414968396,
+      "learning_rate": 9.931090000021218e-06,
+      "loss": 0.364,
+      "step": 1294
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.117366431507436,
+      "learning_rate": 9.930921389693907e-06,
+      "loss": 0.4244,
+      "step": 1295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6985165345579143,
+      "learning_rate": 9.93075257477391e-06,
+      "loss": 0.4044,
+      "step": 1296
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2261173854773073,
+      "learning_rate": 9.930583555268232e-06,
+      "loss": 0.3767,
+      "step": 1297
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8436602713913757,
+      "learning_rate": 9.930414331183883e-06,
+      "loss": 0.4865,
+      "step": 1298
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8361347906103542,
+      "learning_rate": 9.930244902527885e-06,
+      "loss": 0.4039,
+      "step": 1299
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.714770980754513,
+      "learning_rate": 9.930075269307271e-06,
+      "loss": 0.4118,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5300074767503293,
+      "learning_rate": 9.929905431529077e-06,
+      "loss": 0.3933,
+      "step": 1301
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 187.2722523416001,
+      "learning_rate": 9.92973538920035e-06,
+      "loss": 0.42,
+      "step": 1302
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.621836919561685,
+      "learning_rate": 9.929565142328145e-06,
+      "loss": 0.419,
+      "step": 1303
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4384579060836136,
+      "learning_rate": 9.929394690919527e-06,
+      "loss": 0.3842,
+      "step": 1304
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.729187611472591,
+      "learning_rate": 9.929224034981568e-06,
+      "loss": 0.3754,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9428666565866435,
+      "learning_rate": 9.929053174521348e-06,
+      "loss": 0.4653,
+      "step": 1306
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3237029777076876,
+      "learning_rate": 9.928882109545956e-06,
+      "loss": 0.389,
+      "step": 1307
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.655425575128045,
+      "learning_rate": 9.928710840062492e-06,
+      "loss": 0.4253,
+      "step": 1308
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7001924494572114,
+      "learning_rate": 9.92853936607806e-06,
+      "loss": 0.4164,
+      "step": 1309
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9253618092665186,
+      "learning_rate": 9.928367687599775e-06,
+      "loss": 0.3999,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.356637593843065,
+      "learning_rate": 9.928195804634761e-06,
+      "loss": 0.4199,
+      "step": 1311
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6567326503830144,
+      "learning_rate": 9.928023717190152e-06,
+      "loss": 0.4097,
+      "step": 1312
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8788831437638487,
+      "learning_rate": 9.927851425273082e-06,
+      "loss": 0.389,
+      "step": 1313
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.371769965405906,
+      "learning_rate": 9.927678928890707e-06,
+      "loss": 0.3723,
+      "step": 1314
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.353042406999752,
+      "learning_rate": 9.92750622805018e-06,
+      "loss": 0.3935,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.760828452052688,
+      "learning_rate": 9.927333322758665e-06,
+      "loss": 0.4225,
+      "step": 1316
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7590947394628937,
+      "learning_rate": 9.92716021302334e-06,
+      "loss": 0.4391,
+      "step": 1317
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.405071562879864,
+      "learning_rate": 9.926986898851387e-06,
+      "loss": 0.4159,
+      "step": 1318
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0184503357446797,
+      "learning_rate": 9.926813380249995e-06,
+      "loss": 0.3992,
+      "step": 1319
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0824959041518154,
+      "learning_rate": 9.926639657226366e-06,
+      "loss": 0.4168,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7458826976092214,
+      "learning_rate": 9.926465729787707e-06,
+      "loss": 0.378,
+      "step": 1321
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5773154796096445,
+      "learning_rate": 9.926291597941234e-06,
+      "loss": 0.4083,
+      "step": 1322
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.11089775571572,
+      "learning_rate": 9.926117261694171e-06,
+      "loss": 0.375,
+      "step": 1323
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5481490222725935,
+      "learning_rate": 9.925942721053755e-06,
+      "loss": 0.3894,
+      "step": 1324
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.283882783001254,
+      "learning_rate": 9.925767976027226e-06,
+      "loss": 0.4141,
+      "step": 1325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7701549571254545,
+      "learning_rate": 9.925593026621833e-06,
+      "loss": 0.3949,
+      "step": 1326
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7391856293318613,
+      "learning_rate": 9.925417872844838e-06,
+      "loss": 0.3814,
+      "step": 1327
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7757977417711746,
+      "learning_rate": 9.925242514703505e-06,
+      "loss": 0.4038,
+      "step": 1328
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0179859633576873,
+      "learning_rate": 9.925066952205113e-06,
+      "loss": 0.4149,
+      "step": 1329
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.27224864723054,
+      "learning_rate": 9.924891185356946e-06,
+      "loss": 0.3984,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 31.31739209436363,
+      "learning_rate": 9.924715214166297e-06,
+      "loss": 0.401,
+      "step": 1331
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6021015902600118,
+      "learning_rate": 9.924539038640464e-06,
+      "loss": 0.426,
+      "step": 1332
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4224725330505392,
+      "learning_rate": 9.92436265878676e-06,
+      "loss": 0.4028,
+      "step": 1333
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5472933057935412,
+      "learning_rate": 9.924186074612502e-06,
+      "loss": 0.3756,
+      "step": 1334
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.758829975204866,
+      "learning_rate": 9.92400928612502e-06,
+      "loss": 0.3987,
+      "step": 1335
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.868957251687971,
+      "learning_rate": 9.923832293331645e-06,
+      "loss": 0.3917,
+      "step": 1336
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3265652216559642,
+      "learning_rate": 9.923655096239722e-06,
+      "loss": 0.4568,
+      "step": 1337
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8043481729106445,
+      "learning_rate": 9.923477694856605e-06,
+      "loss": 0.4653,
+      "step": 1338
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.755369831885525,
+      "learning_rate": 9.923300089189653e-06,
+      "loss": 0.4351,
+      "step": 1339
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6445669739920281,
+      "learning_rate": 9.923122279246234e-06,
+      "loss": 0.3942,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.275747062318478,
+      "learning_rate": 9.922944265033729e-06,
+      "loss": 0.3919,
+      "step": 1341
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.812507379854062,
+      "learning_rate": 9.922766046559522e-06,
+      "loss": 0.3981,
+      "step": 1342
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6450415429396688,
+      "learning_rate": 9.922587623831007e-06,
+      "loss": 0.3976,
+      "step": 1343
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4341460590153163,
+      "learning_rate": 9.922408996855588e-06,
+      "loss": 0.4085,
+      "step": 1344
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.877096783037955,
+      "learning_rate": 9.922230165640678e-06,
+      "loss": 0.4003,
+      "step": 1345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8133136042326141,
+      "learning_rate": 9.922051130193694e-06,
+      "loss": 0.3784,
+      "step": 1346
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.181148964621575,
+      "learning_rate": 9.921871890522066e-06,
+      "loss": 0.3794,
+      "step": 1347
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6226953846995316,
+      "learning_rate": 9.921692446633233e-06,
+      "loss": 0.3905,
+      "step": 1348
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.190614017631786,
+      "learning_rate": 9.921512798534637e-06,
+      "loss": 0.4035,
+      "step": 1349
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8681966725964875,
+      "learning_rate": 9.921332946233733e-06,
+      "loss": 0.4883,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2761549349465735,
+      "learning_rate": 9.921152889737985e-06,
+      "loss": 0.4457,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.682495836410096,
+      "learning_rate": 9.920972629054862e-06,
+      "loss": 0.4148,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.799774426097016,
+      "learning_rate": 9.920792164191844e-06,
+      "loss": 0.4202,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7054108308219185,
+      "learning_rate": 9.920611495156418e-06,
+      "loss": 0.409,
+      "step": 1354
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7246505563609941,
+      "learning_rate": 9.920430621956082e-06,
+      "loss": 0.5196,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5568292174747451,
+      "learning_rate": 9.92024954459834e-06,
+      "loss": 0.4426,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8441637875609818,
+      "learning_rate": 9.920068263090706e-06,
+      "loss": 0.4272,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.087098202640914,
+      "learning_rate": 9.9198867774407e-06,
+      "loss": 0.4004,
+      "step": 1358
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7063518045266224,
+      "learning_rate": 9.919705087655851e-06,
+      "loss": 0.4794,
+      "step": 1359
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3065403119255796,
+      "learning_rate": 9.919523193743701e-06,
+      "loss": 0.4401,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.032868476159452,
+      "learning_rate": 9.919341095711796e-06,
+      "loss": 0.3968,
+      "step": 1361
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.026238979965795,
+      "learning_rate": 9.91915879356769e-06,
+      "loss": 0.4,
+      "step": 1362
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0441710274220073,
+      "learning_rate": 9.918976287318948e-06,
+      "loss": 0.4487,
+      "step": 1363
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.460569208947006,
+      "learning_rate": 9.918793576973145e-06,
+      "loss": 0.4442,
+      "step": 1364
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7484439196754873,
+      "learning_rate": 9.91861066253786e-06,
+      "loss": 0.3806,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4863483898466465,
+      "learning_rate": 9.918427544020678e-06,
+      "loss": 0.4098,
+      "step": 1366
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.800766859707209,
+      "learning_rate": 9.918244221429205e-06,
+      "loss": 0.4336,
+      "step": 1367
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6647415061590345,
+      "learning_rate": 9.918060694771043e-06,
+      "loss": 0.3888,
+      "step": 1368
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.901222269363619,
+      "learning_rate": 9.917876964053806e-06,
+      "loss": 0.4253,
+      "step": 1369
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2585778539588093,
+      "learning_rate": 9.91769302928512e-06,
+      "loss": 0.4311,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5765041214152822,
+      "learning_rate": 9.917508890472613e-06,
+      "loss": 0.4164,
+      "step": 1371
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.11036074889742,
+      "learning_rate": 9.91732454762393e-06,
+      "loss": 0.3994,
+      "step": 1372
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.676435633402841,
+      "learning_rate": 9.917140000746717e-06,
+      "loss": 0.4027,
+      "step": 1373
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4989157433821845,
+      "learning_rate": 9.916955249848631e-06,
+      "loss": 0.3891,
+      "step": 1374
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1544482520646615,
+      "learning_rate": 9.916770294937339e-06,
+      "loss": 0.3936,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5951856923610135,
+      "learning_rate": 9.916585136020513e-06,
+      "loss": 0.3946,
+      "step": 1376
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.132462663210312,
+      "learning_rate": 9.916399773105839e-06,
+      "loss": 0.3855,
+      "step": 1377
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.869095672777846,
+      "learning_rate": 9.916214206201003e-06,
+      "loss": 0.4398,
+      "step": 1378
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0817447810542613,
+      "learning_rate": 9.91602843531371e-06,
+      "loss": 0.418,
+      "step": 1379
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5619346912924112,
+      "learning_rate": 9.915842460451663e-06,
+      "loss": 0.4257,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6600325364141273,
+      "learning_rate": 9.915656281622584e-06,
+      "loss": 0.3744,
+      "step": 1381
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8280151388735506,
+      "learning_rate": 9.915469898834191e-06,
+      "loss": 0.4255,
+      "step": 1382
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4442944631790735,
+      "learning_rate": 9.915283312094222e-06,
+      "loss": 0.4083,
+      "step": 1383
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6272636478051172,
+      "learning_rate": 9.91509652141042e-06,
+      "loss": 0.3985,
+      "step": 1384
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.187900171096838,
+      "learning_rate": 9.91490952679053e-06,
+      "loss": 0.4175,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.751596329155221,
+      "learning_rate": 9.914722328242316e-06,
+      "loss": 0.395,
+      "step": 1386
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.015177979893566,
+      "learning_rate": 9.914534925773543e-06,
+      "loss": 0.4644,
+      "step": 1387
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5039838659252927,
+      "learning_rate": 9.914347319391987e-06,
+      "loss": 0.4122,
+      "step": 1388
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7764630100827854,
+      "learning_rate": 9.914159509105431e-06,
+      "loss": 0.3779,
+      "step": 1389
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9194266369463542,
+      "learning_rate": 9.913971494921669e-06,
+      "loss": 0.3932,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6374283687756888,
+      "learning_rate": 9.9137832768485e-06,
+      "loss": 0.4016,
+      "step": 1391
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6075015121871845,
+      "learning_rate": 9.913594854893738e-06,
+      "loss": 0.42,
+      "step": 1392
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5945031378142487,
+      "learning_rate": 9.913406229065196e-06,
+      "loss": 0.4224,
+      "step": 1393
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3031052827672984,
+      "learning_rate": 9.913217399370702e-06,
+      "loss": 0.3812,
+      "step": 1394
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2678496994549624,
+      "learning_rate": 9.913028365818092e-06,
+      "loss": 0.3828,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4283198790855907,
+      "learning_rate": 9.912839128415209e-06,
+      "loss": 0.4556,
+      "step": 1396
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5556333121747936,
+      "learning_rate": 9.912649687169901e-06,
+      "loss": 0.4559,
+      "step": 1397
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8638758042854766,
+      "learning_rate": 9.912460042090035e-06,
+      "loss": 0.3946,
+      "step": 1398
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2184827489764256,
+      "learning_rate": 9.912270193183476e-06,
+      "loss": 0.3992,
+      "step": 1399
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2500568925914206,
+      "learning_rate": 9.912080140458102e-06,
+      "loss": 0.3921,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2813042713793945,
+      "learning_rate": 9.911889883921797e-06,
+      "loss": 0.4086,
+      "step": 1401
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.940302937643413,
+      "learning_rate": 9.911699423582457e-06,
+      "loss": 0.3584,
+      "step": 1402
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8593565231209148,
+      "learning_rate": 9.911508759447984e-06,
+      "loss": 0.4025,
+      "step": 1403
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6361669972872965,
+      "learning_rate": 9.911317891526286e-06,
+      "loss": 0.4029,
+      "step": 1404
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0361185029938778,
+      "learning_rate": 9.911126819825287e-06,
+      "loss": 0.4166,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6620122593043212,
+      "learning_rate": 9.910935544352914e-06,
+      "loss": 0.4295,
+      "step": 1406
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9094808654016637,
+      "learning_rate": 9.910744065117101e-06,
+      "loss": 0.4207,
+      "step": 1407
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.115139760675823,
+      "learning_rate": 9.910552382125797e-06,
+      "loss": 0.3801,
+      "step": 1408
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.750216896874437,
+      "learning_rate": 9.91036049538695e-06,
+      "loss": 0.3902,
+      "step": 1409
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.764468872418187,
+      "learning_rate": 9.910168404908525e-06,
+      "loss": 0.409,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8805666909647074,
+      "learning_rate": 9.909976110698491e-06,
+      "loss": 0.414,
+      "step": 1411
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8817569820293847,
+      "learning_rate": 9.909783612764827e-06,
+      "loss": 0.408,
+      "step": 1412
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8955912629798846,
+      "learning_rate": 9.909590911115521e-06,
+      "loss": 0.421,
+      "step": 1413
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8233016733210808,
+      "learning_rate": 9.909398005758567e-06,
+      "loss": 0.5172,
+      "step": 1414
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9509445029106045,
+      "learning_rate": 9.909204896701969e-06,
+      "loss": 0.399,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0487482740909604,
+      "learning_rate": 9.909011583953743e-06,
+      "loss": 0.4208,
+      "step": 1416
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9694569347932842,
+      "learning_rate": 9.908818067521904e-06,
+      "loss": 0.4085,
+      "step": 1417
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0491318225826527,
+      "learning_rate": 9.908624347414486e-06,
+      "loss": 0.423,
+      "step": 1418
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4949193002474532,
+      "learning_rate": 9.908430423639524e-06,
+      "loss": 0.4163,
+      "step": 1419
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.605398834105979,
+      "learning_rate": 9.908236296205066e-06,
+      "loss": 0.4119,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4125172978253353,
+      "learning_rate": 9.908041965119167e-06,
+      "loss": 0.3951,
+      "step": 1421
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.185038540239043,
+      "learning_rate": 9.907847430389887e-06,
+      "loss": 0.3815,
+      "step": 1422
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4387868815608758,
+      "learning_rate": 9.9076526920253e-06,
+      "loss": 0.3979,
+      "step": 1423
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4015483848228638,
+      "learning_rate": 9.907457750033487e-06,
+      "loss": 0.3604,
+      "step": 1424
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.357191981985771,
+      "learning_rate": 9.907262604422537e-06,
+      "loss": 0.3918,
+      "step": 1425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5894567076827006,
+      "learning_rate": 9.907067255200543e-06,
+      "loss": 0.404,
+      "step": 1426
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9072082104415833,
+      "learning_rate": 9.906871702375611e-06,
+      "loss": 0.3818,
+      "step": 1427
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2901148439758,
+      "learning_rate": 9.90667594595586e-06,
+      "loss": 0.4283,
+      "step": 1428
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.327012966175589,
+      "learning_rate": 9.906479985949407e-06,
+      "loss": 0.3966,
+      "step": 1429
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.459879041421819,
+      "learning_rate": 9.906283822364384e-06,
+      "loss": 0.4073,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7313958299477028,
+      "learning_rate": 9.90608745520893e-06,
+      "loss": 0.4236,
+      "step": 1431
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.530695929689069,
+      "learning_rate": 9.905890884491196e-06,
+      "loss": 0.4107,
+      "step": 1432
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4345884043627475,
+      "learning_rate": 9.905694110219335e-06,
+      "loss": 0.4173,
+      "step": 1433
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.250153863573317,
+      "learning_rate": 9.90549713240151e-06,
+      "loss": 0.4001,
+      "step": 1434
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9437826118184052,
+      "learning_rate": 9.905299951045897e-06,
+      "loss": 0.4219,
+      "step": 1435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5513169384343328,
+      "learning_rate": 9.905102566160676e-06,
+      "loss": 0.4077,
+      "step": 1436
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1322290329463036,
+      "learning_rate": 9.904904977754038e-06,
+      "loss": 0.4017,
+      "step": 1437
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6814866941618773,
+      "learning_rate": 9.904707185834178e-06,
+      "loss": 0.4068,
+      "step": 1438
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.386708142763933,
+      "learning_rate": 9.904509190409306e-06,
+      "loss": 0.4049,
+      "step": 1439
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9453388681624317,
+      "learning_rate": 9.904310991487638e-06,
+      "loss": 0.4025,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8745260712880345,
+      "learning_rate": 9.904112589077395e-06,
+      "loss": 0.4031,
+      "step": 1441
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5975200337702475,
+      "learning_rate": 9.90391398318681e-06,
+      "loss": 0.4017,
+      "step": 1442
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.324353289238748,
+      "learning_rate": 9.903715173824123e-06,
+      "loss": 0.393,
+      "step": 1443
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1388241598048117,
+      "learning_rate": 9.903516160997583e-06,
+      "loss": 0.3989,
+      "step": 1444
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.059743073559714,
+      "learning_rate": 9.903316944715449e-06,
+      "loss": 0.3701,
+      "step": 1445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.886145433535198,
+      "learning_rate": 9.903117524985986e-06,
+      "loss": 0.3989,
+      "step": 1446
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9183789478214723,
+      "learning_rate": 9.902917901817466e-06,
+      "loss": 0.519,
+      "step": 1447
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6682113945767112,
+      "learning_rate": 9.902718075218176e-06,
+      "loss": 0.4025,
+      "step": 1448
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.043910702886714,
+      "learning_rate": 9.902518045196404e-06,
+      "loss": 0.4205,
+      "step": 1449
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6660949741835804,
+      "learning_rate": 9.902317811760449e-06,
+      "loss": 0.3856,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6677608029610768,
+      "learning_rate": 9.902117374918623e-06,
+      "loss": 0.4778,
+      "step": 1451
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.945677472625155,
+      "learning_rate": 9.901916734679237e-06,
+      "loss": 0.4131,
+      "step": 1452
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.673107488354125,
+      "learning_rate": 9.901715891050622e-06,
+      "loss": 0.4297,
+      "step": 1453
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.023626774814282,
+      "learning_rate": 9.901514844041107e-06,
+      "loss": 0.3997,
+      "step": 1454
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7819456139503294,
+      "learning_rate": 9.901313593659035e-06,
+      "loss": 0.3899,
+      "step": 1455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.755152624488014,
+      "learning_rate": 9.901112139912757e-06,
+      "loss": 0.4234,
+      "step": 1456
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3820210120304215,
+      "learning_rate": 9.90091048281063e-06,
+      "loss": 0.3778,
+      "step": 1457
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.201837689877226,
+      "learning_rate": 9.90070862236102e-06,
+      "loss": 0.3819,
+      "step": 1458
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.46930200786034,
+      "learning_rate": 9.900506558572309e-06,
+      "loss": 0.4427,
+      "step": 1459
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4506886847437297,
+      "learning_rate": 9.900304291452873e-06,
+      "loss": 0.3964,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7782056866041125,
+      "learning_rate": 9.90010182101111e-06,
+      "loss": 0.38,
+      "step": 1461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1578013332547443,
+      "learning_rate": 9.899899147255418e-06,
+      "loss": 0.3963,
+      "step": 1462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6429588857787716,
+      "learning_rate": 9.899696270194208e-06,
+      "loss": 0.4039,
+      "step": 1463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.782755101220264,
+      "learning_rate": 9.899493189835896e-06,
+      "loss": 0.4021,
+      "step": 1464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.805700991513922,
+      "learning_rate": 9.899289906188909e-06,
+      "loss": 0.3774,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1592057386846453,
+      "learning_rate": 9.899086419261683e-06,
+      "loss": 0.4994,
+      "step": 1466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2501526015494115,
+      "learning_rate": 9.89888272906266e-06,
+      "loss": 0.4112,
+      "step": 1467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.562735614642925,
+      "learning_rate": 9.89867883560029e-06,
+      "loss": 0.4091,
+      "step": 1468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7883783901104329,
+      "learning_rate": 9.898474738883033e-06,
+      "loss": 0.4354,
+      "step": 1469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.86557666058772,
+      "learning_rate": 9.898270438919359e-06,
+      "loss": 0.4203,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3268077301449126,
+      "learning_rate": 9.898065935717746e-06,
+      "loss": 0.3679,
+      "step": 1471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0982417650484964,
+      "learning_rate": 9.897861229286676e-06,
+      "loss": 0.3848,
+      "step": 1472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.484283339522934,
+      "learning_rate": 9.897656319634643e-06,
+      "loss": 0.412,
+      "step": 1473
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.871856135984185,
+      "learning_rate": 9.897451206770152e-06,
+      "loss": 0.3826,
+      "step": 1474
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7713368662430186,
+      "learning_rate": 9.897245890701713e-06,
+      "loss": 0.4779,
+      "step": 1475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1635938435196946,
+      "learning_rate": 9.89704037143784e-06,
+      "loss": 0.4546,
+      "step": 1476
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5875702231075441,
+      "learning_rate": 9.896834648987065e-06,
+      "loss": 0.4266,
+      "step": 1477
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.968822406602112,
+      "learning_rate": 9.896628723357923e-06,
+      "loss": 0.3995,
+      "step": 1478
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.271946906731635,
+      "learning_rate": 9.896422594558957e-06,
+      "loss": 0.4044,
+      "step": 1479
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7425755441872492,
+      "learning_rate": 9.896216262598722e-06,
+      "loss": 0.3874,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.951237218746135,
+      "learning_rate": 9.896009727485778e-06,
+      "loss": 0.4111,
+      "step": 1481
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.664163980488461,
+      "learning_rate": 9.895802989228691e-06,
+      "loss": 0.4666,
+      "step": 1482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1918348287420946,
+      "learning_rate": 9.895596047836045e-06,
+      "loss": 0.416,
+      "step": 1483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.302188351677537,
+      "learning_rate": 9.895388903316424e-06,
+      "loss": 0.3995,
+      "step": 1484
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4967731946965595,
+      "learning_rate": 9.895181555678419e-06,
+      "loss": 0.429,
+      "step": 1485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4109589221032037,
+      "learning_rate": 9.894974004930638e-06,
+      "loss": 0.3997,
+      "step": 1486
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6575259792296093,
+      "learning_rate": 9.894766251081691e-06,
+      "loss": 0.424,
+      "step": 1487
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.621454676112057,
+      "learning_rate": 9.894558294140199e-06,
+      "loss": 0.4205,
+      "step": 1488
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4304440388356046,
+      "learning_rate": 9.894350134114788e-06,
+      "loss": 0.4239,
+      "step": 1489
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.149051462032158,
+      "learning_rate": 9.894141771014098e-06,
+      "loss": 0.439,
+      "step": 1490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.675273168917547,
+      "learning_rate": 9.893933204846772e-06,
+      "loss": 0.395,
+      "step": 1491
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9996833286022,
+      "learning_rate": 9.893724435621466e-06,
+      "loss": 0.4184,
+      "step": 1492
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.554857719359423,
+      "learning_rate": 9.893515463346841e-06,
+      "loss": 0.3705,
+      "step": 1493
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.521627959182956,
+      "learning_rate": 9.893306288031565e-06,
+      "loss": 0.4201,
+      "step": 1494
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.78095549023195,
+      "learning_rate": 9.893096909684323e-06,
+      "loss": 0.4135,
+      "step": 1495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2207693140787756,
+      "learning_rate": 9.892887328313796e-06,
+      "loss": 0.4111,
+      "step": 1496
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.956904065585402,
+      "learning_rate": 9.892677543928687e-06,
+      "loss": 0.4336,
+      "step": 1497
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.407424917915186,
+      "learning_rate": 9.892467556537692e-06,
+      "loss": 0.4286,
+      "step": 1498
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5868108314301557,
+      "learning_rate": 9.89225736614953e-06,
+      "loss": 0.384,
+      "step": 1499
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7259959262127642,
+      "learning_rate": 9.89204697277292e-06,
+      "loss": 0.3981,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.074368901182226,
+      "learning_rate": 9.891836376416593e-06,
+      "loss": 0.3943,
+      "step": 1501
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.731554332407603,
+      "learning_rate": 9.891625577089285e-06,
+      "loss": 0.3714,
+      "step": 1502
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6119072743723795,
+      "learning_rate": 9.891414574799743e-06,
+      "loss": 0.413,
+      "step": 1503
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.301487286293184,
+      "learning_rate": 9.891203369556722e-06,
+      "loss": 0.3408,
+      "step": 1504
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5721496853568406,
+      "learning_rate": 9.890991961368986e-06,
+      "loss": 0.3876,
+      "step": 1505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4907831396596125,
+      "learning_rate": 9.890780350245305e-06,
+      "loss": 0.3909,
+      "step": 1506
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7123762050986583,
+      "learning_rate": 9.890568536194462e-06,
+      "loss": 0.4781,
+      "step": 1507
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.167422005271638,
+      "learning_rate": 9.890356519225244e-06,
+      "loss": 0.3659,
+      "step": 1508
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.645429350292824,
+      "learning_rate": 9.890144299346445e-06,
+      "loss": 0.3883,
+      "step": 1509
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.52323903959301,
+      "learning_rate": 9.889931876566877e-06,
+      "loss": 0.4334,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3712718822586965,
+      "learning_rate": 9.889719250895347e-06,
+      "loss": 0.4106,
+      "step": 1511
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3503318056051734,
+      "learning_rate": 9.88950642234068e-06,
+      "loss": 0.3916,
+      "step": 1512
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6745853673430179,
+      "learning_rate": 9.889293390911708e-06,
+      "loss": 0.4582,
+      "step": 1513
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1754403974935093,
+      "learning_rate": 9.88908015661727e-06,
+      "loss": 0.4488,
+      "step": 1514
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9143321788962888,
+      "learning_rate": 9.88886671946621e-06,
+      "loss": 0.4121,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9017372274543405,
+      "learning_rate": 9.888653079467388e-06,
+      "loss": 0.3977,
+      "step": 1516
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5638253592781546,
+      "learning_rate": 9.888439236629665e-06,
+      "loss": 0.4016,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.771085353001597,
+      "learning_rate": 9.888225190961916e-06,
+      "loss": 0.4062,
+      "step": 1518
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6580835384058848,
+      "learning_rate": 9.888010942473021e-06,
+      "loss": 0.4446,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.725761692401521,
+      "learning_rate": 9.887796491171871e-06,
+      "loss": 0.3749,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.135603827933371,
+      "learning_rate": 9.887581837067362e-06,
+      "loss": 0.4283,
+      "step": 1521
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1041515451167183,
+      "learning_rate": 9.8873669801684e-06,
+      "loss": 0.4041,
+      "step": 1522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.1887339764795035,
+      "learning_rate": 9.887151920483904e-06,
+      "loss": 0.4079,
+      "step": 1523
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1493065584900863,
+      "learning_rate": 9.886936658022792e-06,
+      "loss": 0.42,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0796373023459873,
+      "learning_rate": 9.886721192793998e-06,
+      "loss": 0.4442,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7252849265660066,
+      "learning_rate": 9.886505524806462e-06,
+      "loss": 0.4001,
+      "step": 1526
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.115412855532158,
+      "learning_rate": 9.886289654069134e-06,
+      "loss": 0.3988,
+      "step": 1527
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0289058711966885,
+      "learning_rate": 9.886073580590968e-06,
+      "loss": 0.4022,
+      "step": 1528
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.639479006211727,
+      "learning_rate": 9.88585730438093e-06,
+      "loss": 0.4213,
+      "step": 1529
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.285274508998952,
+      "learning_rate": 9.885640825447995e-06,
+      "loss": 0.4253,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.332633692302388,
+      "learning_rate": 9.885424143801144e-06,
+      "loss": 0.426,
+      "step": 1531
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.067381084276767,
+      "learning_rate": 9.885207259449367e-06,
+      "loss": 0.3872,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.012950955899202,
+      "learning_rate": 9.884990172401664e-06,
+      "loss": 0.3851,
+      "step": 1533
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.509710170829564,
+      "learning_rate": 9.884772882667045e-06,
+      "loss": 0.4018,
+      "step": 1534
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.557531037945831,
+      "learning_rate": 9.88455539025452e-06,
+      "loss": 0.4018,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8023998549226363,
+      "learning_rate": 9.884337695173115e-06,
+      "loss": 0.3882,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.602399167302603,
+      "learning_rate": 9.884119797431864e-06,
+      "loss": 0.357,
+      "step": 1537
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3876713227208,
+      "learning_rate": 9.883901697039809e-06,
+      "loss": 0.3714,
+      "step": 1538
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6313126107514955,
+      "learning_rate": 9.883683394005997e-06,
+      "loss": 0.4047,
+      "step": 1539
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3148597499003523,
+      "learning_rate": 9.883464888339487e-06,
+      "loss": 0.3718,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8969185070682832,
+      "learning_rate": 9.883246180049345e-06,
+      "loss": 0.3797,
+      "step": 1541
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7229688177338228,
+      "learning_rate": 9.883027269144643e-06,
+      "loss": 0.4941,
+      "step": 1542
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7406262608377298,
+      "learning_rate": 9.882808155634469e-06,
+      "loss": 0.4786,
+      "step": 1543
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.130647243420708,
+      "learning_rate": 9.88258883952791e-06,
+      "loss": 0.3824,
+      "step": 1544
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.61938707034376,
+      "learning_rate": 9.882369320834068e-06,
+      "loss": 0.4437,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2788536790741447,
+      "learning_rate": 9.882149599562052e-06,
+      "loss": 0.4287,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.4476349201232,
+      "learning_rate": 9.881929675720976e-06,
+      "loss": 0.4043,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2189743944428164,
+      "learning_rate": 9.881709549319967e-06,
+      "loss": 0.4272,
+      "step": 1548
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4537071417017278,
+      "learning_rate": 9.881489220368159e-06,
+      "loss": 0.424,
+      "step": 1549
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9156323739614187,
+      "learning_rate": 9.881268688874692e-06,
+      "loss": 0.4018,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8939609430497777,
+      "learning_rate": 9.881047954848716e-06,
+      "loss": 0.4109,
+      "step": 1551
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.203243748382527,
+      "learning_rate": 9.880827018299392e-06,
+      "loss": 0.3821,
+      "step": 1552
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6113795149291312,
+      "learning_rate": 9.880605879235885e-06,
+      "loss": 0.4107,
+      "step": 1553
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.55470526097955,
+      "learning_rate": 9.880384537667371e-06,
+      "loss": 0.3983,
+      "step": 1554
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5228408730582013,
+      "learning_rate": 9.880162993603032e-06,
+      "loss": 0.3949,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.716271503272738,
+      "learning_rate": 9.879941247052066e-06,
+      "loss": 0.3875,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.048444172986198,
+      "learning_rate": 9.879719298023669e-06,
+      "loss": 0.3819,
+      "step": 1557
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7746808164630226,
+      "learning_rate": 9.87949714652705e-06,
+      "loss": 0.3964,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0100644144483955,
+      "learning_rate": 9.879274792571427e-06,
+      "loss": 0.4053,
+      "step": 1559
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0803703390667034,
+      "learning_rate": 9.879052236166029e-06,
+      "loss": 0.4091,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0160432550163137,
+      "learning_rate": 9.878829477320085e-06,
+      "loss": 0.3961,
+      "step": 1561
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.542126576885845,
+      "learning_rate": 9.87860651604284e-06,
+      "loss": 0.4289,
+      "step": 1562
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8848961311721117,
+      "learning_rate": 9.878383352343546e-06,
+      "loss": 0.4093,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6335960775558451,
+      "learning_rate": 9.878159986231461e-06,
+      "loss": 0.3961,
+      "step": 1564
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5331882824257386,
+      "learning_rate": 9.877936417715856e-06,
+      "loss": 0.3945,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.216200431204067,
+      "learning_rate": 9.877712646806003e-06,
+      "loss": 0.39,
+      "step": 1566
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 13.218505149285058,
+      "learning_rate": 9.87748867351119e-06,
+      "loss": 0.3896,
+      "step": 1567
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1156303943815904,
+      "learning_rate": 9.877264497840707e-06,
+      "loss": 0.4114,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.875347790840662,
+      "learning_rate": 9.877040119803855e-06,
+      "loss": 0.403,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9594594653352964,
+      "learning_rate": 9.87681553940995e-06,
+      "loss": 0.4144,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7172301747489058,
+      "learning_rate": 9.876590756668303e-06,
+      "loss": 0.3885,
+      "step": 1571
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5643726014145547,
+      "learning_rate": 9.876365771588246e-06,
+      "loss": 0.3906,
+      "step": 1572
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.379988230516699,
+      "learning_rate": 9.876140584179111e-06,
+      "loss": 0.4027,
+      "step": 1573
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4426545999875537,
+      "learning_rate": 9.87591519445024e-06,
+      "loss": 0.3911,
+      "step": 1574
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4506249836419176,
+      "learning_rate": 9.87568960241099e-06,
+      "loss": 0.3985,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3969053387991355,
+      "learning_rate": 9.875463808070715e-06,
+      "loss": 0.395,
+      "step": 1576
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8140566787652515,
+      "learning_rate": 9.87523781143879e-06,
+      "loss": 0.3764,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6049219044618324,
+      "learning_rate": 9.875011612524588e-06,
+      "loss": 0.4054,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6594386474651648,
+      "learning_rate": 9.874785211337495e-06,
+      "loss": 0.4294,
+      "step": 1579
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8489354266790934,
+      "learning_rate": 9.874558607886903e-06,
+      "loss": 0.4106,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9675224339666788,
+      "learning_rate": 9.874331802182218e-06,
+      "loss": 0.4027,
+      "step": 1581
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.438333569282693,
+      "learning_rate": 9.874104794232849e-06,
+      "loss": 0.3833,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.758263080262827,
+      "learning_rate": 9.873877584048215e-06,
+      "loss": 0.4013,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3092880221130323,
+      "learning_rate": 9.873650171637742e-06,
+      "loss": 0.4122,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.797009473110623,
+      "learning_rate": 9.873422557010868e-06,
+      "loss": 0.3947,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 33.2149911845115,
+      "learning_rate": 9.873194740177035e-06,
+      "loss": 0.399,
+      "step": 1586
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8303575239244754,
+      "learning_rate": 9.872966721145696e-06,
+      "loss": 0.4253,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6917698522229314,
+      "learning_rate": 9.872738499926313e-06,
+      "loss": 0.4217,
+      "step": 1588
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.867497128558469,
+      "learning_rate": 9.872510076528354e-06,
+      "loss": 0.39,
+      "step": 1589
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8831302942748656,
+      "learning_rate": 9.872281450961298e-06,
+      "loss": 0.3866,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6060489849243156,
+      "learning_rate": 9.872052623234632e-06,
+      "loss": 0.4132,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1373426595867966,
+      "learning_rate": 9.871823593357847e-06,
+      "loss": 0.4262,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4616934618411233,
+      "learning_rate": 9.871594361340448e-06,
+      "loss": 0.377,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1246185516809977,
+      "learning_rate": 9.871364927191946e-06,
+      "loss": 0.4494,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6667229764513893,
+      "learning_rate": 9.87113529092186e-06,
+      "loss": 0.381,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9040267498341337,
+      "learning_rate": 9.870905452539721e-06,
+      "loss": 0.3983,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.445000136769461,
+      "learning_rate": 9.870675412055061e-06,
+      "loss": 0.3836,
+      "step": 1597
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.98771284005042,
+      "learning_rate": 9.870445169477428e-06,
+      "loss": 0.5325,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.22609115830435,
+      "learning_rate": 9.870214724816373e-06,
+      "loss": 0.3934,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9701962859123494,
+      "learning_rate": 9.869984078081459e-06,
+      "loss": 0.4125,
+      "step": 1600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.816912168703317,
+      "learning_rate": 9.869753229282256e-06,
+      "loss": 0.3964,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8402158447295298,
+      "learning_rate": 9.869522178428342e-06,
+      "loss": 0.3973,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8551094047563308,
+      "learning_rate": 9.869290925529303e-06,
+      "loss": 0.4264,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.053508950539129,
+      "learning_rate": 9.869059470594734e-06,
+      "loss": 0.3979,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.370219472752349,
+      "learning_rate": 9.86882781363424e-06,
+      "loss": 0.3683,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.420278752012178,
+      "learning_rate": 9.868595954657432e-06,
+      "loss": 0.3844,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.094662872243442,
+      "learning_rate": 9.86836389367393e-06,
+      "loss": 0.4059,
+      "step": 1607
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.17522914834035,
+      "learning_rate": 9.868131630693363e-06,
+      "loss": 0.3919,
+      "step": 1608
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0180806861260696,
+      "learning_rate": 9.867899165725367e-06,
+      "loss": 0.3979,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8407740171109623,
+      "learning_rate": 9.867666498779589e-06,
+      "loss": 0.3942,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7393380060798258,
+      "learning_rate": 9.867433629865682e-06,
+      "loss": 0.3783,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.653835467011482,
+      "learning_rate": 9.867200558993308e-06,
+      "loss": 0.4261,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3885446718149328,
+      "learning_rate": 9.866967286172138e-06,
+      "loss": 0.3663,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.768975395850437,
+      "learning_rate": 9.866733811411851e-06,
+      "loss": 0.3793,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6931236979606916,
+      "learning_rate": 9.866500134722135e-06,
+      "loss": 0.3825,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6536041558537826,
+      "learning_rate": 9.866266256112683e-06,
+      "loss": 0.3672,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2573399623354495,
+      "learning_rate": 9.8660321755932e-06,
+      "loss": 0.4195,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4421460492372233,
+      "learning_rate": 9.865797893173398e-06,
+      "loss": 0.3736,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.634157856818414,
+      "learning_rate": 9.865563408863001e-06,
+      "loss": 0.439,
+      "step": 1619
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8969091112606098,
+      "learning_rate": 9.865328722671736e-06,
+      "loss": 0.4869,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.780938074845408,
+      "learning_rate": 9.86509383460934e-06,
+      "loss": 0.3823,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2094474894144644,
+      "learning_rate": 9.86485874468556e-06,
+      "loss": 0.4181,
+      "step": 1622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9213414327350735,
+      "learning_rate": 9.864623452910147e-06,
+      "loss": 0.3897,
+      "step": 1623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.595731521198233,
+      "learning_rate": 9.86438795929287e-06,
+      "loss": 0.4508,
+      "step": 1624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2749829925861649,
+      "learning_rate": 9.864152263843494e-06,
+      "loss": 0.3683,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.931634651898662,
+      "learning_rate": 9.863916366571801e-06,
+      "loss": 0.3681,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.047504369413371,
+      "learning_rate": 9.863680267487579e-06,
+      "loss": 0.3851,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.562617848727528,
+      "learning_rate": 9.863443966600625e-06,
+      "loss": 0.43,
+      "step": 1628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3665948079790065,
+      "learning_rate": 9.863207463920741e-06,
+      "loss": 0.4091,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9577528694811124,
+      "learning_rate": 9.862970759457741e-06,
+      "loss": 0.3883,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7101952810443333,
+      "learning_rate": 9.86273385322145e-06,
+      "loss": 0.3971,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8747489452894384,
+      "learning_rate": 9.862496745221691e-06,
+      "loss": 0.3899,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.399927821968412,
+      "learning_rate": 9.862259435468305e-06,
+      "loss": 0.4037,
+      "step": 1633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.434823447356248,
+      "learning_rate": 9.862021923971139e-06,
+      "loss": 0.4224,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7375833342698581,
+      "learning_rate": 9.861784210740048e-06,
+      "loss": 0.3817,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8843841630356726,
+      "learning_rate": 9.861546295784896e-06,
+      "loss": 0.3869,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8641709311826786,
+      "learning_rate": 9.86130817911555e-06,
+      "loss": 0.3636,
+      "step": 1637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.42880075483143,
+      "learning_rate": 9.861069860741896e-06,
+      "loss": 0.4223,
+      "step": 1638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5093568068595866,
+      "learning_rate": 9.860831340673818e-06,
+      "loss": 0.3893,
+      "step": 1639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.479179518708439,
+      "learning_rate": 9.860592618921213e-06,
+      "loss": 0.3596,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8754934456139913,
+      "learning_rate": 9.860353695493987e-06,
+      "loss": 0.411,
+      "step": 1641
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0307806127913177,
+      "learning_rate": 9.860114570402055e-06,
+      "loss": 0.4007,
+      "step": 1642
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9901093263597351,
+      "learning_rate": 9.859875243655336e-06,
+      "loss": 0.5332,
+      "step": 1643
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6527672581452526,
+      "learning_rate": 9.85963571526376e-06,
+      "loss": 0.4029,
+      "step": 1644
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9962457409802663,
+      "learning_rate": 9.859395985237268e-06,
+      "loss": 0.4079,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8445980379975375,
+      "learning_rate": 9.859156053585805e-06,
+      "loss": 0.4168,
+      "step": 1646
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6673165701862798,
+      "learning_rate": 9.858915920319325e-06,
+      "loss": 0.3907,
+      "step": 1647
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1450276935779082,
+      "learning_rate": 9.858675585447795e-06,
+      "loss": 0.3825,
+      "step": 1648
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9224143513215968,
+      "learning_rate": 9.858435048981184e-06,
+      "loss": 0.4104,
+      "step": 1649
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.099476366730014,
+      "learning_rate": 9.858194310929474e-06,
+      "loss": 0.4048,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7980374577923077,
+      "learning_rate": 9.857953371302651e-06,
+      "loss": 0.3801,
+      "step": 1651
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4202534694887032,
+      "learning_rate": 9.857712230110717e-06,
+      "loss": 0.37,
+      "step": 1652
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7026733049341634,
+      "learning_rate": 9.857470887363672e-06,
+      "loss": 0.3849,
+      "step": 1653
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9840497547579354,
+      "learning_rate": 9.857229343071532e-06,
+      "loss": 0.4376,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.3932046299332677,
+      "learning_rate": 9.85698759724432e-06,
+      "loss": 0.3797,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6382322967133962,
+      "learning_rate": 9.856745649892066e-06,
+      "loss": 0.3596,
+      "step": 1656
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.638171228039363,
+      "learning_rate": 9.856503501024807e-06,
+      "loss": 0.4371,
+      "step": 1657
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0245138275986685,
+      "learning_rate": 9.856261150652593e-06,
+      "loss": 0.3624,
+      "step": 1658
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5752020038104408,
+      "learning_rate": 9.856018598785477e-06,
+      "loss": 0.4202,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6923476571685003,
+      "learning_rate": 9.855775845433527e-06,
+      "loss": 0.4174,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9751350900252784,
+      "learning_rate": 9.855532890606809e-06,
+      "loss": 0.4053,
+      "step": 1661
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6875648058751105,
+      "learning_rate": 9.855289734315407e-06,
+      "loss": 0.4246,
+      "step": 1662
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7199870567275772,
+      "learning_rate": 9.855046376569412e-06,
+      "loss": 0.3807,
+      "step": 1663
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.969451586330859,
+      "learning_rate": 9.854802817378918e-06,
+      "loss": 0.3796,
+      "step": 1664
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6592376380297793,
+      "learning_rate": 9.854559056754031e-06,
+      "loss": 0.4015,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5959033763493164,
+      "learning_rate": 9.85431509470487e-06,
+      "loss": 0.3982,
+      "step": 1666
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7046308592515858,
+      "learning_rate": 9.85407093124155e-06,
+      "loss": 0.3678,
+      "step": 1667
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.102849780940891,
+      "learning_rate": 9.853826566374206e-06,
+      "loss": 0.4168,
+      "step": 1668
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.734985635647761,
+      "learning_rate": 9.853582000112976e-06,
+      "loss": 0.3746,
+      "step": 1669
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.045023648234381,
+      "learning_rate": 9.853337232468008e-06,
+      "loss": 0.5443,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4610084570737127,
+      "learning_rate": 9.85309226344946e-06,
+      "loss": 0.3926,
+      "step": 1671
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8481416710293914,
+      "learning_rate": 9.85284709306749e-06,
+      "loss": 0.4055,
+      "step": 1672
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.965990410125677,
+      "learning_rate": 9.852601721332278e-06,
+      "loss": 0.3732,
+      "step": 1673
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.52394801893013,
+      "learning_rate": 9.852356148253999e-06,
+      "loss": 0.4066,
+      "step": 1674
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.366918402226468,
+      "learning_rate": 9.852110373842846e-06,
+      "loss": 0.3858,
+      "step": 1675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9034868674091014,
+      "learning_rate": 9.851864398109015e-06,
+      "loss": 0.4044,
+      "step": 1676
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.471418975171063,
+      "learning_rate": 9.851618221062712e-06,
+      "loss": 0.3832,
+      "step": 1677
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7960215718083363,
+      "learning_rate": 9.851371842714152e-06,
+      "loss": 0.3868,
+      "step": 1678
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1817685990868085,
+      "learning_rate": 9.851125263073556e-06,
+      "loss": 0.41,
+      "step": 1679
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.971487075980344,
+      "learning_rate": 9.850878482151158e-06,
+      "loss": 0.4223,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.410151505915163,
+      "learning_rate": 9.850631499957196e-06,
+      "loss": 0.3754,
+      "step": 1681
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.067354857684912,
+      "learning_rate": 9.850384316501915e-06,
+      "loss": 0.3854,
+      "step": 1682
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8751145476407016,
+      "learning_rate": 9.850136931795576e-06,
+      "loss": 0.3807,
+      "step": 1683
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7303156103006214,
+      "learning_rate": 9.849889345848438e-06,
+      "loss": 0.4143,
+      "step": 1684
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4127764967959593,
+      "learning_rate": 9.849641558670779e-06,
+      "loss": 0.3912,
+      "step": 1685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9093287900042089,
+      "learning_rate": 9.849393570272875e-06,
+      "loss": 0.4027,
+      "step": 1686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5926673449856141,
+      "learning_rate": 9.849145380665021e-06,
+      "loss": 0.4059,
+      "step": 1687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.00550667338651,
+      "learning_rate": 9.848896989857512e-06,
+      "loss": 0.3613,
+      "step": 1688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7627412396817441,
+      "learning_rate": 9.848648397860654e-06,
+      "loss": 0.3947,
+      "step": 1689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8667088767380863,
+      "learning_rate": 9.848399604684762e-06,
+      "loss": 0.486,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3910336632697318,
+      "learning_rate": 9.848150610340158e-06,
+      "loss": 0.3763,
+      "step": 1691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9004403833278962,
+      "learning_rate": 9.847901414837173e-06,
+      "loss": 0.4266,
+      "step": 1692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.020778797956488,
+      "learning_rate": 9.847652018186149e-06,
+      "loss": 0.3843,
+      "step": 1693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7097985113804344,
+      "learning_rate": 9.847402420397431e-06,
+      "loss": 0.4501,
+      "step": 1694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.795067892387273,
+      "learning_rate": 9.847152621481378e-06,
+      "loss": 0.3937,
+      "step": 1695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.137775892968101,
+      "learning_rate": 9.846902621448354e-06,
+      "loss": 0.4002,
+      "step": 1696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2106505072535425,
+      "learning_rate": 9.846652420308728e-06,
+      "loss": 0.4096,
+      "step": 1697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.252113147118114,
+      "learning_rate": 9.846402018072888e-06,
+      "loss": 0.4264,
+      "step": 1698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8598361009513775,
+      "learning_rate": 9.846151414751217e-06,
+      "loss": 0.3912,
+      "step": 1699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6481025561732026,
+      "learning_rate": 9.845900610354117e-06,
+      "loss": 0.4814,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8954598861269756,
+      "learning_rate": 9.845649604891996e-06,
+      "loss": 0.3972,
+      "step": 1701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6747124277930796,
+      "learning_rate": 9.845398398375264e-06,
+      "loss": 0.386,
+      "step": 1702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9833024773849135,
+      "learning_rate": 9.845146990814345e-06,
+      "loss": 0.3548,
+      "step": 1703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7982434928052577,
+      "learning_rate": 9.844895382219673e-06,
+      "loss": 0.3917,
+      "step": 1704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.620190892161271,
+      "learning_rate": 9.844643572601685e-06,
+      "loss": 0.4271,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9066001799296057,
+      "learning_rate": 9.844391561970831e-06,
+      "loss": 0.379,
+      "step": 1706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0798176899619896,
+      "learning_rate": 9.844139350337566e-06,
+      "loss": 0.3768,
+      "step": 1707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.166746859557011,
+      "learning_rate": 9.843886937712355e-06,
+      "loss": 0.3948,
+      "step": 1708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.285812184538659,
+      "learning_rate": 9.843634324105671e-06,
+      "loss": 0.4038,
+      "step": 1709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.289329268104687,
+      "learning_rate": 9.843381509527997e-06,
+      "loss": 0.3844,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8602726742533615,
+      "learning_rate": 9.84312849398982e-06,
+      "loss": 0.3818,
+      "step": 1711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6256248754386069,
+      "learning_rate": 9.84287527750164e-06,
+      "loss": 0.4762,
+      "step": 1712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8693918376721483,
+      "learning_rate": 9.842621860073963e-06,
+      "loss": 0.3712,
+      "step": 1713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3307457721439757,
+      "learning_rate": 9.842368241717304e-06,
+      "loss": 0.3741,
+      "step": 1714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.146545990434858,
+      "learning_rate": 9.842114422442184e-06,
+      "loss": 0.3798,
+      "step": 1715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6478512616698305,
+      "learning_rate": 9.841860402259139e-06,
+      "loss": 0.4784,
+      "step": 1716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7282611243319868,
+      "learning_rate": 9.841606181178703e-06,
+      "loss": 0.4248,
+      "step": 1717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2365112124158246,
+      "learning_rate": 9.841351759211426e-06,
+      "loss": 0.4063,
+      "step": 1718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9877658777221152,
+      "learning_rate": 9.841097136367868e-06,
+      "loss": 0.3733,
+      "step": 1719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.70774960444382,
+      "learning_rate": 9.84084231265859e-06,
+      "loss": 0.3845,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.967471005270002,
+      "learning_rate": 9.840587288094165e-06,
+      "loss": 0.3592,
+      "step": 1721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.7789218726089935,
+      "learning_rate": 9.840332062685179e-06,
+      "loss": 0.3853,
+      "step": 1722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.882678991935357,
+      "learning_rate": 9.840076636442215e-06,
+      "loss": 0.4165,
+      "step": 1723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.273843033979599,
+      "learning_rate": 9.839821009375876e-06,
+      "loss": 0.3865,
+      "step": 1724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.138827529030236,
+      "learning_rate": 9.839565181496766e-06,
+      "loss": 0.4044,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5108712898976502,
+      "learning_rate": 9.839309152815501e-06,
+      "loss": 0.4047,
+      "step": 1726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.570049755105157,
+      "learning_rate": 9.839052923342704e-06,
+      "loss": 0.3974,
+      "step": 1727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2654150722194646,
+      "learning_rate": 9.838796493089004e-06,
+      "loss": 0.4003,
+      "step": 1728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.524340356891728,
+      "learning_rate": 9.838539862065047e-06,
+      "loss": 0.3893,
+      "step": 1729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7232827143685543,
+      "learning_rate": 9.838283030281472e-06,
+      "loss": 0.4895,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.608871570310679,
+      "learning_rate": 9.838025997748943e-06,
+      "loss": 0.3789,
+      "step": 1731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.332670753315059,
+      "learning_rate": 9.837768764478121e-06,
+      "loss": 0.3698,
+      "step": 1732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.937897982725912,
+      "learning_rate": 9.837511330479683e-06,
+      "loss": 0.4052,
+      "step": 1733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0249010151356908,
+      "learning_rate": 9.837253695764304e-06,
+      "loss": 0.3733,
+      "step": 1734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.201770709371206,
+      "learning_rate": 9.83699586034268e-06,
+      "loss": 0.3763,
+      "step": 1735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6129306947845914,
+      "learning_rate": 9.836737824225504e-06,
+      "loss": 0.4478,
+      "step": 1736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.872094004617166,
+      "learning_rate": 9.836479587423487e-06,
+      "loss": 0.4344,
+      "step": 1737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1390342220198892,
+      "learning_rate": 9.836221149947339e-06,
+      "loss": 0.3946,
+      "step": 1738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.943947248325319,
+      "learning_rate": 9.835962511807786e-06,
+      "loss": 0.3732,
+      "step": 1739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.724561629835258,
+      "learning_rate": 9.835703673015559e-06,
+      "loss": 0.3871,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.4942667460140875,
+      "learning_rate": 9.835444633581398e-06,
+      "loss": 0.3933,
+      "step": 1741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.564816540790656,
+      "learning_rate": 9.83518539351605e-06,
+      "loss": 0.4036,
+      "step": 1742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6389411447725055,
+      "learning_rate": 9.834925952830272e-06,
+      "loss": 0.3921,
+      "step": 1743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.692429649224205,
+      "learning_rate": 9.834666311534828e-06,
+      "loss": 0.3958,
+      "step": 1744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.749031893229669,
+      "learning_rate": 9.834406469640492e-06,
+      "loss": 0.3612,
+      "step": 1745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7740671276584536,
+      "learning_rate": 9.834146427158043e-06,
+      "loss": 0.4254,
+      "step": 1746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7175704351884112,
+      "learning_rate": 9.833886184098273e-06,
+      "loss": 0.4693,
+      "step": 1747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7262648669789762,
+      "learning_rate": 9.833625740471981e-06,
+      "loss": 0.3786,
+      "step": 1748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.563889080426523,
+      "learning_rate": 9.833365096289971e-06,
+      "loss": 0.3518,
+      "step": 1749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.300867437494952,
+      "learning_rate": 9.833104251563058e-06,
+      "loss": 0.411,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.5285836738716725,
+      "learning_rate": 9.832843206302063e-06,
+      "loss": 0.373,
+      "step": 1751
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.806762271498663,
+      "learning_rate": 9.832581960517821e-06,
+      "loss": 0.3953,
+      "step": 1752
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.281968199466739,
+      "learning_rate": 9.83232051422117e-06,
+      "loss": 0.359,
+      "step": 1753
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.617781802791413,
+      "learning_rate": 9.832058867422959e-06,
+      "loss": 0.3946,
+      "step": 1754
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.94586607218607,
+      "learning_rate": 9.831797020134039e-06,
+      "loss": 0.4012,
+      "step": 1755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3379790367971496,
+      "learning_rate": 9.831534972365282e-06,
+      "loss": 0.3921,
+      "step": 1756
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5255728322019193,
+      "learning_rate": 9.831272724127555e-06,
+      "loss": 0.3752,
+      "step": 1757
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2969952038498853,
+      "learning_rate": 9.831010275431743e-06,
+      "loss": 0.3881,
+      "step": 1758
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9045826114081688,
+      "learning_rate": 9.830747626288732e-06,
+      "loss": 0.4727,
+      "step": 1759
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9483183425740722,
+      "learning_rate": 9.830484776709424e-06,
+      "loss": 0.3546,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.637340387906478,
+      "learning_rate": 9.830221726704721e-06,
+      "loss": 0.3681,
+      "step": 1761
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1048434018018094,
+      "learning_rate": 9.82995847628554e-06,
+      "loss": 0.3739,
+      "step": 1762
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3747090243167563,
+      "learning_rate": 9.829695025462803e-06,
+      "loss": 0.3902,
+      "step": 1763
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6126801855514128,
+      "learning_rate": 9.82943137424744e-06,
+      "loss": 0.3728,
+      "step": 1764
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7642456284124133,
+      "learning_rate": 9.82916752265039e-06,
+      "loss": 0.3927,
+      "step": 1765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.211968680808807,
+      "learning_rate": 9.828903470682604e-06,
+      "loss": 0.3728,
+      "step": 1766
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.0141784081619045,
+      "learning_rate": 9.828639218355036e-06,
+      "loss": 0.3732,
+      "step": 1767
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.972918918329857,
+      "learning_rate": 9.82837476567865e-06,
+      "loss": 0.3737,
+      "step": 1768
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4618233814111172,
+      "learning_rate": 9.828110112664417e-06,
+      "loss": 0.3696,
+      "step": 1769
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7775168603784308,
+      "learning_rate": 9.82784525932332e-06,
+      "loss": 0.3962,
+      "step": 1770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6929858049923627,
+      "learning_rate": 9.82758020566635e-06,
+      "loss": 0.3979,
+      "step": 1771
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6325018358882113,
+      "learning_rate": 9.827314951704501e-06,
+      "loss": 0.3647,
+      "step": 1772
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.457398218064355,
+      "learning_rate": 9.827049497448782e-06,
+      "loss": 0.3767,
+      "step": 1773
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.421800879401619,
+      "learning_rate": 9.826783842910203e-06,
+      "loss": 0.3973,
+      "step": 1774
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3614445207751285,
+      "learning_rate": 9.826517988099793e-06,
+      "loss": 0.3866,
+      "step": 1775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1527589195533974,
+      "learning_rate": 9.826251933028574e-06,
+      "loss": 0.3845,
+      "step": 1776
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9785993283026837,
+      "learning_rate": 9.825985677707593e-06,
+      "loss": 0.3833,
+      "step": 1777
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5470676531409917,
+      "learning_rate": 9.825719222147894e-06,
+      "loss": 0.3563,
+      "step": 1778
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.671811416386511,
+      "learning_rate": 9.825452566360533e-06,
+      "loss": 0.5295,
+      "step": 1779
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9049135842674518,
+      "learning_rate": 9.825185710356573e-06,
+      "loss": 0.3834,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9408021248551481,
+      "learning_rate": 9.824918654147088e-06,
+      "loss": 0.3884,
+      "step": 1781
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.364538884082852,
+      "learning_rate": 9.824651397743159e-06,
+      "loss": 0.3815,
+      "step": 1782
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.100834262112139,
+      "learning_rate": 9.824383941155872e-06,
+      "loss": 0.3906,
+      "step": 1783
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.679919936795614,
+      "learning_rate": 9.824116284396328e-06,
+      "loss": 0.3599,
+      "step": 1784
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.890702532650259,
+      "learning_rate": 9.82384842747563e-06,
+      "loss": 0.4019,
+      "step": 1785
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3214655606383343,
+      "learning_rate": 9.823580370404893e-06,
+      "loss": 0.4022,
+      "step": 1786
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.200897581694426,
+      "learning_rate": 9.823312113195238e-06,
+      "loss": 0.3882,
+      "step": 1787
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.072186629286691,
+      "learning_rate": 9.823043655857796e-06,
+      "loss": 0.3611,
+      "step": 1788
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6297660956814268,
+      "learning_rate": 9.822774998403707e-06,
+      "loss": 0.3635,
+      "step": 1789
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.292905578531731,
+      "learning_rate": 9.822506140844118e-06,
+      "loss": 0.3811,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7087918382651415,
+      "learning_rate": 9.82223708319018e-06,
+      "loss": 0.37,
+      "step": 1791
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.146685421380077,
+      "learning_rate": 9.821967825453063e-06,
+      "loss": 0.38,
+      "step": 1792
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2668390265253864,
+      "learning_rate": 9.821698367643936e-06,
+      "loss": 0.3637,
+      "step": 1793
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5321677385444092,
+      "learning_rate": 9.82142870977398e-06,
+      "loss": 0.3579,
+      "step": 1794
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4785359192400516,
+      "learning_rate": 9.821158851854381e-06,
+      "loss": 0.3679,
+      "step": 1795
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7017927566259976,
+      "learning_rate": 9.82088879389634e-06,
+      "loss": 0.3792,
+      "step": 1796
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.231011743947335,
+      "learning_rate": 9.820618535911057e-06,
+      "loss": 0.3807,
+      "step": 1797
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.535052581022571,
+      "learning_rate": 9.820348077909751e-06,
+      "loss": 0.3752,
+      "step": 1798
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2642552869800614,
+      "learning_rate": 9.820077419903642e-06,
+      "loss": 0.3643,
+      "step": 1799
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.856920118871183,
+      "learning_rate": 9.819806561903958e-06,
+      "loss": 0.3782,
+      "step": 1800
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4157727619762503,
+      "learning_rate": 9.819535503921939e-06,
+      "loss": 0.3795,
+      "step": 1801
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6028140208829351,
+      "learning_rate": 9.819264245968831e-06,
+      "loss": 0.36,
+      "step": 1802
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1175706556514085,
+      "learning_rate": 9.81899278805589e-06,
+      "loss": 0.379,
+      "step": 1803
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7873196904004678,
+      "learning_rate": 9.81872113019438e-06,
+      "loss": 0.3611,
+      "step": 1804
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.47258866737647,
+      "learning_rate": 9.818449272395569e-06,
+      "loss": 0.4067,
+      "step": 1805
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6120758077434205,
+      "learning_rate": 9.818177214670742e-06,
+      "loss": 0.3958,
+      "step": 1806
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1997375866342472,
+      "learning_rate": 9.81790495703118e-06,
+      "loss": 0.3687,
+      "step": 1807
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8691936911868914,
+      "learning_rate": 9.817632499488188e-06,
+      "loss": 0.3963,
+      "step": 1808
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.93149599238276,
+      "learning_rate": 9.817359842053064e-06,
+      "loss": 0.3514,
+      "step": 1809
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.8647426600058616,
+      "learning_rate": 9.817086984737126e-06,
+      "loss": 0.378,
+      "step": 1810
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7732346592905521,
+      "learning_rate": 9.81681392755169e-06,
+      "loss": 0.3994,
+      "step": 1811
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7418135546211012,
+      "learning_rate": 9.81654067050809e-06,
+      "loss": 0.491,
+      "step": 1812
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9501015746712647,
+      "learning_rate": 9.816267213617663e-06,
+      "loss": 0.3689,
+      "step": 1813
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1195063900736173,
+      "learning_rate": 9.815993556891753e-06,
+      "loss": 0.3815,
+      "step": 1814
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2836222840638096,
+      "learning_rate": 9.815719700341717e-06,
+      "loss": 0.3979,
+      "step": 1815
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.610703393567781,
+      "learning_rate": 9.815445643978918e-06,
+      "loss": 0.3927,
+      "step": 1816
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.370038951133397,
+      "learning_rate": 9.815171387814726e-06,
+      "loss": 0.411,
+      "step": 1817
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3157353744370717,
+      "learning_rate": 9.814896931860518e-06,
+      "loss": 0.3883,
+      "step": 1818
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9855552743737201,
+      "learning_rate": 9.814622276127685e-06,
+      "loss": 0.3855,
+      "step": 1819
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4350674817463283,
+      "learning_rate": 9.814347420627624e-06,
+      "loss": 0.3797,
+      "step": 1820
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1328802884017537,
+      "learning_rate": 9.814072365371736e-06,
+      "loss": 0.4107,
+      "step": 1821
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8727191776776886,
+      "learning_rate": 9.813797110371435e-06,
+      "loss": 0.4059,
+      "step": 1822
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4758822076628575,
+      "learning_rate": 9.81352165563814e-06,
+      "loss": 0.3964,
+      "step": 1823
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7728905932904224,
+      "learning_rate": 9.813246001183283e-06,
+      "loss": 0.3744,
+      "step": 1824
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3548693281117927,
+      "learning_rate": 9.812970147018301e-06,
+      "loss": 0.387,
+      "step": 1825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7098037795919612,
+      "learning_rate": 9.812694093154637e-06,
+      "loss": 0.4919,
+      "step": 1826
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8816550590224619,
+      "learning_rate": 9.812417839603748e-06,
+      "loss": 0.3631,
+      "step": 1827
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3025814611521094,
+      "learning_rate": 9.812141386377095e-06,
+      "loss": 0.3602,
+      "step": 1828
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8788408297975687,
+      "learning_rate": 9.811864733486148e-06,
+      "loss": 0.3771,
+      "step": 1829
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3724140881944669,
+      "learning_rate": 9.811587880942387e-06,
+      "loss": 0.3961,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4392367378303383,
+      "learning_rate": 9.8113108287573e-06,
+      "loss": 0.3919,
+      "step": 1831
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6599125668086123,
+      "learning_rate": 9.811033576942377e-06,
+      "loss": 0.3523,
+      "step": 1832
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.176707774711695,
+      "learning_rate": 9.81075612550913e-06,
+      "loss": 0.3869,
+      "step": 1833
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0351763040173565,
+      "learning_rate": 9.810478474469063e-06,
+      "loss": 0.3822,
+      "step": 1834
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6228436799116237,
+      "learning_rate": 9.8102006238337e-06,
+      "loss": 0.3755,
+      "step": 1835
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5019206471209272,
+      "learning_rate": 9.80992257361457e-06,
+      "loss": 0.3835,
+      "step": 1836
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0701113391645256,
+      "learning_rate": 9.809644323823208e-06,
+      "loss": 0.3685,
+      "step": 1837
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8122415268301864,
+      "learning_rate": 9.809365874471162e-06,
+      "loss": 0.3677,
+      "step": 1838
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.392152207416042,
+      "learning_rate": 9.809087225569982e-06,
+      "loss": 0.3762,
+      "step": 1839
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.973822000004366,
+      "learning_rate": 9.808808377131232e-06,
+      "loss": 0.3985,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1141198723828953,
+      "learning_rate": 9.80852932916648e-06,
+      "loss": 0.3799,
+      "step": 1841
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.536958406072693,
+      "learning_rate": 9.808250081687307e-06,
+      "loss": 0.3838,
+      "step": 1842
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6693594269266874,
+      "learning_rate": 9.807970634705297e-06,
+      "loss": 0.3677,
+      "step": 1843
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.879146621583926,
+      "learning_rate": 9.807690988232046e-06,
+      "loss": 0.3748,
+      "step": 1844
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7360582201283752,
+      "learning_rate": 9.807411142279155e-06,
+      "loss": 0.3736,
+      "step": 1845
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7292268293298563,
+      "learning_rate": 9.807131096858237e-06,
+      "loss": 0.4865,
+      "step": 1846
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4239736621705856,
+      "learning_rate": 9.806850851980913e-06,
+      "loss": 0.4336,
+      "step": 1847
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6510268875876362,
+      "learning_rate": 9.806570407658807e-06,
+      "loss": 0.3613,
+      "step": 1848
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5781351338102192,
+      "learning_rate": 9.806289763903558e-06,
+      "loss": 0.4012,
+      "step": 1849
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6703472805886768,
+      "learning_rate": 9.80600892072681e-06,
+      "loss": 0.3647,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4835899672197248,
+      "learning_rate": 9.805727878140216e-06,
+      "loss": 0.3848,
+      "step": 1851
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3772710502274328,
+      "learning_rate": 9.805446636155435e-06,
+      "loss": 0.3511,
+      "step": 1852
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8725052006963179,
+      "learning_rate": 9.80516519478414e-06,
+      "loss": 0.3736,
+      "step": 1853
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.783161350617083,
+      "learning_rate": 9.804883554038005e-06,
+      "loss": 0.3717,
+      "step": 1854
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.443154753730953,
+      "learning_rate": 9.804601713928716e-06,
+      "loss": 0.3866,
+      "step": 1855
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4668533782426225,
+      "learning_rate": 9.804319674467969e-06,
+      "loss": 0.3919,
+      "step": 1856
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.026507821088116,
+      "learning_rate": 9.804037435667465e-06,
+      "loss": 0.406,
+      "step": 1857
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9596348999579245,
+      "learning_rate": 9.803754997538915e-06,
+      "loss": 0.3937,
+      "step": 1858
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.821964444594066,
+      "learning_rate": 9.803472360094037e-06,
+      "loss": 0.5198,
+      "step": 1859
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.844427766970074,
+      "learning_rate": 9.80318952334456e-06,
+      "loss": 0.3886,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5945862940221964,
+      "learning_rate": 9.802906487302217e-06,
+      "loss": 0.4021,
+      "step": 1861
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7011221792938005,
+      "learning_rate": 9.802623251978754e-06,
+      "loss": 0.4896,
+      "step": 1862
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.18323762157779,
+      "learning_rate": 9.802339817385921e-06,
+      "loss": 0.3756,
+      "step": 1863
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2646470276799664,
+      "learning_rate": 9.80205618353548e-06,
+      "loss": 0.3962,
+      "step": 1864
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8768123406953179,
+      "learning_rate": 9.801772350439197e-06,
+      "loss": 0.3497,
+      "step": 1865
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8224164058034686,
+      "learning_rate": 9.80148831810885e-06,
+      "loss": 0.3835,
+      "step": 1866
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4532793718077375,
+      "learning_rate": 9.801204086556226e-06,
+      "loss": 0.3818,
+      "step": 1867
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.242903254828831,
+      "learning_rate": 9.800919655793118e-06,
+      "loss": 0.4071,
+      "step": 1868
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7055773310523081,
+      "learning_rate": 9.800635025831323e-06,
+      "loss": 0.3604,
+      "step": 1869
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.368379984554867,
+      "learning_rate": 9.800350196682655e-06,
+      "loss": 0.3688,
+      "step": 1870
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8032910357834164,
+      "learning_rate": 9.800065168358932e-06,
+      "loss": 0.5017,
+      "step": 1871
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.270233253613688,
+      "learning_rate": 9.799779940871978e-06,
+      "loss": 0.3551,
+      "step": 1872
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.424770210701576,
+      "learning_rate": 9.79949451423363e-06,
+      "loss": 0.4159,
+      "step": 1873
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5062834576744304,
+      "learning_rate": 9.799208888455728e-06,
+      "loss": 0.3857,
+      "step": 1874
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9901631549421632,
+      "learning_rate": 9.798923063550126e-06,
+      "loss": 0.394,
+      "step": 1875
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6528928084315886,
+      "learning_rate": 9.798637039528682e-06,
+      "loss": 0.3741,
+      "step": 1876
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.851195106538528,
+      "learning_rate": 9.798350816403264e-06,
+      "loss": 0.3903,
+      "step": 1877
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6243796938564128,
+      "learning_rate": 9.798064394185747e-06,
+      "loss": 0.3806,
+      "step": 1878
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0692094744188174,
+      "learning_rate": 9.797777772888018e-06,
+      "loss": 0.375,
+      "step": 1879
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9722223855442804,
+      "learning_rate": 9.797490952521965e-06,
+      "loss": 0.404,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.323189381334899,
+      "learning_rate": 9.797203933099492e-06,
+      "loss": 0.4047,
+      "step": 1881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8726616653905463,
+      "learning_rate": 9.796916714632507e-06,
+      "loss": 0.4161,
+      "step": 1882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4112081624351538,
+      "learning_rate": 9.796629297132927e-06,
+      "loss": 0.3643,
+      "step": 1883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1861272231532065,
+      "learning_rate": 9.79634168061268e-06,
+      "loss": 0.3979,
+      "step": 1884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.003887017501671,
+      "learning_rate": 9.796053865083694e-06,
+      "loss": 0.3589,
+      "step": 1885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6608401463733822,
+      "learning_rate": 9.795765850557917e-06,
+      "loss": 0.3634,
+      "step": 1886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.719302499213828,
+      "learning_rate": 9.795477637047295e-06,
+      "loss": 0.3713,
+      "step": 1887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5173584641802416,
+      "learning_rate": 9.795189224563788e-06,
+      "loss": 0.3754,
+      "step": 1888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.5218269577982015,
+      "learning_rate": 9.794900613119364e-06,
+      "loss": 0.3594,
+      "step": 1889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8582866394907955,
+      "learning_rate": 9.794611802725997e-06,
+      "loss": 0.3571,
+      "step": 1890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.072569259131529,
+      "learning_rate": 9.794322793395669e-06,
+      "loss": 0.367,
+      "step": 1891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.812020311598828,
+      "learning_rate": 9.794033585140372e-06,
+      "loss": 0.524,
+      "step": 1892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3552908356253797,
+      "learning_rate": 9.793744177972108e-06,
+      "loss": 0.3782,
+      "step": 1893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.422310012617706,
+      "learning_rate": 9.793454571902883e-06,
+      "loss": 0.3586,
+      "step": 1894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1116590567977065,
+      "learning_rate": 9.793164766944714e-06,
+      "loss": 0.4087,
+      "step": 1895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5817790936508511,
+      "learning_rate": 9.792874763109624e-06,
+      "loss": 0.401,
+      "step": 1896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.317936493490959,
+      "learning_rate": 9.792584560409649e-06,
+      "loss": 0.3802,
+      "step": 1897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6673907285813097,
+      "learning_rate": 9.792294158856826e-06,
+      "loss": 0.5049,
+      "step": 1898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5708418207204,
+      "learning_rate": 9.792003558463207e-06,
+      "loss": 0.3849,
+      "step": 1899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8808633162336108,
+      "learning_rate": 9.79171275924085e-06,
+      "loss": 0.3635,
+      "step": 1900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1897425596392917,
+      "learning_rate": 9.791421761201816e-06,
+      "loss": 0.3867,
+      "step": 1901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4173800838434034,
+      "learning_rate": 9.791130564358187e-06,
+      "loss": 0.3594,
+      "step": 1902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4725636900039665,
+      "learning_rate": 9.790839168722037e-06,
+      "loss": 0.3565,
+      "step": 1903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1064058612467136,
+      "learning_rate": 9.790547574305463e-06,
+      "loss": 0.3769,
+      "step": 1904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8853211503689868,
+      "learning_rate": 9.79025578112056e-06,
+      "loss": 0.3813,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.227915359719842,
+      "learning_rate": 9.789963789179438e-06,
+      "loss": 0.3948,
+      "step": 1906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.595272909569428,
+      "learning_rate": 9.789671598494208e-06,
+      "loss": 0.372,
+      "step": 1907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7529055558688227,
+      "learning_rate": 9.789379209076997e-06,
+      "loss": 0.4011,
+      "step": 1908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2201098265594053,
+      "learning_rate": 9.789086620939936e-06,
+      "loss": 0.4142,
+      "step": 1909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6230325565626995,
+      "learning_rate": 9.788793834095165e-06,
+      "loss": 0.3868,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8220788291104209,
+      "learning_rate": 9.788500848554831e-06,
+      "loss": 0.4005,
+      "step": 1911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.86975313357887,
+      "learning_rate": 9.788207664331093e-06,
+      "loss": 0.3862,
+      "step": 1912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5087622500622635,
+      "learning_rate": 9.787914281436112e-06,
+      "loss": 0.4086,
+      "step": 1913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7034975329118294,
+      "learning_rate": 9.787620699882064e-06,
+      "loss": 0.3916,
+      "step": 1914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8567750650849524,
+      "learning_rate": 9.78732691968113e-06,
+      "loss": 0.3843,
+      "step": 1915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9640602234460758,
+      "learning_rate": 9.787032940845499e-06,
+      "loss": 0.3707,
+      "step": 1916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.655332001580452,
+      "learning_rate": 9.786738763387368e-06,
+      "loss": 0.3766,
+      "step": 1917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.033882687241404,
+      "learning_rate": 9.786444387318943e-06,
+      "loss": 0.3894,
+      "step": 1918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.074227083942343,
+      "learning_rate": 9.78614981265244e-06,
+      "loss": 0.3678,
+      "step": 1919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8630051286114978,
+      "learning_rate": 9.785855039400079e-06,
+      "loss": 0.5252,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.519566927333756,
+      "learning_rate": 9.785560067574092e-06,
+      "loss": 0.3803,
+      "step": 1921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6710749180215918,
+      "learning_rate": 9.785264897186718e-06,
+      "loss": 0.38,
+      "step": 1922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0016881036651193,
+      "learning_rate": 9.784969528250204e-06,
+      "loss": 0.3609,
+      "step": 1923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9523850550089148,
+      "learning_rate": 9.784673960776805e-06,
+      "loss": 0.39,
+      "step": 1924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.696981077145076,
+      "learning_rate": 9.784378194778786e-06,
+      "loss": 0.3593,
+      "step": 1925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.303193951213647,
+      "learning_rate": 9.784082230268415e-06,
+      "loss": 0.3688,
+      "step": 1926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.83381667490055,
+      "learning_rate": 9.783786067257976e-06,
+      "loss": 0.3837,
+      "step": 1927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7024276635852529,
+      "learning_rate": 9.783489705759756e-06,
+      "loss": 0.3759,
+      "step": 1928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0647040389760325,
+      "learning_rate": 9.783193145786053e-06,
+      "loss": 0.3857,
+      "step": 1929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.017366550215066,
+      "learning_rate": 9.78289638734917e-06,
+      "loss": 0.3885,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2335354084717443,
+      "learning_rate": 9.78259943046142e-06,
+      "loss": 0.358,
+      "step": 1931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7159171008819794,
+      "learning_rate": 9.782302275135124e-06,
+      "loss": 0.4002,
+      "step": 1932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.280026934917581,
+      "learning_rate": 9.782004921382612e-06,
+      "loss": 0.3735,
+      "step": 1933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1313823366662983,
+      "learning_rate": 9.781707369216224e-06,
+      "loss": 0.3797,
+      "step": 1934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 21.506648057275566,
+      "learning_rate": 9.781409618648303e-06,
+      "loss": 0.3678,
+      "step": 1935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6622627212392396,
+      "learning_rate": 9.781111669691203e-06,
+      "loss": 0.4027,
+      "step": 1936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1157157261458726,
+      "learning_rate": 9.78081352235729e-06,
+      "loss": 0.386,
+      "step": 1937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.771600485678825,
+      "learning_rate": 9.780515176658931e-06,
+      "loss": 0.3738,
+      "step": 1938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0231255387124607,
+      "learning_rate": 9.780216632608505e-06,
+      "loss": 0.3615,
+      "step": 1939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5574669321774992,
+      "learning_rate": 9.779917890218403e-06,
+      "loss": 0.3783,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.11765915699987,
+      "learning_rate": 9.779618949501016e-06,
+      "loss": 0.3744,
+      "step": 1941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.200780852177338,
+      "learning_rate": 9.77931981046875e-06,
+      "loss": 0.3644,
+      "step": 1942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2477496902459144,
+      "learning_rate": 9.779020473134016e-06,
+      "loss": 0.379,
+      "step": 1943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.546558164427262,
+      "learning_rate": 9.778720937509232e-06,
+      "loss": 0.3738,
+      "step": 1944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.606784032500254,
+      "learning_rate": 9.77842120360683e-06,
+      "loss": 0.3791,
+      "step": 1945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.544110328603507,
+      "learning_rate": 9.778121271439244e-06,
+      "loss": 0.4039,
+      "step": 1946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8406293720086553,
+      "learning_rate": 9.777821141018921e-06,
+      "loss": 0.3507,
+      "step": 1947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.305839609769871,
+      "learning_rate": 9.777520812358312e-06,
+      "loss": 0.3889,
+      "step": 1948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.029526921175561,
+      "learning_rate": 9.77722028546988e-06,
+      "loss": 0.3825,
+      "step": 1949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9220591347031635,
+      "learning_rate": 9.776919560366091e-06,
+      "loss": 0.3596,
+      "step": 1950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4369895353485262,
+      "learning_rate": 9.776618637059426e-06,
+      "loss": 0.3793,
+      "step": 1951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1288381073443885,
+      "learning_rate": 9.776317515562368e-06,
+      "loss": 0.3773,
+      "step": 1952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.907481904334835,
+      "learning_rate": 9.776016195887412e-06,
+      "loss": 0.51,
+      "step": 1953
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.881392823035069,
+      "learning_rate": 9.775714678047062e-06,
+      "loss": 0.3707,
+      "step": 1954
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9875774410229394,
+      "learning_rate": 9.775412962053827e-06,
+      "loss": 0.3848,
+      "step": 1955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6703353259398892,
+      "learning_rate": 9.775111047920227e-06,
+      "loss": 0.3707,
+      "step": 1956
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3450663523316644,
+      "learning_rate": 9.774808935658789e-06,
+      "loss": 0.3824,
+      "step": 1957
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2576297293049863,
+      "learning_rate": 9.774506625282045e-06,
+      "loss": 0.3924,
+      "step": 1958
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1682736851014184,
+      "learning_rate": 9.774204116802541e-06,
+      "loss": 0.3861,
+      "step": 1959
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.176652831134348,
+      "learning_rate": 9.77390141023283e-06,
+      "loss": 0.3936,
+      "step": 1960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.462140381311946,
+      "learning_rate": 9.773598505585469e-06,
+      "loss": 0.3875,
+      "step": 1961
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.948576032937403,
+      "learning_rate": 9.773295402873027e-06,
+      "loss": 0.3846,
+      "step": 1962
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.40336693857439,
+      "learning_rate": 9.772992102108081e-06,
+      "loss": 0.372,
+      "step": 1963
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4358865188548455,
+      "learning_rate": 9.772688603303212e-06,
+      "loss": 0.3784,
+      "step": 1964
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1518668362808788,
+      "learning_rate": 9.772384906471019e-06,
+      "loss": 0.3786,
+      "step": 1965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.74123761591902,
+      "learning_rate": 9.7720810116241e-06,
+      "loss": 0.3787,
+      "step": 1966
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0715609386547476,
+      "learning_rate": 9.771776918775062e-06,
+      "loss": 0.3665,
+      "step": 1967
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0775998085029475,
+      "learning_rate": 9.771472627936523e-06,
+      "loss": 0.3756,
+      "step": 1968
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7415421794158161,
+      "learning_rate": 9.771168139121112e-06,
+      "loss": 0.3795,
+      "step": 1969
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.561194303588185,
+      "learning_rate": 9.77086345234146e-06,
+      "loss": 0.3796,
+      "step": 1970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8302956151937808,
+      "learning_rate": 9.770558567610206e-06,
+      "loss": 0.3749,
+      "step": 1971
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.810312018015889,
+      "learning_rate": 9.770253484940006e-06,
+      "loss": 0.3764,
+      "step": 1972
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5425732988747862,
+      "learning_rate": 9.769948204343516e-06,
+      "loss": 0.3728,
+      "step": 1973
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.647153725909506,
+      "learning_rate": 9.769642725833404e-06,
+      "loss": 0.383,
+      "step": 1974
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.149574646060756,
+      "learning_rate": 9.769337049422342e-06,
+      "loss": 0.3739,
+      "step": 1975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0619356098848436,
+      "learning_rate": 9.769031175123014e-06,
+      "loss": 0.3921,
+      "step": 1976
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.1116561985412945,
+      "learning_rate": 9.768725102948114e-06,
+      "loss": 0.3822,
+      "step": 1977
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8387468500880593,
+      "learning_rate": 9.768418832910335e-06,
+      "loss": 0.3498,
+      "step": 1978
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.466877314186575,
+      "learning_rate": 9.768112365022393e-06,
+      "loss": 0.3734,
+      "step": 1979
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6249384200281445,
+      "learning_rate": 9.767805699297e-06,
+      "loss": 0.3763,
+      "step": 1980
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.317160429032891,
+      "learning_rate": 9.767498835746877e-06,
+      "loss": 0.3709,
+      "step": 1981
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.546149542105723,
+      "learning_rate": 9.767191774384762e-06,
+      "loss": 0.3775,
+      "step": 1982
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.392984475719745,
+      "learning_rate": 9.76688451522339e-06,
+      "loss": 0.3851,
+      "step": 1983
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3679476559327224,
+      "learning_rate": 9.766577058275515e-06,
+      "loss": 0.3711,
+      "step": 1984
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4697593364660515,
+      "learning_rate": 9.76626940355389e-06,
+      "loss": 0.3875,
+      "step": 1985
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6129796454028429,
+      "learning_rate": 9.765961551071281e-06,
+      "loss": 0.3509,
+      "step": 1986
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4820353944038447,
+      "learning_rate": 9.765653500840465e-06,
+      "loss": 0.3905,
+      "step": 1987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7673014753661827,
+      "learning_rate": 9.765345252874218e-06,
+      "loss": 0.3686,
+      "step": 1988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.89545528951867,
+      "learning_rate": 9.765036807185333e-06,
+      "loss": 0.3747,
+      "step": 1989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8510555016509191,
+      "learning_rate": 9.764728163786607e-06,
+      "loss": 0.3469,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5776007745789054,
+      "learning_rate": 9.764419322690846e-06,
+      "loss": 0.3916,
+      "step": 1991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6629761168978403,
+      "learning_rate": 9.764110283910864e-06,
+      "loss": 0.3613,
+      "step": 1992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8771437108275397,
+      "learning_rate": 9.763801047459487e-06,
+      "loss": 0.3824,
+      "step": 1993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8611967018793132,
+      "learning_rate": 9.76349161334954e-06,
+      "loss": 0.383,
+      "step": 1994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5387062747787261,
+      "learning_rate": 9.763181981593868e-06,
+      "loss": 0.3686,
+      "step": 1995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5395254340364815,
+      "learning_rate": 9.762872152205313e-06,
+      "loss": 0.3639,
+      "step": 1996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.288502229213549,
+      "learning_rate": 9.762562125196734e-06,
+      "loss": 0.3671,
+      "step": 1997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.357514527059941,
+      "learning_rate": 9.762251900580992e-06,
+      "loss": 0.3594,
+      "step": 1998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7716245546139546,
+      "learning_rate": 9.76194147837096e-06,
+      "loss": 0.3733,
+      "step": 1999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.893383549361695,
+      "learning_rate": 9.76163085857952e-06,
+      "loss": 0.3726,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8859889597204025,
+      "learning_rate": 9.761320041219555e-06,
+      "loss": 0.4947,
+      "step": 2001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.221491499199223,
+      "learning_rate": 9.761009026303968e-06,
+      "loss": 0.3807,
+      "step": 2002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.693554606557953,
+      "learning_rate": 9.760697813845656e-06,
+      "loss": 0.5143,
+      "step": 2003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7756681160884638,
+      "learning_rate": 9.760386403857538e-06,
+      "loss": 0.3808,
+      "step": 2004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.765243284947136,
+      "learning_rate": 9.760074796352532e-06,
+      "loss": 0.3849,
+      "step": 2005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9655707569516148,
+      "learning_rate": 9.759762991343569e-06,
+      "loss": 0.386,
+      "step": 2006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.769282624130448,
+      "learning_rate": 9.759450988843585e-06,
+      "loss": 0.3801,
+      "step": 2007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.712213566302072,
+      "learning_rate": 9.759138788865524e-06,
+      "loss": 0.3818,
+      "step": 2008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.82289568923332,
+      "learning_rate": 9.758826391422343e-06,
+      "loss": 0.3733,
+      "step": 2009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.032705911899892,
+      "learning_rate": 9.758513796527002e-06,
+      "loss": 0.3661,
+      "step": 2010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6822890353301314,
+      "learning_rate": 9.758201004192471e-06,
+      "loss": 0.3604,
+      "step": 2011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.190475743965018,
+      "learning_rate": 9.75788801443173e-06,
+      "loss": 0.3906,
+      "step": 2012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.261698410710128,
+      "learning_rate": 9.757574827257764e-06,
+      "loss": 0.3629,
+      "step": 2013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0359058427099048,
+      "learning_rate": 9.757261442683568e-06,
+      "loss": 0.5211,
+      "step": 2014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6760454853543507,
+      "learning_rate": 9.756947860722143e-06,
+      "loss": 0.3784,
+      "step": 2015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8343607510351445,
+      "learning_rate": 9.756634081386504e-06,
+      "loss": 0.3767,
+      "step": 2016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.226232322669994,
+      "learning_rate": 9.756320104689667e-06,
+      "loss": 0.3815,
+      "step": 2017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.906365266928955,
+      "learning_rate": 9.756005930644662e-06,
+      "loss": 0.3899,
+      "step": 2018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.127075075399167,
+      "learning_rate": 9.755691559264522e-06,
+      "loss": 0.4045,
+      "step": 2019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6405042644879044,
+      "learning_rate": 9.755376990562295e-06,
+      "loss": 0.3806,
+      "step": 2020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.314911126990063,
+      "learning_rate": 9.755062224551026e-06,
+      "loss": 0.4185,
+      "step": 2021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9300617164925287,
+      "learning_rate": 9.754747261243782e-06,
+      "loss": 0.389,
+      "step": 2022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4834698528387698,
+      "learning_rate": 9.754432100653628e-06,
+      "loss": 0.3853,
+      "step": 2023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.236187734459461,
+      "learning_rate": 9.754116742793643e-06,
+      "loss": 0.3711,
+      "step": 2024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.210971605398184,
+      "learning_rate": 9.753801187676908e-06,
+      "loss": 0.3749,
+      "step": 2025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9677015445758617,
+      "learning_rate": 9.753485435316518e-06,
+      "loss": 0.349,
+      "step": 2026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8476003512204442,
+      "learning_rate": 9.753169485725575e-06,
+      "loss": 0.3663,
+      "step": 2027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.368513933261431,
+      "learning_rate": 9.752853338917187e-06,
+      "loss": 0.3689,
+      "step": 2028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0499042881887637,
+      "learning_rate": 9.752536994904473e-06,
+      "loss": 0.3622,
+      "step": 2029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0956243531214254,
+      "learning_rate": 9.752220453700556e-06,
+      "loss": 0.351,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.314003420331242,
+      "learning_rate": 9.751903715318572e-06,
+      "loss": 0.3467,
+      "step": 2031
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.785946372863839,
+      "learning_rate": 9.751586779771663e-06,
+      "loss": 0.3788,
+      "step": 2032
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4905383618900525,
+      "learning_rate": 9.751269647072978e-06,
+      "loss": 0.3885,
+      "step": 2033
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6776188208779383,
+      "learning_rate": 9.750952317235678e-06,
+      "loss": 0.389,
+      "step": 2034
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.164191962494118,
+      "learning_rate": 9.750634790272926e-06,
+      "loss": 0.3705,
+      "step": 2035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.562466612448937,
+      "learning_rate": 9.750317066197899e-06,
+      "loss": 0.3635,
+      "step": 2036
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6978323981070979,
+      "learning_rate": 9.74999914502378e-06,
+      "loss": 0.4722,
+      "step": 2037
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7676229652145903,
+      "learning_rate": 9.749681026763758e-06,
+      "loss": 0.3735,
+      "step": 2038
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7321059484900543,
+      "learning_rate": 9.749362711431034e-06,
+      "loss": 0.3826,
+      "step": 2039
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5780720677869677,
+      "learning_rate": 9.749044199038817e-06,
+      "loss": 0.3683,
+      "step": 2040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7770761131145016,
+      "learning_rate": 9.74872548960032e-06,
+      "loss": 0.3505,
+      "step": 2041
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9147841645237889,
+      "learning_rate": 9.748406583128766e-06,
+      "loss": 0.3468,
+      "step": 2042
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0539811371487953,
+      "learning_rate": 9.748087479637392e-06,
+      "loss": 0.3869,
+      "step": 2043
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1851613542010764,
+      "learning_rate": 9.747768179139433e-06,
+      "loss": 0.3993,
+      "step": 2044
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5271231616726504,
+      "learning_rate": 9.74744868164814e-06,
+      "loss": 0.3727,
+      "step": 2045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7743759366833747,
+      "learning_rate": 9.747128987176768e-06,
+      "loss": 0.3638,
+      "step": 2046
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9534302961238634,
+      "learning_rate": 9.746809095738581e-06,
+      "loss": 0.3793,
+      "step": 2047
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2704836228978835,
+      "learning_rate": 9.746489007346856e-06,
+      "loss": 0.3682,
+      "step": 2048
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0785440079876905,
+      "learning_rate": 9.746168722014871e-06,
+      "loss": 0.3638,
+      "step": 2049
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.380900743695415,
+      "learning_rate": 9.745848239755915e-06,
+      "loss": 0.3559,
+      "step": 2050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1384483816761786,
+      "learning_rate": 9.745527560583285e-06,
+      "loss": 0.3768,
+      "step": 2051
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.75043907337137,
+      "learning_rate": 9.745206684510289e-06,
+      "loss": 0.3926,
+      "step": 2052
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.179888990727422,
+      "learning_rate": 9.744885611550239e-06,
+      "loss": 0.3968,
+      "step": 2053
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9498528642622612,
+      "learning_rate": 9.744564341716455e-06,
+      "loss": 0.3305,
+      "step": 2054
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.414228701125571,
+      "learning_rate": 9.744242875022272e-06,
+      "loss": 0.3717,
+      "step": 2055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4024879159291146,
+      "learning_rate": 9.743921211481024e-06,
+      "loss": 0.3693,
+      "step": 2056
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.637135472120715,
+      "learning_rate": 9.74359935110606e-06,
+      "loss": 0.3602,
+      "step": 2057
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3519461879250865,
+      "learning_rate": 9.743277293910733e-06,
+      "loss": 0.3867,
+      "step": 2058
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7779436887641977,
+      "learning_rate": 9.742955039908406e-06,
+      "loss": 0.3596,
+      "step": 2059
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.943526674621215,
+      "learning_rate": 9.742632589112448e-06,
+      "loss": 0.3486,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.677983464515481,
+      "learning_rate": 9.742309941536243e-06,
+      "loss": 0.368,
+      "step": 2061
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1149395646695925,
+      "learning_rate": 9.741987097193174e-06,
+      "loss": 0.3807,
+      "step": 2062
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9735113730032214,
+      "learning_rate": 9.741664056096637e-06,
+      "loss": 0.3833,
+      "step": 2063
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7984818734853594,
+      "learning_rate": 9.741340818260036e-06,
+      "loss": 0.3558,
+      "step": 2064
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0996088945515368,
+      "learning_rate": 9.741017383696784e-06,
+      "loss": 0.3514,
+      "step": 2065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.512897907985271,
+      "learning_rate": 9.7406937524203e-06,
+      "loss": 0.3621,
+      "step": 2066
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.969422018816204,
+      "learning_rate": 9.740369924444011e-06,
+      "loss": 0.3863,
+      "step": 2067
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.242100778849942,
+      "learning_rate": 9.740045899781353e-06,
+      "loss": 0.3789,
+      "step": 2068
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.046530406817976,
+      "learning_rate": 9.739721678445772e-06,
+      "loss": 0.3726,
+      "step": 2069
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2390276927521047,
+      "learning_rate": 9.73939726045072e-06,
+      "loss": 0.3672,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.730884607148511,
+      "learning_rate": 9.739072645809659e-06,
+      "loss": 0.3649,
+      "step": 2071
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.137534335295253,
+      "learning_rate": 9.738747834536056e-06,
+      "loss": 0.3827,
+      "step": 2072
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1657666343437287,
+      "learning_rate": 9.738422826643385e-06,
+      "loss": 0.3657,
+      "step": 2073
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3417265608812627,
+      "learning_rate": 9.738097622145138e-06,
+      "loss": 0.3587,
+      "step": 2074
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3546133096058433,
+      "learning_rate": 9.737772221054805e-06,
+      "loss": 0.3822,
+      "step": 2075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.162846204439659,
+      "learning_rate": 9.737446623385885e-06,
+      "loss": 0.3858,
+      "step": 2076
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.91068662472612,
+      "learning_rate": 9.737120829151892e-06,
+      "loss": 0.369,
+      "step": 2077
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8461118959609997,
+      "learning_rate": 9.736794838366342e-06,
+      "loss": 0.3679,
+      "step": 2078
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5800370867292317,
+      "learning_rate": 9.736468651042759e-06,
+      "loss": 0.3669,
+      "step": 2079
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.39789416614131,
+      "learning_rate": 9.73614226719468e-06,
+      "loss": 0.362,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6228173046928687,
+      "learning_rate": 9.735815686835644e-06,
+      "loss": 0.3846,
+      "step": 2081
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.861306080305837,
+      "learning_rate": 9.735488909979205e-06,
+      "loss": 0.3842,
+      "step": 2082
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.112625158009115,
+      "learning_rate": 9.735161936638919e-06,
+      "loss": 0.3508,
+      "step": 2083
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.706042145528432,
+      "learning_rate": 9.734834766828355e-06,
+      "loss": 0.3668,
+      "step": 2084
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1558621862255363,
+      "learning_rate": 9.734507400561087e-06,
+      "loss": 0.3602,
+      "step": 2085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.1062051333568,
+      "learning_rate": 9.734179837850695e-06,
+      "loss": 0.3581,
+      "step": 2086
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6518722280133271,
+      "learning_rate": 9.733852078710774e-06,
+      "loss": 0.3534,
+      "step": 2087
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3927538505710038,
+      "learning_rate": 9.733524123154923e-06,
+      "loss": 0.3999,
+      "step": 2088
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.83621811927603,
+      "learning_rate": 9.733195971196747e-06,
+      "loss": 0.3594,
+      "step": 2089
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0494235345449296,
+      "learning_rate": 9.732867622849863e-06,
+      "loss": 0.3865,
+      "step": 2090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8664351300158333,
+      "learning_rate": 9.732539078127895e-06,
+      "loss": 0.3907,
+      "step": 2091
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.274292271643615,
+      "learning_rate": 9.732210337044475e-06,
+      "loss": 0.3833,
+      "step": 2092
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3152591139212695,
+      "learning_rate": 9.731881399613243e-06,
+      "loss": 0.3815,
+      "step": 2093
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6219381423340673,
+      "learning_rate": 9.731552265847847e-06,
+      "loss": 0.3649,
+      "step": 2094
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9066997475563694,
+      "learning_rate": 9.731222935761943e-06,
+      "loss": 0.3697,
+      "step": 2095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9432450165133641,
+      "learning_rate": 9.730893409369197e-06,
+      "loss": 0.3576,
+      "step": 2096
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.014531214349162,
+      "learning_rate": 9.730563686683278e-06,
+      "loss": 0.3605,
+      "step": 2097
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4821851517309583,
+      "learning_rate": 9.730233767717872e-06,
+      "loss": 0.3792,
+      "step": 2098
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5019454300568817,
+      "learning_rate": 9.729903652486664e-06,
+      "loss": 0.352,
+      "step": 2099
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2452438071156124,
+      "learning_rate": 9.72957334100335e-06,
+      "loss": 0.3746,
+      "step": 2100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.347600971039136,
+      "learning_rate": 9.72924283328164e-06,
+      "loss": 0.3582,
+      "step": 2101
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5775369431503403,
+      "learning_rate": 9.728912129335243e-06,
+      "loss": 0.3446,
+      "step": 2102
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.758077460841644,
+      "learning_rate": 9.728581229177884e-06,
+      "loss": 0.5081,
+      "step": 2103
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7767864893793226,
+      "learning_rate": 9.72825013282329e-06,
+      "loss": 0.3555,
+      "step": 2104
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.174707671827372,
+      "learning_rate": 9.7279188402852e-06,
+      "loss": 0.349,
+      "step": 2105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.286915943993544,
+      "learning_rate": 9.72758735157736e-06,
+      "loss": 0.3704,
+      "step": 2106
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.095936787943565,
+      "learning_rate": 9.727255666713524e-06,
+      "loss": 0.3737,
+      "step": 2107
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4735221401501195,
+      "learning_rate": 9.726923785707451e-06,
+      "loss": 0.366,
+      "step": 2108
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9593873396119232,
+      "learning_rate": 9.726591708572916e-06,
+      "loss": 0.3597,
+      "step": 2109
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.164677544404085,
+      "learning_rate": 9.726259435323698e-06,
+      "loss": 0.3707,
+      "step": 2110
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9366566055794823,
+      "learning_rate": 9.725926965973579e-06,
+      "loss": 0.3761,
+      "step": 2111
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6178898052831814,
+      "learning_rate": 9.725594300536355e-06,
+      "loss": 0.3591,
+      "step": 2112
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.830151832650409,
+      "learning_rate": 9.72526143902583e-06,
+      "loss": 0.3657,
+      "step": 2113
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.153213668681711,
+      "learning_rate": 9.724928381455817e-06,
+      "loss": 0.3519,
+      "step": 2114
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0002205611991255,
+      "learning_rate": 9.724595127840131e-06,
+      "loss": 0.3581,
+      "step": 2115
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.641998053257897,
+      "learning_rate": 9.724261678192602e-06,
+      "loss": 0.3505,
+      "step": 2116
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9981520047442376,
+      "learning_rate": 9.723928032527066e-06,
+      "loss": 0.3714,
+      "step": 2117
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5665521763553363,
+      "learning_rate": 9.723594190857363e-06,
+      "loss": 0.3701,
+      "step": 2118
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0869060682613885,
+      "learning_rate": 9.72326015319735e-06,
+      "loss": 0.353,
+      "step": 2119
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2564321768009163,
+      "learning_rate": 9.72292591956088e-06,
+      "loss": 0.3621,
+      "step": 2120
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8311405247028496,
+      "learning_rate": 9.722591489961829e-06,
+      "loss": 0.3469,
+      "step": 2121
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9637519786487785,
+      "learning_rate": 9.722256864414065e-06,
+      "loss": 0.3475,
+      "step": 2122
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.885360005957735,
+      "learning_rate": 9.721922042931478e-06,
+      "loss": 0.3653,
+      "step": 2123
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.228766380795061,
+      "learning_rate": 9.721587025527957e-06,
+      "loss": 0.3675,
+      "step": 2124
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8081692880168567,
+      "learning_rate": 9.721251812217405e-06,
+      "loss": 0.3461,
+      "step": 2125
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.592096217280288,
+      "learning_rate": 9.720916403013729e-06,
+      "loss": 0.3507,
+      "step": 2126
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9566894768797642,
+      "learning_rate": 9.720580797930845e-06,
+      "loss": 0.3679,
+      "step": 2127
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.810191299703153,
+      "learning_rate": 9.720244996982683e-06,
+      "loss": 0.3338,
+      "step": 2128
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.345368455037903,
+      "learning_rate": 9.719909000183167e-06,
+      "loss": 0.3882,
+      "step": 2129
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8941777637996409,
+      "learning_rate": 9.719572807546246e-06,
+      "loss": 0.3543,
+      "step": 2130
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.094741052343396,
+      "learning_rate": 9.719236419085866e-06,
+      "loss": 0.3667,
+      "step": 2131
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6390743301002137,
+      "learning_rate": 9.718899834815984e-06,
+      "loss": 0.3394,
+      "step": 2132
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.913875356846735,
+      "learning_rate": 9.718563054750566e-06,
+      "loss": 0.3695,
+      "step": 2133
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.482946254818429,
+      "learning_rate": 9.718226078903586e-06,
+      "loss": 0.3589,
+      "step": 2134
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7318896115244018,
+      "learning_rate": 9.717888907289026e-06,
+      "loss": 0.3977,
+      "step": 2135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6801088251335015,
+      "learning_rate": 9.717551539920875e-06,
+      "loss": 0.3653,
+      "step": 2136
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.584021393663589,
+      "learning_rate": 9.717213976813131e-06,
+      "loss": 0.386,
+      "step": 2137
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6385926170889076,
+      "learning_rate": 9.7168762179798e-06,
+      "loss": 0.3489,
+      "step": 2138
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9959221926480015,
+      "learning_rate": 9.716538263434899e-06,
+      "loss": 0.3412,
+      "step": 2139
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.781919752728311,
+      "learning_rate": 9.716200113192445e-06,
+      "loss": 0.3427,
+      "step": 2140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.660943842445923,
+      "learning_rate": 9.715861767266472e-06,
+      "loss": 0.3367,
+      "step": 2141
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2532444999278076,
+      "learning_rate": 9.715523225671019e-06,
+      "loss": 0.3718,
+      "step": 2142
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5405682742241553,
+      "learning_rate": 9.715184488420132e-06,
+      "loss": 0.3913,
+      "step": 2143
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3165325107589307,
+      "learning_rate": 9.714845555527865e-06,
+      "loss": 0.3552,
+      "step": 2144
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.514725721717478,
+      "learning_rate": 9.714506427008282e-06,
+      "loss": 0.3434,
+      "step": 2145
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9133066620117767,
+      "learning_rate": 9.714167102875452e-06,
+      "loss": 0.3465,
+      "step": 2146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8019517890381582,
+      "learning_rate": 9.713827583143455e-06,
+      "loss": 0.3542,
+      "step": 2147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.231724150089947,
+      "learning_rate": 9.71348786782638e-06,
+      "loss": 0.3729,
+      "step": 2148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2836644862699473,
+      "learning_rate": 9.713147956938322e-06,
+      "loss": 0.3694,
+      "step": 2149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.874461379345722,
+      "learning_rate": 9.712807850493382e-06,
+      "loss": 0.345,
+      "step": 2150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0606832864004367,
+      "learning_rate": 9.712467548505675e-06,
+      "loss": 0.3876,
+      "step": 2151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5345752718763965,
+      "learning_rate": 9.712127050989319e-06,
+      "loss": 0.3635,
+      "step": 2152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.781510323719983,
+      "learning_rate": 9.711786357958442e-06,
+      "loss": 0.3499,
+      "step": 2153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2632539701353243,
+      "learning_rate": 9.711445469427179e-06,
+      "loss": 0.3659,
+      "step": 2154
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.497717047183541,
+      "learning_rate": 9.711104385409676e-06,
+      "loss": 0.4012,
+      "step": 2155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.013333543854991,
+      "learning_rate": 9.710763105920082e-06,
+      "loss": 0.3674,
+      "step": 2156
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8311470053069792,
+      "learning_rate": 9.710421630972563e-06,
+      "loss": 0.4983,
+      "step": 2157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.534554032587132,
+      "learning_rate": 9.71007996058128e-06,
+      "loss": 0.368,
+      "step": 2158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7369760274841606,
+      "learning_rate": 9.709738094760415e-06,
+      "loss": 0.3525,
+      "step": 2159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8680377874275687,
+      "learning_rate": 9.709396033524153e-06,
+      "loss": 0.3587,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8702498262496199,
+      "learning_rate": 9.709053776886683e-06,
+      "loss": 0.3524,
+      "step": 2161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.874239426872966,
+      "learning_rate": 9.708711324862208e-06,
+      "loss": 0.3459,
+      "step": 2162
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7926988622182338,
+      "learning_rate": 9.708368677464936e-06,
+      "loss": 0.3364,
+      "step": 2163
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.068595156773728,
+      "learning_rate": 9.708025834709085e-06,
+      "loss": 0.3882,
+      "step": 2164
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.090259503267306,
+      "learning_rate": 9.707682796608879e-06,
+      "loss": 0.3667,
+      "step": 2165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.543607895466217,
+      "learning_rate": 9.707339563178554e-06,
+      "loss": 0.3809,
+      "step": 2166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.355073725762526,
+      "learning_rate": 9.706996134432346e-06,
+      "loss": 0.3698,
+      "step": 2167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.039427524367994,
+      "learning_rate": 9.70665251038451e-06,
+      "loss": 0.3572,
+      "step": 2168
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3888312239997838,
+      "learning_rate": 9.706308691049302e-06,
+      "loss": 0.3552,
+      "step": 2169
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.188509015657696,
+      "learning_rate": 9.705964676440984e-06,
+      "loss": 0.3437,
+      "step": 2170
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.681347300694205,
+      "learning_rate": 9.705620466573837e-06,
+      "loss": 0.3547,
+      "step": 2171
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4445920081855586,
+      "learning_rate": 9.705276061462135e-06,
+      "loss": 0.3979,
+      "step": 2172
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.107750685951658,
+      "learning_rate": 9.704931461120173e-06,
+      "loss": 0.3716,
+      "step": 2173
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6290719881406557,
+      "learning_rate": 9.70458666556225e-06,
+      "loss": 0.371,
+      "step": 2174
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3135326378460883,
+      "learning_rate": 9.704241674802668e-06,
+      "loss": 0.3849,
+      "step": 2175
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9109486876018453,
+      "learning_rate": 9.703896488855742e-06,
+      "loss": 0.3744,
+      "step": 2176
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.09958236067567,
+      "learning_rate": 9.703551107735798e-06,
+      "loss": 0.3471,
+      "step": 2177
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8349254683229255,
+      "learning_rate": 9.703205531457163e-06,
+      "loss": 0.4838,
+      "step": 2178
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.459660001768668,
+      "learning_rate": 9.702859760034177e-06,
+      "loss": 0.3497,
+      "step": 2179
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.090096725810043,
+      "learning_rate": 9.702513793481186e-06,
+      "loss": 0.3532,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.560318138826427,
+      "learning_rate": 9.702167631812544e-06,
+      "loss": 0.3704,
+      "step": 2181
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.5780875389561455,
+      "learning_rate": 9.701821275042618e-06,
+      "loss": 0.3443,
+      "step": 2182
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.047438538653314,
+      "learning_rate": 9.701474723185774e-06,
+      "loss": 0.3573,
+      "step": 2183
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.001446806845423,
+      "learning_rate": 9.701127976256392e-06,
+      "loss": 0.3436,
+      "step": 2184
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6806415294516435,
+      "learning_rate": 9.700781034268861e-06,
+      "loss": 0.3663,
+      "step": 2185
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.992265753593225,
+      "learning_rate": 9.700433897237576e-06,
+      "loss": 0.3769,
+      "step": 2186
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.383394810883486,
+      "learning_rate": 9.70008656517694e-06,
+      "loss": 0.3643,
+      "step": 2187
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.130913273587383,
+      "learning_rate": 9.699739038101363e-06,
+      "loss": 0.3518,
+      "step": 2188
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.142319729777803,
+      "learning_rate": 9.699391316025266e-06,
+      "loss": 0.3516,
+      "step": 2189
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9270841797880551,
+      "learning_rate": 9.699043398963075e-06,
+      "loss": 0.3514,
+      "step": 2190
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5876699053379104,
+      "learning_rate": 9.69869528692923e-06,
+      "loss": 0.37,
+      "step": 2191
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.541893933981502,
+      "learning_rate": 9.698346979938169e-06,
+      "loss": 0.3654,
+      "step": 2192
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.134315845942576,
+      "learning_rate": 9.697998478004347e-06,
+      "loss": 0.3506,
+      "step": 2193
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6481892197476746,
+      "learning_rate": 9.697649781142225e-06,
+      "loss": 0.3597,
+      "step": 2194
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.612137458004614,
+      "learning_rate": 9.697300889366268e-06,
+      "loss": 0.3576,
+      "step": 2195
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9316800321828856,
+      "learning_rate": 9.696951802690955e-06,
+      "loss": 0.5251,
+      "step": 2196
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.172566635186495,
+      "learning_rate": 9.696602521130768e-06,
+      "loss": 0.3546,
+      "step": 2197
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.960716584993781,
+      "learning_rate": 9.6962530447002e-06,
+      "loss": 0.3566,
+      "step": 2198
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.32410744425751,
+      "learning_rate": 9.695903373413753e-06,
+      "loss": 0.3782,
+      "step": 2199
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.665659509685868,
+      "learning_rate": 9.695553507285934e-06,
+      "loss": 0.3784,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5038958624637533,
+      "learning_rate": 9.695203446331258e-06,
+      "loss": 0.3638,
+      "step": 2201
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5654988510252177,
+      "learning_rate": 9.694853190564253e-06,
+      "loss": 0.3836,
+      "step": 2202
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.274045728265552,
+      "learning_rate": 9.694502739999449e-06,
+      "loss": 0.3323,
+      "step": 2203
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3439694332278393,
+      "learning_rate": 9.69415209465139e-06,
+      "loss": 0.3605,
+      "step": 2204
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.560716690652859,
+      "learning_rate": 9.693801254534622e-06,
+      "loss": 0.3624,
+      "step": 2205
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.409334241591523,
+      "learning_rate": 9.693450219663703e-06,
+      "loss": 0.3699,
+      "step": 2206
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9714270733155934,
+      "learning_rate": 9.693098990053197e-06,
+      "loss": 0.3708,
+      "step": 2207
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0923282592001042,
+      "learning_rate": 9.692747565717677e-06,
+      "loss": 0.3682,
+      "step": 2208
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3440626411450176,
+      "learning_rate": 9.692395946671727e-06,
+      "loss": 0.3488,
+      "step": 2209
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5333082990512574,
+      "learning_rate": 9.692044132929934e-06,
+      "loss": 0.3774,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0748073568895737,
+      "learning_rate": 9.691692124506896e-06,
+      "loss": 0.3603,
+      "step": 2211
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2534062520128604,
+      "learning_rate": 9.691339921417219e-06,
+      "loss": 0.3377,
+      "step": 2212
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5344521511625326,
+      "learning_rate": 9.690987523675514e-06,
+      "loss": 0.3578,
+      "step": 2213
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.335457593172534,
+      "learning_rate": 9.690634931296408e-06,
+      "loss": 0.3637,
+      "step": 2214
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.446251914770414,
+      "learning_rate": 9.690282144294524e-06,
+      "loss": 0.333,
+      "step": 2215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2504853190219287,
+      "learning_rate": 9.689929162684503e-06,
+      "loss": 0.3596,
+      "step": 2216
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9979319485051117,
+      "learning_rate": 9.689575986480992e-06,
+      "loss": 0.3674,
+      "step": 2217
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.147606422679075,
+      "learning_rate": 9.689222615698643e-06,
+      "loss": 0.3513,
+      "step": 2218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0713409710626527,
+      "learning_rate": 9.688869050352119e-06,
+      "loss": 0.3708,
+      "step": 2219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.697857059745171,
+      "learning_rate": 9.688515290456092e-06,
+      "loss": 0.3462,
+      "step": 2220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.905978614169605,
+      "learning_rate": 9.688161336025234e-06,
+      "loss": 0.3429,
+      "step": 2221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.909014222345689,
+      "learning_rate": 9.687807187074238e-06,
+      "loss": 0.3435,
+      "step": 2222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.615511745018682,
+      "learning_rate": 9.687452843617792e-06,
+      "loss": 0.3758,
+      "step": 2223
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8141319252177568,
+      "learning_rate": 9.687098305670606e-06,
+      "loss": 0.3673,
+      "step": 2224
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3395920323086177,
+      "learning_rate": 9.686743573247383e-06,
+      "loss": 0.3605,
+      "step": 2225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6540624458603483,
+      "learning_rate": 9.686388646362846e-06,
+      "loss": 0.3421,
+      "step": 2226
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.120099167484428,
+      "learning_rate": 9.68603352503172e-06,
+      "loss": 0.3772,
+      "step": 2227
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8111315477744756,
+      "learning_rate": 9.685678209268738e-06,
+      "loss": 0.3679,
+      "step": 2228
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.121181310744283,
+      "learning_rate": 9.685322699088647e-06,
+      "loss": 0.372,
+      "step": 2229
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3190016218223635,
+      "learning_rate": 9.684966994506193e-06,
+      "loss": 0.3582,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0008797873312196,
+      "learning_rate": 9.684611095536137e-06,
+      "loss": 0.513,
+      "step": 2231
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.085802656987168,
+      "learning_rate": 9.684255002193246e-06,
+      "loss": 0.3413,
+      "step": 2232
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.842011045685869,
+      "learning_rate": 9.683898714492296e-06,
+      "loss": 0.3636,
+      "step": 2233
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.345497910221064,
+      "learning_rate": 9.683542232448068e-06,
+      "loss": 0.356,
+      "step": 2234
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3031539572997266,
+      "learning_rate": 9.683185556075354e-06,
+      "loss": 0.3674,
+      "step": 2235
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5441288379330023,
+      "learning_rate": 9.682828685388954e-06,
+      "loss": 0.3426,
+      "step": 2236
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.07700230917055,
+      "learning_rate": 9.682471620403673e-06,
+      "loss": 0.3548,
+      "step": 2237
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4081556992150213,
+      "learning_rate": 9.682114361134327e-06,
+      "loss": 0.3308,
+      "step": 2238
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7716307869559373,
+      "learning_rate": 9.681756907595741e-06,
+      "loss": 0.3497,
+      "step": 2239
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1025799805934735,
+      "learning_rate": 9.681399259802744e-06,
+      "loss": 0.3453,
+      "step": 2240
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5764295638939267,
+      "learning_rate": 9.681041417770176e-06,
+      "loss": 0.3626,
+      "step": 2241
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.239281158943012,
+      "learning_rate": 9.680683381512888e-06,
+      "loss": 0.3478,
+      "step": 2242
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2584329750144967,
+      "learning_rate": 9.68032515104573e-06,
+      "loss": 0.3585,
+      "step": 2243
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.780756543442347,
+      "learning_rate": 9.679966726383569e-06,
+      "loss": 0.382,
+      "step": 2244
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.641132909899936,
+      "learning_rate": 9.679608107541278e-06,
+      "loss": 0.376,
+      "step": 2245
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9373484888075962,
+      "learning_rate": 9.679249294533733e-06,
+      "loss": 0.3515,
+      "step": 2246
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7470182000533456,
+      "learning_rate": 9.678890287375823e-06,
+      "loss": 0.349,
+      "step": 2247
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.075101042853636,
+      "learning_rate": 9.678531086082444e-06,
+      "loss": 0.3598,
+      "step": 2248
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.545818294862518,
+      "learning_rate": 9.6781716906685e-06,
+      "loss": 0.3414,
+      "step": 2249
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.72612675981877,
+      "learning_rate": 9.677812101148906e-06,
+      "loss": 0.3542,
+      "step": 2250
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.599238197206029,
+      "learning_rate": 9.677452317538576e-06,
+      "loss": 0.3667,
+      "step": 2251
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.552295386551471,
+      "learning_rate": 9.677092339852443e-06,
+      "loss": 0.3757,
+      "step": 2252
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2749637261026314,
+      "learning_rate": 9.676732168105443e-06,
+      "loss": 0.3502,
+      "step": 2253
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8524830075707692,
+      "learning_rate": 9.676371802312515e-06,
+      "loss": 0.361,
+      "step": 2254
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5399726269706826,
+      "learning_rate": 9.676011242488616e-06,
+      "loss": 0.3845,
+      "step": 2255
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.354759199309399,
+      "learning_rate": 9.675650488648707e-06,
+      "loss": 0.3341,
+      "step": 2256
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.603553884786818,
+      "learning_rate": 9.675289540807752e-06,
+      "loss": 0.3333,
+      "step": 2257
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.773587318579438,
+      "learning_rate": 9.674928398980729e-06,
+      "loss": 0.3362,
+      "step": 2258
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.67176691595643,
+      "learning_rate": 9.674567063182626e-06,
+      "loss": 0.3625,
+      "step": 2259
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.908384262448215,
+      "learning_rate": 9.674205533428431e-06,
+      "loss": 0.3691,
+      "step": 2260
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.307049662295108,
+      "learning_rate": 9.673843809733145e-06,
+      "loss": 0.3554,
+      "step": 2261
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1999270013192764,
+      "learning_rate": 9.67348189211178e-06,
+      "loss": 0.3623,
+      "step": 2262
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.79505183529914,
+      "learning_rate": 9.67311978057935e-06,
+      "loss": 0.3797,
+      "step": 2263
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.568944247674684,
+      "learning_rate": 9.672757475150878e-06,
+      "loss": 0.3605,
+      "step": 2264
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 67.47337375924475,
+      "learning_rate": 9.6723949758414e-06,
+      "loss": 0.3194,
+      "step": 2265
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1499024496376413,
+      "learning_rate": 9.672032282665954e-06,
+      "loss": 0.3772,
+      "step": 2266
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4091814983918054,
+      "learning_rate": 9.67166939563959e-06,
+      "loss": 0.3843,
+      "step": 2267
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3335453182494446,
+      "learning_rate": 9.671306314777367e-06,
+      "loss": 0.345,
+      "step": 2268
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.160702802736921,
+      "learning_rate": 9.670943040094347e-06,
+      "loss": 0.3515,
+      "step": 2269
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4360182720795938,
+      "learning_rate": 9.670579571605605e-06,
+      "loss": 0.3471,
+      "step": 2270
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.653755317861475,
+      "learning_rate": 9.670215909326219e-06,
+      "loss": 0.3304,
+      "step": 2271
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1601480854090656,
+      "learning_rate": 9.669852053271278e-06,
+      "loss": 0.3534,
+      "step": 2272
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.951268372737922,
+      "learning_rate": 9.669488003455884e-06,
+      "loss": 0.3548,
+      "step": 2273
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1361984311785482,
+      "learning_rate": 9.669123759895137e-06,
+      "loss": 0.372,
+      "step": 2274
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.46221679531022,
+      "learning_rate": 9.668759322604154e-06,
+      "loss": 0.354,
+      "step": 2275
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9824662415726244,
+      "learning_rate": 9.66839469159805e-06,
+      "loss": 0.3596,
+      "step": 2276
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.800468707625609,
+      "learning_rate": 9.668029866891962e-06,
+      "loss": 0.3483,
+      "step": 2277
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0650565663973084,
+      "learning_rate": 9.667664848501022e-06,
+      "loss": 0.5287,
+      "step": 2278
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8209529394633437,
+      "learning_rate": 9.667299636440377e-06,
+      "loss": 0.3604,
+      "step": 2279
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6744587582900383,
+      "learning_rate": 9.66693423072518e-06,
+      "loss": 0.3607,
+      "step": 2280
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5962459095432746,
+      "learning_rate": 9.666568631370592e-06,
+      "loss": 0.3614,
+      "step": 2281
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6639755639194774,
+      "learning_rate": 9.666202838391783e-06,
+      "loss": 0.4927,
+      "step": 2282
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2951901857763573,
+      "learning_rate": 9.66583685180393e-06,
+      "loss": 0.3445,
+      "step": 2283
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.511161265817613,
+      "learning_rate": 9.66547067162222e-06,
+      "loss": 0.3773,
+      "step": 2284
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.901182692439293,
+      "learning_rate": 9.665104297861842e-06,
+      "loss": 0.3601,
+      "step": 2285
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5004267910064106,
+      "learning_rate": 9.664737730538003e-06,
+      "loss": 0.3425,
+      "step": 2286
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.699872788028884,
+      "learning_rate": 9.66437096966591e-06,
+      "loss": 0.3646,
+      "step": 2287
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.466253276055321,
+      "learning_rate": 9.66400401526078e-06,
+      "loss": 0.3685,
+      "step": 2288
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.974486155869181,
+      "learning_rate": 9.663636867337838e-06,
+      "loss": 0.348,
+      "step": 2289
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.009561889838298,
+      "learning_rate": 9.66326952591232e-06,
+      "loss": 0.3456,
+      "step": 2290
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3282365927121345,
+      "learning_rate": 9.662901990999468e-06,
+      "loss": 0.3474,
+      "step": 2291
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7540195970671437,
+      "learning_rate": 9.662534262614528e-06,
+      "loss": 0.3526,
+      "step": 2292
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.031315959580994,
+      "learning_rate": 9.66216634077276e-06,
+      "loss": 0.3614,
+      "step": 2293
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.290872175706192,
+      "learning_rate": 9.66179822548943e-06,
+      "loss": 0.3535,
+      "step": 2294
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.08464816871641,
+      "learning_rate": 9.661429916779812e-06,
+      "loss": 0.3481,
+      "step": 2295
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.618922565460359,
+      "learning_rate": 9.661061414659185e-06,
+      "loss": 0.3652,
+      "step": 2296
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8224276494957194,
+      "learning_rate": 9.660692719142843e-06,
+      "loss": 0.3428,
+      "step": 2297
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.68191708088931,
+      "learning_rate": 9.660323830246082e-06,
+      "loss": 0.3411,
+      "step": 2298
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.328870414668892,
+      "learning_rate": 9.659954747984206e-06,
+      "loss": 0.3403,
+      "step": 2299
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.38353695337482,
+      "learning_rate": 9.65958547237253e-06,
+      "loss": 0.3736,
+      "step": 2300
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.749184432940049,
+      "learning_rate": 9.659216003426378e-06,
+      "loss": 0.3624,
+      "step": 2301
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.169476479975519,
+      "learning_rate": 9.658846341161079e-06,
+      "loss": 0.3761,
+      "step": 2302
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.365985932567745,
+      "learning_rate": 9.658476485591968e-06,
+      "loss": 0.3781,
+      "step": 2303
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8827023001781615,
+      "learning_rate": 9.658106436734395e-06,
+      "loss": 0.3669,
+      "step": 2304
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3501941336783827,
+      "learning_rate": 9.657736194603711e-06,
+      "loss": 0.3702,
+      "step": 2305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8538506973369806,
+      "learning_rate": 9.657365759215281e-06,
+      "loss": 0.3449,
+      "step": 2306
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7980927854541333,
+      "learning_rate": 9.656995130584473e-06,
+      "loss": 0.3454,
+      "step": 2307
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.245172410954305,
+      "learning_rate": 9.656624308726662e-06,
+      "loss": 0.3634,
+      "step": 2308
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.389572450307798,
+      "learning_rate": 9.65625329365724e-06,
+      "loss": 0.3397,
+      "step": 2309
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3100204934540214,
+      "learning_rate": 9.6558820853916e-06,
+      "loss": 0.3358,
+      "step": 2310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0980354364598153,
+      "learning_rate": 9.655510683945139e-06,
+      "loss": 0.3616,
+      "step": 2311
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.034610519930271,
+      "learning_rate": 9.655139089333272e-06,
+      "loss": 0.3441,
+      "step": 2312
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5320154288031897,
+      "learning_rate": 9.654767301571418e-06,
+      "loss": 0.3595,
+      "step": 2313
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9739325212177743,
+      "learning_rate": 9.654395320674998e-06,
+      "loss": 0.3398,
+      "step": 2314
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2320474183680483,
+      "learning_rate": 9.65402314665945e-06,
+      "loss": 0.3235,
+      "step": 2315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2347126569903,
+      "learning_rate": 9.653650779540214e-06,
+      "loss": 0.3631,
+      "step": 2316
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8442530371937012,
+      "learning_rate": 9.653278219332742e-06,
+      "loss": 0.3616,
+      "step": 2317
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.60449634011267,
+      "learning_rate": 9.652905466052492e-06,
+      "loss": 0.3279,
+      "step": 2318
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8604555700909606,
+      "learning_rate": 9.65253251971493e-06,
+      "loss": 0.357,
+      "step": 2319
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.248440473110048,
+      "learning_rate": 9.65215938033553e-06,
+      "loss": 0.4123,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2557147594249787,
+      "learning_rate": 9.651786047929772e-06,
+      "loss": 0.3519,
+      "step": 2321
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.337233149359728,
+      "learning_rate": 9.651412522513151e-06,
+      "loss": 0.3498,
+      "step": 2322
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.08644851302118,
+      "learning_rate": 9.651038804101162e-06,
+      "loss": 0.368,
+      "step": 2323
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.201117362555605,
+      "learning_rate": 9.650664892709311e-06,
+      "loss": 0.3862,
+      "step": 2324
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.785244373287499,
+      "learning_rate": 9.650290788353114e-06,
+      "loss": 0.3688,
+      "step": 2325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.887624913253619,
+      "learning_rate": 9.649916491048092e-06,
+      "loss": 0.3399,
+      "step": 2326
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0446341964073884,
+      "learning_rate": 9.649542000809775e-06,
+      "loss": 0.3469,
+      "step": 2327
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2234333392104157,
+      "learning_rate": 9.649167317653703e-06,
+      "loss": 0.5292,
+      "step": 2328
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0099881681755503,
+      "learning_rate": 9.64879244159542e-06,
+      "loss": 0.3513,
+      "step": 2329
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.855983876377842,
+      "learning_rate": 9.648417372650482e-06,
+      "loss": 0.3457,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7475013771571395,
+      "learning_rate": 9.64804211083445e-06,
+      "loss": 0.3537,
+      "step": 2331
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1653810440796653,
+      "learning_rate": 9.647666656162898e-06,
+      "loss": 0.3517,
+      "step": 2332
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.02294797819566,
+      "learning_rate": 9.647291008651398e-06,
+      "loss": 0.3651,
+      "step": 2333
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.048615989165704,
+      "learning_rate": 9.64691516831554e-06,
+      "loss": 0.3682,
+      "step": 2334
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.654696700868606,
+      "learning_rate": 9.646539135170919e-06,
+      "loss": 0.366,
+      "step": 2335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9975112521803116,
+      "learning_rate": 9.646162909233135e-06,
+      "loss": 0.3726,
+      "step": 2336
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1020659732556655,
+      "learning_rate": 9.6457864905178e-06,
+      "loss": 0.3499,
+      "step": 2337
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4810057879316187,
+      "learning_rate": 9.64540987904053e-06,
+      "loss": 0.3301,
+      "step": 2338
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.880311443001001,
+      "learning_rate": 9.645033074816955e-06,
+      "loss": 0.3508,
+      "step": 2339
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9626301859520199,
+      "learning_rate": 9.644656077862706e-06,
+      "loss": 0.3628,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6082585080628293,
+      "learning_rate": 9.644278888193427e-06,
+      "loss": 0.3801,
+      "step": 2341
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.508594621300968,
+      "learning_rate": 9.64390150582477e-06,
+      "loss": 0.3827,
+      "step": 2342
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.162779175873289,
+      "learning_rate": 9.643523930772388e-06,
+      "loss": 0.3826,
+      "step": 2343
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9706900900386937,
+      "learning_rate": 9.64314616305195e-06,
+      "loss": 0.372,
+      "step": 2344
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.306415379539258,
+      "learning_rate": 9.64276820267913e-06,
+      "loss": 0.3443,
+      "step": 2345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.345837477259276,
+      "learning_rate": 9.642390049669614e-06,
+      "loss": 0.345,
+      "step": 2346
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0960833003109123,
+      "learning_rate": 9.642011704039087e-06,
+      "loss": 0.3802,
+      "step": 2347
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1174944807098237,
+      "learning_rate": 9.64163316580325e-06,
+      "loss": 0.3522,
+      "step": 2348
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5986618743569525,
+      "learning_rate": 9.64125443497781e-06,
+      "loss": 0.3747,
+      "step": 2349
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.178705716855747,
+      "learning_rate": 9.640875511578475e-06,
+      "loss": 0.3789,
+      "step": 2350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7108012864525584,
+      "learning_rate": 9.640496395620976e-06,
+      "loss": 0.3437,
+      "step": 2351
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9976088571999921,
+      "learning_rate": 9.640117087121038e-06,
+      "loss": 0.3555,
+      "step": 2352
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.332575910936503,
+      "learning_rate": 9.6397375860944e-06,
+      "loss": 0.3602,
+      "step": 2353
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0413358711022536,
+      "learning_rate": 9.63935789255681e-06,
+      "loss": 0.3297,
+      "step": 2354
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9357488207033153,
+      "learning_rate": 9.638978006524017e-06,
+      "loss": 0.3331,
+      "step": 2355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.794251389142506,
+      "learning_rate": 9.63859792801179e-06,
+      "loss": 0.3618,
+      "step": 2356
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0938584713205217,
+      "learning_rate": 9.638217657035895e-06,
+      "loss": 0.3472,
+      "step": 2357
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1667692379618164,
+      "learning_rate": 9.637837193612112e-06,
+      "loss": 0.3462,
+      "step": 2358
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2927168392100215,
+      "learning_rate": 9.637456537756224e-06,
+      "loss": 0.353,
+      "step": 2359
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7197863118466372,
+      "learning_rate": 9.637075689484027e-06,
+      "loss": 0.3371,
+      "step": 2360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.817247900294978,
+      "learning_rate": 9.636694648811326e-06,
+      "loss": 0.3552,
+      "step": 2361
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.543037062907734,
+      "learning_rate": 9.636313415753927e-06,
+      "loss": 0.3741,
+      "step": 2362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.536416547673591,
+      "learning_rate": 9.635931990327649e-06,
+      "loss": 0.3268,
+      "step": 2363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6625163197933777,
+      "learning_rate": 9.635550372548317e-06,
+      "loss": 0.3682,
+      "step": 2364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3342808485981523,
+      "learning_rate": 9.635168562431769e-06,
+      "loss": 0.3543,
+      "step": 2365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7700670597111325,
+      "learning_rate": 9.634786559993842e-06,
+      "loss": 0.3591,
+      "step": 2366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.35645116878552,
+      "learning_rate": 9.634404365250391e-06,
+      "loss": 0.3506,
+      "step": 2367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.707460131914564,
+      "learning_rate": 9.63402197821727e-06,
+      "loss": 0.3716,
+      "step": 2368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5427111309804353,
+      "learning_rate": 9.633639398910346e-06,
+      "loss": 0.3174,
+      "step": 2369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.715698775498241,
+      "learning_rate": 9.633256627345494e-06,
+      "loss": 0.366,
+      "step": 2370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.0358262155339855,
+      "learning_rate": 9.632873663538594e-06,
+      "loss": 0.3549,
+      "step": 2371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.52722156177181,
+      "learning_rate": 9.632490507505536e-06,
+      "loss": 0.3644,
+      "step": 2372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0780316152770864,
+      "learning_rate": 9.632107159262218e-06,
+      "loss": 0.3586,
+      "step": 2373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.64635097432633,
+      "learning_rate": 9.631723618824549e-06,
+      "loss": 0.3482,
+      "step": 2374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.197373002378108,
+      "learning_rate": 9.63133988620844e-06,
+      "loss": 0.3509,
+      "step": 2375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.87685837698477,
+      "learning_rate": 9.63095596142981e-06,
+      "loss": 0.3944,
+      "step": 2376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.500786724373431,
+      "learning_rate": 9.630571844504594e-06,
+      "loss": 0.3502,
+      "step": 2377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7455177113536733,
+      "learning_rate": 9.630187535448727e-06,
+      "loss": 0.3756,
+      "step": 2378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.23001573548013,
+      "learning_rate": 9.629803034278155e-06,
+      "loss": 0.4042,
+      "step": 2379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.94647314913899,
+      "learning_rate": 9.629418341008831e-06,
+      "loss": 0.3496,
+      "step": 2380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.537475468068543,
+      "learning_rate": 9.62903345565672e-06,
+      "loss": 0.3839,
+      "step": 2381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4068648716576693,
+      "learning_rate": 9.628648378237786e-06,
+      "loss": 0.3421,
+      "step": 2382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9439316442881507,
+      "learning_rate": 9.62826310876801e-06,
+      "loss": 0.3703,
+      "step": 2383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9368852642831618,
+      "learning_rate": 9.627877647263378e-06,
+      "loss": 0.3556,
+      "step": 2384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.335718471294046,
+      "learning_rate": 9.62749199373988e-06,
+      "loss": 0.3522,
+      "step": 2385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.029347325183008,
+      "learning_rate": 9.627106148213521e-06,
+      "loss": 0.3527,
+      "step": 2386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.49920180217814,
+      "learning_rate": 9.62672011070031e-06,
+      "loss": 0.3331,
+      "step": 2387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.666386157382629,
+      "learning_rate": 9.626333881216263e-06,
+      "loss": 0.3868,
+      "step": 2388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.453130005653644,
+      "learning_rate": 9.625947459777408e-06,
+      "loss": 0.3699,
+      "step": 2389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.799579787938819,
+      "learning_rate": 9.625560846399774e-06,
+      "loss": 0.3519,
+      "step": 2390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0096918035158344,
+      "learning_rate": 9.625174041099403e-06,
+      "loss": 0.3553,
+      "step": 2391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.00932756878222,
+      "learning_rate": 9.624787043892349e-06,
+      "loss": 0.381,
+      "step": 2392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2037870386656717,
+      "learning_rate": 9.624399854794664e-06,
+      "loss": 0.3331,
+      "step": 2393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.667139834720969,
+      "learning_rate": 9.624012473822417e-06,
+      "loss": 0.3477,
+      "step": 2394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.955719936010595,
+      "learning_rate": 9.623624900991676e-06,
+      "loss": 0.3683,
+      "step": 2395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5910109694969887,
+      "learning_rate": 9.623237136318529e-06,
+      "loss": 0.3476,
+      "step": 2396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.44330547353958,
+      "learning_rate": 9.62284917981906e-06,
+      "loss": 0.3623,
+      "step": 2397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7897407289814233,
+      "learning_rate": 9.622461031509366e-06,
+      "loss": 0.3634,
+      "step": 2398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.556571649402105,
+      "learning_rate": 9.622072691405557e-06,
+      "loss": 0.3448,
+      "step": 2399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9887895139227743,
+      "learning_rate": 9.621684159523739e-06,
+      "loss": 0.3528,
+      "step": 2400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.848350239825674,
+      "learning_rate": 9.621295435880038e-06,
+      "loss": 0.3416,
+      "step": 2401
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.979998864781763,
+      "learning_rate": 9.62090652049058e-06,
+      "loss": 0.3527,
+      "step": 2402
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.741250462625021,
+      "learning_rate": 9.620517413371503e-06,
+      "loss": 0.3682,
+      "step": 2403
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4301657781684147,
+      "learning_rate": 9.62012811453895e-06,
+      "loss": 0.3721,
+      "step": 2404
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.741096371782502,
+      "learning_rate": 9.619738624009078e-06,
+      "loss": 0.359,
+      "step": 2405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.18560320803686,
+      "learning_rate": 9.619348941798044e-06,
+      "loss": 0.345,
+      "step": 2406
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1573392761835413,
+      "learning_rate": 9.618959067922019e-06,
+      "loss": 0.3554,
+      "step": 2407
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.900530906012863,
+      "learning_rate": 9.618569002397176e-06,
+      "loss": 0.3414,
+      "step": 2408
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8469436713944822,
+      "learning_rate": 9.618178745239701e-06,
+      "loss": 0.3523,
+      "step": 2409
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.61986651265615,
+      "learning_rate": 9.617788296465789e-06,
+      "loss": 0.3599,
+      "step": 2410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.172438301853682,
+      "learning_rate": 9.617397656091637e-06,
+      "loss": 0.3371,
+      "step": 2411
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6323418059838244,
+      "learning_rate": 9.617006824133455e-06,
+      "loss": 0.3428,
+      "step": 2412
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8478925125461014,
+      "learning_rate": 9.61661580060746e-06,
+      "loss": 0.366,
+      "step": 2413
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.964172206535194,
+      "learning_rate": 9.616224585529873e-06,
+      "loss": 0.3687,
+      "step": 2414
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1357494876913283,
+      "learning_rate": 9.615833178916932e-06,
+      "loss": 0.3491,
+      "step": 2415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.510423121848369,
+      "learning_rate": 9.615441580784873e-06,
+      "loss": 0.3324,
+      "step": 2416
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4890764516931316,
+      "learning_rate": 9.615049791149944e-06,
+      "loss": 0.3411,
+      "step": 2417
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.176409776466373,
+      "learning_rate": 9.614657810028402e-06,
+      "loss": 0.3902,
+      "step": 2418
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.276996482228377,
+      "learning_rate": 9.614265637436511e-06,
+      "loss": 0.3526,
+      "step": 2419
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3215123930986823,
+      "learning_rate": 9.613873273390544e-06,
+      "loss": 0.3386,
+      "step": 2420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.603729331200219,
+      "learning_rate": 9.613480717906778e-06,
+      "loss": 0.3224,
+      "step": 2421
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5575212249276613,
+      "learning_rate": 9.613087971001502e-06,
+      "loss": 0.3416,
+      "step": 2422
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.24107590249232,
+      "learning_rate": 9.612695032691013e-06,
+      "loss": 0.3392,
+      "step": 2423
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.012508516067043,
+      "learning_rate": 9.612301902991615e-06,
+      "loss": 0.3495,
+      "step": 2424
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.909970317601092,
+      "learning_rate": 9.611908581919618e-06,
+      "loss": 0.3626,
+      "step": 2425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.230590477505716,
+      "learning_rate": 9.611515069491342e-06,
+      "loss": 0.37,
+      "step": 2426
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9191284574500584,
+      "learning_rate": 9.611121365723115e-06,
+      "loss": 0.3696,
+      "step": 2427
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8037932763176585,
+      "learning_rate": 9.610727470631273e-06,
+      "loss": 0.3561,
+      "step": 2428
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8363081561004386,
+      "learning_rate": 9.610333384232158e-06,
+      "loss": 0.3554,
+      "step": 2429
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.137207765860285,
+      "learning_rate": 9.609939106542123e-06,
+      "loss": 0.3614,
+      "step": 2430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7826653475158007,
+      "learning_rate": 9.609544637577524e-06,
+      "loss": 0.3588,
+      "step": 2431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.958046642291602,
+      "learning_rate": 9.609149977354733e-06,
+      "loss": 0.3699,
+      "step": 2432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2142013270687104,
+      "learning_rate": 9.608755125890121e-06,
+      "loss": 0.3577,
+      "step": 2433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7252581614357747,
+      "learning_rate": 9.608360083200074e-06,
+      "loss": 0.3535,
+      "step": 2434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.9804888275379895,
+      "learning_rate": 9.607964849300981e-06,
+      "loss": 0.3517,
+      "step": 2435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1916292446784316,
+      "learning_rate": 9.607569424209243e-06,
+      "loss": 0.3332,
+      "step": 2436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.127344557367979,
+      "learning_rate": 9.607173807941263e-06,
+      "loss": 0.3447,
+      "step": 2437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.290307544206036,
+      "learning_rate": 9.60677800051346e-06,
+      "loss": 0.3452,
+      "step": 2438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7397610296303663,
+      "learning_rate": 9.606382001942256e-06,
+      "loss": 0.3292,
+      "step": 2439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.187356183486216,
+      "learning_rate": 9.605985812244079e-06,
+      "loss": 0.3511,
+      "step": 2440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.865261940931568,
+      "learning_rate": 9.605589431435371e-06,
+      "loss": 0.3339,
+      "step": 2441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4530789950310563,
+      "learning_rate": 9.605192859532577e-06,
+      "loss": 0.3564,
+      "step": 2442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7145727865835654,
+      "learning_rate": 9.604796096552151e-06,
+      "loss": 0.3763,
+      "step": 2443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.854123911001881,
+      "learning_rate": 9.604399142510557e-06,
+      "loss": 0.3717,
+      "step": 2444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4863620391474894,
+      "learning_rate": 9.604001997424261e-06,
+      "loss": 0.3624,
+      "step": 2445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8969110954008894,
+      "learning_rate": 9.603604661309747e-06,
+      "loss": 0.3415,
+      "step": 2446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 21.640684111149046,
+      "learning_rate": 9.6032071341835e-06,
+      "loss": 0.3714,
+      "step": 2447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.682823158355087,
+      "learning_rate": 9.602809416062011e-06,
+      "loss": 0.3616,
+      "step": 2448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4358330308884466,
+      "learning_rate": 9.602411506961784e-06,
+      "loss": 0.3422,
+      "step": 2449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.193203650915292,
+      "learning_rate": 9.602013406899328e-06,
+      "loss": 0.364,
+      "step": 2450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2529084116181983,
+      "learning_rate": 9.601615115891164e-06,
+      "loss": 0.3626,
+      "step": 2451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.456621999420369,
+      "learning_rate": 9.601216633953813e-06,
+      "loss": 0.3517,
+      "step": 2452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8859910782301066,
+      "learning_rate": 9.600817961103812e-06,
+      "loss": 0.383,
+      "step": 2453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.699870473647478,
+      "learning_rate": 9.600419097357703e-06,
+      "loss": 0.3199,
+      "step": 2454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5162567919215273,
+      "learning_rate": 9.600020042732032e-06,
+      "loss": 0.3282,
+      "step": 2455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7765206890967875,
+      "learning_rate": 9.599620797243361e-06,
+      "loss": 0.3416,
+      "step": 2456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.122703075758099,
+      "learning_rate": 9.599221360908252e-06,
+      "loss": 0.3719,
+      "step": 2457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4466795969818937,
+      "learning_rate": 9.59882173374328e-06,
+      "loss": 0.3518,
+      "step": 2458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2239881683850915,
+      "learning_rate": 9.598421915765026e-06,
+      "loss": 0.3514,
+      "step": 2459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8689568773882375,
+      "learning_rate": 9.598021906990079e-06,
+      "loss": 0.3669,
+      "step": 2460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9236755291344516,
+      "learning_rate": 9.597621707435036e-06,
+      "loss": 0.3709,
+      "step": 2461
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3783941980244863,
+      "learning_rate": 9.5972213171165e-06,
+      "loss": 0.3667,
+      "step": 2462
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2096218671045524,
+      "learning_rate": 9.59682073605109e-06,
+      "loss": 0.3389,
+      "step": 2463
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.333596723331247,
+      "learning_rate": 9.59641996425542e-06,
+      "loss": 0.3414,
+      "step": 2464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.812144092220335,
+      "learning_rate": 9.596019001746122e-06,
+      "loss": 0.3302,
+      "step": 2465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.86242633104935,
+      "learning_rate": 9.595617848539834e-06,
+      "loss": 0.3439,
+      "step": 2466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.066804945858839,
+      "learning_rate": 9.595216504653197e-06,
+      "loss": 0.347,
+      "step": 2467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.222549658608177,
+      "learning_rate": 9.594814970102865e-06,
+      "loss": 0.3337,
+      "step": 2468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1083775393708573,
+      "learning_rate": 9.594413244905499e-06,
+      "loss": 0.3313,
+      "step": 2469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.299993580756389,
+      "learning_rate": 9.594011329077765e-06,
+      "loss": 0.343,
+      "step": 2470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.6414229614092255,
+      "learning_rate": 9.593609222636344e-06,
+      "loss": 0.3643,
+      "step": 2471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.23002238737333,
+      "learning_rate": 9.593206925597916e-06,
+      "loss": 0.3566,
+      "step": 2472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3584540697125704,
+      "learning_rate": 9.592804437979175e-06,
+      "loss": 0.3536,
+      "step": 2473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5745857917975177,
+      "learning_rate": 9.592401759796818e-06,
+      "loss": 0.3418,
+      "step": 2474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.021634240799391,
+      "learning_rate": 9.591998891067558e-06,
+      "loss": 0.3403,
+      "step": 2475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.69197132094815,
+      "learning_rate": 9.591595831808105e-06,
+      "loss": 0.334,
+      "step": 2476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1549057972336096,
+      "learning_rate": 9.591192582035187e-06,
+      "loss": 0.346,
+      "step": 2477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1599262429004344,
+      "learning_rate": 9.590789141765534e-06,
+      "loss": 0.3469,
+      "step": 2478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8986599195289973,
+      "learning_rate": 9.590385511015885e-06,
+      "loss": 0.3643,
+      "step": 2479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.043228553739762,
+      "learning_rate": 9.589981689802988e-06,
+      "loss": 0.3316,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9831795768658242,
+      "learning_rate": 9.589577678143596e-06,
+      "loss": 0.3271,
+      "step": 2481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4642250151960425,
+      "learning_rate": 9.589173476054476e-06,
+      "loss": 0.3533,
+      "step": 2482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.002826439586185,
+      "learning_rate": 9.588769083552396e-06,
+      "loss": 0.3752,
+      "step": 2483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1579029014664104,
+      "learning_rate": 9.588364500654137e-06,
+      "loss": 0.3288,
+      "step": 2484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.905411452132041,
+      "learning_rate": 9.587959727376485e-06,
+      "loss": 0.4955,
+      "step": 2485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.367883538218573,
+      "learning_rate": 9.587554763736236e-06,
+      "loss": 0.3361,
+      "step": 2486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.728250185300285,
+      "learning_rate": 9.58714960975019e-06,
+      "loss": 0.3318,
+      "step": 2487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.701796077702341,
+      "learning_rate": 9.586744265435158e-06,
+      "loss": 0.3359,
+      "step": 2488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.930769574319089,
+      "learning_rate": 9.58633873080796e-06,
+      "loss": 0.3655,
+      "step": 2489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8947631935626261,
+      "learning_rate": 9.585933005885423e-06,
+      "loss": 0.3234,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8732111665478097,
+      "learning_rate": 9.58552709068438e-06,
+      "loss": 0.3421,
+      "step": 2491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5178262424855333,
+      "learning_rate": 9.585120985221672e-06,
+      "loss": 0.3552,
+      "step": 2492
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.138368178767361,
+      "learning_rate": 9.58471468951415e-06,
+      "loss": 0.373,
+      "step": 2493
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.180941829419746,
+      "learning_rate": 9.584308203578674e-06,
+      "loss": 0.3418,
+      "step": 2494
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.64143411198683,
+      "learning_rate": 9.583901527432106e-06,
+      "loss": 0.3507,
+      "step": 2495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0856897197685846,
+      "learning_rate": 9.583494661091324e-06,
+      "loss": 0.3384,
+      "step": 2496
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.774938131234399,
+      "learning_rate": 9.583087604573206e-06,
+      "loss": 0.3668,
+      "step": 2497
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7117203023785232,
+      "learning_rate": 9.582680357894643e-06,
+      "loss": 0.3745,
+      "step": 2498
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8687654925308135,
+      "learning_rate": 9.58227292107253e-06,
+      "loss": 0.3309,
+      "step": 2499
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0499200104652617,
+      "learning_rate": 9.581865294123778e-06,
+      "loss": 0.331,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4279129559113004,
+      "learning_rate": 9.581457477065294e-06,
+      "loss": 0.3434,
+      "step": 2501
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.934042091373923,
+      "learning_rate": 9.581049469914004e-06,
+      "loss": 0.3533,
+      "step": 2502
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9453778743274504,
+      "learning_rate": 9.580641272686833e-06,
+      "loss": 0.3558,
+      "step": 2503
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8543505270470636,
+      "learning_rate": 9.58023288540072e-06,
+      "loss": 0.5233,
+      "step": 2504
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2962050006236105,
+      "learning_rate": 9.57982430807261e-06,
+      "loss": 0.3444,
+      "step": 2505
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.39168303113425,
+      "learning_rate": 9.579415540719453e-06,
+      "loss": 0.36,
+      "step": 2506
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6872424447087488,
+      "learning_rate": 9.579006583358212e-06,
+      "loss": 0.3264,
+      "step": 2507
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.727876161396312,
+      "learning_rate": 9.578597436005854e-06,
+      "loss": 0.3498,
+      "step": 2508
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6006231337472312,
+      "learning_rate": 9.578188098679357e-06,
+      "loss": 0.3436,
+      "step": 2509
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2575156671701886,
+      "learning_rate": 9.577778571395704e-06,
+      "loss": 0.3469,
+      "step": 2510
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1869701324813695,
+      "learning_rate": 9.577368854171887e-06,
+      "loss": 0.365,
+      "step": 2511
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1311656418139058,
+      "learning_rate": 9.576958947024906e-06,
+      "loss": 0.347,
+      "step": 2512
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3141995316655706,
+      "learning_rate": 9.576548849971767e-06,
+      "loss": 0.3475,
+      "step": 2513
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3770366881392695,
+      "learning_rate": 9.576138563029489e-06,
+      "loss": 0.3466,
+      "step": 2514
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7840172310317712,
+      "learning_rate": 9.575728086215093e-06,
+      "loss": 0.3618,
+      "step": 2515
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.222578822306946,
+      "learning_rate": 9.575317419545611e-06,
+      "loss": 0.3115,
+      "step": 2516
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.216072415252696,
+      "learning_rate": 9.574906563038084e-06,
+      "loss": 0.3426,
+      "step": 2517
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4519382156869867,
+      "learning_rate": 9.574495516709557e-06,
+      "loss": 0.336,
+      "step": 2518
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.205717967051008,
+      "learning_rate": 9.574084280577085e-06,
+      "loss": 0.34,
+      "step": 2519
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9249991143022704,
+      "learning_rate": 9.573672854657734e-06,
+      "loss": 0.3676,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6796598181338873,
+      "learning_rate": 9.57326123896857e-06,
+      "loss": 0.3558,
+      "step": 2521
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9809568005933895,
+      "learning_rate": 9.572849433526677e-06,
+      "loss": 0.3659,
+      "step": 2522
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.563760658607657,
+      "learning_rate": 9.572437438349136e-06,
+      "loss": 0.3477,
+      "step": 2523
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9663139792349593,
+      "learning_rate": 9.572025253453045e-06,
+      "loss": 0.3672,
+      "step": 2524
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.103341722871928,
+      "learning_rate": 9.571612878855505e-06,
+      "loss": 0.339,
+      "step": 2525
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9121819361216263,
+      "learning_rate": 9.571200314573628e-06,
+      "loss": 0.3688,
+      "step": 2526
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0710557211190554,
+      "learning_rate": 9.57078756062453e-06,
+      "loss": 0.3456,
+      "step": 2527
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.092516506635837,
+      "learning_rate": 9.570374617025336e-06,
+      "loss": 0.3876,
+      "step": 2528
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5375152893594675,
+      "learning_rate": 9.569961483793183e-06,
+      "loss": 0.3618,
+      "step": 2529
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9619645359300384,
+      "learning_rate": 9.56954816094521e-06,
+      "loss": 0.3596,
+      "step": 2530
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.156338035702917,
+      "learning_rate": 9.569134648498568e-06,
+      "loss": 0.3652,
+      "step": 2531
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1336542173655793,
+      "learning_rate": 9.568720946470414e-06,
+      "loss": 0.3316,
+      "step": 2532
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9472305196009703,
+      "learning_rate": 9.568307054877911e-06,
+      "loss": 0.3587,
+      "step": 2533
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5438049737756938,
+      "learning_rate": 9.567892973738236e-06,
+      "loss": 0.3428,
+      "step": 2534
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.768489667835759,
+      "learning_rate": 9.567478703068567e-06,
+      "loss": 0.3454,
+      "step": 2535
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7685372092996738,
+      "learning_rate": 9.567064242886095e-06,
+      "loss": 0.3357,
+      "step": 2536
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.305059977430131,
+      "learning_rate": 9.566649593208015e-06,
+      "loss": 0.3534,
+      "step": 2537
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.666618246425068,
+      "learning_rate": 9.566234754051531e-06,
+      "loss": 0.3724,
+      "step": 2538
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.009465063885724,
+      "learning_rate": 9.565819725433857e-06,
+      "loss": 0.3502,
+      "step": 2539
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.48785389289798,
+      "learning_rate": 9.565404507372213e-06,
+      "loss": 0.33,
+      "step": 2540
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.241466167236867,
+      "learning_rate": 9.564989099883828e-06,
+      "loss": 0.3404,
+      "step": 2541
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.588467178710788,
+      "learning_rate": 9.564573502985936e-06,
+      "loss": 0.3305,
+      "step": 2542
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8390386030689703,
+      "learning_rate": 9.564157716695783e-06,
+      "loss": 0.3479,
+      "step": 2543
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7269386059475975,
+      "learning_rate": 9.563741741030616e-06,
+      "loss": 0.3628,
+      "step": 2544
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.156759376096128,
+      "learning_rate": 9.563325576007702e-06,
+      "loss": 0.3594,
+      "step": 2545
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6332586990458544,
+      "learning_rate": 9.562909221644303e-06,
+      "loss": 0.3368,
+      "step": 2546
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.44282817679179,
+      "learning_rate": 9.562492677957695e-06,
+      "loss": 0.3506,
+      "step": 2547
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.377567667525139,
+      "learning_rate": 9.562075944965163e-06,
+      "loss": 0.3544,
+      "step": 2548
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.488996642594074,
+      "learning_rate": 9.561659022683996e-06,
+      "loss": 0.3544,
+      "step": 2549
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.429938374982071,
+      "learning_rate": 9.561241911131494e-06,
+      "loss": 0.3443,
+      "step": 2550
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8974550576283076,
+      "learning_rate": 9.560824610324964e-06,
+      "loss": 0.3361,
+      "step": 2551
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1326873919114324,
+      "learning_rate": 9.560407120281718e-06,
+      "loss": 0.359,
+      "step": 2552
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.242227199761406,
+      "learning_rate": 9.559989441019081e-06,
+      "loss": 0.3543,
+      "step": 2553
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0913298144477195,
+      "learning_rate": 9.559571572554385e-06,
+      "loss": 0.3724,
+      "step": 2554
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.890813144751611,
+      "learning_rate": 9.559153514904964e-06,
+      "loss": 0.3664,
+      "step": 2555
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.425751552170047,
+      "learning_rate": 9.558735268088167e-06,
+      "loss": 0.3593,
+      "step": 2556
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7633058242306126,
+      "learning_rate": 9.558316832121346e-06,
+      "loss": 0.3535,
+      "step": 2557
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5866897233691692,
+      "learning_rate": 9.55789820702186e-06,
+      "loss": 0.3379,
+      "step": 2558
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.733393093973504,
+      "learning_rate": 9.557479392807085e-06,
+      "loss": 0.3447,
+      "step": 2559
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.073121911444511,
+      "learning_rate": 9.557060389494395e-06,
+      "loss": 0.3395,
+      "step": 2560
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 13.08060753241703,
+      "learning_rate": 9.556641197101173e-06,
+      "loss": 0.3654,
+      "step": 2561
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8624715183558385,
+      "learning_rate": 9.556221815644818e-06,
+      "loss": 0.3428,
+      "step": 2562
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.209289265708224,
+      "learning_rate": 9.555802245142724e-06,
+      "loss": 0.3388,
+      "step": 2563
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6628985536653489,
+      "learning_rate": 9.555382485612304e-06,
+      "loss": 0.3393,
+      "step": 2564
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5411508399725578,
+      "learning_rate": 9.554962537070973e-06,
+      "loss": 0.3606,
+      "step": 2565
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.727935169955974,
+      "learning_rate": 9.554542399536156e-06,
+      "loss": 0.3413,
+      "step": 2566
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3995116338867413,
+      "learning_rate": 9.554122073025284e-06,
+      "loss": 0.338,
+      "step": 2567
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.866370812255361,
+      "learning_rate": 9.5537015575558e-06,
+      "loss": 0.3582,
+      "step": 2568
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7772709331670047,
+      "learning_rate": 9.553280853145148e-06,
+      "loss": 0.3868,
+      "step": 2569
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.133986095637687,
+      "learning_rate": 9.552859959810787e-06,
+      "loss": 0.3481,
+      "step": 2570
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.649019599880264,
+      "learning_rate": 9.55243887757018e-06,
+      "loss": 0.3532,
+      "step": 2571
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4470511545477005,
+      "learning_rate": 9.552017606440798e-06,
+      "loss": 0.3649,
+      "step": 2572
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2246959892360683,
+      "learning_rate": 9.551596146440119e-06,
+      "loss": 0.3537,
+      "step": 2573
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2384490679719873,
+      "learning_rate": 9.551174497585632e-06,
+      "loss": 0.3638,
+      "step": 2574
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7628758447553016,
+      "learning_rate": 9.550752659894831e-06,
+      "loss": 0.4955,
+      "step": 2575
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6820409910663758,
+      "learning_rate": 9.550330633385218e-06,
+      "loss": 0.3521,
+      "step": 2576
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.116288676880126,
+      "learning_rate": 9.549908418074307e-06,
+      "loss": 0.3675,
+      "step": 2577
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8147406939372055,
+      "learning_rate": 9.549486013979614e-06,
+      "loss": 0.355,
+      "step": 2578
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.07027267824064,
+      "learning_rate": 9.549063421118664e-06,
+      "loss": 0.3554,
+      "step": 2579
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.179566494282651,
+      "learning_rate": 9.548640639508994e-06,
+      "loss": 0.3263,
+      "step": 2580
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.950898328802244,
+      "learning_rate": 9.548217669168144e-06,
+      "loss": 0.3655,
+      "step": 2581
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6431588628993141,
+      "learning_rate": 9.547794510113663e-06,
+      "loss": 0.3521,
+      "step": 2582
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.463704793114517,
+      "learning_rate": 9.547371162363112e-06,
+      "loss": 0.3729,
+      "step": 2583
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.530652400296561,
+      "learning_rate": 9.546947625934055e-06,
+      "loss": 0.3831,
+      "step": 2584
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4571569208762076,
+      "learning_rate": 9.546523900844063e-06,
+      "loss": 0.3703,
+      "step": 2585
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.049199584590343,
+      "learning_rate": 9.54609998711072e-06,
+      "loss": 0.3654,
+      "step": 2586
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.087287113882566,
+      "learning_rate": 9.545675884751611e-06,
+      "loss": 0.3658,
+      "step": 2587
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5926436517104823,
+      "learning_rate": 9.545251593784339e-06,
+      "loss": 0.3583,
+      "step": 2588
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2027156831411783,
+      "learning_rate": 9.544827114226502e-06,
+      "loss": 0.372,
+      "step": 2589
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8152647557611538,
+      "learning_rate": 9.544402446095718e-06,
+      "loss": 0.5014,
+      "step": 2590
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9393017128013679,
+      "learning_rate": 9.543977589409603e-06,
+      "loss": 0.3533,
+      "step": 2591
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3638423578958885,
+      "learning_rate": 9.543552544185788e-06,
+      "loss": 0.3458,
+      "step": 2592
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1573274962314226,
+      "learning_rate": 9.54312731044191e-06,
+      "loss": 0.3417,
+      "step": 2593
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.6799178455487676,
+      "learning_rate": 9.542701888195606e-06,
+      "loss": 0.3437,
+      "step": 2594
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6581331359518916,
+      "learning_rate": 9.542276277464534e-06,
+      "loss": 0.3349,
+      "step": 2595
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.012567633353498,
+      "learning_rate": 9.54185047826635e-06,
+      "loss": 0.3543,
+      "step": 2596
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7440730857683153,
+      "learning_rate": 9.541424490618724e-06,
+      "loss": 0.3475,
+      "step": 2597
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0455347129235357,
+      "learning_rate": 9.540998314539327e-06,
+      "loss": 0.3222,
+      "step": 2598
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5701933075545593,
+      "learning_rate": 9.540571950045847e-06,
+      "loss": 0.3627,
+      "step": 2599
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.105918678269197,
+      "learning_rate": 9.540145397155972e-06,
+      "loss": 0.3538,
+      "step": 2600
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1305296537632055,
+      "learning_rate": 9.539718655887398e-06,
+      "loss": 0.3482,
+      "step": 2601
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.3103267626157065,
+      "learning_rate": 9.539291726257835e-06,
+      "loss": 0.3614,
+      "step": 2602
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0582360901291676,
+      "learning_rate": 9.538864608284994e-06,
+      "loss": 0.3414,
+      "step": 2603
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8434645933427474,
+      "learning_rate": 9.5384373019866e-06,
+      "loss": 0.3451,
+      "step": 2604
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1976945757225415,
+      "learning_rate": 9.538009807380381e-06,
+      "loss": 0.3503,
+      "step": 2605
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7150175096102385,
+      "learning_rate": 9.537582124484074e-06,
+      "loss": 0.348,
+      "step": 2606
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6675833627117203,
+      "learning_rate": 9.537154253315426e-06,
+      "loss": 0.3355,
+      "step": 2607
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.916166557382315,
+      "learning_rate": 9.53672619389219e-06,
+      "loss": 0.3631,
+      "step": 2608
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8562654869946944,
+      "learning_rate": 9.536297946232124e-06,
+      "loss": 0.3449,
+      "step": 2609
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.148763267097097,
+      "learning_rate": 9.535869510353e-06,
+      "loss": 0.3324,
+      "step": 2610
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.035107214695197,
+      "learning_rate": 9.535440886272592e-06,
+      "loss": 0.3491,
+      "step": 2611
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.074690254764154,
+      "learning_rate": 9.535012074008688e-06,
+      "loss": 0.3475,
+      "step": 2612
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2345198467273626,
+      "learning_rate": 9.534583073579076e-06,
+      "loss": 0.3642,
+      "step": 2613
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.0103357340379375,
+      "learning_rate": 9.534153885001557e-06,
+      "loss": 0.3656,
+      "step": 2614
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4933549283783316,
+      "learning_rate": 9.533724508293942e-06,
+      "loss": 0.3364,
+      "step": 2615
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.846978499674889,
+      "learning_rate": 9.533294943474044e-06,
+      "loss": 0.3725,
+      "step": 2616
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.751904603471427,
+      "learning_rate": 9.532865190559686e-06,
+      "loss": 0.3667,
+      "step": 2617
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5832888425694582,
+      "learning_rate": 9.532435249568701e-06,
+      "loss": 0.3582,
+      "step": 2618
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8667365545365433,
+      "learning_rate": 9.532005120518927e-06,
+      "loss": 0.3301,
+      "step": 2619
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.050198970183284,
+      "learning_rate": 9.53157480342821e-06,
+      "loss": 0.328,
+      "step": 2620
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8338078372465996,
+      "learning_rate": 9.531144298314406e-06,
+      "loss": 0.3569,
+      "step": 2621
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0190805522209225,
+      "learning_rate": 9.530713605195377e-06,
+      "loss": 0.3611,
+      "step": 2622
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.744709556009685,
+      "learning_rate": 9.530282724088991e-06,
+      "loss": 0.3411,
+      "step": 2623
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7538256351208994,
+      "learning_rate": 9.52985165501313e-06,
+      "loss": 0.3534,
+      "step": 2624
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.853596636209183,
+      "learning_rate": 9.529420397985678e-06,
+      "loss": 0.3498,
+      "step": 2625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8416926794641617,
+      "learning_rate": 9.528988953024528e-06,
+      "loss": 0.5177,
+      "step": 2626
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.037586542004035,
+      "learning_rate": 9.528557320147583e-06,
+      "loss": 0.3797,
+      "step": 2627
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.91942641742022,
+      "learning_rate": 9.52812549937275e-06,
+      "loss": 0.3596,
+      "step": 2628
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.939099554594765,
+      "learning_rate": 9.527693490717946e-06,
+      "loss": 0.3371,
+      "step": 2629
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1548338000960006,
+      "learning_rate": 9.527261294201098e-06,
+      "loss": 0.3316,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.740173607989743,
+      "learning_rate": 9.52682890984014e-06,
+      "loss": 0.3512,
+      "step": 2631
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.152701064283289,
+      "learning_rate": 9.526396337653008e-06,
+      "loss": 0.3423,
+      "step": 2632
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.988591564166526,
+      "learning_rate": 9.525963577657651e-06,
+      "loss": 0.3387,
+      "step": 2633
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5276796406272264,
+      "learning_rate": 9.525530629872027e-06,
+      "loss": 0.3593,
+      "step": 2634
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.950624673873196,
+      "learning_rate": 9.5250974943141e-06,
+      "loss": 0.3651,
+      "step": 2635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.785631448046799,
+      "learning_rate": 9.524664171001839e-06,
+      "loss": 0.3434,
+      "step": 2636
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.6251489728593995,
+      "learning_rate": 9.524230659953227e-06,
+      "loss": 0.34,
+      "step": 2637
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6580689955724477,
+      "learning_rate": 9.523796961186247e-06,
+      "loss": 0.3601,
+      "step": 2638
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3618623929564544,
+      "learning_rate": 9.523363074718895e-06,
+      "loss": 0.3278,
+      "step": 2639
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.126048441695036,
+      "learning_rate": 9.522929000569177e-06,
+      "loss": 0.3733,
+      "step": 2640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.632442609393323,
+      "learning_rate": 9.522494738755099e-06,
+      "loss": 0.349,
+      "step": 2641
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.849686572228049,
+      "learning_rate": 9.522060289294683e-06,
+      "loss": 0.3543,
+      "step": 2642
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8878210352938096,
+      "learning_rate": 9.521625652205954e-06,
+      "loss": 0.3485,
+      "step": 2643
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7771626486961767,
+      "learning_rate": 9.521190827506944e-06,
+      "loss": 0.3576,
+      "step": 2644
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8613438572064482,
+      "learning_rate": 9.520755815215697e-06,
+      "loss": 0.3407,
+      "step": 2645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6765864240324504,
+      "learning_rate": 9.52032061535026e-06,
+      "loss": 0.3447,
+      "step": 2646
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5914867880010433,
+      "learning_rate": 9.519885227928693e-06,
+      "loss": 0.3449,
+      "step": 2647
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9024737081700733,
+      "learning_rate": 9.519449652969059e-06,
+      "loss": 0.3508,
+      "step": 2648
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8534869736610464,
+      "learning_rate": 9.519013890489433e-06,
+      "loss": 0.3361,
+      "step": 2649
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.054828248159896,
+      "learning_rate": 9.518577940507893e-06,
+      "loss": 0.3442,
+      "step": 2650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.422687525071993,
+      "learning_rate": 9.518141803042528e-06,
+      "loss": 0.3325,
+      "step": 2651
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.20343090953775,
+      "learning_rate": 9.517705478111434e-06,
+      "loss": 0.3451,
+      "step": 2652
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7751071814947603,
+      "learning_rate": 9.517268965732716e-06,
+      "loss": 0.3689,
+      "step": 2653
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3940516617122967,
+      "learning_rate": 9.516832265924485e-06,
+      "loss": 0.365,
+      "step": 2654
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.354668871640598,
+      "learning_rate": 9.516395378704862e-06,
+      "loss": 0.3412,
+      "step": 2655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.353812747703441,
+      "learning_rate": 9.51595830409197e-06,
+      "loss": 0.3643,
+      "step": 2656
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.076640095928003,
+      "learning_rate": 9.515521042103948e-06,
+      "loss": 0.3303,
+      "step": 2657
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.880974707584823,
+      "learning_rate": 9.515083592758939e-06,
+      "loss": 0.3404,
+      "step": 2658
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.592343016773818,
+      "learning_rate": 9.514645956075088e-06,
+      "loss": 0.3489,
+      "step": 2659
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3784402938981795,
+      "learning_rate": 9.514208132070561e-06,
+      "loss": 0.3714,
+      "step": 2660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4593005782718267,
+      "learning_rate": 9.513770120763518e-06,
+      "loss": 0.3369,
+      "step": 2661
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.361727502900006,
+      "learning_rate": 9.513331922172138e-06,
+      "loss": 0.3442,
+      "step": 2662
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.366551211810116,
+      "learning_rate": 9.512893536314598e-06,
+      "loss": 0.3334,
+      "step": 2663
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0336864622878514,
+      "learning_rate": 9.51245496320909e-06,
+      "loss": 0.3557,
+      "step": 2664
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3405514413183313,
+      "learning_rate": 9.51201620287381e-06,
+      "loss": 0.3605,
+      "step": 2665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6840921552846995,
+      "learning_rate": 9.511577255326965e-06,
+      "loss": 0.3413,
+      "step": 2666
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.749522367654022,
+      "learning_rate": 9.511138120586765e-06,
+      "loss": 0.3463,
+      "step": 2667
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6912371977457925,
+      "learning_rate": 9.510698798671432e-06,
+      "loss": 0.3545,
+      "step": 2668
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.436818083944917,
+      "learning_rate": 9.510259289599195e-06,
+      "loss": 0.3544,
+      "step": 2669
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.82972619419692,
+      "learning_rate": 9.509819593388288e-06,
+      "loss": 0.3525,
+      "step": 2670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.082100565723151,
+      "learning_rate": 9.509379710056954e-06,
+      "loss": 0.3475,
+      "step": 2671
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.864480199932987,
+      "learning_rate": 9.508939639623448e-06,
+      "loss": 0.5275,
+      "step": 2672
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9609948489310076,
+      "learning_rate": 9.508499382106028e-06,
+      "loss": 0.3582,
+      "step": 2673
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7104948444024886,
+      "learning_rate": 9.508058937522959e-06,
+      "loss": 0.3544,
+      "step": 2674
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8965084469653606,
+      "learning_rate": 9.50761830589252e-06,
+      "loss": 0.3704,
+      "step": 2675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.239225738722652,
+      "learning_rate": 9.507177487232989e-06,
+      "loss": 0.3597,
+      "step": 2676
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.542830523225034,
+      "learning_rate": 9.506736481562659e-06,
+      "loss": 0.3436,
+      "step": 2677
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.401912160955188,
+      "learning_rate": 9.506295288899827e-06,
+      "loss": 0.3635,
+      "step": 2678
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.600686461753694,
+      "learning_rate": 9.5058539092628e-06,
+      "loss": 0.3507,
+      "step": 2679
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6601039617645533,
+      "learning_rate": 9.505412342669891e-06,
+      "loss": 0.4829,
+      "step": 2680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.66433923637712,
+      "learning_rate": 9.504970589139422e-06,
+      "loss": 0.3541,
+      "step": 2681
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.709006116533518,
+      "learning_rate": 9.50452864868972e-06,
+      "loss": 0.3532,
+      "step": 2682
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.122040693803943,
+      "learning_rate": 9.504086521339124e-06,
+      "loss": 0.342,
+      "step": 2683
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.271705062000997,
+      "learning_rate": 9.50364420710598e-06,
+      "loss": 0.3539,
+      "step": 2684
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0816573387018766,
+      "learning_rate": 9.503201706008636e-06,
+      "loss": 0.3534,
+      "step": 2685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.8107531371915115,
+      "learning_rate": 9.502759018065455e-06,
+      "loss": 0.3736,
+      "step": 2686
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9865703063707727,
+      "learning_rate": 9.502316143294806e-06,
+      "loss": 0.3456,
+      "step": 2687
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.389427019947758,
+      "learning_rate": 9.501873081715062e-06,
+      "loss": 0.3439,
+      "step": 2688
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.81338473799617,
+      "learning_rate": 9.501429833344608e-06,
+      "loss": 0.3494,
+      "step": 2689
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.352215648096498,
+      "learning_rate": 9.500986398201835e-06,
+      "loss": 0.3599,
+      "step": 2690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.099168036639196,
+      "learning_rate": 9.50054277630514e-06,
+      "loss": 0.3332,
+      "step": 2691
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.577477281100561,
+      "learning_rate": 9.500098967672933e-06,
+      "loss": 0.3376,
+      "step": 2692
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.00425205734045,
+      "learning_rate": 9.499654972323627e-06,
+      "loss": 0.3238,
+      "step": 2693
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4047158583404316,
+      "learning_rate": 9.499210790275642e-06,
+      "loss": 0.3317,
+      "step": 2694
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.481222274156482,
+      "learning_rate": 9.498766421547412e-06,
+      "loss": 0.3344,
+      "step": 2695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.760444148294617,
+      "learning_rate": 9.498321866157372e-06,
+      "loss": 0.3257,
+      "step": 2696
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.342022172020348,
+      "learning_rate": 9.497877124123967e-06,
+      "loss": 0.3182,
+      "step": 2697
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1425706342728876,
+      "learning_rate": 9.497432195465652e-06,
+      "loss": 0.3347,
+      "step": 2698
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.284001068161913,
+      "learning_rate": 9.496987080200886e-06,
+      "loss": 0.3541,
+      "step": 2699
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1254002365975726,
+      "learning_rate": 9.496541778348138e-06,
+      "loss": 0.344,
+      "step": 2700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9013143896783475,
+      "learning_rate": 9.496096289925886e-06,
+      "loss": 0.3475,
+      "step": 2701
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3387279893590986,
+      "learning_rate": 9.495650614952612e-06,
+      "loss": 0.332,
+      "step": 2702
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.953452241927364,
+      "learning_rate": 9.495204753446809e-06,
+      "loss": 0.3452,
+      "step": 2703
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.111188700003503,
+      "learning_rate": 9.494758705426978e-06,
+      "loss": 0.3236,
+      "step": 2704
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.7961239694588595,
+      "learning_rate": 9.494312470911622e-06,
+      "loss": 0.3573,
+      "step": 2705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9370592498352033,
+      "learning_rate": 9.493866049919261e-06,
+      "loss": 0.3477,
+      "step": 2706
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.006058893233341,
+      "learning_rate": 9.493419442468414e-06,
+      "loss": 0.3378,
+      "step": 2707
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.65946956367097,
+      "learning_rate": 9.492972648577616e-06,
+      "loss": 0.3736,
+      "step": 2708
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.3844249311446575,
+      "learning_rate": 9.4925256682654e-06,
+      "loss": 0.3471,
+      "step": 2709
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.856291878012339,
+      "learning_rate": 9.492078501550314e-06,
+      "loss": 0.3516,
+      "step": 2710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.264358389602798,
+      "learning_rate": 9.491631148450914e-06,
+      "loss": 0.345,
+      "step": 2711
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.857814904464563,
+      "learning_rate": 9.491183608985757e-06,
+      "loss": 0.3792,
+      "step": 2712
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6986824116877774,
+      "learning_rate": 9.490735883173417e-06,
+      "loss": 0.3626,
+      "step": 2713
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7352081879720826,
+      "learning_rate": 9.490287971032468e-06,
+      "loss": 0.475,
+      "step": 2714
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1577051020112794,
+      "learning_rate": 9.489839872581497e-06,
+      "loss": 0.3453,
+      "step": 2715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2301244478097497,
+      "learning_rate": 9.489391587839091e-06,
+      "loss": 0.3643,
+      "step": 2716
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.677422097230705,
+      "learning_rate": 9.488943116823857e-06,
+      "loss": 0.3687,
+      "step": 2717
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.126018716941779,
+      "learning_rate": 9.488494459554401e-06,
+      "loss": 0.3197,
+      "step": 2718
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.7465849147696,
+      "learning_rate": 9.488045616049335e-06,
+      "loss": 0.3424,
+      "step": 2719
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.6874308654596755,
+      "learning_rate": 9.487596586327286e-06,
+      "loss": 0.3431,
+      "step": 2720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2281008504858373,
+      "learning_rate": 9.487147370406882e-06,
+      "loss": 0.3361,
+      "step": 2721
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6901945952652017,
+      "learning_rate": 9.486697968306764e-06,
+      "loss": 0.4635,
+      "step": 2722
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.953759134002954,
+      "learning_rate": 9.48624838004558e-06,
+      "loss": 0.3449,
+      "step": 2723
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.082152399780204,
+      "learning_rate": 9.48579860564198e-06,
+      "loss": 0.3783,
+      "step": 2724
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.703853281066354,
+      "learning_rate": 9.48534864511463e-06,
+      "loss": 0.3493,
+      "step": 2725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8544233592956116,
+      "learning_rate": 9.484898498482195e-06,
+      "loss": 0.3318,
+      "step": 2726
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9885102183229755,
+      "learning_rate": 9.484448165763358e-06,
+      "loss": 0.3502,
+      "step": 2727
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9964802042385448,
+      "learning_rate": 9.483997646976802e-06,
+      "loss": 0.342,
+      "step": 2728
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2126700481589343,
+      "learning_rate": 9.483546942141216e-06,
+      "loss": 0.3218,
+      "step": 2729
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.412135230470621,
+      "learning_rate": 9.483096051275305e-06,
+      "loss": 0.3721,
+      "step": 2730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8492547414066736,
+      "learning_rate": 9.482644974397776e-06,
+      "loss": 0.339,
+      "step": 2731
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.569498592151541,
+      "learning_rate": 9.482193711527345e-06,
+      "loss": 0.3215,
+      "step": 2732
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6170889960445285,
+      "learning_rate": 9.481742262682738e-06,
+      "loss": 0.3352,
+      "step": 2733
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.202934192988548,
+      "learning_rate": 9.48129062788268e-06,
+      "loss": 0.3705,
+      "step": 2734
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.605696492639625,
+      "learning_rate": 9.480838807145916e-06,
+      "loss": 0.3302,
+      "step": 2735
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.185066278772632,
+      "learning_rate": 9.480386800491192e-06,
+      "loss": 0.3633,
+      "step": 2736
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.228770585131004,
+      "learning_rate": 9.47993460793726e-06,
+      "loss": 0.3279,
+      "step": 2737
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.401999316003607,
+      "learning_rate": 9.479482229502886e-06,
+      "loss": 0.3476,
+      "step": 2738
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4559655797869313,
+      "learning_rate": 9.479029665206837e-06,
+      "loss": 0.3466,
+      "step": 2739
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3456006468757646,
+      "learning_rate": 9.47857691506789e-06,
+      "loss": 0.3439,
+      "step": 2740
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4428161643087463,
+      "learning_rate": 9.478123979104831e-06,
+      "loss": 0.3432,
+      "step": 2741
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9358941894391037,
+      "learning_rate": 9.477670857336458e-06,
+      "loss": 0.3302,
+      "step": 2742
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6408629330984094,
+      "learning_rate": 9.477217549781565e-06,
+      "loss": 0.512,
+      "step": 2743
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.42865217078952,
+      "learning_rate": 9.476764056458964e-06,
+      "loss": 0.3511,
+      "step": 2744
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.6952496177437375,
+      "learning_rate": 9.476310377387473e-06,
+      "loss": 0.3379,
+      "step": 2745
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2930089427411766,
+      "learning_rate": 9.47585651258591e-06,
+      "loss": 0.3298,
+      "step": 2746
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.833766862752183,
+      "learning_rate": 9.475402462073111e-06,
+      "loss": 0.3575,
+      "step": 2747
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.956849764735481,
+      "learning_rate": 9.474948225867916e-06,
+      "loss": 0.3262,
+      "step": 2748
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.730707992760973,
+      "learning_rate": 9.47449380398917e-06,
+      "loss": 0.3422,
+      "step": 2749
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2083692841088185,
+      "learning_rate": 9.474039196455729e-06,
+      "loss": 0.3314,
+      "step": 2750
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6240525352921156,
+      "learning_rate": 9.473584403286454e-06,
+      "loss": 0.4988,
+      "step": 2751
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.187405850798595,
+      "learning_rate": 9.473129424500218e-06,
+      "loss": 0.3522,
+      "step": 2752
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7106370564114695,
+      "learning_rate": 9.472674260115895e-06,
+      "loss": 0.3372,
+      "step": 2753
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.443339382493905,
+      "learning_rate": 9.472218910152374e-06,
+      "loss": 0.3364,
+      "step": 2754
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4018084146984333,
+      "learning_rate": 9.471763374628546e-06,
+      "loss": 0.3498,
+      "step": 2755
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7353046132565118,
+      "learning_rate": 9.471307653563313e-06,
+      "loss": 0.3442,
+      "step": 2756
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.083106695960372,
+      "learning_rate": 9.470851746975582e-06,
+      "loss": 0.341,
+      "step": 2757
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.237154015534022,
+      "learning_rate": 9.470395654884273e-06,
+      "loss": 0.3456,
+      "step": 2758
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.690322370540809,
+      "learning_rate": 9.469939377308308e-06,
+      "loss": 0.3349,
+      "step": 2759
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4584919329566106,
+      "learning_rate": 9.469482914266618e-06,
+      "loss": 0.3625,
+      "step": 2760
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.093674191456435,
+      "learning_rate": 9.469026265778142e-06,
+      "loss": 0.3294,
+      "step": 2761
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7750403189206274,
+      "learning_rate": 9.468569431861831e-06,
+      "loss": 0.3288,
+      "step": 2762
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7542207612319314,
+      "learning_rate": 9.468112412536636e-06,
+      "loss": 0.3534,
+      "step": 2763
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9043950895979895,
+      "learning_rate": 9.46765520782152e-06,
+      "loss": 0.3499,
+      "step": 2764
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.7052852220722436,
+      "learning_rate": 9.467197817735455e-06,
+      "loss": 0.3219,
+      "step": 2765
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.767424423941762,
+      "learning_rate": 9.466740242297418e-06,
+      "loss": 0.3533,
+      "step": 2766
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7762048598563815,
+      "learning_rate": 9.466282481526393e-06,
+      "loss": 0.3347,
+      "step": 2767
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2167111051395056,
+      "learning_rate": 9.465824535441377e-06,
+      "loss": 0.3383,
+      "step": 2768
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.126082207266447,
+      "learning_rate": 9.465366404061366e-06,
+      "loss": 0.3593,
+      "step": 2769
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.933674012815756,
+      "learning_rate": 9.464908087405374e-06,
+      "loss": 0.3248,
+      "step": 2770
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3307450476425386,
+      "learning_rate": 9.464449585492415e-06,
+      "loss": 0.3469,
+      "step": 2771
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.5991292608208285,
+      "learning_rate": 9.463990898341511e-06,
+      "loss": 0.3588,
+      "step": 2772
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0305170395619663,
+      "learning_rate": 9.463532025971696e-06,
+      "loss": 0.3189,
+      "step": 2773
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.243510270280099,
+      "learning_rate": 9.46307296840201e-06,
+      "loss": 0.319,
+      "step": 2774
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0991391084486075,
+      "learning_rate": 9.4626137256515e-06,
+      "loss": 0.3376,
+      "step": 2775
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.942842649520549,
+      "learning_rate": 9.46215429773922e-06,
+      "loss": 0.3388,
+      "step": 2776
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.977117474931626,
+      "learning_rate": 9.46169468468423e-06,
+      "loss": 0.3439,
+      "step": 2777
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.224291263022979,
+      "learning_rate": 9.461234886505604e-06,
+      "loss": 0.3355,
+      "step": 2778
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0031455465503547,
+      "learning_rate": 9.46077490322242e-06,
+      "loss": 0.3531,
+      "step": 2779
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.874284615010175,
+      "learning_rate": 9.46031473485376e-06,
+      "loss": 0.3409,
+      "step": 2780
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3548662817068347,
+      "learning_rate": 9.45985438141872e-06,
+      "loss": 0.3325,
+      "step": 2781
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.455369618580227,
+      "learning_rate": 9.4593938429364e-06,
+      "loss": 0.3349,
+      "step": 2782
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6616707836133435,
+      "learning_rate": 9.45893311942591e-06,
+      "loss": 0.3372,
+      "step": 2783
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.265891540437684,
+      "learning_rate": 9.458472210906363e-06,
+      "loss": 0.3339,
+      "step": 2784
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.51312814291431,
+      "learning_rate": 9.458011117396886e-06,
+      "loss": 0.3346,
+      "step": 2785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6527253674339057,
+      "learning_rate": 9.45754983891661e-06,
+      "loss": 0.332,
+      "step": 2786
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.38085799438359,
+      "learning_rate": 9.457088375484671e-06,
+      "loss": 0.3386,
+      "step": 2787
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1564390974614422,
+      "learning_rate": 9.45662672712022e-06,
+      "loss": 0.3395,
+      "step": 2788
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.613545889420841,
+      "learning_rate": 9.456164893842411e-06,
+      "loss": 0.3485,
+      "step": 2789
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3711434342955076,
+      "learning_rate": 9.455702875670405e-06,
+      "loss": 0.3349,
+      "step": 2790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.663328043729727,
+      "learning_rate": 9.455240672623373e-06,
+      "loss": 0.3498,
+      "step": 2791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.981069481642985,
+      "learning_rate": 9.454778284720492e-06,
+      "loss": 0.3459,
+      "step": 2792
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.414719483762963,
+      "learning_rate": 9.454315711980947e-06,
+      "loss": 0.342,
+      "step": 2793
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8841543131580942,
+      "learning_rate": 9.453852954423931e-06,
+      "loss": 0.3418,
+      "step": 2794
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0040390069744674,
+      "learning_rate": 9.453390012068645e-06,
+      "loss": 0.3222,
+      "step": 2795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.13952994394859,
+      "learning_rate": 9.452926884934299e-06,
+      "loss": 0.3434,
+      "step": 2796
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.353655141473368,
+      "learning_rate": 9.452463573040105e-06,
+      "loss": 0.3338,
+      "step": 2797
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9606795539717619,
+      "learning_rate": 9.45200007640529e-06,
+      "loss": 0.3514,
+      "step": 2798
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.886733353083376,
+      "learning_rate": 9.451536395049083e-06,
+      "loss": 0.3471,
+      "step": 2799
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4776331690200872,
+      "learning_rate": 9.451072528990726e-06,
+      "loss": 0.3449,
+      "step": 2800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.722283122940009,
+      "learning_rate": 9.450608478249463e-06,
+      "loss": 0.3412,
+      "step": 2801
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.666023026044664,
+      "learning_rate": 9.45014424284455e-06,
+      "loss": 0.3582,
+      "step": 2802
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2828030174161635,
+      "learning_rate": 9.449679822795246e-06,
+      "loss": 0.3409,
+      "step": 2803
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1537365589222173,
+      "learning_rate": 9.449215218120823e-06,
+      "loss": 0.3452,
+      "step": 2804
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.648680999644945,
+      "learning_rate": 9.448750428840558e-06,
+      "loss": 0.3366,
+      "step": 2805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2211160690738936,
+      "learning_rate": 9.448285454973739e-06,
+      "loss": 0.3302,
+      "step": 2806
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5726546881622903,
+      "learning_rate": 9.447820296539651e-06,
+      "loss": 0.3401,
+      "step": 2807
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4645098836059254,
+      "learning_rate": 9.447354953557601e-06,
+      "loss": 0.3373,
+      "step": 2808
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.460336111103413,
+      "learning_rate": 9.446889426046894e-06,
+      "loss": 0.3601,
+      "step": 2809
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4755813323162292,
+      "learning_rate": 9.446423714026846e-06,
+      "loss": 0.3774,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.121433864366657,
+      "learning_rate": 9.44595781751678e-06,
+      "loss": 0.3458,
+      "step": 2811
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.359935062868799,
+      "learning_rate": 9.445491736536027e-06,
+      "loss": 0.3517,
+      "step": 2812
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8225412692039424,
+      "learning_rate": 9.445025471103928e-06,
+      "loss": 0.3444,
+      "step": 2813
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3172362395571358,
+      "learning_rate": 9.444559021239826e-06,
+      "loss": 0.3294,
+      "step": 2814
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2337656213902073,
+      "learning_rate": 9.444092386963075e-06,
+      "loss": 0.3488,
+      "step": 2815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6685777243671267,
+      "learning_rate": 9.443625568293038e-06,
+      "loss": 0.3562,
+      "step": 2816
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0430253762704766,
+      "learning_rate": 9.443158565249082e-06,
+      "loss": 0.3217,
+      "step": 2817
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.4861147230698055,
+      "learning_rate": 9.442691377850585e-06,
+      "loss": 0.3563,
+      "step": 2818
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2111236880069094,
+      "learning_rate": 9.442224006116935e-06,
+      "loss": 0.3338,
+      "step": 2819
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.163038200171697,
+      "learning_rate": 9.441756450067519e-06,
+      "loss": 0.3481,
+      "step": 2820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.91268043030134,
+      "learning_rate": 9.441288709721737e-06,
+      "loss": 0.3443,
+      "step": 2821
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.726556643663211,
+      "learning_rate": 9.440820785098998e-06,
+      "loss": 0.3482,
+      "step": 2822
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2242300931295373,
+      "learning_rate": 9.440352676218718e-06,
+      "loss": 0.3519,
+      "step": 2823
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.962656578743003,
+      "learning_rate": 9.439884383100319e-06,
+      "loss": 0.3454,
+      "step": 2824
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.4737964293811485,
+      "learning_rate": 9.43941590576323e-06,
+      "loss": 0.3574,
+      "step": 2825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1128971723349954,
+      "learning_rate": 9.438947244226889e-06,
+      "loss": 0.3269,
+      "step": 2826
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.592958046916075,
+      "learning_rate": 9.438478398510744e-06,
+      "loss": 0.3734,
+      "step": 2827
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5672723601376353,
+      "learning_rate": 9.438009368634244e-06,
+      "loss": 0.3451,
+      "step": 2828
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.442651213825842,
+      "learning_rate": 9.437540154616856e-06,
+      "loss": 0.3383,
+      "step": 2829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6106401136534685,
+      "learning_rate": 9.437070756478043e-06,
+      "loss": 0.3195,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.27165960013486,
+      "learning_rate": 9.436601174237283e-06,
+      "loss": 0.3486,
+      "step": 2831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.241876553490455,
+      "learning_rate": 9.43613140791406e-06,
+      "loss": 0.3774,
+      "step": 2832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.192553518065512,
+      "learning_rate": 9.435661457527867e-06,
+      "loss": 0.3465,
+      "step": 2833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.755475426359888,
+      "learning_rate": 9.435191323098201e-06,
+      "loss": 0.3265,
+      "step": 2834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.520657051379427,
+      "learning_rate": 9.434721004644567e-06,
+      "loss": 0.3434,
+      "step": 2835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.363693194886535,
+      "learning_rate": 9.434250502186483e-06,
+      "loss": 0.3301,
+      "step": 2836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.654449150747247,
+      "learning_rate": 9.43377981574347e-06,
+      "loss": 0.3455,
+      "step": 2837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0885392601630985,
+      "learning_rate": 9.433308945335058e-06,
+      "loss": 0.3423,
+      "step": 2838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.274065645708655,
+      "learning_rate": 9.432837890980781e-06,
+      "loss": 0.3465,
+      "step": 2839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.544574334105781,
+      "learning_rate": 9.432366652700189e-06,
+      "loss": 0.3329,
+      "step": 2840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8224492961804457,
+      "learning_rate": 9.43189523051283e-06,
+      "loss": 0.3466,
+      "step": 2841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.262078457758155,
+      "learning_rate": 9.431423624438264e-06,
+      "loss": 0.3265,
+      "step": 2842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.0143890497233015,
+      "learning_rate": 9.430951834496064e-06,
+      "loss": 0.3302,
+      "step": 2843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.49727120577982,
+      "learning_rate": 9.430479860705802e-06,
+      "loss": 0.3324,
+      "step": 2844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1745162216049105,
+      "learning_rate": 9.43000770308706e-06,
+      "loss": 0.3195,
+      "step": 2845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.5148866182805305,
+      "learning_rate": 9.42953536165943e-06,
+      "loss": 0.3487,
+      "step": 2846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0211553893290013,
+      "learning_rate": 9.429062836442512e-06,
+      "loss": 0.3315,
+      "step": 2847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.42692257157526,
+      "learning_rate": 9.428590127455908e-06,
+      "loss": 0.3252,
+      "step": 2848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.06852227211664,
+      "learning_rate": 9.428117234719234e-06,
+      "loss": 0.3433,
+      "step": 2849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9072314545965625,
+      "learning_rate": 9.427644158252109e-06,
+      "loss": 0.3688,
+      "step": 2850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.852432088608627,
+      "learning_rate": 9.427170898074166e-06,
+      "loss": 0.3392,
+      "step": 2851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7173015240382314,
+      "learning_rate": 9.426697454205039e-06,
+      "loss": 0.3377,
+      "step": 2852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 31.45710822411603,
+      "learning_rate": 9.42622382666437e-06,
+      "loss": 0.3548,
+      "step": 2853
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.633232496009611,
+      "learning_rate": 9.425750015471813e-06,
+      "loss": 0.3437,
+      "step": 2854
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.542315376695413,
+      "learning_rate": 9.425276020647026e-06,
+      "loss": 0.3293,
+      "step": 2855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.719098152199475,
+      "learning_rate": 9.42480184220968e-06,
+      "loss": 0.3383,
+      "step": 2856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.007903046452314,
+      "learning_rate": 9.424327480179443e-06,
+      "loss": 0.3383,
+      "step": 2857
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6998075779836991,
+      "learning_rate": 9.423852934576003e-06,
+      "loss": 0.5152,
+      "step": 2858
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.315523560050483,
+      "learning_rate": 9.423378205419044e-06,
+      "loss": 0.3383,
+      "step": 2859
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.80912279664292,
+      "learning_rate": 9.42290329272827e-06,
+      "loss": 0.351,
+      "step": 2860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.657376800209197,
+      "learning_rate": 9.422428196523382e-06,
+      "loss": 0.3621,
+      "step": 2861
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3861887157105075,
+      "learning_rate": 9.42195291682409e-06,
+      "loss": 0.3439,
+      "step": 2862
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.104540718554345,
+      "learning_rate": 9.421477453650118e-06,
+      "loss": 0.3406,
+      "step": 2863
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.9449944011930014,
+      "learning_rate": 9.421001807021196e-06,
+      "loss": 0.3325,
+      "step": 2864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.203217123864811,
+      "learning_rate": 9.420525976957053e-06,
+      "loss": 0.3271,
+      "step": 2865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5999976239510811,
+      "learning_rate": 9.420049963477437e-06,
+      "loss": 0.4892,
+      "step": 2866
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.444723589831954,
+      "learning_rate": 9.419573766602097e-06,
+      "loss": 0.3451,
+      "step": 2867
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.796242662466096,
+      "learning_rate": 9.41909738635079e-06,
+      "loss": 0.3507,
+      "step": 2868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.75746100509902,
+      "learning_rate": 9.418620822743284e-06,
+      "loss": 0.3328,
+      "step": 2869
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.923284656401906,
+      "learning_rate": 9.418144075799353e-06,
+      "loss": 0.345,
+      "step": 2870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.192724605693848,
+      "learning_rate": 9.417667145538778e-06,
+      "loss": 0.3493,
+      "step": 2871
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.947398314030883,
+      "learning_rate": 9.417190031981343e-06,
+      "loss": 0.3264,
+      "step": 2872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.4545985955900775,
+      "learning_rate": 9.41671273514685e-06,
+      "loss": 0.3547,
+      "step": 2873
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.281984482714142,
+      "learning_rate": 9.416235255055099e-06,
+      "loss": 0.3592,
+      "step": 2874
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.176301761464211,
+      "learning_rate": 9.415757591725905e-06,
+      "loss": 0.366,
+      "step": 2875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.554313991084236,
+      "learning_rate": 9.415279745179084e-06,
+      "loss": 0.3332,
+      "step": 2876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.742543879468014,
+      "learning_rate": 9.414801715434464e-06,
+      "loss": 0.3368,
+      "step": 2877
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.879691455524878,
+      "learning_rate": 9.41432350251188e-06,
+      "loss": 0.3397,
+      "step": 2878
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.792333491669887,
+      "learning_rate": 9.413845106431171e-06,
+      "loss": 0.3477,
+      "step": 2879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.6793008193667434,
+      "learning_rate": 9.41336652721219e-06,
+      "loss": 0.3394,
+      "step": 2880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.5621857549984695,
+      "learning_rate": 9.412887764874794e-06,
+      "loss": 0.3129,
+      "step": 2881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3834765786758307,
+      "learning_rate": 9.412408819438847e-06,
+      "loss": 0.3301,
+      "step": 2882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.512591564980338,
+      "learning_rate": 9.41192969092422e-06,
+      "loss": 0.3561,
+      "step": 2883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4747121695984333,
+      "learning_rate": 9.411450379350791e-06,
+      "loss": 0.3456,
+      "step": 2884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5570404248386294,
+      "learning_rate": 9.410970884738453e-06,
+      "loss": 0.3432,
+      "step": 2885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.039301780760711,
+      "learning_rate": 9.410491207107097e-06,
+      "loss": 0.3325,
+      "step": 2886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0699921469434797,
+      "learning_rate": 9.410011346476628e-06,
+      "loss": 0.3136,
+      "step": 2887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.257155959011966,
+      "learning_rate": 9.409531302866954e-06,
+      "loss": 0.3161,
+      "step": 2888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4186503244613218,
+      "learning_rate": 9.409051076297994e-06,
+      "loss": 0.3277,
+      "step": 2889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.690839806498641,
+      "learning_rate": 9.408570666789674e-06,
+      "loss": 0.3273,
+      "step": 2890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5094390086689167,
+      "learning_rate": 9.408090074361927e-06,
+      "loss": 0.3208,
+      "step": 2891
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 26.004381051412743,
+      "learning_rate": 9.407609299034693e-06,
+      "loss": 0.3309,
+      "step": 2892
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.498975777808764,
+      "learning_rate": 9.40712834082792e-06,
+      "loss": 0.3423,
+      "step": 2893
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.55535006750138,
+      "learning_rate": 9.406647199761564e-06,
+      "loss": 0.3042,
+      "step": 2894
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.251041822218146,
+      "learning_rate": 9.406165875855589e-06,
+      "loss": 0.3514,
+      "step": 2895
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6971310108682597,
+      "learning_rate": 9.405684369129965e-06,
+      "loss": 0.5189,
+      "step": 2896
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.029326151582228,
+      "learning_rate": 9.40520267960467e-06,
+      "loss": 0.3469,
+      "step": 2897
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0755822619381115,
+      "learning_rate": 9.404720807299694e-06,
+      "loss": 0.3337,
+      "step": 2898
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.829548529739491,
+      "learning_rate": 9.404238752235028e-06,
+      "loss": 0.3392,
+      "step": 2899
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.697211143258505,
+      "learning_rate": 9.40375651443067e-06,
+      "loss": 0.3278,
+      "step": 2900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.828282563810434,
+      "learning_rate": 9.403274093906635e-06,
+      "loss": 0.3295,
+      "step": 2901
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.471500784682395,
+      "learning_rate": 9.402791490682938e-06,
+      "loss": 0.3103,
+      "step": 2902
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.8394804349192406,
+      "learning_rate": 9.4023087047796e-06,
+      "loss": 0.3445,
+      "step": 2903
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1678344497176716,
+      "learning_rate": 9.401825736216654e-06,
+      "loss": 0.3342,
+      "step": 2904
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.012326991399161,
+      "learning_rate": 9.401342585014141e-06,
+      "loss": 0.319,
+      "step": 2905
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0374766040889085,
+      "learning_rate": 9.400859251192104e-06,
+      "loss": 0.3354,
+      "step": 2906
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.121481576580878,
+      "learning_rate": 9.400375734770603e-06,
+      "loss": 0.3411,
+      "step": 2907
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.287520775902825,
+      "learning_rate": 9.399892035769696e-06,
+      "loss": 0.3343,
+      "step": 2908
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.992660820053152,
+      "learning_rate": 9.399408154209452e-06,
+      "loss": 0.3436,
+      "step": 2909
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.734115944816152,
+      "learning_rate": 9.39892409010995e-06,
+      "loss": 0.3756,
+      "step": 2910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.892929914048489,
+      "learning_rate": 9.398439843491273e-06,
+      "loss": 0.3462,
+      "step": 2911
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.99508133042363,
+      "learning_rate": 9.397955414373518e-06,
+      "loss": 0.331,
+      "step": 2912
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.587144228356296,
+      "learning_rate": 9.397470802776777e-06,
+      "loss": 0.3445,
+      "step": 2913
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.323607839701405,
+      "learning_rate": 9.396986008721165e-06,
+      "loss": 0.3257,
+      "step": 2914
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.71452420951083,
+      "learning_rate": 9.396501032226793e-06,
+      "loss": 0.3573,
+      "step": 2915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.412411037982219,
+      "learning_rate": 9.396015873313781e-06,
+      "loss": 0.361,
+      "step": 2916
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.392461801012857,
+      "learning_rate": 9.395530532002265e-06,
+      "loss": 0.3419,
+      "step": 2917
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.1243772104218155,
+      "learning_rate": 9.39504500831238e-06,
+      "loss": 0.333,
+      "step": 2918
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.900944207336713,
+      "learning_rate": 9.39455930226427e-06,
+      "loss": 0.3153,
+      "step": 2919
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4305900779326794,
+      "learning_rate": 9.394073413878089e-06,
+      "loss": 0.3236,
+      "step": 2920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.136392600562325,
+      "learning_rate": 9.393587343173998e-06,
+      "loss": 0.3565,
+      "step": 2921
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7350195142850207,
+      "learning_rate": 9.393101090172164e-06,
+      "loss": 0.327,
+      "step": 2922
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.63063528563335,
+      "learning_rate": 9.392614654892761e-06,
+      "loss": 0.3185,
+      "step": 2923
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.539628583735849,
+      "learning_rate": 9.392128037355977e-06,
+      "loss": 0.317,
+      "step": 2924
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.74064779015083,
+      "learning_rate": 9.391641237581998e-06,
+      "loss": 0.3446,
+      "step": 2925
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6063991643633355,
+      "learning_rate": 9.391154255591025e-06,
+      "loss": 0.3148,
+      "step": 2926
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.9856885299036335,
+      "learning_rate": 9.39066709140326e-06,
+      "loss": 0.3384,
+      "step": 2927
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.995242282639124,
+      "learning_rate": 9.39017974503892e-06,
+      "loss": 0.3352,
+      "step": 2928
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4789716544967804,
+      "learning_rate": 9.389692216518224e-06,
+      "loss": 0.3461,
+      "step": 2929
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.6322536561450733,
+      "learning_rate": 9.3892045058614e-06,
+      "loss": 0.3515,
+      "step": 2930
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.069745363414444,
+      "learning_rate": 9.388716613088686e-06,
+      "loss": 0.3188,
+      "step": 2931
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8862009541003968,
+      "learning_rate": 9.388228538220326e-06,
+      "loss": 0.3243,
+      "step": 2932
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.801954251400872,
+      "learning_rate": 9.387740281276568e-06,
+      "loss": 0.3479,
+      "step": 2933
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.189852651221901,
+      "learning_rate": 9.387251842277672e-06,
+      "loss": 0.3114,
+      "step": 2934
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.32992466654144,
+      "learning_rate": 9.386763221243905e-06,
+      "loss": 0.346,
+      "step": 2935
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4940048590995523,
+      "learning_rate": 9.38627441819554e-06,
+      "loss": 0.3467,
+      "step": 2936
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9183728794530075,
+      "learning_rate": 9.385785433152857e-06,
+      "loss": 0.5135,
+      "step": 2937
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.947386359204182,
+      "learning_rate": 9.385296266136148e-06,
+      "loss": 0.3439,
+      "step": 2938
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.008728640681746,
+      "learning_rate": 9.384806917165707e-06,
+      "loss": 0.3548,
+      "step": 2939
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5869169313035396,
+      "learning_rate": 9.384317386261841e-06,
+      "loss": 0.3093,
+      "step": 2940
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7542506859998492,
+      "learning_rate": 9.383827673444856e-06,
+      "loss": 0.3358,
+      "step": 2941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8796394186241903,
+      "learning_rate": 9.383337778735076e-06,
+      "loss": 0.3122,
+      "step": 2942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.122534408531552,
+      "learning_rate": 9.382847702152827e-06,
+      "loss": 0.3351,
+      "step": 2943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.567870346776088,
+      "learning_rate": 9.382357443718439e-06,
+      "loss": 0.3158,
+      "step": 2944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.072062774682705,
+      "learning_rate": 9.381867003452258e-06,
+      "loss": 0.3413,
+      "step": 2945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.79819236834924,
+      "learning_rate": 9.381376381374634e-06,
+      "loss": 0.3347,
+      "step": 2946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8536223395008538,
+      "learning_rate": 9.380885577505919e-06,
+      "loss": 0.3458,
+      "step": 2947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.572644934308234,
+      "learning_rate": 9.380394591866482e-06,
+      "loss": 0.3623,
+      "step": 2948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4068019338376945,
+      "learning_rate": 9.379903424476692e-06,
+      "loss": 0.3634,
+      "step": 2949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3813495653493635,
+      "learning_rate": 9.379412075356928e-06,
+      "loss": 0.3271,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.677791592204469,
+      "learning_rate": 9.37892054452758e-06,
+      "loss": 0.3262,
+      "step": 2951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9352185064751115,
+      "learning_rate": 9.37842883200904e-06,
+      "loss": 0.3411,
+      "step": 2952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0548663831312037,
+      "learning_rate": 9.377936937821712e-06,
+      "loss": 0.336,
+      "step": 2953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5100520512670617,
+      "learning_rate": 9.377444861986004e-06,
+      "loss": 0.3397,
+      "step": 2954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.136633990219711,
+      "learning_rate": 9.376952604522333e-06,
+      "loss": 0.3238,
+      "step": 2955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 18.523231953088995,
+      "learning_rate": 9.376460165451125e-06,
+      "loss": 0.3339,
+      "step": 2956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7411311506306433,
+      "learning_rate": 9.37596754479281e-06,
+      "loss": 0.3264,
+      "step": 2957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8916559043431933,
+      "learning_rate": 9.37547474256783e-06,
+      "loss": 0.3646,
+      "step": 2958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.2563683169537105,
+      "learning_rate": 9.37498175879663e-06,
+      "loss": 0.3352,
+      "step": 2959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.762369507271014,
+      "learning_rate": 9.374488593499666e-06,
+      "loss": 0.3786,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.92035920480478,
+      "learning_rate": 9.3739952466974e-06,
+      "loss": 0.3291,
+      "step": 2961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.137139730565293,
+      "learning_rate": 9.373501718410303e-06,
+      "loss": 0.3303,
+      "step": 2962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.367799803671223,
+      "learning_rate": 9.373008008658852e-06,
+      "loss": 0.3405,
+      "step": 2963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.0506847971138855,
+      "learning_rate": 9.372514117463531e-06,
+      "loss": 0.3706,
+      "step": 2964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.00407829290628,
+      "learning_rate": 9.372020044844832e-06,
+      "loss": 0.3437,
+      "step": 2965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7504908868335254,
+      "learning_rate": 9.371525790823256e-06,
+      "loss": 0.3101,
+      "step": 2966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.506023327800912,
+      "learning_rate": 9.371031355419311e-06,
+      "loss": 0.3482,
+      "step": 2967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.564355648983167,
+      "learning_rate": 9.370536738653513e-06,
+      "loss": 0.3386,
+      "step": 2968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 38.763433171693926,
+      "learning_rate": 9.37004194054638e-06,
+      "loss": 0.3493,
+      "step": 2969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.712320349004185,
+      "learning_rate": 9.369546961118446e-06,
+      "loss": 0.3124,
+      "step": 2970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9352975918037343,
+      "learning_rate": 9.369051800390245e-06,
+      "loss": 0.345,
+      "step": 2971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.836155204661295,
+      "learning_rate": 9.368556458382327e-06,
+      "loss": 0.3134,
+      "step": 2972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.151021719818088,
+      "learning_rate": 9.368060935115243e-06,
+      "loss": 0.3484,
+      "step": 2973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6961994500940625,
+      "learning_rate": 9.36756523060955e-06,
+      "loss": 0.3389,
+      "step": 2974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.963762113080279,
+      "learning_rate": 9.367069344885818e-06,
+      "loss": 0.3335,
+      "step": 2975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.356519212038357,
+      "learning_rate": 9.366573277964623e-06,
+      "loss": 0.3423,
+      "step": 2976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.219726325252524,
+      "learning_rate": 9.366077029866546e-06,
+      "loss": 0.3258,
+      "step": 2977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.647013140102176,
+      "learning_rate": 9.365580600612178e-06,
+      "loss": 0.3241,
+      "step": 2978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7452563532522922,
+      "learning_rate": 9.365083990222118e-06,
+      "loss": 0.3339,
+      "step": 2979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.204300651698719,
+      "learning_rate": 9.36458719871697e-06,
+      "loss": 0.3233,
+      "step": 2980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.127717263896908,
+      "learning_rate": 9.364090226117346e-06,
+      "loss": 0.3514,
+      "step": 2981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.179443951254958,
+      "learning_rate": 9.363593072443865e-06,
+      "loss": 0.353,
+      "step": 2982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.417908955575161,
+      "learning_rate": 9.36309573771716e-06,
+      "loss": 0.3516,
+      "step": 2983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9608663473809073,
+      "learning_rate": 9.362598221957862e-06,
+      "loss": 0.49,
+      "step": 2984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.707532724763154,
+      "learning_rate": 9.362100525186616e-06,
+      "loss": 0.3401,
+      "step": 2985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.223660909851831,
+      "learning_rate": 9.361602647424069e-06,
+      "loss": 0.3273,
+      "step": 2986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.732123712149332,
+      "learning_rate": 9.361104588690882e-06,
+      "loss": 0.3412,
+      "step": 2987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3554936274931797,
+      "learning_rate": 9.36060634900772e-06,
+      "loss": 0.3259,
+      "step": 2988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.429438026735335,
+      "learning_rate": 9.360107928395254e-06,
+      "loss": 0.3459,
+      "step": 2989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4984931577952993,
+      "learning_rate": 9.359609326874169e-06,
+      "loss": 0.3276,
+      "step": 2990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.173541636929663,
+      "learning_rate": 9.359110544465146e-06,
+      "loss": 0.3234,
+      "step": 2991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5013455005168916,
+      "learning_rate": 9.358611581188884e-06,
+      "loss": 0.3576,
+      "step": 2992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.090817040137613,
+      "learning_rate": 9.358112437066088e-06,
+      "loss": 0.3734,
+      "step": 2993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7682740799018313,
+      "learning_rate": 9.357613112117464e-06,
+      "loss": 0.3278,
+      "step": 2994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9360354284112907,
+      "learning_rate": 9.357113606363732e-06,
+      "loss": 0.487,
+      "step": 2995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3217461974291056,
+      "learning_rate": 9.356613919825619e-06,
+      "loss": 0.3453,
+      "step": 2996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8950696661048922,
+      "learning_rate": 9.356114052523854e-06,
+      "loss": 0.3449,
+      "step": 2997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0136953082508855,
+      "learning_rate": 9.355614004479182e-06,
+      "loss": 0.3282,
+      "step": 2998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.2662429306810665,
+      "learning_rate": 9.355113775712348e-06,
+      "loss": 0.3294,
+      "step": 2999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.899287811153719,
+      "learning_rate": 9.354613366244108e-06,
+      "loss": 0.3294,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.660311033876325,
+      "learning_rate": 9.354112776095224e-06,
+      "loss": 0.3208,
+      "step": 3001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.027784651214224,
+      "learning_rate": 9.353612005286467e-06,
+      "loss": 0.332,
+      "step": 3002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.437953303219593,
+      "learning_rate": 9.353111053838616e-06,
+      "loss": 0.3424,
+      "step": 3003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.37422599106699,
+      "learning_rate": 9.352609921772453e-06,
+      "loss": 0.3381,
+      "step": 3004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.086888289084342,
+      "learning_rate": 9.352108609108775e-06,
+      "loss": 0.3681,
+      "step": 3005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.817461296972661,
+      "learning_rate": 9.351607115868379e-06,
+      "loss": 0.3162,
+      "step": 3006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6842867164610493,
+      "learning_rate": 9.351105442072077e-06,
+      "loss": 0.3308,
+      "step": 3007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.074557487684647,
+      "learning_rate": 9.350603587740681e-06,
+      "loss": 0.3518,
+      "step": 3008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.494142419307336,
+      "learning_rate": 9.350101552895015e-06,
+      "loss": 0.336,
+      "step": 3009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.646759380098731,
+      "learning_rate": 9.349599337555908e-06,
+      "loss": 0.3509,
+      "step": 3010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.209083844362035,
+      "learning_rate": 9.3490969417442e-06,
+      "loss": 0.3559,
+      "step": 3011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8798719763971381,
+      "learning_rate": 9.348594365480731e-06,
+      "loss": 0.5078,
+      "step": 3012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.850350505727774,
+      "learning_rate": 9.348091608786362e-06,
+      "loss": 0.3135,
+      "step": 3013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3426099933361617,
+      "learning_rate": 9.347588671681949e-06,
+      "loss": 0.3241,
+      "step": 3014
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.129191082772908,
+      "learning_rate": 9.347085554188358e-06,
+      "loss": 0.3431,
+      "step": 3015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3763864590358255,
+      "learning_rate": 9.346582256326466e-06,
+      "loss": 0.3512,
+      "step": 3016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.5564846425568994,
+      "learning_rate": 9.346078778117157e-06,
+      "loss": 0.3552,
+      "step": 3017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6375082356956074,
+      "learning_rate": 9.34557511958132e-06,
+      "loss": 0.3562,
+      "step": 3018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.675383635989742,
+      "learning_rate": 9.345071280739853e-06,
+      "loss": 0.3266,
+      "step": 3019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8312994945653984,
+      "learning_rate": 9.34456726161366e-06,
+      "loss": 0.3541,
+      "step": 3020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.981447370141572,
+      "learning_rate": 9.344063062223653e-06,
+      "loss": 0.3316,
+      "step": 3021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3336506962385037,
+      "learning_rate": 9.343558682590757e-06,
+      "loss": 0.3407,
+      "step": 3022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3114342319151735,
+      "learning_rate": 9.343054122735894e-06,
+      "loss": 0.3312,
+      "step": 3023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9229177179769605,
+      "learning_rate": 9.342549382680002e-06,
+      "loss": 0.3311,
+      "step": 3024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.629641199016785,
+      "learning_rate": 9.342044462444023e-06,
+      "loss": 0.3182,
+      "step": 3025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.004324983753988,
+      "learning_rate": 9.341539362048906e-06,
+      "loss": 0.3424,
+      "step": 3026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.57432765267654,
+      "learning_rate": 9.34103408151561e-06,
+      "loss": 0.3384,
+      "step": 3027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6515450597736896,
+      "learning_rate": 9.340528620865099e-06,
+      "loss": 0.3425,
+      "step": 3028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.024417049623289,
+      "learning_rate": 9.340022980118346e-06,
+      "loss": 0.3551,
+      "step": 3029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.823128998136369,
+      "learning_rate": 9.33951715929633e-06,
+      "loss": 0.307,
+      "step": 3030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.302761290109694,
+      "learning_rate": 9.339011158420042e-06,
+      "loss": 0.321,
+      "step": 3031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7861581439453387,
+      "learning_rate": 9.338504977510471e-06,
+      "loss": 0.3333,
+      "step": 3032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.255436023938514,
+      "learning_rate": 9.337998616588624e-06,
+      "loss": 0.3197,
+      "step": 3033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7986968576143894,
+      "learning_rate": 9.337492075675509e-06,
+      "loss": 0.354,
+      "step": 3034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.427175316812047,
+      "learning_rate": 9.336985354792143e-06,
+      "loss": 0.3422,
+      "step": 3035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.262349587614101,
+      "learning_rate": 9.336478453959552e-06,
+      "loss": 0.3455,
+      "step": 3036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.892854706867287,
+      "learning_rate": 9.335971373198768e-06,
+      "loss": 0.345,
+      "step": 3037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.66963503363804,
+      "learning_rate": 9.33546411253083e-06,
+      "loss": 0.333,
+      "step": 3038
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.724888961257912,
+      "learning_rate": 9.334956671976784e-06,
+      "loss": 0.3304,
+      "step": 3039
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5438924357371304,
+      "learning_rate": 9.334449051557687e-06,
+      "loss": 0.3718,
+      "step": 3040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.697884305935806,
+      "learning_rate": 9.3339412512946e-06,
+      "loss": 0.3305,
+      "step": 3041
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.46906054185119,
+      "learning_rate": 9.333433271208592e-06,
+      "loss": 0.3239,
+      "step": 3042
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9475025359327924,
+      "learning_rate": 9.332925111320741e-06,
+      "loss": 0.3392,
+      "step": 3043
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.746327794218007,
+      "learning_rate": 9.33241677165213e-06,
+      "loss": 0.3452,
+      "step": 3044
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7648031217775895,
+      "learning_rate": 9.331908252223853e-06,
+      "loss": 0.3245,
+      "step": 3045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.079432984670926,
+      "learning_rate": 9.331399553057008e-06,
+      "loss": 0.3239,
+      "step": 3046
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.08175775219486,
+      "learning_rate": 9.330890674172703e-06,
+      "loss": 0.3445,
+      "step": 3047
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8930370994396265,
+      "learning_rate": 9.33038161559205e-06,
+      "loss": 0.5207,
+      "step": 3048
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.035597248186252,
+      "learning_rate": 9.329872377336173e-06,
+      "loss": 0.3343,
+      "step": 3049
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.783789110657642,
+      "learning_rate": 9.3293629594262e-06,
+      "loss": 0.3643,
+      "step": 3050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.452415922961428,
+      "learning_rate": 9.328853361883268e-06,
+      "loss": 0.3556,
+      "step": 3051
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4411137902450526,
+      "learning_rate": 9.32834358472852e-06,
+      "loss": 0.3363,
+      "step": 3052
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.742630188402976,
+      "learning_rate": 9.327833627983111e-06,
+      "loss": 0.3372,
+      "step": 3053
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1818245706896975,
+      "learning_rate": 9.327323491668197e-06,
+      "loss": 0.3173,
+      "step": 3054
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.266553037326733,
+      "learning_rate": 9.326813175804943e-06,
+      "loss": 0.3431,
+      "step": 3055
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3925988723286973,
+      "learning_rate": 9.326302680414527e-06,
+      "loss": 0.3535,
+      "step": 3056
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8704980127405775,
+      "learning_rate": 9.325792005518128e-06,
+      "loss": 0.3413,
+      "step": 3057
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.718238107609266,
+      "learning_rate": 9.325281151136936e-06,
+      "loss": 0.3143,
+      "step": 3058
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0086727732829504,
+      "learning_rate": 9.324770117292146e-06,
+      "loss": 0.3355,
+      "step": 3059
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2553592507183766,
+      "learning_rate": 9.324258904004961e-06,
+      "loss": 0.3232,
+      "step": 3060
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2727486642764254,
+      "learning_rate": 9.323747511296594e-06,
+      "loss": 0.3667,
+      "step": 3061
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.191618268891413,
+      "learning_rate": 9.323235939188265e-06,
+      "loss": 0.3226,
+      "step": 3062
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8422933989115813,
+      "learning_rate": 9.322724187701195e-06,
+      "loss": 0.3238,
+      "step": 3063
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.052104265411854,
+      "learning_rate": 9.322212256856622e-06,
+      "loss": 0.3272,
+      "step": 3064
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.0366610133851966,
+      "learning_rate": 9.321700146675785e-06,
+      "loss": 0.5136,
+      "step": 3065
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.899692312448913,
+      "learning_rate": 9.321187857179932e-06,
+      "loss": 0.3569,
+      "step": 3066
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.692567049351729,
+      "learning_rate": 9.320675388390321e-06,
+      "loss": 0.3685,
+      "step": 3067
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.750752151923033,
+      "learning_rate": 9.320162740328211e-06,
+      "loss": 0.3474,
+      "step": 3068
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9846018748256262,
+      "learning_rate": 9.319649913014878e-06,
+      "loss": 0.3264,
+      "step": 3069
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.545901753362208,
+      "learning_rate": 9.319136906471598e-06,
+      "loss": 0.3371,
+      "step": 3070
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.407767762102129,
+      "learning_rate": 9.318623720719654e-06,
+      "loss": 0.3256,
+      "step": 3071
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6079027392657275,
+      "learning_rate": 9.318110355780342e-06,
+      "loss": 0.3315,
+      "step": 3072
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.657985241783449,
+      "learning_rate": 9.317596811674962e-06,
+      "loss": 0.3201,
+      "step": 3073
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.4748325443645935,
+      "learning_rate": 9.317083088424822e-06,
+      "loss": 0.3421,
+      "step": 3074
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.736417483683428,
+      "learning_rate": 9.316569186051234e-06,
+      "loss": 0.316,
+      "step": 3075
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.767032320678666,
+      "learning_rate": 9.316055104575525e-06,
+      "loss": 0.353,
+      "step": 3076
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.42877438132356,
+      "learning_rate": 9.315540844019025e-06,
+      "loss": 0.3299,
+      "step": 3077
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1262551084630785,
+      "learning_rate": 9.315026404403068e-06,
+      "loss": 0.3619,
+      "step": 3078
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8639683136116036,
+      "learning_rate": 9.314511785749001e-06,
+      "loss": 0.3223,
+      "step": 3079
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.035077716440803,
+      "learning_rate": 9.313996988078178e-06,
+      "loss": 0.3354,
+      "step": 3080
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.395584925167177,
+      "learning_rate": 9.313482011411957e-06,
+      "loss": 0.332,
+      "step": 3081
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.851017873781992,
+      "learning_rate": 9.312966855771704e-06,
+      "loss": 0.3354,
+      "step": 3082
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.727301729924524,
+      "learning_rate": 9.312451521178798e-06,
+      "loss": 0.3383,
+      "step": 3083
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5206288984760876,
+      "learning_rate": 9.311936007654618e-06,
+      "loss": 0.3465,
+      "step": 3084
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5301215803483834,
+      "learning_rate": 9.311420315220553e-06,
+      "loss": 0.3325,
+      "step": 3085
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.107286661096052,
+      "learning_rate": 9.310904443898e-06,
+      "loss": 0.3394,
+      "step": 3086
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.430047693361223,
+      "learning_rate": 9.310388393708368e-06,
+      "loss": 0.3134,
+      "step": 3087
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0517604275704704,
+      "learning_rate": 9.309872164673063e-06,
+      "loss": 0.3392,
+      "step": 3088
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8432433407508314,
+      "learning_rate": 9.30935575681351e-06,
+      "loss": 0.4864,
+      "step": 3089
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.094975924775508,
+      "learning_rate": 9.30883917015113e-06,
+      "loss": 0.3317,
+      "step": 3090
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5260535063997995,
+      "learning_rate": 9.308322404707359e-06,
+      "loss": 0.3371,
+      "step": 3091
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.746233518227412,
+      "learning_rate": 9.307805460503639e-06,
+      "loss": 0.343,
+      "step": 3092
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.940903599970333,
+      "learning_rate": 9.30728833756142e-06,
+      "loss": 0.3205,
+      "step": 3093
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.4481602653343,
+      "learning_rate": 9.306771035902155e-06,
+      "loss": 0.315,
+      "step": 3094
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6580255121552712,
+      "learning_rate": 9.306253555547313e-06,
+      "loss": 0.4848,
+      "step": 3095
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9432766208942893,
+      "learning_rate": 9.30573589651836e-06,
+      "loss": 0.3398,
+      "step": 3096
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5327925577779693,
+      "learning_rate": 9.305218058836778e-06,
+      "loss": 0.341,
+      "step": 3097
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6333683879285672,
+      "learning_rate": 9.304700042524052e-06,
+      "loss": 0.4821,
+      "step": 3098
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6309725759518167,
+      "learning_rate": 9.304181847601675e-06,
+      "loss": 0.5041,
+      "step": 3099
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.026104691477374,
+      "learning_rate": 9.303663474091146e-06,
+      "loss": 0.3304,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.704467245912126,
+      "learning_rate": 9.303144922013979e-06,
+      "loss": 0.3496,
+      "step": 3101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9587455227385404,
+      "learning_rate": 9.302626191391684e-06,
+      "loss": 0.3582,
+      "step": 3102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1080186806649497,
+      "learning_rate": 9.302107282245785e-06,
+      "loss": 0.3252,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.911883259041764,
+      "learning_rate": 9.301588194597815e-06,
+      "loss": 0.3305,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.475237511109905,
+      "learning_rate": 9.30106892846931e-06,
+      "loss": 0.3291,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.438939667144132,
+      "learning_rate": 9.300549483881816e-06,
+      "loss": 0.3584,
+      "step": 3106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.738246580062574,
+      "learning_rate": 9.300029860856886e-06,
+      "loss": 0.3348,
+      "step": 3107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 43.85318841066162,
+      "learning_rate": 9.299510059416077e-06,
+      "loss": 0.3466,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.173025583594784,
+      "learning_rate": 9.298990079580959e-06,
+      "loss": 0.3364,
+      "step": 3109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.399942685315112,
+      "learning_rate": 9.298469921373108e-06,
+      "loss": 0.3367,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.2835717012918515,
+      "learning_rate": 9.297949584814105e-06,
+      "loss": 0.3471,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.326600887265767,
+      "learning_rate": 9.297429069925539e-06,
+      "loss": 0.3504,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.999845651609561,
+      "learning_rate": 9.296908376729009e-06,
+      "loss": 0.352,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3929357335956967,
+      "learning_rate": 9.296387505246116e-06,
+      "loss": 0.3213,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3982066984923813,
+      "learning_rate": 9.295866455498477e-06,
+      "loss": 0.3705,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.6702218223034135,
+      "learning_rate": 9.295345227507707e-06,
+      "loss": 0.3333,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.803485576760056,
+      "learning_rate": 9.294823821295433e-06,
+      "loss": 0.3195,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9781165440263238,
+      "learning_rate": 9.294302236883293e-06,
+      "loss": 0.3333,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9132682386226283,
+      "learning_rate": 9.293780474292923e-06,
+      "loss": 0.5514,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9893112971431743,
+      "learning_rate": 9.293258533545975e-06,
+      "loss": 0.3381,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.947785338266833,
+      "learning_rate": 9.292736414664105e-06,
+      "loss": 0.3055,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.423557293362017,
+      "learning_rate": 9.292214117668979e-06,
+      "loss": 0.3434,
+      "step": 3122
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0290706973690344,
+      "learning_rate": 9.291691642582262e-06,
+      "loss": 0.3491,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 22.13754401247412,
+      "learning_rate": 9.291168989425636e-06,
+      "loss": 0.3557,
+      "step": 3124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.111758513326185,
+      "learning_rate": 9.290646158220789e-06,
+      "loss": 0.3426,
+      "step": 3125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1929741948548553,
+      "learning_rate": 9.29012314898941e-06,
+      "loss": 0.3313,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.386553889091693,
+      "learning_rate": 9.289599961753201e-06,
+      "loss": 0.33,
+      "step": 3127
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.792576308994678,
+      "learning_rate": 9.289076596533873e-06,
+      "loss": 0.3516,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7360230492857958,
+      "learning_rate": 9.288553053353136e-06,
+      "loss": 0.3513,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.696616546068349,
+      "learning_rate": 9.288029332232718e-06,
+      "loss": 0.3338,
+      "step": 3130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.458912832502881,
+      "learning_rate": 9.287505433194344e-06,
+      "loss": 0.3319,
+      "step": 3131
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5319191980373486,
+      "learning_rate": 9.286981356259756e-06,
+      "loss": 0.3549,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5404540188621425,
+      "learning_rate": 9.286457101450695e-06,
+      "loss": 0.333,
+      "step": 3133
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8531424140218881,
+      "learning_rate": 9.285932668788917e-06,
+      "loss": 0.507,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.499563968202511,
+      "learning_rate": 9.28540805829618e-06,
+      "loss": 0.3394,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.601357275161658,
+      "learning_rate": 9.284883269994249e-06,
+      "loss": 0.3343,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4033310464057345,
+      "learning_rate": 9.284358303904902e-06,
+      "loss": 0.3346,
+      "step": 3137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5510444402774155,
+      "learning_rate": 9.28383316004992e-06,
+      "loss": 0.3308,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6581716057280045,
+      "learning_rate": 9.28330783845109e-06,
+      "loss": 0.5153,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.766367521467626,
+      "learning_rate": 9.28278233913021e-06,
+      "loss": 0.3639,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.648176503325645,
+      "learning_rate": 9.282256662109082e-06,
+      "loss": 0.3282,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.252889421194652,
+      "learning_rate": 9.281730807409522e-06,
+      "loss": 0.345,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.567440367216447,
+      "learning_rate": 9.281204775053342e-06,
+      "loss": 0.3296,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.164396048956933,
+      "learning_rate": 9.280678565062375e-06,
+      "loss": 0.3431,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8443810127063882,
+      "learning_rate": 9.280152177458449e-06,
+      "loss": 0.3378,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.347160350923574,
+      "learning_rate": 9.279625612263407e-06,
+      "loss": 0.3405,
+      "step": 3146
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.193585439101035,
+      "learning_rate": 9.279098869499097e-06,
+      "loss": 0.3184,
+      "step": 3147
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1915934188963817,
+      "learning_rate": 9.278571949187376e-06,
+      "loss": 0.3387,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.981672700854984,
+      "learning_rate": 9.278044851350103e-06,
+      "loss": 0.3314,
+      "step": 3149
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8244792300051936,
+      "learning_rate": 9.277517576009152e-06,
+      "loss": 0.5116,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0340519588206303,
+      "learning_rate": 9.276990123186397e-06,
+      "loss": 0.3521,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6617316786693868,
+      "learning_rate": 9.276462492903726e-06,
+      "loss": 0.4951,
+      "step": 3152
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3169855087654603,
+      "learning_rate": 9.275934685183032e-06,
+      "loss": 0.3192,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0251337233818814,
+      "learning_rate": 9.275406700046211e-06,
+      "loss": 0.3288,
+      "step": 3154
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1191920647426374,
+      "learning_rate": 9.274878537515173e-06,
+      "loss": 0.358,
+      "step": 3155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0121696385723533,
+      "learning_rate": 9.274350197611832e-06,
+      "loss": 0.3382,
+      "step": 3156
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0404721997473247,
+      "learning_rate": 9.273821680358107e-06,
+      "loss": 0.3168,
+      "step": 3157
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.273886983149378,
+      "learning_rate": 9.27329298577593e-06,
+      "loss": 0.3313,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.057467819196444,
+      "learning_rate": 9.272764113887237e-06,
+      "loss": 0.3255,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2025889870920823,
+      "learning_rate": 9.272235064713974e-06,
+      "loss": 0.3302,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.935982182351442,
+      "learning_rate": 9.271705838278086e-06,
+      "loss": 0.3437,
+      "step": 3161
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7691037434104837,
+      "learning_rate": 9.27117643460154e-06,
+      "loss": 0.3425,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.309836296686852,
+      "learning_rate": 9.270646853706293e-06,
+      "loss": 0.3575,
+      "step": 3163
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9099254762908546,
+      "learning_rate": 9.270117095614324e-06,
+      "loss": 0.3441,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3344908505501163,
+      "learning_rate": 9.269587160347612e-06,
+      "loss": 0.3233,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9357850985698077,
+      "learning_rate": 9.269057047928144e-06,
+      "loss": 0.3472,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8801315563559973,
+      "learning_rate": 9.268526758377919e-06,
+      "loss": 0.3638,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7230291838491256,
+      "learning_rate": 9.267996291718936e-06,
+      "loss": 0.3573,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.07079599260868,
+      "learning_rate": 9.267465647973206e-06,
+      "loss": 0.3545,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1648255080740424,
+      "learning_rate": 9.266934827162746e-06,
+      "loss": 0.34,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1379391469547526,
+      "learning_rate": 9.26640382930958e-06,
+      "loss": 0.3466,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8543325603940297,
+      "learning_rate": 9.265872654435743e-06,
+      "loss": 0.3387,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8881740805926435,
+      "learning_rate": 9.26534130256327e-06,
+      "loss": 0.3239,
+      "step": 3173
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.49069838949244,
+      "learning_rate": 9.264809773714214e-06,
+      "loss": 0.3218,
+      "step": 3174
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6883041794621128,
+      "learning_rate": 9.264278067910625e-06,
+      "loss": 0.3517,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.842922373876776,
+      "learning_rate": 9.263746185174562e-06,
+      "loss": 0.3417,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.839745658010722,
+      "learning_rate": 9.263214125528097e-06,
+      "loss": 0.3349,
+      "step": 3177
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.713975034051698,
+      "learning_rate": 9.262681888993306e-06,
+      "loss": 0.3444,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4326683224951275,
+      "learning_rate": 9.262149475592272e-06,
+      "loss": 0.3554,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1903551994351953,
+      "learning_rate": 9.261616885347087e-06,
+      "loss": 0.3523,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.988480030316767,
+      "learning_rate": 9.261084118279846e-06,
+      "loss": 0.3268,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1729441910723652,
+      "learning_rate": 9.26055117441266e-06,
+      "loss": 0.3306,
+      "step": 3182
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8335747055425395,
+      "learning_rate": 9.260018053767634e-06,
+      "loss": 0.3294,
+      "step": 3183
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.483853626654878,
+      "learning_rate": 9.259484756366894e-06,
+      "loss": 0.3334,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8398749503857479,
+      "learning_rate": 9.258951282232567e-06,
+      "loss": 0.3409,
+      "step": 3185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3869859868712218,
+      "learning_rate": 9.258417631386784e-06,
+      "loss": 0.3438,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.067702151113323,
+      "learning_rate": 9.257883803851692e-06,
+      "loss": 0.3668,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1551763649208757,
+      "learning_rate": 9.257349799649437e-06,
+      "loss": 0.3519,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8092883233789556,
+      "learning_rate": 9.256815618802178e-06,
+      "loss": 0.3489,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8746975925359735,
+      "learning_rate": 9.256281261332076e-06,
+      "loss": 0.3201,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.940141382151419,
+      "learning_rate": 9.255746727261305e-06,
+      "loss": 0.3537,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6974583371840613,
+      "learning_rate": 9.255212016612044e-06,
+      "loss": 0.3213,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.844741808366749,
+      "learning_rate": 9.254677129406477e-06,
+      "loss": 0.3473,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2017702725927175,
+      "learning_rate": 9.254142065666802e-06,
+      "loss": 0.3399,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8349854964968766,
+      "learning_rate": 9.253606825415213e-06,
+      "loss": 0.3329,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.461347680543548,
+      "learning_rate": 9.253071408673924e-06,
+      "loss": 0.3349,
+      "step": 3196
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.728796034177238,
+      "learning_rate": 9.252535815465146e-06,
+      "loss": 0.3277,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1476735913302596,
+      "learning_rate": 9.252000045811105e-06,
+      "loss": 0.3488,
+      "step": 3198
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.572846878344725,
+      "learning_rate": 9.25146409973403e-06,
+      "loss": 0.3329,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 47.92774792019559,
+      "learning_rate": 9.250927977256155e-06,
+      "loss": 0.3573,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.197579392795628,
+      "learning_rate": 9.25039167839973e-06,
+      "loss": 0.3226,
+      "step": 3201
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9942998432191135,
+      "learning_rate": 9.249855203187007e-06,
+      "loss": 0.3523,
+      "step": 3202
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.009663714699753,
+      "learning_rate": 9.249318551640238e-06,
+      "loss": 0.3222,
+      "step": 3203
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.592879992005864,
+      "learning_rate": 9.2487817237817e-06,
+      "loss": 0.3509,
+      "step": 3204
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7727777628097505,
+      "learning_rate": 9.248244719633659e-06,
+      "loss": 0.3264,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.90503630217772,
+      "learning_rate": 9.247707539218398e-06,
+      "loss": 0.3486,
+      "step": 3206
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.495555896714801,
+      "learning_rate": 9.247170182558207e-06,
+      "loss": 0.3121,
+      "step": 3207
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.186933314109556,
+      "learning_rate": 9.246632649675382e-06,
+      "loss": 0.3618,
+      "step": 3208
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0725383590343203,
+      "learning_rate": 9.246094940592224e-06,
+      "loss": 0.2995,
+      "step": 3209
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7157495868341974,
+      "learning_rate": 9.245557055331046e-06,
+      "loss": 0.346,
+      "step": 3210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1595383518209115,
+      "learning_rate": 9.245018993914166e-06,
+      "loss": 0.32,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.779679331049103,
+      "learning_rate": 9.244480756363904e-06,
+      "loss": 0.3803,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.372475177234385,
+      "learning_rate": 9.243942342702601e-06,
+      "loss": 0.3582,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2102588382653683,
+      "learning_rate": 9.24340375295259e-06,
+      "loss": 0.3208,
+      "step": 3214
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1949738710810096,
+      "learning_rate": 9.24286498713622e-06,
+      "loss": 0.3499,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.516471386721431,
+      "learning_rate": 9.242326045275846e-06,
+      "loss": 0.3384,
+      "step": 3216
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.246970065640399,
+      "learning_rate": 9.24178692739383e-06,
+      "loss": 0.32,
+      "step": 3217
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8097854465160834,
+      "learning_rate": 9.241247633512539e-06,
+      "loss": 0.3363,
+      "step": 3218
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9042000808275117,
+      "learning_rate": 9.240708163654351e-06,
+      "loss": 0.3487,
+      "step": 3219
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6299503296182896,
+      "learning_rate": 9.240168517841648e-06,
+      "loss": 0.3395,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.5934262865465665,
+      "learning_rate": 9.239628696096823e-06,
+      "loss": 0.3383,
+      "step": 3221
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4141524686623534,
+      "learning_rate": 9.239088698442272e-06,
+      "loss": 0.3247,
+      "step": 3222
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.7875869828206374,
+      "learning_rate": 9.238548524900401e-06,
+      "loss": 0.3478,
+      "step": 3223
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.2898484715310765,
+      "learning_rate": 9.238008175493625e-06,
+      "loss": 0.3232,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.4207614453299655,
+      "learning_rate": 9.23746765024436e-06,
+      "loss": 0.3275,
+      "step": 3225
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.932672407280655,
+      "learning_rate": 9.236926949175037e-06,
+      "loss": 0.3167,
+      "step": 3226
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8429470125541183,
+      "learning_rate": 9.23638607230809e-06,
+      "loss": 0.3219,
+      "step": 3227
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.862949427711048,
+      "learning_rate": 9.235845019665959e-06,
+      "loss": 0.3273,
+      "step": 3228
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.320392874054271,
+      "learning_rate": 9.235303791271094e-06,
+      "loss": 0.3384,
+      "step": 3229
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.5602851974905425,
+      "learning_rate": 9.234762387145952e-06,
+      "loss": 0.3367,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0776500161744,
+      "learning_rate": 9.234220807312998e-06,
+      "loss": 0.3207,
+      "step": 3231
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8843792434976,
+      "learning_rate": 9.233679051794701e-06,
+      "loss": 0.36,
+      "step": 3232
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4680316916760128,
+      "learning_rate": 9.23313712061354e-06,
+      "loss": 0.3544,
+      "step": 3233
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0833265338170346,
+      "learning_rate": 9.232595013792004e-06,
+      "loss": 0.3251,
+      "step": 3234
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.291851370050809,
+      "learning_rate": 9.232052731352578e-06,
+      "loss": 0.3223,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4115139121529108,
+      "learning_rate": 9.23151027331777e-06,
+      "loss": 0.3164,
+      "step": 3236
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2953978123927525,
+      "learning_rate": 9.230967639710085e-06,
+      "loss": 0.3322,
+      "step": 3237
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.742944667903736,
+      "learning_rate": 9.230424830552035e-06,
+      "loss": 0.3339,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.666546241009362,
+      "learning_rate": 9.229881845866148e-06,
+      "loss": 0.3319,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.7040286816864745,
+      "learning_rate": 9.229338685674948e-06,
+      "loss": 0.341,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.363329064621435,
+      "learning_rate": 9.228795350000977e-06,
+      "loss": 0.3384,
+      "step": 3241
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.861605510467439,
+      "learning_rate": 9.228251838866772e-06,
+      "loss": 0.3223,
+      "step": 3242
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.940652153111325,
+      "learning_rate": 9.227708152294889e-06,
+      "loss": 0.3407,
+      "step": 3243
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.22711923987941,
+      "learning_rate": 9.227164290307887e-06,
+      "loss": 0.3381,
+      "step": 3244
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.673798403637111,
+      "learning_rate": 9.22662025292833e-06,
+      "loss": 0.3444,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.555285733453422,
+      "learning_rate": 9.226076040178788e-06,
+      "loss": 0.3213,
+      "step": 3246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.820065874417691,
+      "learning_rate": 9.225531652081848e-06,
+      "loss": 0.3215,
+      "step": 3247
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.12254801547705,
+      "learning_rate": 9.224987088660094e-06,
+      "loss": 0.3258,
+      "step": 3248
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7767350258221444,
+      "learning_rate": 9.22444234993612e-06,
+      "loss": 0.353,
+      "step": 3249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.323233219968319,
+      "learning_rate": 9.223897435932532e-06,
+      "loss": 0.3495,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0002888207855123,
+      "learning_rate": 9.223352346671935e-06,
+      "loss": 0.3204,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4286053134398724,
+      "learning_rate": 9.222807082176948e-06,
+      "loss": 0.3241,
+      "step": 3252
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0042722335846075,
+      "learning_rate": 9.222261642470194e-06,
+      "loss": 0.3569,
+      "step": 3253
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.923639477056454,
+      "learning_rate": 9.221716027574306e-06,
+      "loss": 0.3627,
+      "step": 3254
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0732575508603124,
+      "learning_rate": 9.22117023751192e-06,
+      "loss": 0.3112,
+      "step": 3255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.941078040252632,
+      "learning_rate": 9.220624272305683e-06,
+      "loss": 0.3574,
+      "step": 3256
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6451251005631082,
+      "learning_rate": 9.22007813197825e-06,
+      "loss": 0.3392,
+      "step": 3257
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.28083926606095,
+      "learning_rate": 9.219531816552279e-06,
+      "loss": 0.3649,
+      "step": 3258
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.9430400688118254,
+      "learning_rate": 9.218985326050439e-06,
+      "loss": 0.3463,
+      "step": 3259
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 26.166018466888865,
+      "learning_rate": 9.218438660495401e-06,
+      "loss": 0.3411,
+      "step": 3260
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6872472537936742,
+      "learning_rate": 9.217891819909854e-06,
+      "loss": 0.3653,
+      "step": 3261
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.436469040818058,
+      "learning_rate": 9.217344804316482e-06,
+      "loss": 0.3364,
+      "step": 3262
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.1852007080247144,
+      "learning_rate": 9.216797613737983e-06,
+      "loss": 0.3385,
+      "step": 3263
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3268107476088984,
+      "learning_rate": 9.21625024819706e-06,
+      "loss": 0.3381,
+      "step": 3264
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.110258063796547,
+      "learning_rate": 9.215702707716427e-06,
+      "loss": 0.3365,
+      "step": 3265
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7578030950227537,
+      "learning_rate": 9.2151549923188e-06,
+      "loss": 0.3405,
+      "step": 3266
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.871857191905456,
+      "learning_rate": 9.214607102026905e-06,
+      "loss": 0.3576,
+      "step": 3267
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.68667289240282,
+      "learning_rate": 9.214059036863477e-06,
+      "loss": 0.3373,
+      "step": 3268
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.376630280488469,
+      "learning_rate": 9.213510796851253e-06,
+      "loss": 0.3487,
+      "step": 3269
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6182677959069554,
+      "learning_rate": 9.212962382012981e-06,
+      "loss": 0.3455,
+      "step": 3270
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.791047510562209,
+      "learning_rate": 9.21241379237142e-06,
+      "loss": 0.3583,
+      "step": 3271
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.172415132107194,
+      "learning_rate": 9.211865027949328e-06,
+      "loss": 0.3175,
+      "step": 3272
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5181523258848157,
+      "learning_rate": 9.211316088769473e-06,
+      "loss": 0.3381,
+      "step": 3273
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.622769907928998,
+      "learning_rate": 9.210766974854634e-06,
+      "loss": 0.3399,
+      "step": 3274
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3280229125343714,
+      "learning_rate": 9.210217686227593e-06,
+      "loss": 0.3402,
+      "step": 3275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9442166229888342,
+      "learning_rate": 9.209668222911143e-06,
+      "loss": 0.341,
+      "step": 3276
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0759456934942504,
+      "learning_rate": 9.209118584928082e-06,
+      "loss": 0.3301,
+      "step": 3277
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1493635590756597,
+      "learning_rate": 9.208568772301213e-06,
+      "loss": 0.3369,
+      "step": 3278
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8006361457377222,
+      "learning_rate": 9.208018785053353e-06,
+      "loss": 0.3372,
+      "step": 3279
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5474933450527315,
+      "learning_rate": 9.207468623207317e-06,
+      "loss": 0.3464,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7352724708460774,
+      "learning_rate": 9.206918286785936e-06,
+      "loss": 0.3218,
+      "step": 3281
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.254842152358999,
+      "learning_rate": 9.206367775812042e-06,
+      "loss": 0.3443,
+      "step": 3282
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4955303744267425,
+      "learning_rate": 9.205817090308478e-06,
+      "loss": 0.3364,
+      "step": 3283
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0817601413132634,
+      "learning_rate": 9.205266230298092e-06,
+      "loss": 0.3414,
+      "step": 3284
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.109554213401208,
+      "learning_rate": 9.204715195803742e-06,
+      "loss": 0.3199,
+      "step": 3285
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.138672698305992,
+      "learning_rate": 9.20416398684829e-06,
+      "loss": 0.3211,
+      "step": 3286
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1371287351578165,
+      "learning_rate": 9.203612603454605e-06,
+      "loss": 0.3382,
+      "step": 3287
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2386771843454425,
+      "learning_rate": 9.203061045645567e-06,
+      "loss": 0.3231,
+      "step": 3288
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.523871963082305,
+      "learning_rate": 9.202509313444061e-06,
+      "loss": 0.3278,
+      "step": 3289
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.824293203598094,
+      "learning_rate": 9.20195740687298e-06,
+      "loss": 0.3449,
+      "step": 3290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.593839876042038,
+      "learning_rate": 9.201405325955222e-06,
+      "loss": 0.3364,
+      "step": 3291
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.112760574710892,
+      "learning_rate": 9.200853070713695e-06,
+      "loss": 0.3203,
+      "step": 3292
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.540605316940034,
+      "learning_rate": 9.20030064117131e-06,
+      "loss": 0.3193,
+      "step": 3293
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.32568423100732,
+      "learning_rate": 9.199748037350996e-06,
+      "loss": 0.3432,
+      "step": 3294
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.773215132307938,
+      "learning_rate": 9.199195259275673e-06,
+      "loss": 0.3374,
+      "step": 3295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.438571219906153,
+      "learning_rate": 9.19864230696828e-06,
+      "loss": 0.3266,
+      "step": 3296
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.87962527297526,
+      "learning_rate": 9.198089180451761e-06,
+      "loss": 0.3335,
+      "step": 3297
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.452510535287162,
+      "learning_rate": 9.197535879749065e-06,
+      "loss": 0.3382,
+      "step": 3298
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7645794132909876,
+      "learning_rate": 9.196982404883147e-06,
+      "loss": 0.3291,
+      "step": 3299
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1315230713321487,
+      "learning_rate": 9.196428755876978e-06,
+      "loss": 0.3582,
+      "step": 3300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4407918219457003,
+      "learning_rate": 9.195874932753525e-06,
+      "loss": 0.3521,
+      "step": 3301
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.401129747125721,
+      "learning_rate": 9.195320935535769e-06,
+      "loss": 0.3602,
+      "step": 3302
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.532072527604524,
+      "learning_rate": 9.194766764246694e-06,
+      "loss": 0.3534,
+      "step": 3303
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.412806962110028,
+      "learning_rate": 9.194212418909296e-06,
+      "loss": 0.3307,
+      "step": 3304
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5199638819278407,
+      "learning_rate": 9.193657899546575e-06,
+      "loss": 0.3348,
+      "step": 3305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5887591176661275,
+      "learning_rate": 9.19310320618154e-06,
+      "loss": 0.3424,
+      "step": 3306
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.593860703339445,
+      "learning_rate": 9.192548338837204e-06,
+      "loss": 0.3307,
+      "step": 3307
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7497839312493255,
+      "learning_rate": 9.19199329753659e-06,
+      "loss": 0.3141,
+      "step": 3308
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4448666963415697,
+      "learning_rate": 9.191438082302731e-06,
+      "loss": 0.3187,
+      "step": 3309
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2215136403363744,
+      "learning_rate": 9.190882693158658e-06,
+      "loss": 0.3298,
+      "step": 3310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.554905590613441,
+      "learning_rate": 9.19032713012742e-06,
+      "loss": 0.3416,
+      "step": 3311
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8985134435188125,
+      "learning_rate": 9.189771393232065e-06,
+      "loss": 0.3219,
+      "step": 3312
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8571913602771155,
+      "learning_rate": 9.189215482495655e-06,
+      "loss": 0.3188,
+      "step": 3313
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0283925225733554,
+      "learning_rate": 9.188659397941252e-06,
+      "loss": 0.3381,
+      "step": 3314
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5622267881856042,
+      "learning_rate": 9.188103139591934e-06,
+      "loss": 0.3366,
+      "step": 3315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6204691471689936,
+      "learning_rate": 9.187546707470773e-06,
+      "loss": 0.3257,
+      "step": 3316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.479359532002313,
+      "learning_rate": 9.186990101600865e-06,
+      "loss": 0.3587,
+      "step": 3317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5156339811433504,
+      "learning_rate": 9.186433322005298e-06,
+      "loss": 0.3252,
+      "step": 3318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0677644046278445,
+      "learning_rate": 9.185876368707178e-06,
+      "loss": 0.3484,
+      "step": 3319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8363778250072302,
+      "learning_rate": 9.185319241729614e-06,
+      "loss": 0.3283,
+      "step": 3320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5325217068777581,
+      "learning_rate": 9.184761941095716e-06,
+      "loss": 0.3328,
+      "step": 3321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1078622392122868,
+      "learning_rate": 9.184204466828618e-06,
+      "loss": 0.3491,
+      "step": 3322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7331955765680322,
+      "learning_rate": 9.18364681895144e-06,
+      "loss": 0.3464,
+      "step": 3323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1285879808933634,
+      "learning_rate": 9.183088997487326e-06,
+      "loss": 0.3405,
+      "step": 3324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.188071343440662,
+      "learning_rate": 9.18253100245942e-06,
+      "loss": 0.3391,
+      "step": 3325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.56313134255,
+      "learning_rate": 9.181972833890875e-06,
+      "loss": 0.3324,
+      "step": 3326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0520794722060756,
+      "learning_rate": 9.181414491804846e-06,
+      "loss": 0.3396,
+      "step": 3327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4612046989301044,
+      "learning_rate": 9.180855976224505e-06,
+      "loss": 0.3285,
+      "step": 3328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2254198126880214,
+      "learning_rate": 9.180297287173022e-06,
+      "loss": 0.3317,
+      "step": 3329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.358969469056055,
+      "learning_rate": 9.179738424673582e-06,
+      "loss": 0.3415,
+      "step": 3330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3580050157462966,
+      "learning_rate": 9.179179388749369e-06,
+      "loss": 0.3421,
+      "step": 3331
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.7782296233398425,
+      "learning_rate": 9.17862017942358e-06,
+      "loss": 0.3424,
+      "step": 3332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.562563165541226,
+      "learning_rate": 9.178060796719417e-06,
+      "loss": 0.3355,
+      "step": 3333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1177647447941483,
+      "learning_rate": 9.177501240660091e-06,
+      "loss": 0.3427,
+      "step": 3334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.871763419278553,
+      "learning_rate": 9.176941511268818e-06,
+      "loss": 0.3622,
+      "step": 3335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.036188243176669,
+      "learning_rate": 9.176381608568824e-06,
+      "loss": 0.3325,
+      "step": 3336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.487483995921712,
+      "learning_rate": 9.175821532583338e-06,
+      "loss": 0.3647,
+      "step": 3337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7326747344651108,
+      "learning_rate": 9.1752612833356e-06,
+      "loss": 0.3313,
+      "step": 3338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.40782656536387,
+      "learning_rate": 9.174700860848855e-06,
+      "loss": 0.3335,
+      "step": 3339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9838042980440616,
+      "learning_rate": 9.174140265146356e-06,
+      "loss": 0.3451,
+      "step": 3340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7016867447401482,
+      "learning_rate": 9.173579496251363e-06,
+      "loss": 0.3469,
+      "step": 3341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.090011723773526,
+      "learning_rate": 9.173018554187145e-06,
+      "loss": 0.3482,
+      "step": 3342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8980960308910082,
+      "learning_rate": 9.172457438976974e-06,
+      "loss": 0.3282,
+      "step": 3343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4699340183754797,
+      "learning_rate": 9.171896150644132e-06,
+      "loss": 0.3615,
+      "step": 3344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.797645045459058,
+      "learning_rate": 9.17133468921191e-06,
+      "loss": 0.3462,
+      "step": 3345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4314848202489756,
+      "learning_rate": 9.170773054703603e-06,
+      "loss": 0.3203,
+      "step": 3346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.872175357632315,
+      "learning_rate": 9.170211247142514e-06,
+      "loss": 0.3468,
+      "step": 3347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2198033083294213,
+      "learning_rate": 9.169649266551951e-06,
+      "loss": 0.3393,
+      "step": 3348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.512652006046951,
+      "learning_rate": 9.169087112955234e-06,
+      "loss": 0.3391,
+      "step": 3349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6463453697273636,
+      "learning_rate": 9.168524786375689e-06,
+      "loss": 0.3312,
+      "step": 3350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.908131989900112,
+      "learning_rate": 9.167962286836647e-06,
+      "loss": 0.3357,
+      "step": 3351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8293700890744862,
+      "learning_rate": 9.167399614361445e-06,
+      "loss": 0.3225,
+      "step": 3352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0266428828740053,
+      "learning_rate": 9.166836768973434e-06,
+      "loss": 0.3274,
+      "step": 3353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.943899273866206,
+      "learning_rate": 9.16627375069596e-06,
+      "loss": 0.3369,
+      "step": 3354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.884931678526371,
+      "learning_rate": 9.16571055955239e-06,
+      "loss": 0.3465,
+      "step": 3355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.400698594126364,
+      "learning_rate": 9.165147195566089e-06,
+      "loss": 0.3294,
+      "step": 3356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1125453260908458,
+      "learning_rate": 9.164583658760432e-06,
+      "loss": 0.3434,
+      "step": 3357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.468269633182379,
+      "learning_rate": 9.164019949158804e-06,
+      "loss": 0.3235,
+      "step": 3358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0730537988756113,
+      "learning_rate": 9.163456066784591e-06,
+      "loss": 0.3275,
+      "step": 3359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.284549859365693,
+      "learning_rate": 9.162892011661192e-06,
+      "loss": 0.3543,
+      "step": 3360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5003264599554074,
+      "learning_rate": 9.162327783812008e-06,
+      "loss": 0.389,
+      "step": 3361
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9787320970232836,
+      "learning_rate": 9.161763383260452e-06,
+      "loss": 0.3177,
+      "step": 3362
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1673964662883662,
+      "learning_rate": 9.16119881002994e-06,
+      "loss": 0.5297,
+      "step": 3363
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.815700730125225,
+      "learning_rate": 9.160634064143899e-06,
+      "loss": 0.3383,
+      "step": 3364
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7103598931713948,
+      "learning_rate": 9.16006914562576e-06,
+      "loss": 0.3523,
+      "step": 3365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.024844366786342,
+      "learning_rate": 9.159504054498964e-06,
+      "loss": 0.37,
+      "step": 3366
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.676963080534276,
+      "learning_rate": 9.158938790786955e-06,
+      "loss": 0.3721,
+      "step": 3367
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2917037062206203,
+      "learning_rate": 9.15837335451319e-06,
+      "loss": 0.3256,
+      "step": 3368
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.657558902634187,
+      "learning_rate": 9.157807745701128e-06,
+      "loss": 0.3164,
+      "step": 3369
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6724387004788803,
+      "learning_rate": 9.157241964374237e-06,
+      "loss": 0.3387,
+      "step": 3370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.503820732107395,
+      "learning_rate": 9.156676010555993e-06,
+      "loss": 0.3328,
+      "step": 3371
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3457756669202174,
+      "learning_rate": 9.15610988426988e-06,
+      "loss": 0.3407,
+      "step": 3372
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.548715836200099,
+      "learning_rate": 9.155543585539384e-06,
+      "loss": 0.3289,
+      "step": 3373
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0279051529499923,
+      "learning_rate": 9.154977114388003e-06,
+      "loss": 0.328,
+      "step": 3374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4596567994902307,
+      "learning_rate": 9.154410470839243e-06,
+      "loss": 0.3438,
+      "step": 3375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.859275170217395,
+      "learning_rate": 9.153843654916611e-06,
+      "loss": 0.3806,
+      "step": 3376
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.131478239727446,
+      "learning_rate": 9.15327666664363e-06,
+      "loss": 0.3191,
+      "step": 3377
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7706683433185906,
+      "learning_rate": 9.152709506043823e-06,
+      "loss": 0.3402,
+      "step": 3378
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8437933644698665,
+      "learning_rate": 9.152142173140722e-06,
+      "loss": 0.3376,
+      "step": 3379
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.7778771841816985,
+      "learning_rate": 9.151574667957868e-06,
+      "loss": 0.3266,
+      "step": 3380
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5977096609640595,
+      "learning_rate": 9.151006990518806e-06,
+      "loss": 0.3291,
+      "step": 3381
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.266092286796118,
+      "learning_rate": 9.150439140847091e-06,
+      "loss": 0.3347,
+      "step": 3382
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.818289659645148,
+      "learning_rate": 9.149871118966285e-06,
+      "loss": 0.3379,
+      "step": 3383
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.215748932020619,
+      "learning_rate": 9.149302924899954e-06,
+      "loss": 0.3164,
+      "step": 3384
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.570108372727355,
+      "learning_rate": 9.148734558671675e-06,
+      "loss": 0.3327,
+      "step": 3385
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.71923744775369,
+      "learning_rate": 9.14816602030503e-06,
+      "loss": 0.3755,
+      "step": 3386
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4901020425627403,
+      "learning_rate": 9.14759730982361e-06,
+      "loss": 0.3154,
+      "step": 3387
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.661112022598833,
+      "learning_rate": 9.14702842725101e-06,
+      "loss": 0.3379,
+      "step": 3388
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9211450406357264,
+      "learning_rate": 9.146459372610834e-06,
+      "loss": 0.3458,
+      "step": 3389
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.7073832687348816,
+      "learning_rate": 9.145890145926695e-06,
+      "loss": 0.3221,
+      "step": 3390
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9851121593898338,
+      "learning_rate": 9.145320747222208e-06,
+      "loss": 0.3396,
+      "step": 3391
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8435272556335773,
+      "learning_rate": 9.144751176521002e-06,
+      "loss": 0.3525,
+      "step": 3392
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7941329282709573,
+      "learning_rate": 9.144181433846707e-06,
+      "loss": 0.3337,
+      "step": 3393
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.659776119343045,
+      "learning_rate": 9.143611519222964e-06,
+      "loss": 0.3274,
+      "step": 3394
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.398221496478616,
+      "learning_rate": 9.143041432673419e-06,
+      "loss": 0.336,
+      "step": 3395
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.213436868912288,
+      "learning_rate": 9.142471174221726e-06,
+      "loss": 0.3309,
+      "step": 3396
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9721671511681025,
+      "learning_rate": 9.141900743891546e-06,
+      "loss": 0.5126,
+      "step": 3397
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8562501935600255,
+      "learning_rate": 9.141330141706546e-06,
+      "loss": 0.3318,
+      "step": 3398
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0973830541994762,
+      "learning_rate": 9.140759367690404e-06,
+      "loss": 0.3279,
+      "step": 3399
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7071700560578132,
+      "learning_rate": 9.1401884218668e-06,
+      "loss": 0.5134,
+      "step": 3400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.110391149165387,
+      "learning_rate": 9.139617304259427e-06,
+      "loss": 0.3299,
+      "step": 3401
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.5283307902572165,
+      "learning_rate": 9.139046014891977e-06,
+      "loss": 0.3416,
+      "step": 3402
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.180471716055632,
+      "learning_rate": 9.138474553788157e-06,
+      "loss": 0.3394,
+      "step": 3403
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6996761485645479,
+      "learning_rate": 9.137902920971678e-06,
+      "loss": 0.5086,
+      "step": 3404
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.551640849621617,
+      "learning_rate": 9.137331116466256e-06,
+      "loss": 0.3267,
+      "step": 3405
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.8730634452546617,
+      "learning_rate": 9.136759140295615e-06,
+      "loss": 0.3462,
+      "step": 3406
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.467497394945697,
+      "learning_rate": 9.136186992483492e-06,
+      "loss": 0.3318,
+      "step": 3407
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8178472384582345,
+      "learning_rate": 9.135614673053624e-06,
+      "loss": 0.3283,
+      "step": 3408
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2951332013735337,
+      "learning_rate": 9.135042182029757e-06,
+      "loss": 0.3313,
+      "step": 3409
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.26418890154553,
+      "learning_rate": 9.134469519435646e-06,
+      "loss": 0.3418,
+      "step": 3410
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.818220638789112,
+      "learning_rate": 9.13389668529505e-06,
+      "loss": 0.3337,
+      "step": 3411
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.86822808388928,
+      "learning_rate": 9.133323679631738e-06,
+      "loss": 0.3174,
+      "step": 3412
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.109364057932612,
+      "learning_rate": 9.132750502469485e-06,
+      "loss": 0.3165,
+      "step": 3413
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.672558504126033,
+      "learning_rate": 9.132177153832074e-06,
+      "loss": 0.311,
+      "step": 3414
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.832940662614624,
+      "learning_rate": 9.131603633743292e-06,
+      "loss": 0.3322,
+      "step": 3415
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.204792192716843,
+      "learning_rate": 9.131029942226937e-06,
+      "loss": 0.3317,
+      "step": 3416
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.832418154868519,
+      "learning_rate": 9.130456079306814e-06,
+      "loss": 0.3822,
+      "step": 3417
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5062226053206063,
+      "learning_rate": 9.12988204500673e-06,
+      "loss": 0.3308,
+      "step": 3418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1590126938251,
+      "learning_rate": 9.129307839350504e-06,
+      "loss": 0.3274,
+      "step": 3419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2142266209046824,
+      "learning_rate": 9.128733462361963e-06,
+      "loss": 0.3387,
+      "step": 3420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6472285423146584,
+      "learning_rate": 9.128158914064934e-06,
+      "loss": 0.317,
+      "step": 3421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5113775531399516,
+      "learning_rate": 9.127584194483262e-06,
+      "loss": 0.3489,
+      "step": 3422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.943572123418341,
+      "learning_rate": 9.12700930364079e-06,
+      "loss": 0.319,
+      "step": 3423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8812612659283072,
+      "learning_rate": 9.12643424156137e-06,
+      "loss": 0.3208,
+      "step": 3424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9013054774185956,
+      "learning_rate": 9.125859008268867e-06,
+      "loss": 0.3423,
+      "step": 3425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.544791406628796,
+      "learning_rate": 9.125283603787142e-06,
+      "loss": 0.3342,
+      "step": 3426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.394809981961453,
+      "learning_rate": 9.124708028140075e-06,
+      "loss": 0.3356,
+      "step": 3427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0129142018826496,
+      "learning_rate": 9.124132281351545e-06,
+      "loss": 0.3326,
+      "step": 3428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.7907079098252705,
+      "learning_rate": 9.123556363445442e-06,
+      "loss": 0.3419,
+      "step": 3429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.423517417702465,
+      "learning_rate": 9.12298027444566e-06,
+      "loss": 0.3238,
+      "step": 3430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2696674535860373,
+      "learning_rate": 9.122404014376104e-06,
+      "loss": 0.341,
+      "step": 3431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8037064937903338,
+      "learning_rate": 9.121827583260686e-06,
+      "loss": 0.3079,
+      "step": 3432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7462740294945083,
+      "learning_rate": 9.121250981123315e-06,
+      "loss": 0.3332,
+      "step": 3433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.9452868801654155,
+      "learning_rate": 9.120674207987923e-06,
+      "loss": 0.3197,
+      "step": 3434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.483170471661308,
+      "learning_rate": 9.12009726387844e-06,
+      "loss": 0.3365,
+      "step": 3435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7605779973586477,
+      "learning_rate": 9.119520148818804e-06,
+      "loss": 0.311,
+      "step": 3436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.403593592969136,
+      "learning_rate": 9.118942862832958e-06,
+      "loss": 0.3307,
+      "step": 3437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.670852267455348,
+      "learning_rate": 9.118365405944856e-06,
+      "loss": 0.3207,
+      "step": 3438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.939563130698033,
+      "learning_rate": 9.11778777817846e-06,
+      "loss": 0.3319,
+      "step": 3439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3762713484256435,
+      "learning_rate": 9.117209979557734e-06,
+      "loss": 0.3312,
+      "step": 3440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9976177286620933,
+      "learning_rate": 9.116632010106654e-06,
+      "loss": 0.3313,
+      "step": 3441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2074607433805937,
+      "learning_rate": 9.116053869849198e-06,
+      "loss": 0.3562,
+      "step": 3442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0660503344787537,
+      "learning_rate": 9.115475558809358e-06,
+      "loss": 0.3273,
+      "step": 3443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.01375776371647,
+      "learning_rate": 9.114897077011128e-06,
+      "loss": 0.3725,
+      "step": 3444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6250311791907452,
+      "learning_rate": 9.114318424478506e-06,
+      "loss": 0.3419,
+      "step": 3445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.186364353903956,
+      "learning_rate": 9.113739601235508e-06,
+      "loss": 0.3446,
+      "step": 3446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3887747665661734,
+      "learning_rate": 9.113160607306143e-06,
+      "loss": 0.3601,
+      "step": 3447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2462392757443417,
+      "learning_rate": 9.112581442714443e-06,
+      "loss": 0.3427,
+      "step": 3448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2132050586897387,
+      "learning_rate": 9.112002107484433e-06,
+      "loss": 0.2929,
+      "step": 3449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.778137670186096,
+      "learning_rate": 9.111422601640151e-06,
+      "loss": 0.3448,
+      "step": 3450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9218415945345637,
+      "learning_rate": 9.110842925205642e-06,
+      "loss": 0.5113,
+      "step": 3451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3952053390492414,
+      "learning_rate": 9.11026307820496e-06,
+      "loss": 0.3367,
+      "step": 3452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.006458909704897,
+      "learning_rate": 9.10968306066216e-06,
+      "loss": 0.3262,
+      "step": 3453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5686362626453807,
+      "learning_rate": 9.109102872601312e-06,
+      "loss": 0.3307,
+      "step": 3454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2922075069105805,
+      "learning_rate": 9.108522514046487e-06,
+      "loss": 0.3395,
+      "step": 3455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.548510923432122,
+      "learning_rate": 9.107941985021765e-06,
+      "loss": 0.3599,
+      "step": 3456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0702866979160897,
+      "learning_rate": 9.107361285551236e-06,
+      "loss": 0.3282,
+      "step": 3457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.821491197807495,
+      "learning_rate": 9.106780415658989e-06,
+      "loss": 0.3466,
+      "step": 3458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2321418725937776,
+      "learning_rate": 9.106199375369127e-06,
+      "loss": 0.3603,
+      "step": 3459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6049074226834605,
+      "learning_rate": 9.10561816470576e-06,
+      "loss": 0.4935,
+      "step": 3460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.659252169270551,
+      "learning_rate": 9.105036783693006e-06,
+      "loss": 0.3379,
+      "step": 3461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.256271392693812,
+      "learning_rate": 9.104455232354982e-06,
+      "loss": 0.3505,
+      "step": 3462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5900481103157382,
+      "learning_rate": 9.10387351071582e-06,
+      "loss": 0.334,
+      "step": 3463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7640226305940168,
+      "learning_rate": 9.103291618799657e-06,
+      "loss": 0.3398,
+      "step": 3464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.739483356634692,
+      "learning_rate": 9.102709556630639e-06,
+      "loss": 0.3349,
+      "step": 3465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.902883040974735,
+      "learning_rate": 9.10212732423291e-06,
+      "loss": 0.342,
+      "step": 3466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.323800145239479,
+      "learning_rate": 9.101544921630634e-06,
+      "loss": 0.3215,
+      "step": 3467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.354977132274946,
+      "learning_rate": 9.100962348847974e-06,
+      "loss": 0.3277,
+      "step": 3468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.436653894832212,
+      "learning_rate": 9.100379605909102e-06,
+      "loss": 0.3378,
+      "step": 3469
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5648961244173885,
+      "learning_rate": 9.099796692838198e-06,
+      "loss": 0.3047,
+      "step": 3470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 102.54707431872112,
+      "learning_rate": 9.099213609659448e-06,
+      "loss": 0.3161,
+      "step": 3471
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.952786479893375,
+      "learning_rate": 9.09863035639704e-06,
+      "loss": 0.3645,
+      "step": 3472
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.18717403910133,
+      "learning_rate": 9.098046933075182e-06,
+      "loss": 0.3426,
+      "step": 3473
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.055273260646196,
+      "learning_rate": 9.097463339718077e-06,
+      "loss": 0.3182,
+      "step": 3474
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.137222206697158,
+      "learning_rate": 9.09687957634994e-06,
+      "loss": 0.3281,
+      "step": 3475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.610557892693063,
+      "learning_rate": 9.096295642994993e-06,
+      "loss": 0.3319,
+      "step": 3476
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.246057627288268,
+      "learning_rate": 9.095711539677464e-06,
+      "loss": 0.3434,
+      "step": 3477
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.653807832279,
+      "learning_rate": 9.095127266421589e-06,
+      "loss": 0.3567,
+      "step": 3478
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5642670646223147,
+      "learning_rate": 9.094542823251609e-06,
+      "loss": 0.3241,
+      "step": 3479
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.681253858790611,
+      "learning_rate": 9.093958210191773e-06,
+      "loss": 0.3339,
+      "step": 3480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.795870686069218,
+      "learning_rate": 9.09337342726634e-06,
+      "loss": 0.3071,
+      "step": 3481
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.649140128270586,
+      "learning_rate": 9.092788474499574e-06,
+      "loss": 0.5054,
+      "step": 3482
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.564914198833503,
+      "learning_rate": 9.092203351915744e-06,
+      "loss": 0.3149,
+      "step": 3483
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.054874636347387,
+      "learning_rate": 9.091618059539129e-06,
+      "loss": 0.3367,
+      "step": 3484
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.561306503066033,
+      "learning_rate": 9.091032597394012e-06,
+      "loss": 0.333,
+      "step": 3485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.352116762755251,
+      "learning_rate": 9.090446965504687e-06,
+      "loss": 0.3292,
+      "step": 3486
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6444201086478323,
+      "learning_rate": 9.089861163895453e-06,
+      "loss": 0.3343,
+      "step": 3487
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.831919652606544,
+      "learning_rate": 9.089275192590613e-06,
+      "loss": 0.3517,
+      "step": 3488
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5929553608429483,
+      "learning_rate": 9.088689051614483e-06,
+      "loss": 0.3397,
+      "step": 3489
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6375140527785934,
+      "learning_rate": 9.08810274099138e-06,
+      "loss": 0.3196,
+      "step": 3490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.36569335735749,
+      "learning_rate": 9.087516260745635e-06,
+      "loss": 0.3104,
+      "step": 3491
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.002990115600427,
+      "learning_rate": 9.08692961090158e-06,
+      "loss": 0.3313,
+      "step": 3492
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.842323738404551,
+      "learning_rate": 9.086342791483555e-06,
+      "loss": 0.3316,
+      "step": 3493
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5964259177477587,
+      "learning_rate": 9.08575580251591e-06,
+      "loss": 0.3253,
+      "step": 3494
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.552320116874668,
+      "learning_rate": 9.085168644022999e-06,
+      "loss": 0.326,
+      "step": 3495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7945823372035328,
+      "learning_rate": 9.084581316029186e-06,
+      "loss": 0.3226,
+      "step": 3496
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.597765238121262,
+      "learning_rate": 9.08399381855884e-06,
+      "loss": 0.3434,
+      "step": 3497
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3106496510742263,
+      "learning_rate": 9.083406151636334e-06,
+      "loss": 0.3027,
+      "step": 3498
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.31908087309126,
+      "learning_rate": 9.082818315286054e-06,
+      "loss": 0.34,
+      "step": 3499
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.601157452757643,
+      "learning_rate": 9.082230309532393e-06,
+      "loss": 0.3334,
+      "step": 3500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.223736593445799,
+      "learning_rate": 9.081642134399744e-06,
+      "loss": 0.338,
+      "step": 3501
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.344115079313927,
+      "learning_rate": 9.081053789912513e-06,
+      "loss": 0.3377,
+      "step": 3502
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.746027179157278,
+      "learning_rate": 9.080465276095112e-06,
+      "loss": 0.3448,
+      "step": 3503
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.61044875039412,
+      "learning_rate": 9.079876592971957e-06,
+      "loss": 0.357,
+      "step": 3504
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.74052503939075,
+      "learning_rate": 9.079287740567478e-06,
+      "loss": 0.3373,
+      "step": 3505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1027324826146008,
+      "learning_rate": 9.078698718906103e-06,
+      "loss": 0.3181,
+      "step": 3506
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6315058772643531,
+      "learning_rate": 9.078109528012274e-06,
+      "loss": 0.5071,
+      "step": 3507
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2257699242738753,
+      "learning_rate": 9.077520167910438e-06,
+      "loss": 0.3336,
+      "step": 3508
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.594941072026912,
+      "learning_rate": 9.076930638625047e-06,
+      "loss": 0.3089,
+      "step": 3509
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.551073750864412,
+      "learning_rate": 9.076340940180563e-06,
+      "loss": 0.3648,
+      "step": 3510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.045605593319252,
+      "learning_rate": 9.075751072601453e-06,
+      "loss": 0.3124,
+      "step": 3511
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.974039112999473,
+      "learning_rate": 9.07516103591219e-06,
+      "loss": 0.3474,
+      "step": 3512
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7088514065386926,
+      "learning_rate": 9.07457083013726e-06,
+      "loss": 0.3248,
+      "step": 3513
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.516938561201477,
+      "learning_rate": 9.073980455301148e-06,
+      "loss": 0.3181,
+      "step": 3514
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.129209581638368,
+      "learning_rate": 9.07338991142835e-06,
+      "loss": 0.3183,
+      "step": 3515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8965342970940835,
+      "learning_rate": 9.072799198543369e-06,
+      "loss": 0.3116,
+      "step": 3516
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.422203672342523,
+      "learning_rate": 9.072208316670716e-06,
+      "loss": 0.3219,
+      "step": 3517
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.006018969271595,
+      "learning_rate": 9.071617265834907e-06,
+      "loss": 0.3378,
+      "step": 3518
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4732222221762674,
+      "learning_rate": 9.071026046060465e-06,
+      "loss": 0.3313,
+      "step": 3519
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5821227700492675,
+      "learning_rate": 9.070434657371923e-06,
+      "loss": 0.3191,
+      "step": 3520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1388317281603095,
+      "learning_rate": 9.069843099793815e-06,
+      "loss": 0.3446,
+      "step": 3521
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4189128178981054,
+      "learning_rate": 9.069251373350689e-06,
+      "loss": 0.3223,
+      "step": 3522
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.516764703725452,
+      "learning_rate": 9.068659478067096e-06,
+      "loss": 0.3122,
+      "step": 3523
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.839666827460726,
+      "learning_rate": 9.068067413967594e-06,
+      "loss": 0.3333,
+      "step": 3524
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.475461679296307,
+      "learning_rate": 9.067475181076751e-06,
+      "loss": 0.3266,
+      "step": 3525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.280856693851178,
+      "learning_rate": 9.066882779419135e-06,
+      "loss": 0.3116,
+      "step": 3526
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.340628060280135,
+      "learning_rate": 9.066290209019331e-06,
+      "loss": 0.3141,
+      "step": 3527
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.84375309229679,
+      "learning_rate": 9.065697469901923e-06,
+      "loss": 0.3334,
+      "step": 3528
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.915250938557479,
+      "learning_rate": 9.065104562091506e-06,
+      "loss": 0.3349,
+      "step": 3529
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2244600292329144,
+      "learning_rate": 9.064511485612679e-06,
+      "loss": 0.3215,
+      "step": 3530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.027082296129365,
+      "learning_rate": 9.063918240490052e-06,
+      "loss": 0.321,
+      "step": 3531
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5432314854030387,
+      "learning_rate": 9.063324826748239e-06,
+      "loss": 0.3452,
+      "step": 3532
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.511747029055402,
+      "learning_rate": 9.062731244411862e-06,
+      "loss": 0.3438,
+      "step": 3533
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.897502735314697,
+      "learning_rate": 9.062137493505548e-06,
+      "loss": 0.3331,
+      "step": 3534
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7969701159775067,
+      "learning_rate": 9.061543574053936e-06,
+      "loss": 0.3351,
+      "step": 3535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9887297670828126,
+      "learning_rate": 9.060949486081665e-06,
+      "loss": 0.3247,
+      "step": 3536
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.526286054079545,
+      "learning_rate": 9.060355229613389e-06,
+      "loss": 0.3153,
+      "step": 3537
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.94897289070156,
+      "learning_rate": 9.059760804673761e-06,
+      "loss": 0.355,
+      "step": 3538
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.96042756630533,
+      "learning_rate": 9.059166211287447e-06,
+      "loss": 0.328,
+      "step": 3539
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2121761861262934,
+      "learning_rate": 9.058571449479117e-06,
+      "loss": 0.3284,
+      "step": 3540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.092016249111747,
+      "learning_rate": 9.057976519273448e-06,
+      "loss": 0.3637,
+      "step": 3541
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3464536840370003,
+      "learning_rate": 9.057381420695126e-06,
+      "loss": 0.3295,
+      "step": 3542
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5024058788918464,
+      "learning_rate": 9.056786153768841e-06,
+      "loss": 0.3358,
+      "step": 3543
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.243955832177578,
+      "learning_rate": 9.056190718519295e-06,
+      "loss": 0.3442,
+      "step": 3544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5049918892965763,
+      "learning_rate": 9.05559511497119e-06,
+      "loss": 0.3402,
+      "step": 3545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.46510198512261,
+      "learning_rate": 9.05499934314924e-06,
+      "loss": 0.3071,
+      "step": 3546
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.90580788485331,
+      "learning_rate": 9.054403403078164e-06,
+      "loss": 0.3401,
+      "step": 3547
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.320993086896735,
+      "learning_rate": 9.053807294782692e-06,
+      "loss": 0.3142,
+      "step": 3548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7345491157967383,
+      "learning_rate": 9.053211018287553e-06,
+      "loss": 0.3236,
+      "step": 3549
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.879192789824122,
+      "learning_rate": 9.05261457361749e-06,
+      "loss": 0.3221,
+      "step": 3550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.852462718096176,
+      "learning_rate": 9.05201796079725e-06,
+      "loss": 0.338,
+      "step": 3551
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.012162404812537,
+      "learning_rate": 9.051421179851588e-06,
+      "loss": 0.3126,
+      "step": 3552
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3622073026504316,
+      "learning_rate": 9.050824230805266e-06,
+      "loss": 0.3185,
+      "step": 3553
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.122895476222348,
+      "learning_rate": 9.05022711368305e-06,
+      "loss": 0.3386,
+      "step": 3554
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.988696057976078,
+      "learning_rate": 9.049629828509719e-06,
+      "loss": 0.3559,
+      "step": 3555
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4607113623648877,
+      "learning_rate": 9.04903237531005e-06,
+      "loss": 0.3178,
+      "step": 3556
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.988982176083395,
+      "learning_rate": 9.04843475410884e-06,
+      "loss": 0.3264,
+      "step": 3557
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.371250697174492,
+      "learning_rate": 9.047836964930877e-06,
+      "loss": 0.3374,
+      "step": 3558
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8809763407763955,
+      "learning_rate": 9.047239007800972e-06,
+      "loss": 0.3463,
+      "step": 3559
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.393894186510673,
+      "learning_rate": 9.04664088274393e-06,
+      "loss": 0.3185,
+      "step": 3560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.970437004367653,
+      "learning_rate": 9.04604258978457e-06,
+      "loss": 0.3444,
+      "step": 3561
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.156822946079777,
+      "learning_rate": 9.045444128947719e-06,
+      "loss": 0.3195,
+      "step": 3562
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.363840755745807,
+      "learning_rate": 9.0448455002582e-06,
+      "loss": 0.3487,
+      "step": 3563
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9172663491165216,
+      "learning_rate": 9.044246703740863e-06,
+      "loss": 0.3386,
+      "step": 3564
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.385849456951373,
+      "learning_rate": 9.043647739420543e-06,
+      "loss": 0.3249,
+      "step": 3565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.663101178950992,
+      "learning_rate": 9.043048607322097e-06,
+      "loss": 0.3222,
+      "step": 3566
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.854934296067239,
+      "learning_rate": 9.042449307470384e-06,
+      "loss": 0.3484,
+      "step": 3567
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.6198768133298485,
+      "learning_rate": 9.041849839890267e-06,
+      "loss": 0.34,
+      "step": 3568
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.411133402728843,
+      "learning_rate": 9.041250204606623e-06,
+      "loss": 0.3426,
+      "step": 3569
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.885057626932522,
+      "learning_rate": 9.040650401644329e-06,
+      "loss": 0.3252,
+      "step": 3570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.09092358706227,
+      "learning_rate": 9.040050431028273e-06,
+      "loss": 0.3179,
+      "step": 3571
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.794735501343648,
+      "learning_rate": 9.039450292783349e-06,
+      "loss": 0.3531,
+      "step": 3572
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.384471046150093,
+      "learning_rate": 9.038849986934457e-06,
+      "loss": 0.3264,
+      "step": 3573
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.685652594496571,
+      "learning_rate": 9.038249513506506e-06,
+      "loss": 0.3161,
+      "step": 3574
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6532738353491947,
+      "learning_rate": 9.03764887252441e-06,
+      "loss": 0.3322,
+      "step": 3575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.452544866358713,
+      "learning_rate": 9.037048064013088e-06,
+      "loss": 0.3356,
+      "step": 3576
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 24.05446308649078,
+      "learning_rate": 9.036447087997473e-06,
+      "loss": 0.3459,
+      "step": 3577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.84714724386355,
+      "learning_rate": 9.0358459445025e-06,
+      "loss": 0.3222,
+      "step": 3578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.087490678264344,
+      "learning_rate": 9.035244633553109e-06,
+      "loss": 0.3316,
+      "step": 3579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6993876837380792,
+      "learning_rate": 9.034643155174251e-06,
+      "loss": 0.3244,
+      "step": 3580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8286101056562893,
+      "learning_rate": 9.034041509390884e-06,
+      "loss": 0.3337,
+      "step": 3581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.402972408534468,
+      "learning_rate": 9.033439696227966e-06,
+      "loss": 0.3333,
+      "step": 3582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.200789197971586,
+      "learning_rate": 9.032837715710472e-06,
+      "loss": 0.3221,
+      "step": 3583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.571183544346358,
+      "learning_rate": 9.032235567863379e-06,
+      "loss": 0.328,
+      "step": 3584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6601985659672188,
+      "learning_rate": 9.031633252711669e-06,
+      "loss": 0.5005,
+      "step": 3585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1690557004051634,
+      "learning_rate": 9.031030770280335e-06,
+      "loss": 0.3284,
+      "step": 3586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.490221960830418,
+      "learning_rate": 9.030428120594375e-06,
+      "loss": 0.3033,
+      "step": 3587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.145002496929343,
+      "learning_rate": 9.029825303678794e-06,
+      "loss": 0.325,
+      "step": 3588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.538829016865508,
+      "learning_rate": 9.029222319558602e-06,
+      "loss": 0.3326,
+      "step": 3589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6067011415421141,
+      "learning_rate": 9.028619168258818e-06,
+      "loss": 0.5025,
+      "step": 3590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.701706051279161,
+      "learning_rate": 9.028015849804473e-06,
+      "loss": 0.334,
+      "step": 3591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.2478422724832585,
+      "learning_rate": 9.027412364220592e-06,
+      "loss": 0.3281,
+      "step": 3592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.079626237912594,
+      "learning_rate": 9.02680871153222e-06,
+      "loss": 0.3066,
+      "step": 3593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6004598064573669,
+      "learning_rate": 9.026204891764402e-06,
+      "loss": 0.5165,
+      "step": 3594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.2952808171753825,
+      "learning_rate": 9.025600904942192e-06,
+      "loss": 0.3151,
+      "step": 3595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4729453103217804,
+      "learning_rate": 9.024996751090652e-06,
+      "loss": 0.2955,
+      "step": 3596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.162442025468631,
+      "learning_rate": 9.024392430234843e-06,
+      "loss": 0.3127,
+      "step": 3597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.857648722786493,
+      "learning_rate": 9.023787942399847e-06,
+      "loss": 0.3066,
+      "step": 3598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.573597314277269,
+      "learning_rate": 9.023183287610742e-06,
+      "loss": 0.3216,
+      "step": 3599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7757878378595167,
+      "learning_rate": 9.022578465892616e-06,
+      "loss": 0.3224,
+      "step": 3600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.496563184280213,
+      "learning_rate": 9.021973477270564e-06,
+      "loss": 0.3266,
+      "step": 3601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.343982204643135,
+      "learning_rate": 9.02136832176969e-06,
+      "loss": 0.3165,
+      "step": 3602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9868406128698726,
+      "learning_rate": 9.0207629994151e-06,
+      "loss": 0.3195,
+      "step": 3603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8329817214190003,
+      "learning_rate": 9.020157510231913e-06,
+      "loss": 0.3146,
+      "step": 3604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.514187088615563,
+      "learning_rate": 9.019551854245252e-06,
+      "loss": 0.3298,
+      "step": 3605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.818872750892389,
+      "learning_rate": 9.018946031480242e-06,
+      "loss": 0.3017,
+      "step": 3606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.50794779055897,
+      "learning_rate": 9.018340041962023e-06,
+      "loss": 0.3221,
+      "step": 3607
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6856148215037243,
+      "learning_rate": 9.017733885715738e-06,
+      "loss": 0.5393,
+      "step": 3608
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.187059545194669,
+      "learning_rate": 9.01712756276654e-06,
+      "loss": 0.3584,
+      "step": 3609
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.237515072571624,
+      "learning_rate": 9.01652107313958e-06,
+      "loss": 0.3429,
+      "step": 3610
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.8302884012032035,
+      "learning_rate": 9.01591441686003e-06,
+      "loss": 0.3252,
+      "step": 3611
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8054598641119983,
+      "learning_rate": 9.015307593953058e-06,
+      "loss": 0.3379,
+      "step": 3612
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8162869027273643,
+      "learning_rate": 9.014700604443841e-06,
+      "loss": 0.3201,
+      "step": 3613
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.565828862736785,
+      "learning_rate": 9.014093448357565e-06,
+      "loss": 0.3054,
+      "step": 3614
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.235441939588098,
+      "learning_rate": 9.013486125719421e-06,
+      "loss": 0.3294,
+      "step": 3615
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.50756348779826,
+      "learning_rate": 9.012878636554612e-06,
+      "loss": 0.3289,
+      "step": 3616
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.150400408841957,
+      "learning_rate": 9.012270980888339e-06,
+      "loss": 0.324,
+      "step": 3617
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.8499623137831405,
+      "learning_rate": 9.011663158745815e-06,
+      "loss": 0.3419,
+      "step": 3618
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.099733398191777,
+      "learning_rate": 9.011055170152262e-06,
+      "loss": 0.318,
+      "step": 3619
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0350593646917723,
+      "learning_rate": 9.010447015132906e-06,
+      "loss": 0.3322,
+      "step": 3620
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.38263128775083,
+      "learning_rate": 9.009838693712977e-06,
+      "loss": 0.3161,
+      "step": 3621
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.93563306779475,
+      "learning_rate": 9.009230205917723e-06,
+      "loss": 0.347,
+      "step": 3622
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.199331672094639,
+      "learning_rate": 9.008621551772384e-06,
+      "loss": 0.2994,
+      "step": 3623
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.090170592346973,
+      "learning_rate": 9.008012731302218e-06,
+      "loss": 0.3171,
+      "step": 3624
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6784359594335939,
+      "learning_rate": 9.007403744532483e-06,
+      "loss": 0.507,
+      "step": 3625
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.347774684910878,
+      "learning_rate": 9.00679459148845e-06,
+      "loss": 0.3225,
+      "step": 3626
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.125561745303137,
+      "learning_rate": 9.006185272195392e-06,
+      "loss": 0.3219,
+      "step": 3627
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6468054495308965,
+      "learning_rate": 9.005575786678591e-06,
+      "loss": 0.5253,
+      "step": 3628
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.800250621248189,
+      "learning_rate": 9.004966134963336e-06,
+      "loss": 0.3316,
+      "step": 3629
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.103353805935026,
+      "learning_rate": 9.004356317074923e-06,
+      "loss": 0.3657,
+      "step": 3630
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.928793129887084,
+      "learning_rate": 9.003746333038654e-06,
+      "loss": 0.3305,
+      "step": 3631
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8442002566843962,
+      "learning_rate": 9.003136182879836e-06,
+      "loss": 0.3017,
+      "step": 3632
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.138988018592437,
+      "learning_rate": 9.00252586662379e-06,
+      "loss": 0.3098,
+      "step": 3633
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.17121237081852,
+      "learning_rate": 9.001915384295836e-06,
+      "loss": 0.3109,
+      "step": 3634
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.181123988311583,
+      "learning_rate": 9.001304735921304e-06,
+      "loss": 0.3384,
+      "step": 3635
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.976924379649157,
+      "learning_rate": 9.000693921525532e-06,
+      "loss": 0.3378,
+      "step": 3636
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.55450110106337,
+      "learning_rate": 9.000082941133864e-06,
+      "loss": 0.3183,
+      "step": 3637
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.263468258046588,
+      "learning_rate": 8.999471794771648e-06,
+      "loss": 0.3129,
+      "step": 3638
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2530543059965544,
+      "learning_rate": 8.998860482464243e-06,
+      "loss": 0.3411,
+      "step": 3639
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.108798511236559,
+      "learning_rate": 8.998249004237015e-06,
+      "loss": 0.3337,
+      "step": 3640
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.657061460978529,
+      "learning_rate": 8.997637360115334e-06,
+      "loss": 0.3208,
+      "step": 3641
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.294467745471402,
+      "learning_rate": 8.99702555012458e-06,
+      "loss": 0.3128,
+      "step": 3642
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.162573474260642,
+      "learning_rate": 8.996413574290132e-06,
+      "loss": 0.3148,
+      "step": 3643
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.229211969470677,
+      "learning_rate": 8.99580143263739e-06,
+      "loss": 0.3322,
+      "step": 3644
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3248378481328436,
+      "learning_rate": 8.995189125191747e-06,
+      "loss": 0.3311,
+      "step": 3645
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.46492374513933,
+      "learning_rate": 8.994576651978613e-06,
+      "loss": 0.3179,
+      "step": 3646
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.584422035067047,
+      "learning_rate": 8.993964013023398e-06,
+      "loss": 0.3129,
+      "step": 3647
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8053558546913938,
+      "learning_rate": 8.993351208351521e-06,
+      "loss": 0.4946,
+      "step": 3648
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.749542234067599,
+      "learning_rate": 8.99273823798841e-06,
+      "loss": 0.3234,
+      "step": 3649
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.943498660557402,
+      "learning_rate": 8.992125101959499e-06,
+      "loss": 0.3296,
+      "step": 3650
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.275118711234532,
+      "learning_rate": 8.991511800290224e-06,
+      "loss": 0.351,
+      "step": 3651
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.236552152549496,
+      "learning_rate": 8.990898333006038e-06,
+      "loss": 0.3288,
+      "step": 3652
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.566987854465521,
+      "learning_rate": 8.990284700132388e-06,
+      "loss": 0.3435,
+      "step": 3653
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.753969343465564,
+      "learning_rate": 8.98967090169474e-06,
+      "loss": 0.3323,
+      "step": 3654
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.09232575087877,
+      "learning_rate": 8.98905693771856e-06,
+      "loss": 0.3177,
+      "step": 3655
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.73294460088611,
+      "learning_rate": 8.988442808229321e-06,
+      "loss": 0.3312,
+      "step": 3656
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.124225989766791,
+      "learning_rate": 8.987828513252508e-06,
+      "loss": 0.3328,
+      "step": 3657
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6993878960059834,
+      "learning_rate": 8.987214052813605e-06,
+      "loss": 0.321,
+      "step": 3658
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.571417924358675,
+      "learning_rate": 8.98659942693811e-06,
+      "loss": 0.3178,
+      "step": 3659
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.062422683062402,
+      "learning_rate": 8.985984635651523e-06,
+      "loss": 0.3215,
+      "step": 3660
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.53089332379965,
+      "learning_rate": 8.985369678979351e-06,
+      "loss": 0.3241,
+      "step": 3661
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9476753349579776,
+      "learning_rate": 8.984754556947116e-06,
+      "loss": 0.3145,
+      "step": 3662
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5988612497606813,
+      "learning_rate": 8.984139269580337e-06,
+      "loss": 0.3628,
+      "step": 3663
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.463845682359028,
+      "learning_rate": 8.98352381690454e-06,
+      "loss": 0.328,
+      "step": 3664
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9701219402967818,
+      "learning_rate": 8.982908198945266e-06,
+      "loss": 0.3132,
+      "step": 3665
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.189989455839819,
+      "learning_rate": 8.982292415728057e-06,
+      "loss": 0.3458,
+      "step": 3666
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1810603676710274,
+      "learning_rate": 8.981676467278461e-06,
+      "loss": 0.309,
+      "step": 3667
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.188915244416057,
+      "learning_rate": 8.981060353622037e-06,
+      "loss": 0.3099,
+      "step": 3668
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8310137010128612,
+      "learning_rate": 8.980444074784347e-06,
+      "loss": 0.5004,
+      "step": 3669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.259419586692642,
+      "learning_rate": 8.979827630790962e-06,
+      "loss": 0.3066,
+      "step": 3670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6841209023908836,
+      "learning_rate": 8.979211021667462e-06,
+      "loss": 0.4819,
+      "step": 3671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6469296979890933,
+      "learning_rate": 8.978594247439428e-06,
+      "loss": 0.3148,
+      "step": 3672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 94.27647612541318,
+      "learning_rate": 8.977977308132451e-06,
+      "loss": 0.3161,
+      "step": 3673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.418441586277652,
+      "learning_rate": 8.97736020377213e-06,
+      "loss": 0.3237,
+      "step": 3674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2618596391224317,
+      "learning_rate": 8.976742934384069e-06,
+      "loss": 0.3235,
+      "step": 3675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.700137943136939,
+      "learning_rate": 8.97612549999388e-06,
+      "loss": 0.328,
+      "step": 3676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.443272284273608,
+      "learning_rate": 8.975507900627183e-06,
+      "loss": 0.3398,
+      "step": 3677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.320011875026783,
+      "learning_rate": 8.9748901363096e-06,
+      "loss": 0.3062,
+      "step": 3678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.802621666852393,
+      "learning_rate": 8.974272207066767e-06,
+      "loss": 0.3297,
+      "step": 3679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.487813321902356,
+      "learning_rate": 8.973654112924321e-06,
+      "loss": 0.3389,
+      "step": 3680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.811909074531975,
+      "learning_rate": 8.973035853907906e-06,
+      "loss": 0.3192,
+      "step": 3681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.903283211290379,
+      "learning_rate": 8.972417430043178e-06,
+      "loss": 0.3363,
+      "step": 3682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.701398877082495,
+      "learning_rate": 8.971798841355794e-06,
+      "loss": 0.3427,
+      "step": 3683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.833770578560621,
+      "learning_rate": 8.971180087871423e-06,
+      "loss": 0.3268,
+      "step": 3684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.305994630544555,
+      "learning_rate": 8.970561169615734e-06,
+      "loss": 0.3375,
+      "step": 3685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 26.902374538022954,
+      "learning_rate": 8.969942086614413e-06,
+      "loss": 0.3362,
+      "step": 3686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.1060702742579815,
+      "learning_rate": 8.969322838893141e-06,
+      "loss": 0.3261,
+      "step": 3687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.6775137361151065,
+      "learning_rate": 8.968703426477614e-06,
+      "loss": 0.3114,
+      "step": 3688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3155985014505402,
+      "learning_rate": 8.968083849393535e-06,
+      "loss": 0.3194,
+      "step": 3689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.1287358994957526,
+      "learning_rate": 8.967464107666605e-06,
+      "loss": 0.5333,
+      "step": 3690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3079687843565235,
+      "learning_rate": 8.966844201322546e-06,
+      "loss": 0.3094,
+      "step": 3691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.262970277809321,
+      "learning_rate": 8.966224130387073e-06,
+      "loss": 0.3257,
+      "step": 3692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.5092325938675017,
+      "learning_rate": 8.965603894885917e-06,
+      "loss": 0.3101,
+      "step": 3693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.107537115365177,
+      "learning_rate": 8.96498349484481e-06,
+      "loss": 0.3248,
+      "step": 3694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.978604594760016,
+      "learning_rate": 8.964362930289497e-06,
+      "loss": 0.3308,
+      "step": 3695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.581691378008034,
+      "learning_rate": 8.963742201245725e-06,
+      "loss": 0.3307,
+      "step": 3696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.270998623950874,
+      "learning_rate": 8.963121307739246e-06,
+      "loss": 0.3145,
+      "step": 3697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.328146910350842,
+      "learning_rate": 8.962500249795826e-06,
+      "loss": 0.3495,
+      "step": 3698
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4659047660069,
+      "learning_rate": 8.961879027441234e-06,
+      "loss": 0.3082,
+      "step": 3699
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6379598399266184,
+      "learning_rate": 8.961257640701243e-06,
+      "loss": 0.3328,
+      "step": 3700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.803712893457548,
+      "learning_rate": 8.960636089601636e-06,
+      "loss": 0.309,
+      "step": 3701
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.910081057150349,
+      "learning_rate": 8.960014374168203e-06,
+      "loss": 0.3428,
+      "step": 3702
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.2453245711636765,
+      "learning_rate": 8.959392494426739e-06,
+      "loss": 0.324,
+      "step": 3703
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6834998604875886,
+      "learning_rate": 8.958770450403049e-06,
+      "loss": 0.3078,
+      "step": 3704
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0581316057051184,
+      "learning_rate": 8.958148242122941e-06,
+      "loss": 0.3152,
+      "step": 3705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5118843044790875,
+      "learning_rate": 8.957525869612234e-06,
+      "loss": 0.3574,
+      "step": 3706
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8920246205566102,
+      "learning_rate": 8.956903332896747e-06,
+      "loss": 0.3389,
+      "step": 3707
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.730116226629794,
+      "learning_rate": 8.956280632002312e-06,
+      "loss": 0.312,
+      "step": 3708
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.230495785228844,
+      "learning_rate": 8.955657766954768e-06,
+      "loss": 0.3301,
+      "step": 3709
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.865760992518893,
+      "learning_rate": 8.955034737779955e-06,
+      "loss": 0.3301,
+      "step": 3710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.083156251365915,
+      "learning_rate": 8.95441154450373e-06,
+      "loss": 0.3169,
+      "step": 3711
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8244629017007736,
+      "learning_rate": 8.953788187151941e-06,
+      "loss": 0.3368,
+      "step": 3712
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.459738031389702,
+      "learning_rate": 8.953164665750462e-06,
+      "loss": 0.3271,
+      "step": 3713
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7308846485377833,
+      "learning_rate": 8.952540980325158e-06,
+      "loss": 0.327,
+      "step": 3714
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.181800187862495,
+      "learning_rate": 8.951917130901906e-06,
+      "loss": 0.3441,
+      "step": 3715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1652979984756446,
+      "learning_rate": 8.951293117506595e-06,
+      "loss": 0.3309,
+      "step": 3716
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.781485351811903,
+      "learning_rate": 8.950668940165112e-06,
+      "loss": 0.3367,
+      "step": 3717
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.629968117730134,
+      "learning_rate": 8.95004459890336e-06,
+      "loss": 0.3476,
+      "step": 3718
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1850867382302868,
+      "learning_rate": 8.94942009374724e-06,
+      "loss": 0.3261,
+      "step": 3719
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3544394777063706,
+      "learning_rate": 8.948795424722667e-06,
+      "loss": 0.3375,
+      "step": 3720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.028530381368752,
+      "learning_rate": 8.948170591855556e-06,
+      "loss": 0.3539,
+      "step": 3721
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4090725307856715,
+      "learning_rate": 8.947545595171836e-06,
+      "loss": 0.3172,
+      "step": 3722
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.185929210270183,
+      "learning_rate": 8.946920434697438e-06,
+      "loss": 0.3359,
+      "step": 3723
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6740142148931216,
+      "learning_rate": 8.946295110458298e-06,
+      "loss": 0.3222,
+      "step": 3724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9662346786765055,
+      "learning_rate": 8.945669622480366e-06,
+      "loss": 0.3276,
+      "step": 3725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0089739848021453,
+      "learning_rate": 8.945043970789595e-06,
+      "loss": 0.3392,
+      "step": 3726
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.270920344430561,
+      "learning_rate": 8.94441815541194e-06,
+      "loss": 0.3474,
+      "step": 3727
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2934456477702465,
+      "learning_rate": 8.94379217637337e-06,
+      "loss": 0.3351,
+      "step": 3728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.182996384810266,
+      "learning_rate": 8.943166033699858e-06,
+      "loss": 0.3303,
+      "step": 3729
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6661865474886857,
+      "learning_rate": 8.942539727417383e-06,
+      "loss": 0.3328,
+      "step": 3730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8438046535209291,
+      "learning_rate": 8.941913257551933e-06,
+      "loss": 0.5067,
+      "step": 3731
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.816218547431086,
+      "learning_rate": 8.941286624129499e-06,
+      "loss": 0.3015,
+      "step": 3732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.781419759224102,
+      "learning_rate": 8.940659827176083e-06,
+      "loss": 0.3221,
+      "step": 3733
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.139232330368408,
+      "learning_rate": 8.940032866717691e-06,
+      "loss": 0.3431,
+      "step": 3734
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.040457050157037,
+      "learning_rate": 8.939405742780339e-06,
+      "loss": 0.3329,
+      "step": 3735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.425416944842065,
+      "learning_rate": 8.938778455390044e-06,
+      "loss": 0.3168,
+      "step": 3736
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.55777911347417,
+      "learning_rate": 8.938151004572836e-06,
+      "loss": 0.3279,
+      "step": 3737
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3790259567541567,
+      "learning_rate": 8.937523390354746e-06,
+      "loss": 0.3341,
+      "step": 3738
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4216656743961242,
+      "learning_rate": 8.936895612761817e-06,
+      "loss": 0.3277,
+      "step": 3739
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.976394989991298,
+      "learning_rate": 8.936267671820097e-06,
+      "loss": 0.3328,
+      "step": 3740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.830408235172937,
+      "learning_rate": 8.935639567555639e-06,
+      "loss": 0.3122,
+      "step": 3741
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3137290923268266,
+      "learning_rate": 8.935011299994506e-06,
+      "loss": 0.3483,
+      "step": 3742
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6892055719287272,
+      "learning_rate": 8.934382869162763e-06,
+      "loss": 0.4885,
+      "step": 3743
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3636777971339535,
+      "learning_rate": 8.933754275086487e-06,
+      "loss": 0.3242,
+      "step": 3744
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5710501860164143,
+      "learning_rate": 8.93312551779176e-06,
+      "loss": 0.3491,
+      "step": 3745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2249085380746747,
+      "learning_rate": 8.93249659730467e-06,
+      "loss": 0.3356,
+      "step": 3746
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5807819832528485,
+      "learning_rate": 8.931867513651311e-06,
+      "loss": 0.3285,
+      "step": 3747
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8940150187139984,
+      "learning_rate": 8.931238266857783e-06,
+      "loss": 0.3306,
+      "step": 3748
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.766635051508676,
+      "learning_rate": 8.930608856950198e-06,
+      "loss": 0.313,
+      "step": 3749
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.104603320745064,
+      "learning_rate": 8.92997928395467e-06,
+      "loss": 0.3398,
+      "step": 3750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.647511585459074,
+      "learning_rate": 8.929349547897321e-06,
+      "loss": 0.3167,
+      "step": 3751
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7658197810386245,
+      "learning_rate": 8.92871964880428e-06,
+      "loss": 0.3431,
+      "step": 3752
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5170280748470266,
+      "learning_rate": 8.928089586701682e-06,
+      "loss": 0.3142,
+      "step": 3753
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8799127059400216,
+      "learning_rate": 8.927459361615672e-06,
+      "loss": 0.3149,
+      "step": 3754
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1001750807614474,
+      "learning_rate": 8.926828973572396e-06,
+      "loss": 0.3292,
+      "step": 3755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.834904750639392,
+      "learning_rate": 8.92619842259801e-06,
+      "loss": 0.3167,
+      "step": 3756
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6708523043172825,
+      "learning_rate": 8.925567708718679e-06,
+      "loss": 0.3283,
+      "step": 3757
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.647499695013446,
+      "learning_rate": 8.92493683196057e-06,
+      "loss": 0.3223,
+      "step": 3758
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0279041702713605,
+      "learning_rate": 8.924305792349861e-06,
+      "loss": 0.333,
+      "step": 3759
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.7332960540866695,
+      "learning_rate": 8.923674589912735e-06,
+      "loss": 0.3194,
+      "step": 3760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.432950887421614,
+      "learning_rate": 8.92304322467538e-06,
+      "loss": 0.341,
+      "step": 3761
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.082062228111898,
+      "learning_rate": 8.922411696663992e-06,
+      "loss": 0.3478,
+      "step": 3762
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9000167492809568,
+      "learning_rate": 8.921780005904779e-06,
+      "loss": 0.3429,
+      "step": 3763
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2337721272758104,
+      "learning_rate": 8.921148152423946e-06,
+      "loss": 0.3257,
+      "step": 3764
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.471092327159333,
+      "learning_rate": 8.920516136247712e-06,
+      "loss": 0.3332,
+      "step": 3765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.743527301229908,
+      "learning_rate": 8.9198839574023e-06,
+      "loss": 0.3247,
+      "step": 3766
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.899637854207825,
+      "learning_rate": 8.91925161591394e-06,
+      "loss": 0.3148,
+      "step": 3767
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6936980063567434,
+      "learning_rate": 8.918619111808869e-06,
+      "loss": 0.5216,
+      "step": 3768
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.055760857112053,
+      "learning_rate": 8.917986445113329e-06,
+      "loss": 0.317,
+      "step": 3769
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2410123659660615,
+      "learning_rate": 8.917353615853575e-06,
+      "loss": 0.3434,
+      "step": 3770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.233610095788776,
+      "learning_rate": 8.91672062405586e-06,
+      "loss": 0.3511,
+      "step": 3771
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.05950930626397,
+      "learning_rate": 8.916087469746448e-06,
+      "loss": 0.3216,
+      "step": 3772
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.964420344525788,
+      "learning_rate": 8.915454152951613e-06,
+      "loss": 0.3484,
+      "step": 3773
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.502269758118046,
+      "learning_rate": 8.91482067369763e-06,
+      "loss": 0.3218,
+      "step": 3774
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.546544722679174,
+      "learning_rate": 8.914187032010786e-06,
+      "loss": 0.3439,
+      "step": 3775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9001461345661297,
+      "learning_rate": 8.913553227917366e-06,
+      "loss": 0.3146,
+      "step": 3776
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.895905726844477,
+      "learning_rate": 8.912919261443674e-06,
+      "loss": 0.341,
+      "step": 3777
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.256103512767372,
+      "learning_rate": 8.912285132616012e-06,
+      "loss": 0.3249,
+      "step": 3778
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7257512595344169,
+      "learning_rate": 8.911650841460688e-06,
+      "loss": 0.3137,
+      "step": 3779
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.555245107250307,
+      "learning_rate": 8.911016388004026e-06,
+      "loss": 0.3293,
+      "step": 3780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.247794459675173,
+      "learning_rate": 8.910381772272345e-06,
+      "loss": 0.3238,
+      "step": 3781
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.205686012722772,
+      "learning_rate": 8.90974699429198e-06,
+      "loss": 0.3413,
+      "step": 3782
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.756368157426908,
+      "learning_rate": 8.909112054089266e-06,
+      "loss": 0.3684,
+      "step": 3783
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8866906639618424,
+      "learning_rate": 8.90847695169055e-06,
+      "loss": 0.3259,
+      "step": 3784
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.379658212804618,
+      "learning_rate": 8.907841687122185e-06,
+      "loss": 0.3373,
+      "step": 3785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1938073812990284,
+      "learning_rate": 8.907206260410527e-06,
+      "loss": 0.3446,
+      "step": 3786
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0231368009515407,
+      "learning_rate": 8.906570671581937e-06,
+      "loss": 0.3385,
+      "step": 3787
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.854565265644447,
+      "learning_rate": 8.905934920662796e-06,
+      "loss": 0.3425,
+      "step": 3788
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6267245134019084,
+      "learning_rate": 8.905299007679475e-06,
+      "loss": 0.3313,
+      "step": 3789
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.18832562818506,
+      "learning_rate": 8.904662932658363e-06,
+      "loss": 0.3223,
+      "step": 3790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2337071872214906,
+      "learning_rate": 8.90402669562585e-06,
+      "loss": 0.3307,
+      "step": 3791
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.152753565770728,
+      "learning_rate": 8.903390296608334e-06,
+      "loss": 0.3188,
+      "step": 3792
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4122914444398,
+      "learning_rate": 8.902753735632221e-06,
+      "loss": 0.3407,
+      "step": 3793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.65155176540431,
+      "learning_rate": 8.902117012723926e-06,
+      "loss": 0.3396,
+      "step": 3794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2503548434494247,
+      "learning_rate": 8.901480127909862e-06,
+      "loss": 0.3302,
+      "step": 3795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.316943240218811,
+      "learning_rate": 8.900843081216462e-06,
+      "loss": 0.3291,
+      "step": 3796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2810976801956016,
+      "learning_rate": 8.90020587267015e-06,
+      "loss": 0.3135,
+      "step": 3797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7027236895256603,
+      "learning_rate": 8.89956850229737e-06,
+      "loss": 0.318,
+      "step": 3798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9369146122029532,
+      "learning_rate": 8.898930970124567e-06,
+      "loss": 0.3121,
+      "step": 3799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.284425396522207,
+      "learning_rate": 8.898293276178191e-06,
+      "loss": 0.3415,
+      "step": 3800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.813046511380066,
+      "learning_rate": 8.897655420484705e-06,
+      "loss": 0.3289,
+      "step": 3801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7255512268285473,
+      "learning_rate": 8.897017403070572e-06,
+      "loss": 0.3326,
+      "step": 3802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.884150913605871,
+      "learning_rate": 8.896379223962265e-06,
+      "loss": 0.3316,
+      "step": 3803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4795682828721763,
+      "learning_rate": 8.895740883186263e-06,
+      "loss": 0.3576,
+      "step": 3804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6271732721104906,
+      "learning_rate": 8.895102380769051e-06,
+      "loss": 0.3586,
+      "step": 3805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.08513026610504,
+      "learning_rate": 8.894463716737125e-06,
+      "loss": 0.337,
+      "step": 3806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4815764338416195,
+      "learning_rate": 8.89382489111698e-06,
+      "loss": 0.3277,
+      "step": 3807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.491311962319985,
+      "learning_rate": 8.893185903935125e-06,
+      "loss": 0.3257,
+      "step": 3808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3811724499469147,
+      "learning_rate": 8.892546755218071e-06,
+      "loss": 0.3281,
+      "step": 3809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.7815819883152955,
+      "learning_rate": 8.89190744499234e-06,
+      "loss": 0.3304,
+      "step": 3810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.149020440309964,
+      "learning_rate": 8.891267973284457e-06,
+      "loss": 0.3349,
+      "step": 3811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9390393546693092,
+      "learning_rate": 8.890628340120951e-06,
+      "loss": 0.3374,
+      "step": 3812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.97863421238819,
+      "learning_rate": 8.889988545528366e-06,
+      "loss": 0.3251,
+      "step": 3813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.164845331444898,
+      "learning_rate": 8.889348589533247e-06,
+      "loss": 0.3195,
+      "step": 3814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3626702090731966,
+      "learning_rate": 8.888708472162147e-06,
+      "loss": 0.3169,
+      "step": 3815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7538277936932547,
+      "learning_rate": 8.888068193441625e-06,
+      "loss": 0.3222,
+      "step": 3816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8797303039884392,
+      "learning_rate": 8.887427753398249e-06,
+      "loss": 0.3289,
+      "step": 3817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6560216591031136,
+      "learning_rate": 8.886787152058586e-06,
+      "loss": 0.3268,
+      "step": 3818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9276721444233782,
+      "learning_rate": 8.886146389449226e-06,
+      "loss": 0.3232,
+      "step": 3819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.730994382511737,
+      "learning_rate": 8.885505465596747e-06,
+      "loss": 0.3362,
+      "step": 3820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2764880557056943,
+      "learning_rate": 8.884864380527744e-06,
+      "loss": 0.3371,
+      "step": 3821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.450776452083643,
+      "learning_rate": 8.884223134268818e-06,
+      "loss": 0.3482,
+      "step": 3822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.20623220073423,
+      "learning_rate": 8.883581726846577e-06,
+      "loss": 0.3249,
+      "step": 3823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.187844803081438,
+      "learning_rate": 8.882940158287632e-06,
+      "loss": 0.3042,
+      "step": 3824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7681831053019352,
+      "learning_rate": 8.882298428618601e-06,
+      "loss": 0.3337,
+      "step": 3825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.651769413743395,
+      "learning_rate": 8.881656537866114e-06,
+      "loss": 0.33,
+      "step": 3826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.684097477318379,
+      "learning_rate": 8.8810144860568e-06,
+      "loss": 0.3258,
+      "step": 3827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8215029317151425,
+      "learning_rate": 8.880372273217307e-06,
+      "loss": 0.3073,
+      "step": 3828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.582169091594926,
+      "learning_rate": 8.879729899374269e-06,
+      "loss": 0.3372,
+      "step": 3829
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6347330332007806,
+      "learning_rate": 8.879087364554351e-06,
+      "loss": 0.3361,
+      "step": 3830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2103303600187734,
+      "learning_rate": 8.878444668784208e-06,
+      "loss": 0.3396,
+      "step": 3831
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8936190746169632,
+      "learning_rate": 8.877801812090505e-06,
+      "loss": 0.3371,
+      "step": 3832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.024179134237282,
+      "learning_rate": 8.877158794499919e-06,
+      "loss": 0.3271,
+      "step": 3833
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.577105719299173,
+      "learning_rate": 8.876515616039126e-06,
+      "loss": 0.3583,
+      "step": 3834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7980166390455867,
+      "learning_rate": 8.875872276734816e-06,
+      "loss": 0.3328,
+      "step": 3835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.718324015467852,
+      "learning_rate": 8.87522877661368e-06,
+      "loss": 0.3248,
+      "step": 3836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6030182399426876,
+      "learning_rate": 8.874585115702419e-06,
+      "loss": 0.3337,
+      "step": 3837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.13675666787467,
+      "learning_rate": 8.873941294027742e-06,
+      "loss": 0.3441,
+      "step": 3838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.92834574895222,
+      "learning_rate": 8.873297311616356e-06,
+      "loss": 0.3284,
+      "step": 3839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.09250773906718,
+      "learning_rate": 8.872653168494988e-06,
+      "loss": 0.3543,
+      "step": 3840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0568890325978706,
+      "learning_rate": 8.872008864690358e-06,
+      "loss": 0.3658,
+      "step": 3841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.629916235669576,
+      "learning_rate": 8.871364400229206e-06,
+      "loss": 0.3304,
+      "step": 3842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4120061975792972,
+      "learning_rate": 8.870719775138266e-06,
+      "loss": 0.3273,
+      "step": 3843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4307263050769583,
+      "learning_rate": 8.870074989444289e-06,
+      "loss": 0.3571,
+      "step": 3844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.631956928757783,
+      "learning_rate": 8.869430043174027e-06,
+      "loss": 0.3187,
+      "step": 3845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9659698865753197,
+      "learning_rate": 8.868784936354239e-06,
+      "loss": 0.3204,
+      "step": 3846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3322765690960945,
+      "learning_rate": 8.868139669011693e-06,
+      "loss": 0.3383,
+      "step": 3847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.497011087121863,
+      "learning_rate": 8.86749424117316e-06,
+      "loss": 0.3261,
+      "step": 3848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2073216602945696,
+      "learning_rate": 8.866848652865422e-06,
+      "loss": 0.3473,
+      "step": 3849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.733412826783298,
+      "learning_rate": 8.866202904115265e-06,
+      "loss": 0.3377,
+      "step": 3850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.996459300023647,
+      "learning_rate": 8.865556994949484e-06,
+      "loss": 0.344,
+      "step": 3851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7861627929439978,
+      "learning_rate": 8.864910925394875e-06,
+      "loss": 0.5338,
+      "step": 3852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8191845219998062,
+      "learning_rate": 8.864264695478249e-06,
+      "loss": 0.3341,
+      "step": 3853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6348075695485664,
+      "learning_rate": 8.863618305226415e-06,
+      "loss": 0.3345,
+      "step": 3854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3203016489538917,
+      "learning_rate": 8.862971754666197e-06,
+      "loss": 0.3778,
+      "step": 3855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.023482839826505,
+      "learning_rate": 8.862325043824417e-06,
+      "loss": 0.3353,
+      "step": 3856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0528876842471475,
+      "learning_rate": 8.861678172727912e-06,
+      "loss": 0.3191,
+      "step": 3857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.890524971239111,
+      "learning_rate": 8.861031141403521e-06,
+      "loss": 0.3361,
+      "step": 3858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9241296618676054,
+      "learning_rate": 8.86038394987809e-06,
+      "loss": 0.3414,
+      "step": 3859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.095547267068001,
+      "learning_rate": 8.859736598178472e-06,
+      "loss": 0.3347,
+      "step": 3860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.4156849155693427,
+      "learning_rate": 8.859089086331525e-06,
+      "loss": 0.3499,
+      "step": 3861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.41831697144385,
+      "learning_rate": 8.85844141436412e-06,
+      "loss": 0.3265,
+      "step": 3862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.005906095641753,
+      "learning_rate": 8.857793582303126e-06,
+      "loss": 0.3405,
+      "step": 3863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.968545434221177,
+      "learning_rate": 8.857145590175424e-06,
+      "loss": 0.3602,
+      "step": 3864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7047470613912522,
+      "learning_rate": 8.8564974380079e-06,
+      "loss": 0.4768,
+      "step": 3865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9544057181880778,
+      "learning_rate": 8.855849125827449e-06,
+      "loss": 0.331,
+      "step": 3866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 26.940024139412586,
+      "learning_rate": 8.855200653660968e-06,
+      "loss": 0.3507,
+      "step": 3867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.0242492038546995,
+      "learning_rate": 8.854552021535364e-06,
+      "loss": 0.349,
+      "step": 3868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.741353559421358,
+      "learning_rate": 8.853903229477551e-06,
+      "loss": 0.3585,
+      "step": 3869
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6038561947185022,
+      "learning_rate": 8.853254277514448e-06,
+      "loss": 0.5322,
+      "step": 3870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1505945163444187,
+      "learning_rate": 8.852605165672978e-06,
+      "loss": 0.325,
+      "step": 3871
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.921557449567707,
+      "learning_rate": 8.85195589398008e-06,
+      "loss": 0.3531,
+      "step": 3872
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8701218681116438,
+      "learning_rate": 8.851306462462689e-06,
+      "loss": 0.3349,
+      "step": 3873
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.742681666858432,
+      "learning_rate": 8.850656871147751e-06,
+      "loss": 0.3391,
+      "step": 3874
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2987565301394373,
+      "learning_rate": 8.850007120062222e-06,
+      "loss": 0.328,
+      "step": 3875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.915574876506993,
+      "learning_rate": 8.849357209233058e-06,
+      "loss": 0.3304,
+      "step": 3876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.592669304463846,
+      "learning_rate": 8.848707138687227e-06,
+      "loss": 0.3223,
+      "step": 3877
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.59090898074836,
+      "learning_rate": 8.848056908451698e-06,
+      "loss": 0.3329,
+      "step": 3878
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.986947695461738,
+      "learning_rate": 8.847406518553456e-06,
+      "loss": 0.3285,
+      "step": 3879
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0903116730798996,
+      "learning_rate": 8.846755969019483e-06,
+      "loss": 0.3604,
+      "step": 3880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.156612702728645,
+      "learning_rate": 8.846105259876771e-06,
+      "loss": 0.3292,
+      "step": 3881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.6536210372053626,
+      "learning_rate": 8.845454391152322e-06,
+      "loss": 0.3352,
+      "step": 3882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.264987913288185,
+      "learning_rate": 8.84480336287314e-06,
+      "loss": 0.3223,
+      "step": 3883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.925664752201105,
+      "learning_rate": 8.844152175066236e-06,
+      "loss": 0.3229,
+      "step": 3884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2169062660615215,
+      "learning_rate": 8.84350082775863e-06,
+      "loss": 0.3523,
+      "step": 3885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.766125838423153,
+      "learning_rate": 8.84284932097735e-06,
+      "loss": 0.3239,
+      "step": 3886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.894999079590576,
+      "learning_rate": 8.842197654749424e-06,
+      "loss": 0.3214,
+      "step": 3887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.061528927370333,
+      "learning_rate": 8.841545829101895e-06,
+      "loss": 0.3276,
+      "step": 3888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.705359907076887,
+      "learning_rate": 8.840893844061804e-06,
+      "loss": 0.3557,
+      "step": 3889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.401023642036005,
+      "learning_rate": 8.840241699656207e-06,
+      "loss": 0.3414,
+      "step": 3890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5443030138097336,
+      "learning_rate": 8.839589395912159e-06,
+      "loss": 0.3158,
+      "step": 3891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.336497941852103,
+      "learning_rate": 8.838936932856727e-06,
+      "loss": 0.3442,
+      "step": 3892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 24.628950073411175,
+      "learning_rate": 8.838284310516985e-06,
+      "loss": 0.3462,
+      "step": 3893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9694358642533767,
+      "learning_rate": 8.837631528920008e-06,
+      "loss": 0.3383,
+      "step": 3894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9458972961783314,
+      "learning_rate": 8.836978588092883e-06,
+      "loss": 0.3158,
+      "step": 3895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8284781128041392,
+      "learning_rate": 8.8363254880627e-06,
+      "loss": 0.315,
+      "step": 3896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.378205330626439,
+      "learning_rate": 8.83567222885656e-06,
+      "loss": 0.3201,
+      "step": 3897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.530432467524255,
+      "learning_rate": 8.835018810501565e-06,
+      "loss": 0.3473,
+      "step": 3898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0424361820949244,
+      "learning_rate": 8.834365233024829e-06,
+      "loss": 0.3436,
+      "step": 3899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1196827075666516,
+      "learning_rate": 8.833711496453468e-06,
+      "loss": 0.3393,
+      "step": 3900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.952618656269322,
+      "learning_rate": 8.833057600814607e-06,
+      "loss": 0.3624,
+      "step": 3901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.440008519259911,
+      "learning_rate": 8.832403546135379e-06,
+      "loss": 0.3372,
+      "step": 3902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.41174185418652,
+      "learning_rate": 8.831749332442921e-06,
+      "loss": 0.3237,
+      "step": 3903
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2010395801925373,
+      "learning_rate": 8.831094959764377e-06,
+      "loss": 0.3481,
+      "step": 3904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.583312141968117,
+      "learning_rate": 8.830440428126898e-06,
+      "loss": 0.3238,
+      "step": 3905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6722910462704026,
+      "learning_rate": 8.829785737557642e-06,
+      "loss": 0.4975,
+      "step": 3906
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8359111684459553,
+      "learning_rate": 8.829130888083774e-06,
+      "loss": 0.3305,
+      "step": 3907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0463744849386925,
+      "learning_rate": 8.828475879732463e-06,
+      "loss": 0.3269,
+      "step": 3908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.855492911832764,
+      "learning_rate": 8.827820712530888e-06,
+      "loss": 0.3249,
+      "step": 3909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4209041408111671,
+      "learning_rate": 8.827165386506233e-06,
+      "loss": 0.3198,
+      "step": 3910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.925380030191963,
+      "learning_rate": 8.826509901685689e-06,
+      "loss": 0.3197,
+      "step": 3911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2511658113980753,
+      "learning_rate": 8.82585425809645e-06,
+      "loss": 0.3525,
+      "step": 3912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.460079825444177,
+      "learning_rate": 8.825198455765724e-06,
+      "loss": 0.3389,
+      "step": 3913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9491557044021177,
+      "learning_rate": 8.824542494720721e-06,
+      "loss": 0.3192,
+      "step": 3914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9619002074501823,
+      "learning_rate": 8.823886374988655e-06,
+      "loss": 0.349,
+      "step": 3915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5782371474866366,
+      "learning_rate": 8.823230096596751e-06,
+      "loss": 0.3208,
+      "step": 3916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7775910976015772,
+      "learning_rate": 8.82257365957224e-06,
+      "loss": 0.341,
+      "step": 3917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.419382613452219,
+      "learning_rate": 8.821917063942359e-06,
+      "loss": 0.3168,
+      "step": 3918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2680541421652127,
+      "learning_rate": 8.82126030973435e-06,
+      "loss": 0.3577,
+      "step": 3919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3986296824244648,
+      "learning_rate": 8.820603396975463e-06,
+      "loss": 0.3231,
+      "step": 3920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.792165517815148,
+      "learning_rate": 8.819946325692955e-06,
+      "loss": 0.3388,
+      "step": 3921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9036276897324675,
+      "learning_rate": 8.81928909591409e-06,
+      "loss": 0.3421,
+      "step": 3922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5755263412413703,
+      "learning_rate": 8.818631707666136e-06,
+      "loss": 0.3314,
+      "step": 3923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.905918617029896,
+      "learning_rate": 8.817974160976368e-06,
+      "loss": 0.3293,
+      "step": 3924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.074687240108824,
+      "learning_rate": 8.817316455872073e-06,
+      "loss": 0.3397,
+      "step": 3925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6556156451209951,
+      "learning_rate": 8.816658592380538e-06,
+      "loss": 0.3581,
+      "step": 3926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5225123949991586,
+      "learning_rate": 8.816000570529057e-06,
+      "loss": 0.3395,
+      "step": 3927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8796581244586301,
+      "learning_rate": 8.815342390344937e-06,
+      "loss": 0.3456,
+      "step": 3928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6006706301784175,
+      "learning_rate": 8.814684051855482e-06,
+      "loss": 0.339,
+      "step": 3929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.756539450428205,
+      "learning_rate": 8.814025555088011e-06,
+      "loss": 0.3544,
+      "step": 3930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0715619432674304,
+      "learning_rate": 8.813366900069844e-06,
+      "loss": 0.3456,
+      "step": 3931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.210278623420081,
+      "learning_rate": 8.812708086828312e-06,
+      "loss": 0.3466,
+      "step": 3932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6949532304163624,
+      "learning_rate": 8.81204911539075e-06,
+      "loss": 0.337,
+      "step": 3933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4190388446630333,
+      "learning_rate": 8.8113899857845e-06,
+      "loss": 0.3304,
+      "step": 3934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.012773969913957,
+      "learning_rate": 8.810730698036906e-06,
+      "loss": 0.3242,
+      "step": 3935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0562977370739435,
+      "learning_rate": 8.81007125217533e-06,
+      "loss": 0.3797,
+      "step": 3936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.130834954945286,
+      "learning_rate": 8.809411648227129e-06,
+      "loss": 0.3616,
+      "step": 3937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.227304738078117,
+      "learning_rate": 8.808751886219673e-06,
+      "loss": 0.349,
+      "step": 3938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.147724538836497,
+      "learning_rate": 8.808091966180337e-06,
+      "loss": 0.3456,
+      "step": 3939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.554994194092664,
+      "learning_rate": 8.8074318881365e-06,
+      "loss": 0.3311,
+      "step": 3940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.822323489634922,
+      "learning_rate": 8.80677165211555e-06,
+      "loss": 0.3178,
+      "step": 3941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.60007168743747,
+      "learning_rate": 8.806111258144885e-06,
+      "loss": 0.3416,
+      "step": 3942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2015769905986544,
+      "learning_rate": 8.805450706251903e-06,
+      "loss": 0.3459,
+      "step": 3943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2836085210108785,
+      "learning_rate": 8.804789996464012e-06,
+      "loss": 0.3469,
+      "step": 3944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.745469579523758,
+      "learning_rate": 8.804129128808626e-06,
+      "loss": 0.3458,
+      "step": 3945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5482381798668035,
+      "learning_rate": 8.803468103313165e-06,
+      "loss": 0.3205,
+      "step": 3946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9689575422993166,
+      "learning_rate": 8.802806920005055e-06,
+      "loss": 0.3585,
+      "step": 3947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.139807856050084,
+      "learning_rate": 8.802145578911733e-06,
+      "loss": 0.3171,
+      "step": 3948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5883947555878855,
+      "learning_rate": 8.801484080060638e-06,
+      "loss": 0.3435,
+      "step": 3949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9604793929364766,
+      "learning_rate": 8.800822423479217e-06,
+      "loss": 0.3577,
+      "step": 3950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.927768363210243,
+      "learning_rate": 8.800160609194922e-06,
+      "loss": 0.3428,
+      "step": 3951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8353010270690588,
+      "learning_rate": 8.799498637235213e-06,
+      "loss": 0.3635,
+      "step": 3952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.047035962116984,
+      "learning_rate": 8.798836507627556e-06,
+      "loss": 0.3295,
+      "step": 3953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.977250737729621,
+      "learning_rate": 8.798174220399427e-06,
+      "loss": 0.361,
+      "step": 3954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8191323300886468,
+      "learning_rate": 8.7975117755783e-06,
+      "loss": 0.322,
+      "step": 3955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2833073536674253,
+      "learning_rate": 8.796849173191669e-06,
+      "loss": 0.3257,
+      "step": 3956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.653580360973125,
+      "learning_rate": 8.796186413267018e-06,
+      "loss": 0.3586,
+      "step": 3957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5959191441071698,
+      "learning_rate": 8.79552349583185e-06,
+      "loss": 0.3352,
+      "step": 3958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.68755513942267,
+      "learning_rate": 8.794860420913672e-06,
+      "loss": 0.3353,
+      "step": 3959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.290039827976815,
+      "learning_rate": 8.794197188539994e-06,
+      "loss": 0.3505,
+      "step": 3960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6982893771975944,
+      "learning_rate": 8.793533798738335e-06,
+      "loss": 0.3631,
+      "step": 3961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.368724845872284,
+      "learning_rate": 8.792870251536223e-06,
+      "loss": 0.3364,
+      "step": 3962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3437852662567056,
+      "learning_rate": 8.792206546961187e-06,
+      "loss": 0.3339,
+      "step": 3963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.379021111034225,
+      "learning_rate": 8.791542685040764e-06,
+      "loss": 0.331,
+      "step": 3964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.468420002441763,
+      "learning_rate": 8.7908786658025e-06,
+      "loss": 0.3697,
+      "step": 3965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.40442348480613,
+      "learning_rate": 8.790214489273948e-06,
+      "loss": 0.3239,
+      "step": 3966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.443752448014104,
+      "learning_rate": 8.789550155482665e-06,
+      "loss": 0.336,
+      "step": 3967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.423170706382684,
+      "learning_rate": 8.788885664456212e-06,
+      "loss": 0.3522,
+      "step": 3968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0159703741815638,
+      "learning_rate": 8.788221016222167e-06,
+      "loss": 0.327,
+      "step": 3969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.413832410196219,
+      "learning_rate": 8.787556210808101e-06,
+      "loss": 0.3159,
+      "step": 3970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.229140339814073,
+      "learning_rate": 8.786891248241602e-06,
+      "loss": 0.3182,
+      "step": 3971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8559889221850456,
+      "learning_rate": 8.786226128550257e-06,
+      "loss": 0.3232,
+      "step": 3972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7735259003246646,
+      "learning_rate": 8.785560851761666e-06,
+      "loss": 0.3398,
+      "step": 3973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1257374329945917,
+      "learning_rate": 8.784895417903431e-06,
+      "loss": 0.318,
+      "step": 3974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.20304958642847,
+      "learning_rate": 8.784229827003163e-06,
+      "loss": 0.3598,
+      "step": 3975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8481307769447963,
+      "learning_rate": 8.783564079088478e-06,
+      "loss": 0.3366,
+      "step": 3976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4349804126816066,
+      "learning_rate": 8.782898174186998e-06,
+      "loss": 0.3247,
+      "step": 3977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6125703437228829,
+      "learning_rate": 8.782232112326353e-06,
+      "loss": 0.3405,
+      "step": 3978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7563070234189557,
+      "learning_rate": 8.78156589353418e-06,
+      "loss": 0.3517,
+      "step": 3979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.158906279170565,
+      "learning_rate": 8.780899517838122e-06,
+      "loss": 0.3371,
+      "step": 3980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.023815966204676,
+      "learning_rate": 8.78023298526583e-06,
+      "loss": 0.3413,
+      "step": 3981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7695617599975126,
+      "learning_rate": 8.779566295844953e-06,
+      "loss": 0.3358,
+      "step": 3982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8670330030267603,
+      "learning_rate": 8.778899449603158e-06,
+      "loss": 0.345,
+      "step": 3983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7119260387214954,
+      "learning_rate": 8.778232446568114e-06,
+      "loss": 0.3423,
+      "step": 3984
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6557380312202785,
+      "learning_rate": 8.777565286767493e-06,
+      "loss": 0.3149,
+      "step": 3985
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.031829735179775,
+      "learning_rate": 8.776897970228979e-06,
+      "loss": 0.3229,
+      "step": 3986
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6405112659555778,
+      "learning_rate": 8.77623049698026e-06,
+      "loss": 0.3264,
+      "step": 3987
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.650907873991567,
+      "learning_rate": 8.775562867049033e-06,
+      "loss": 0.3352,
+      "step": 3988
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.377611140084689,
+      "learning_rate": 8.774895080462993e-06,
+      "loss": 0.3464,
+      "step": 3989
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0190881235297744,
+      "learning_rate": 8.774227137249851e-06,
+      "loss": 0.3336,
+      "step": 3990
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8384397193429323,
+      "learning_rate": 8.773559037437323e-06,
+      "loss": 0.3244,
+      "step": 3991
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.938548809288756,
+      "learning_rate": 8.772890781053127e-06,
+      "loss": 0.3227,
+      "step": 3992
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.931703895266037,
+      "learning_rate": 8.772222368124992e-06,
+      "loss": 0.3223,
+      "step": 3993
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6433144475496635,
+      "learning_rate": 8.77155379868065e-06,
+      "loss": 0.3265,
+      "step": 3994
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0306364090676436,
+      "learning_rate": 8.770885072747843e-06,
+      "loss": 0.3393,
+      "step": 3995
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7653982096852402,
+      "learning_rate": 8.770216190354316e-06,
+      "loss": 0.515,
+      "step": 3996
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.755430314122129,
+      "learning_rate": 8.769547151527821e-06,
+      "loss": 0.3149,
+      "step": 3997
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1910043438408513,
+      "learning_rate": 8.76887795629612e-06,
+      "loss": 0.3399,
+      "step": 3998
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1466922075478108,
+      "learning_rate": 8.768208604686979e-06,
+      "loss": 0.3409,
+      "step": 3999
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.321892658536272,
+      "learning_rate": 8.767539096728171e-06,
+      "loss": 0.3238,
+      "step": 4000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.00755075469908,
+      "learning_rate": 8.766869432447473e-06,
+      "loss": 0.3231,
+      "step": 4001
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4972781413131402,
+      "learning_rate": 8.76619961187267e-06,
+      "loss": 0.3367,
+      "step": 4002
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.700095470221592,
+      "learning_rate": 8.765529635031556e-06,
+      "loss": 0.3504,
+      "step": 4003
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.124431337435667,
+      "learning_rate": 8.76485950195193e-06,
+      "loss": 0.321,
+      "step": 4004
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.177660901477263,
+      "learning_rate": 8.764189212661597e-06,
+      "loss": 0.3545,
+      "step": 4005
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9543874288575571,
+      "learning_rate": 8.763518767188367e-06,
+      "loss": 0.3304,
+      "step": 4006
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.040240741145905,
+      "learning_rate": 8.762848165560057e-06,
+      "loss": 0.3569,
+      "step": 4007
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1462098143685076,
+      "learning_rate": 8.762177407804495e-06,
+      "loss": 0.3365,
+      "step": 4008
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3557725363827984,
+      "learning_rate": 8.761506493949508e-06,
+      "loss": 0.3272,
+      "step": 4009
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.049154870315858,
+      "learning_rate": 8.760835424022938e-06,
+      "loss": 0.3453,
+      "step": 4010
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.909236488879323,
+      "learning_rate": 8.760164198052625e-06,
+      "loss": 0.3437,
+      "step": 4011
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.986681260659995,
+      "learning_rate": 8.75949281606642e-06,
+      "loss": 0.3473,
+      "step": 4012
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4724205262807475,
+      "learning_rate": 8.75882127809218e-06,
+      "loss": 0.3649,
+      "step": 4013
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3375703560766286,
+      "learning_rate": 8.758149584157772e-06,
+      "loss": 0.3439,
+      "step": 4014
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7335786317428861,
+      "learning_rate": 8.757477734291059e-06,
+      "loss": 0.5261,
+      "step": 4015
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.69490251316695,
+      "learning_rate": 8.756805728519922e-06,
+      "loss": 0.3438,
+      "step": 4016
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.163874147734496,
+      "learning_rate": 8.756133566872242e-06,
+      "loss": 0.3544,
+      "step": 4017
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7208233723521968,
+      "learning_rate": 8.75546124937591e-06,
+      "loss": 0.3189,
+      "step": 4018
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8147174898790515,
+      "learning_rate": 8.754788776058818e-06,
+      "loss": 0.3648,
+      "step": 4019
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.965910224778552,
+      "learning_rate": 8.754116146948873e-06,
+      "loss": 0.31,
+      "step": 4020
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.515679447901728,
+      "learning_rate": 8.753443362073978e-06,
+      "loss": 0.3281,
+      "step": 4021
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3864591791292553,
+      "learning_rate": 8.752770421462053e-06,
+      "loss": 0.3678,
+      "step": 4022
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.511879566043592,
+      "learning_rate": 8.752097325141017e-06,
+      "loss": 0.3656,
+      "step": 4023
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1900113369588854,
+      "learning_rate": 8.751424073138801e-06,
+      "loss": 0.3317,
+      "step": 4024
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9244560198108775,
+      "learning_rate": 8.750750665483333e-06,
+      "loss": 0.3381,
+      "step": 4025
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6220897907902792,
+      "learning_rate": 8.750077102202562e-06,
+      "loss": 0.3326,
+      "step": 4026
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.205144572196654,
+      "learning_rate": 8.749403383324428e-06,
+      "loss": 0.3267,
+      "step": 4027
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.67607270989675,
+      "learning_rate": 8.748729508876889e-06,
+      "loss": 0.3331,
+      "step": 4028
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6842191927654142,
+      "learning_rate": 8.748055478887905e-06,
+      "loss": 0.5013,
+      "step": 4029
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4214965308143919,
+      "learning_rate": 8.74738129338544e-06,
+      "loss": 0.347,
+      "step": 4030
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8626450857396741,
+      "learning_rate": 8.74670695239747e-06,
+      "loss": 0.3354,
+      "step": 4031
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7129575257256846,
+      "learning_rate": 8.746032455951972e-06,
+      "loss": 0.3427,
+      "step": 4032
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.894176120050417,
+      "learning_rate": 8.745357804076937e-06,
+      "loss": 0.3323,
+      "step": 4033
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4032652273676638,
+      "learning_rate": 8.744682996800351e-06,
+      "loss": 0.3469,
+      "step": 4034
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8921606911624178,
+      "learning_rate": 8.744008034150217e-06,
+      "loss": 0.3188,
+      "step": 4035
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4596220130329085,
+      "learning_rate": 8.743332916154541e-06,
+      "loss": 0.3198,
+      "step": 4036
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6352768292601403,
+      "learning_rate": 8.742657642841333e-06,
+      "loss": 0.3366,
+      "step": 4037
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3039448632668107,
+      "learning_rate": 8.741982214238611e-06,
+      "loss": 0.3437,
+      "step": 4038
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0927802688428327,
+      "learning_rate": 8.7413066303744e-06,
+      "loss": 0.3442,
+      "step": 4039
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.535874207421677,
+      "learning_rate": 8.740630891276733e-06,
+      "loss": 0.3189,
+      "step": 4040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.388702468093148,
+      "learning_rate": 8.739954996973646e-06,
+      "loss": 0.3353,
+      "step": 4041
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6144431932867787,
+      "learning_rate": 8.739278947493182e-06,
+      "loss": 0.3352,
+      "step": 4042
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.568850958839546,
+      "learning_rate": 8.738602742863394e-06,
+      "loss": 0.3593,
+      "step": 4043
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5634723197923737,
+      "learning_rate": 8.737926383112336e-06,
+      "loss": 0.3541,
+      "step": 4044
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9645900527362579,
+      "learning_rate": 8.737249868268076e-06,
+      "loss": 0.3212,
+      "step": 4045
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8098296651161077,
+      "learning_rate": 8.736573198358677e-06,
+      "loss": 0.3475,
+      "step": 4046
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.7698301785311354,
+      "learning_rate": 8.735896373412223e-06,
+      "loss": 0.333,
+      "step": 4047
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.849821522211834,
+      "learning_rate": 8.735219393456791e-06,
+      "loss": 0.3535,
+      "step": 4048
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.854147709811618,
+      "learning_rate": 8.73454225852047e-06,
+      "loss": 0.3459,
+      "step": 4049
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.4135048040720535,
+      "learning_rate": 8.73386496863136e-06,
+      "loss": 0.3489,
+      "step": 4050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2122085838660452,
+      "learning_rate": 8.73318752381756e-06,
+      "loss": 0.3629,
+      "step": 4051
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5692931435847661,
+      "learning_rate": 8.732509924107178e-06,
+      "loss": 0.3483,
+      "step": 4052
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.278172906750455,
+      "learning_rate": 8.731832169528331e-06,
+      "loss": 0.3405,
+      "step": 4053
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6080014153388025,
+      "learning_rate": 8.731154260109137e-06,
+      "loss": 0.3407,
+      "step": 4054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.348106601953248,
+      "learning_rate": 8.730476195877726e-06,
+      "loss": 0.3567,
+      "step": 4055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0510569170967754,
+      "learning_rate": 8.729797976862231e-06,
+      "loss": 0.3328,
+      "step": 4056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5894392806231072,
+      "learning_rate": 8.729119603090794e-06,
+      "loss": 0.3264,
+      "step": 4057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.073645775213012,
+      "learning_rate": 8.728441074591561e-06,
+      "loss": 0.3173,
+      "step": 4058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.54799909527397,
+      "learning_rate": 8.727762391392685e-06,
+      "loss": 0.3499,
+      "step": 4059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4898635989281481,
+      "learning_rate": 8.727083553522326e-06,
+      "loss": 0.3271,
+      "step": 4060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6195957426803977,
+      "learning_rate": 8.726404561008652e-06,
+      "loss": 0.34,
+      "step": 4061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4569605547413564,
+      "learning_rate": 8.725725413879833e-06,
+      "loss": 0.3474,
+      "step": 4062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.63672464437792,
+      "learning_rate": 8.725046112164048e-06,
+      "loss": 0.3409,
+      "step": 4063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.07783384634387,
+      "learning_rate": 8.724366655889484e-06,
+      "loss": 0.339,
+      "step": 4064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5052048789526022,
+      "learning_rate": 8.723687045084332e-06,
+      "loss": 0.3447,
+      "step": 4065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0756908514939045,
+      "learning_rate": 8.72300727977679e-06,
+      "loss": 0.3434,
+      "step": 4066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5933030953941294,
+      "learning_rate": 8.722327359995064e-06,
+      "loss": 0.3338,
+      "step": 4067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.953815711101553,
+      "learning_rate": 8.721647285767364e-06,
+      "loss": 0.3369,
+      "step": 4068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0628212915236666,
+      "learning_rate": 8.720967057121908e-06,
+      "loss": 0.3524,
+      "step": 4069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1094032642905916,
+      "learning_rate": 8.720286674086919e-06,
+      "loss": 0.3364,
+      "step": 4070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.29791883779948,
+      "learning_rate": 8.719606136690628e-06,
+      "loss": 0.335,
+      "step": 4071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9666908533427165,
+      "learning_rate": 8.718925444961273e-06,
+      "loss": 0.3611,
+      "step": 4072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5340981209677524,
+      "learning_rate": 8.718244598927095e-06,
+      "loss": 0.361,
+      "step": 4073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5327835452053369,
+      "learning_rate": 8.717563598616342e-06,
+      "loss": 0.3152,
+      "step": 4074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8868073105187637,
+      "learning_rate": 8.716882444057275e-06,
+      "loss": 0.356,
+      "step": 4075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0230935669880217,
+      "learning_rate": 8.716201135278152e-06,
+      "loss": 0.3268,
+      "step": 4076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0560875190386207,
+      "learning_rate": 8.715519672307245e-06,
+      "loss": 0.3375,
+      "step": 4077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0900475278362407,
+      "learning_rate": 8.714838055172825e-06,
+      "loss": 0.3439,
+      "step": 4078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0781631564196195,
+      "learning_rate": 8.714156283903177e-06,
+      "loss": 0.3391,
+      "step": 4079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.270042035954629,
+      "learning_rate": 8.713474358526588e-06,
+      "loss": 0.332,
+      "step": 4080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7474073994956407,
+      "learning_rate": 8.712792279071351e-06,
+      "loss": 0.3173,
+      "step": 4081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6378876500682436,
+      "learning_rate": 8.712110045565768e-06,
+      "loss": 0.3154,
+      "step": 4082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4541175416986114,
+      "learning_rate": 8.711427658038146e-06,
+      "loss": 0.3295,
+      "step": 4083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7196100230078464,
+      "learning_rate": 8.710745116516798e-06,
+      "loss": 0.351,
+      "step": 4084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3541906105401913,
+      "learning_rate": 8.710062421030043e-06,
+      "loss": 0.3602,
+      "step": 4085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1103349288170503,
+      "learning_rate": 8.709379571606211e-06,
+      "loss": 0.3392,
+      "step": 4086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.676024096888773,
+      "learning_rate": 8.708696568273631e-06,
+      "loss": 0.3126,
+      "step": 4087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.769694033779992,
+      "learning_rate": 8.708013411060642e-06,
+      "loss": 0.3628,
+      "step": 4088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.932547398377358,
+      "learning_rate": 8.70733009999559e-06,
+      "loss": 0.3451,
+      "step": 4089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.156018069717261,
+      "learning_rate": 8.706646635106828e-06,
+      "loss": 0.3339,
+      "step": 4090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.480207759245538,
+      "learning_rate": 8.705963016422713e-06,
+      "loss": 0.3145,
+      "step": 4091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9827084793169916,
+      "learning_rate": 8.705279243971612e-06,
+      "loss": 0.3582,
+      "step": 4092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8554133354964357,
+      "learning_rate": 8.704595317781892e-06,
+      "loss": 0.3173,
+      "step": 4093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.399401452220164,
+      "learning_rate": 8.703911237881932e-06,
+      "loss": 0.3295,
+      "step": 4094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.775914267033628,
+      "learning_rate": 8.703227004300117e-06,
+      "loss": 0.3348,
+      "step": 4095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7040139261096963,
+      "learning_rate": 8.702542617064835e-06,
+      "loss": 0.3288,
+      "step": 4096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0923667917542046,
+      "learning_rate": 8.701858076204484e-06,
+      "loss": 0.3464,
+      "step": 4097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7426022535152126,
+      "learning_rate": 8.701173381747465e-06,
+      "loss": 0.3281,
+      "step": 4098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7515229375498405,
+      "learning_rate": 8.700488533722189e-06,
+      "loss": 0.3328,
+      "step": 4099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.783669089310867,
+      "learning_rate": 8.699803532157071e-06,
+      "loss": 0.3502,
+      "step": 4100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3042027517051666,
+      "learning_rate": 8.699118377080534e-06,
+      "loss": 0.3361,
+      "step": 4101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1278033478288436,
+      "learning_rate": 8.698433068521005e-06,
+      "loss": 0.3432,
+      "step": 4102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7118264209442484,
+      "learning_rate": 8.697747606506917e-06,
+      "loss": 0.355,
+      "step": 4103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8498136601044612,
+      "learning_rate": 8.697061991066712e-06,
+      "loss": 0.3486,
+      "step": 4104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.732509917874962,
+      "learning_rate": 8.696376222228841e-06,
+      "loss": 0.3322,
+      "step": 4105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3786558315606947,
+      "learning_rate": 8.695690300021755e-06,
+      "loss": 0.3463,
+      "step": 4106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6395221922436516,
+      "learning_rate": 8.695004224473912e-06,
+      "loss": 0.3354,
+      "step": 4107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2973049883443966,
+      "learning_rate": 8.69431799561378e-06,
+      "loss": 0.3417,
+      "step": 4108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8760467049793943,
+      "learning_rate": 8.693631613469837e-06,
+      "loss": 0.325,
+      "step": 4109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.086697428287031,
+      "learning_rate": 8.692945078070555e-06,
+      "loss": 0.3413,
+      "step": 4110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7362402259724723,
+      "learning_rate": 8.692258389444422e-06,
+      "loss": 0.3261,
+      "step": 4111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7640477171220377,
+      "learning_rate": 8.69157154761993e-06,
+      "loss": 0.3454,
+      "step": 4112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.701332166606028,
+      "learning_rate": 8.690884552625579e-06,
+      "loss": 0.3194,
+      "step": 4113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1595312033463734,
+      "learning_rate": 8.690197404489871e-06,
+      "loss": 0.3383,
+      "step": 4114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.520544858801723,
+      "learning_rate": 8.689510103241318e-06,
+      "loss": 0.3359,
+      "step": 4115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5493797661696136,
+      "learning_rate": 8.68882264890844e-06,
+      "loss": 0.3332,
+      "step": 4116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3290404208472024,
+      "learning_rate": 8.688135041519756e-06,
+      "loss": 0.358,
+      "step": 4117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1260846658752963,
+      "learning_rate": 8.687447281103799e-06,
+      "loss": 0.3466,
+      "step": 4118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3654767099177816,
+      "learning_rate": 8.686759367689102e-06,
+      "loss": 0.3299,
+      "step": 4119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6931618242551096,
+      "learning_rate": 8.686071301304212e-06,
+      "loss": 0.3394,
+      "step": 4120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6070622475886372,
+      "learning_rate": 8.685383081977678e-06,
+      "loss": 0.5224,
+      "step": 4121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9085627802921634,
+      "learning_rate": 8.684694709738053e-06,
+      "loss": 0.3409,
+      "step": 4122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.823691904252609,
+      "learning_rate": 8.6840061846139e-06,
+      "loss": 0.331,
+      "step": 4123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.55624305014329,
+      "learning_rate": 8.683317506633787e-06,
+      "loss": 0.3309,
+      "step": 4124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.531029991410397,
+      "learning_rate": 8.682628675826288e-06,
+      "loss": 0.3762,
+      "step": 4125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6466895504634198,
+      "learning_rate": 8.681939692219984e-06,
+      "loss": 0.314,
+      "step": 4126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.205852777188346,
+      "learning_rate": 8.681250555843462e-06,
+      "loss": 0.3475,
+      "step": 4127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.253520849435823,
+      "learning_rate": 8.680561266725316e-06,
+      "loss": 0.3345,
+      "step": 4128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7651739332864027,
+      "learning_rate": 8.679871824894146e-06,
+      "loss": 0.3313,
+      "step": 4129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2964360891628264,
+      "learning_rate": 8.679182230378558e-06,
+      "loss": 0.3585,
+      "step": 4130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6070789575432157,
+      "learning_rate": 8.678492483207163e-06,
+      "loss": 0.4925,
+      "step": 4131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0468637162024867,
+      "learning_rate": 8.677802583408583e-06,
+      "loss": 0.366,
+      "step": 4132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2241862210350836,
+      "learning_rate": 8.67711253101144e-06,
+      "loss": 0.3831,
+      "step": 4133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.990513516603685,
+      "learning_rate": 8.676422326044366e-06,
+      "loss": 0.353,
+      "step": 4134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.21598960991638,
+      "learning_rate": 8.675731968536004e-06,
+      "loss": 0.3336,
+      "step": 4135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.687426845091582,
+      "learning_rate": 8.67504145851499e-06,
+      "loss": 0.3341,
+      "step": 4136
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5886330691574577,
+      "learning_rate": 8.674350796009979e-06,
+      "loss": 0.3206,
+      "step": 4137
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.1784018677232813,
+      "learning_rate": 8.673659981049627e-06,
+      "loss": 0.341,
+      "step": 4138
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8005510641457785,
+      "learning_rate": 8.672969013662597e-06,
+      "loss": 0.3098,
+      "step": 4139
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9983621755265903,
+      "learning_rate": 8.67227789387756e-06,
+      "loss": 0.3355,
+      "step": 4140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6094853938636869,
+      "learning_rate": 8.671586621723186e-06,
+      "loss": 0.4963,
+      "step": 4141
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8654529172301353,
+      "learning_rate": 8.670895197228166e-06,
+      "loss": 0.3246,
+      "step": 4142
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5184671360323554,
+      "learning_rate": 8.670203620421183e-06,
+      "loss": 0.3675,
+      "step": 4143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8400839084450171,
+      "learning_rate": 8.669511891330935e-06,
+      "loss": 0.3382,
+      "step": 4144
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6928610919616545,
+      "learning_rate": 8.668820009986116e-06,
+      "loss": 0.3207,
+      "step": 4145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.825620037479355,
+      "learning_rate": 8.668127976415442e-06,
+      "loss": 0.3346,
+      "step": 4146
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4993675507647217,
+      "learning_rate": 8.667435790647623e-06,
+      "loss": 0.3233,
+      "step": 4147
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.683817712185597,
+      "learning_rate": 8.666743452711377e-06,
+      "loss": 0.3343,
+      "step": 4148
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2707252503978688,
+      "learning_rate": 8.666050962635433e-06,
+      "loss": 0.3503,
+      "step": 4149
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.765477733405839,
+      "learning_rate": 8.665358320448522e-06,
+      "loss": 0.3297,
+      "step": 4150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.764676704257325,
+      "learning_rate": 8.664665526179385e-06,
+      "loss": 0.321,
+      "step": 4151
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4444340401680011,
+      "learning_rate": 8.663972579856767e-06,
+      "loss": 0.3356,
+      "step": 4152
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2214300610166555,
+      "learning_rate": 8.663279481509418e-06,
+      "loss": 0.3376,
+      "step": 4153
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5631225824079764,
+      "learning_rate": 8.662586231166095e-06,
+      "loss": 0.3251,
+      "step": 4154
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.829923085490244,
+      "learning_rate": 8.661892828855565e-06,
+      "loss": 0.3238,
+      "step": 4155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.807054781236766,
+      "learning_rate": 8.661199274606597e-06,
+      "loss": 0.32,
+      "step": 4156
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2096343244508474,
+      "learning_rate": 8.66050556844797e-06,
+      "loss": 0.3481,
+      "step": 4157
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4568628902542375,
+      "learning_rate": 8.659811710408464e-06,
+      "loss": 0.3326,
+      "step": 4158
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.070347369135558,
+      "learning_rate": 8.659117700516869e-06,
+      "loss": 0.3417,
+      "step": 4159
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7880604823176935,
+      "learning_rate": 8.65842353880198e-06,
+      "loss": 0.3596,
+      "step": 4160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8212989764591963,
+      "learning_rate": 8.657729225292601e-06,
+      "loss": 0.3374,
+      "step": 4161
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3296034565064767,
+      "learning_rate": 8.657034760017542e-06,
+      "loss": 0.3242,
+      "step": 4162
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.671318243491196,
+      "learning_rate": 8.656340143005613e-06,
+      "loss": 0.343,
+      "step": 4163
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5460127523977807,
+      "learning_rate": 8.655645374285637e-06,
+      "loss": 0.3277,
+      "step": 4164
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.5970223843268796,
+      "learning_rate": 8.654950453886443e-06,
+      "loss": 0.3515,
+      "step": 4165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4468730029500596,
+      "learning_rate": 8.65425538183686e-06,
+      "loss": 0.3288,
+      "step": 4166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.317300326240315,
+      "learning_rate": 8.653560158165732e-06,
+      "loss": 0.3548,
+      "step": 4167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5643651382379886,
+      "learning_rate": 8.652864782901904e-06,
+      "loss": 0.3272,
+      "step": 4168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4015717564371584,
+      "learning_rate": 8.652169256074228e-06,
+      "loss": 0.3353,
+      "step": 4169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0544115789224655,
+      "learning_rate": 8.651473577711562e-06,
+      "loss": 0.3255,
+      "step": 4170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.794751494229252,
+      "learning_rate": 8.650777747842772e-06,
+      "loss": 0.3206,
+      "step": 4171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.483213570874632,
+      "learning_rate": 8.65008176649673e-06,
+      "loss": 0.3298,
+      "step": 4172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.244015847694425,
+      "learning_rate": 8.64938563370231e-06,
+      "loss": 0.3627,
+      "step": 4173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5315204706072374,
+      "learning_rate": 8.648689349488398e-06,
+      "loss": 0.3274,
+      "step": 4174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0745904701481184,
+      "learning_rate": 8.647992913883885e-06,
+      "loss": 0.3253,
+      "step": 4175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9800660707088678,
+      "learning_rate": 8.647296326917667e-06,
+      "loss": 0.3457,
+      "step": 4176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9515521489351753,
+      "learning_rate": 8.646599588618645e-06,
+      "loss": 0.3346,
+      "step": 4177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9156005196757473,
+      "learning_rate": 8.645902699015729e-06,
+      "loss": 0.3266,
+      "step": 4178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4545223038385888,
+      "learning_rate": 8.645205658137834e-06,
+      "loss": 0.3307,
+      "step": 4179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0370436764504043,
+      "learning_rate": 8.644508466013884e-06,
+      "loss": 0.3282,
+      "step": 4180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0859785489733227,
+      "learning_rate": 8.6438111226728e-06,
+      "loss": 0.3272,
+      "step": 4181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0441706463861964,
+      "learning_rate": 8.643113628143524e-06,
+      "loss": 0.3334,
+      "step": 4182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6773270507292503,
+      "learning_rate": 8.64241598245499e-06,
+      "loss": 0.4833,
+      "step": 4183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2394590502165745,
+      "learning_rate": 8.64171818563615e-06,
+      "loss": 0.3534,
+      "step": 4184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.302595319260231,
+      "learning_rate": 8.641020237715953e-06,
+      "loss": 0.3266,
+      "step": 4185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8034574015766647,
+      "learning_rate": 8.640322138723357e-06,
+      "loss": 0.3289,
+      "step": 4186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.699338701431057,
+      "learning_rate": 8.639623888687335e-06,
+      "loss": 0.3412,
+      "step": 4187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8050538947985593,
+      "learning_rate": 8.638925487636847e-06,
+      "loss": 0.3431,
+      "step": 4188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.361753746662144,
+      "learning_rate": 8.638226935600881e-06,
+      "loss": 0.3219,
+      "step": 4189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6258431357704101,
+      "learning_rate": 8.637528232608415e-06,
+      "loss": 0.5307,
+      "step": 4190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9084641772079904,
+      "learning_rate": 8.636829378688443e-06,
+      "loss": 0.3455,
+      "step": 4191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.213081098267743,
+      "learning_rate": 8.63613037386996e-06,
+      "loss": 0.3346,
+      "step": 4192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9435176600534825,
+      "learning_rate": 8.635431218181969e-06,
+      "loss": 0.3315,
+      "step": 4193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.360284991032729,
+      "learning_rate": 8.63473191165348e-06,
+      "loss": 0.3636,
+      "step": 4194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9172780629234418,
+      "learning_rate": 8.634032454313507e-06,
+      "loss": 0.3324,
+      "step": 4195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4100594044399137,
+      "learning_rate": 8.633332846191074e-06,
+      "loss": 0.3392,
+      "step": 4196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5594277567110817,
+      "learning_rate": 8.632633087315207e-06,
+      "loss": 0.3202,
+      "step": 4197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.2788181000457977,
+      "learning_rate": 8.631933177714942e-06,
+      "loss": 0.3098,
+      "step": 4198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1170957465696976,
+      "learning_rate": 8.631233117419317e-06,
+      "loss": 0.3259,
+      "step": 4199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3692229325064487,
+      "learning_rate": 8.630532906457381e-06,
+      "loss": 0.3363,
+      "step": 4200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6538040463328463,
+      "learning_rate": 8.629832544858186e-06,
+      "loss": 0.3235,
+      "step": 4201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.110178494042075,
+      "learning_rate": 8.62913203265079e-06,
+      "loss": 0.338,
+      "step": 4202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.633671884984901,
+      "learning_rate": 8.628431369864262e-06,
+      "loss": 0.3267,
+      "step": 4203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4854503434641324,
+      "learning_rate": 8.62773055652767e-06,
+      "loss": 0.3124,
+      "step": 4204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7229283162260696,
+      "learning_rate": 8.627029592670093e-06,
+      "loss": 0.325,
+      "step": 4205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5823103223056096,
+      "learning_rate": 8.626328478320617e-06,
+      "loss": 0.3279,
+      "step": 4206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.4211654887395215,
+      "learning_rate": 8.625627213508332e-06,
+      "loss": 0.333,
+      "step": 4207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.332135911111867,
+      "learning_rate": 8.624925798262335e-06,
+      "loss": 0.371,
+      "step": 4208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8388218575572008,
+      "learning_rate": 8.624224232611726e-06,
+      "loss": 0.332,
+      "step": 4209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8924836693249167,
+      "learning_rate": 8.623522516585618e-06,
+      "loss": 0.324,
+      "step": 4210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.458712333451677,
+      "learning_rate": 8.622820650213122e-06,
+      "loss": 0.3298,
+      "step": 4211
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7460136933060542,
+      "learning_rate": 8.622118633523364e-06,
+      "loss": 0.3249,
+      "step": 4212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5571048992183727,
+      "learning_rate": 8.621416466545472e-06,
+      "loss": 0.3261,
+      "step": 4213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.219814818474756,
+      "learning_rate": 8.620714149308575e-06,
+      "loss": 0.3262,
+      "step": 4214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5310169645285376,
+      "learning_rate": 8.62001168184182e-06,
+      "loss": 0.3346,
+      "step": 4215
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0374915820909085,
+      "learning_rate": 8.619309064174349e-06,
+      "loss": 0.3604,
+      "step": 4216
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4260085703202003,
+      "learning_rate": 8.618606296335314e-06,
+      "loss": 0.3424,
+      "step": 4217
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2833963435369493,
+      "learning_rate": 8.61790337835388e-06,
+      "loss": 0.3269,
+      "step": 4218
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4732389178520906,
+      "learning_rate": 8.617200310259206e-06,
+      "loss": 0.3445,
+      "step": 4219
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.580656104579461,
+      "learning_rate": 8.616497092080469e-06,
+      "loss": 0.3287,
+      "step": 4220
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5432302970182048,
+      "learning_rate": 8.615793723846842e-06,
+      "loss": 0.3295,
+      "step": 4221
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9865712153793176,
+      "learning_rate": 8.615090205587513e-06,
+      "loss": 0.353,
+      "step": 4222
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.682272812913298,
+      "learning_rate": 8.614386537331667e-06,
+      "loss": 0.4891,
+      "step": 4223
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4364191794279937,
+      "learning_rate": 8.613682719108507e-06,
+      "loss": 0.3552,
+      "step": 4224
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.985457930299239,
+      "learning_rate": 8.612978750947232e-06,
+      "loss": 0.3408,
+      "step": 4225
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.666392022632377,
+      "learning_rate": 8.61227463287705e-06,
+      "loss": 0.3176,
+      "step": 4226
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0680902210910026,
+      "learning_rate": 8.611570364927178e-06,
+      "loss": 0.31,
+      "step": 4227
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3050551553530574,
+      "learning_rate": 8.610865947126836e-06,
+      "loss": 0.3189,
+      "step": 4228
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.90337570792768,
+      "learning_rate": 8.610161379505253e-06,
+      "loss": 0.3434,
+      "step": 4229
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.771438139668405,
+      "learning_rate": 8.60945666209166e-06,
+      "loss": 0.34,
+      "step": 4230
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3298283604768903,
+      "learning_rate": 8.6087517949153e-06,
+      "loss": 0.3202,
+      "step": 4231
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.916247704002772,
+      "learning_rate": 8.608046778005419e-06,
+      "loss": 0.3261,
+      "step": 4232
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.030769209932224,
+      "learning_rate": 8.607341611391268e-06,
+      "loss": 0.3622,
+      "step": 4233
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1697410164179995,
+      "learning_rate": 8.606636295102106e-06,
+      "loss": 0.3252,
+      "step": 4234
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.605232603316513,
+      "learning_rate": 8.6059308291672e-06,
+      "loss": 0.3363,
+      "step": 4235
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6585366486403808,
+      "learning_rate": 8.605225213615818e-06,
+      "loss": 0.3221,
+      "step": 4236
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2632132517410937,
+      "learning_rate": 8.604519448477236e-06,
+      "loss": 0.3261,
+      "step": 4237
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7575344780553146,
+      "learning_rate": 8.603813533780742e-06,
+      "loss": 0.3425,
+      "step": 4238
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8349045762388945,
+      "learning_rate": 8.603107469555623e-06,
+      "loss": 0.3225,
+      "step": 4239
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4451097932181565,
+      "learning_rate": 8.602401255831176e-06,
+      "loss": 0.3448,
+      "step": 4240
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2508351057350167,
+      "learning_rate": 8.601694892636701e-06,
+      "loss": 0.3168,
+      "step": 4241
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6972953706707048,
+      "learning_rate": 8.600988380001508e-06,
+      "loss": 0.3391,
+      "step": 4242
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.1794147541088265,
+      "learning_rate": 8.60028171795491e-06,
+      "loss": 0.3366,
+      "step": 4243
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5023694633565843,
+      "learning_rate": 8.599574906526231e-06,
+      "loss": 0.3464,
+      "step": 4244
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6034641172287127,
+      "learning_rate": 8.598867945744794e-06,
+      "loss": 0.3569,
+      "step": 4245
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6657183866249199,
+      "learning_rate": 8.598160835639935e-06,
+      "loss": 0.3222,
+      "step": 4246
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5283025124717415,
+      "learning_rate": 8.59745357624099e-06,
+      "loss": 0.3339,
+      "step": 4247
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2966176158873703,
+      "learning_rate": 8.596746167577308e-06,
+      "loss": 0.3289,
+      "step": 4248
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5372664240922145,
+      "learning_rate": 8.596038609678236e-06,
+      "loss": 0.3139,
+      "step": 4249
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3840010217867644,
+      "learning_rate": 8.595330902573137e-06,
+      "loss": 0.305,
+      "step": 4250
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6570800328309143,
+      "learning_rate": 8.594623046291374e-06,
+      "loss": 0.3161,
+      "step": 4251
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.392855303032146,
+      "learning_rate": 8.593915040862315e-06,
+      "loss": 0.3046,
+      "step": 4252
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.553084091586555,
+      "learning_rate": 8.593206886315338e-06,
+      "loss": 0.3393,
+      "step": 4253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7687996769929684,
+      "learning_rate": 8.592498582679823e-06,
+      "loss": 0.347,
+      "step": 4254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7340705972518944,
+      "learning_rate": 8.591790129985162e-06,
+      "loss": 0.3371,
+      "step": 4255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3751645187296844,
+      "learning_rate": 8.59108152826075e-06,
+      "loss": 0.3306,
+      "step": 4256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.558167259182883,
+      "learning_rate": 8.590372777535984e-06,
+      "loss": 0.3582,
+      "step": 4257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9325602192645501,
+      "learning_rate": 8.589663877840278e-06,
+      "loss": 0.3659,
+      "step": 4258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7099919053117987,
+      "learning_rate": 8.588954829203039e-06,
+      "loss": 0.4824,
+      "step": 4259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.894137370774947,
+      "learning_rate": 8.58824563165369e-06,
+      "loss": 0.3661,
+      "step": 4260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7094474668222037,
+      "learning_rate": 8.587536285221656e-06,
+      "loss": 0.3231,
+      "step": 4261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6262659477981198,
+      "learning_rate": 8.58682678993637e-06,
+      "loss": 0.3357,
+      "step": 4262
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1584387794679487,
+      "learning_rate": 8.58611714582727e-06,
+      "loss": 0.3448,
+      "step": 4263
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.5630222924968535,
+      "learning_rate": 8.585407352923799e-06,
+      "loss": 0.3483,
+      "step": 4264
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9131127117402358,
+      "learning_rate": 8.584697411255409e-06,
+      "loss": 0.3447,
+      "step": 4265
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.8341610832471895,
+      "learning_rate": 8.583987320851556e-06,
+      "loss": 0.3329,
+      "step": 4266
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3496653343170317,
+      "learning_rate": 8.583277081741703e-06,
+      "loss": 0.3336,
+      "step": 4267
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.607165837648559,
+      "learning_rate": 8.58256669395532e-06,
+      "loss": 0.3509,
+      "step": 4268
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9243460639183332,
+      "learning_rate": 8.581856157521882e-06,
+      "loss": 0.3428,
+      "step": 4269
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.80555944204013,
+      "learning_rate": 8.58114547247087e-06,
+      "loss": 0.3495,
+      "step": 4270
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2423985483323439,
+      "learning_rate": 8.580434638831769e-06,
+      "loss": 0.3277,
+      "step": 4271
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5004579684570956,
+      "learning_rate": 8.579723656634077e-06,
+      "loss": 0.3033,
+      "step": 4272
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5256845265497303,
+      "learning_rate": 8.579012525907292e-06,
+      "loss": 0.3342,
+      "step": 4273
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6350629540576957,
+      "learning_rate": 8.57830124668092e-06,
+      "loss": 0.3616,
+      "step": 4274
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.07337339387284,
+      "learning_rate": 8.577589818984473e-06,
+      "loss": 0.3241,
+      "step": 4275
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4436272941552504,
+      "learning_rate": 8.576878242847472e-06,
+      "loss": 0.327,
+      "step": 4276
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5680174375865998,
+      "learning_rate": 8.576166518299438e-06,
+      "loss": 0.3427,
+      "step": 4277
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9773442544469557,
+      "learning_rate": 8.575454645369904e-06,
+      "loss": 0.3532,
+      "step": 4278
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9291989336041662,
+      "learning_rate": 8.574742624088403e-06,
+      "loss": 0.3336,
+      "step": 4279
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1602198116166584,
+      "learning_rate": 8.574030454484486e-06,
+      "loss": 0.3252,
+      "step": 4280
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8307352894774007,
+      "learning_rate": 8.573318136587694e-06,
+      "loss": 0.3255,
+      "step": 4281
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9561684405910387,
+      "learning_rate": 8.572605670427584e-06,
+      "loss": 0.3391,
+      "step": 4282
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6164287395728145,
+      "learning_rate": 8.571893056033722e-06,
+      "loss": 0.3404,
+      "step": 4283
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8667852039769386,
+      "learning_rate": 8.571180293435673e-06,
+      "loss": 0.3273,
+      "step": 4284
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3175344965314466,
+      "learning_rate": 8.570467382663009e-06,
+      "loss": 0.3239,
+      "step": 4285
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.789391909186617,
+      "learning_rate": 8.569754323745311e-06,
+      "loss": 0.3325,
+      "step": 4286
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.674364036570318,
+      "learning_rate": 8.569041116712168e-06,
+      "loss": 0.3414,
+      "step": 4287
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8334698952624648,
+      "learning_rate": 8.568327761593169e-06,
+      "loss": 0.3276,
+      "step": 4288
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4038681936410702,
+      "learning_rate": 8.567614258417911e-06,
+      "loss": 0.3397,
+      "step": 4289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2040393515936567,
+      "learning_rate": 8.566900607216001e-06,
+      "loss": 0.3368,
+      "step": 4290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4904611649416002,
+      "learning_rate": 8.566186808017052e-06,
+      "loss": 0.33,
+      "step": 4291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1832334506308713,
+      "learning_rate": 8.565472860850675e-06,
+      "loss": 0.3358,
+      "step": 4292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5032988899297923,
+      "learning_rate": 8.564758765746499e-06,
+      "loss": 0.3702,
+      "step": 4293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8313876251831345,
+      "learning_rate": 8.564044522734147e-06,
+      "loss": 0.3223,
+      "step": 4294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8462600345655764,
+      "learning_rate": 8.563330131843259e-06,
+      "loss": 0.3358,
+      "step": 4295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.173645721342238,
+      "learning_rate": 8.562615593103474e-06,
+      "loss": 0.3745,
+      "step": 4296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6391342795380124,
+      "learning_rate": 8.56190090654444e-06,
+      "loss": 0.3325,
+      "step": 4297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4289099612583376,
+      "learning_rate": 8.561186072195812e-06,
+      "loss": 0.3388,
+      "step": 4298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2782260343317648,
+      "learning_rate": 8.560471090087247e-06,
+      "loss": 0.3629,
+      "step": 4299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.204730944478129,
+      "learning_rate": 8.559755960248414e-06,
+      "loss": 0.3572,
+      "step": 4300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.948083818207941,
+      "learning_rate": 8.559040682708982e-06,
+      "loss": 0.3281,
+      "step": 4301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.274866949173534,
+      "learning_rate": 8.558325257498632e-06,
+      "loss": 0.3469,
+      "step": 4302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2081337728099282,
+      "learning_rate": 8.557609684647045e-06,
+      "loss": 0.3526,
+      "step": 4303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9668900755299428,
+      "learning_rate": 8.556893964183913e-06,
+      "loss": 0.3348,
+      "step": 4304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.424996006234006,
+      "learning_rate": 8.556178096138933e-06,
+      "loss": 0.3227,
+      "step": 4305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4852727013421994,
+      "learning_rate": 8.555462080541809e-06,
+      "loss": 0.3303,
+      "step": 4306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5745261843264677,
+      "learning_rate": 8.554745917422247e-06,
+      "loss": 0.3625,
+      "step": 4307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7543657370706096,
+      "learning_rate": 8.554029606809962e-06,
+      "loss": 0.3403,
+      "step": 4308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6253439350883727,
+      "learning_rate": 8.553313148734676e-06,
+      "loss": 0.3302,
+      "step": 4309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8840392671545938,
+      "learning_rate": 8.552596543226118e-06,
+      "loss": 0.3388,
+      "step": 4310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.295478319147859,
+      "learning_rate": 8.551879790314018e-06,
+      "loss": 0.3543,
+      "step": 4311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2614115819624998,
+      "learning_rate": 8.551162890028118e-06,
+      "loss": 0.3293,
+      "step": 4312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.88470133706015,
+      "learning_rate": 8.55044584239816e-06,
+      "loss": 0.3529,
+      "step": 4313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4376100337365636,
+      "learning_rate": 8.5497286474539e-06,
+      "loss": 0.3379,
+      "step": 4314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1295369507884203,
+      "learning_rate": 8.54901130522509e-06,
+      "loss": 0.354,
+      "step": 4315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5413742819786713,
+      "learning_rate": 8.5482938157415e-06,
+      "loss": 0.3353,
+      "step": 4316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1450275082854895,
+      "learning_rate": 8.547576179032896e-06,
+      "loss": 0.3509,
+      "step": 4317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.768370691279072,
+      "learning_rate": 8.546858395129055e-06,
+      "loss": 0.3514,
+      "step": 4318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2362246613404333,
+      "learning_rate": 8.54614046405976e-06,
+      "loss": 0.3244,
+      "step": 4319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1640030660592486,
+      "learning_rate": 8.545422385854798e-06,
+      "loss": 0.3587,
+      "step": 4320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5859332241452266,
+      "learning_rate": 8.544704160543964e-06,
+      "loss": 0.3406,
+      "step": 4321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7424619456056605,
+      "learning_rate": 8.543985788157058e-06,
+      "loss": 0.3434,
+      "step": 4322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3040625766961984,
+      "learning_rate": 8.543267268723887e-06,
+      "loss": 0.3342,
+      "step": 4323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7915454490679616,
+      "learning_rate": 8.542548602274262e-06,
+      "loss": 0.3421,
+      "step": 4324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5245026450484587,
+      "learning_rate": 8.541829788838007e-06,
+      "loss": 0.3603,
+      "step": 4325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8410130517323098,
+      "learning_rate": 8.54111082844494e-06,
+      "loss": 0.3339,
+      "step": 4326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4665300091391573,
+      "learning_rate": 8.540391721124897e-06,
+      "loss": 0.3287,
+      "step": 4327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9614137067521844,
+      "learning_rate": 8.53967246690771e-06,
+      "loss": 0.3522,
+      "step": 4328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.169946975148502,
+      "learning_rate": 8.538953065823229e-06,
+      "loss": 0.3423,
+      "step": 4329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4254970564272986,
+      "learning_rate": 8.538233517901298e-06,
+      "loss": 0.335,
+      "step": 4330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8760149928771839,
+      "learning_rate": 8.537513823171773e-06,
+      "loss": 0.3336,
+      "step": 4331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.527033180686907,
+      "learning_rate": 8.536793981664515e-06,
+      "loss": 0.3316,
+      "step": 4332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9056575167826721,
+      "learning_rate": 8.536073993409394e-06,
+      "loss": 0.3512,
+      "step": 4333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7523770080774033,
+      "learning_rate": 8.535353858436284e-06,
+      "loss": 0.3435,
+      "step": 4334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4986535924131184,
+      "learning_rate": 8.534633576775059e-06,
+      "loss": 0.326,
+      "step": 4335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1885721986240636,
+      "learning_rate": 8.533913148455613e-06,
+      "loss": 0.3488,
+      "step": 4336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4865190194148143,
+      "learning_rate": 8.533192573507831e-06,
+      "loss": 0.3345,
+      "step": 4337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7807621071293849,
+      "learning_rate": 8.532471851961612e-06,
+      "loss": 0.3287,
+      "step": 4338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7371039780349717,
+      "learning_rate": 8.531750983846864e-06,
+      "loss": 0.347,
+      "step": 4339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6273966515263565,
+      "learning_rate": 8.531029969193494e-06,
+      "loss": 0.5233,
+      "step": 4340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9254934974362559,
+      "learning_rate": 8.53030880803142e-06,
+      "loss": 0.3349,
+      "step": 4341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.726887305529337,
+      "learning_rate": 8.529587500390561e-06,
+      "loss": 0.358,
+      "step": 4342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7184910973631984,
+      "learning_rate": 8.528866046300848e-06,
+      "loss": 0.3321,
+      "step": 4343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7275129845163628,
+      "learning_rate": 8.528144445792215e-06,
+      "loss": 0.3351,
+      "step": 4344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5879895951632277,
+      "learning_rate": 8.527422698894602e-06,
+      "loss": 0.3278,
+      "step": 4345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0130951180578815,
+      "learning_rate": 8.526700805637955e-06,
+      "loss": 0.327,
+      "step": 4346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.2594256716898915,
+      "learning_rate": 8.52597876605223e-06,
+      "loss": 0.3464,
+      "step": 4347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8173665610318812,
+      "learning_rate": 8.52525658016738e-06,
+      "loss": 0.3337,
+      "step": 4348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1933980716705372,
+      "learning_rate": 8.524534248013374e-06,
+      "loss": 0.3506,
+      "step": 4349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4328666112013297,
+      "learning_rate": 8.523811769620183e-06,
+      "loss": 0.3288,
+      "step": 4350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3862717693388513,
+      "learning_rate": 8.523089145017781e-06,
+      "loss": 0.3382,
+      "step": 4351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.963060191769809,
+      "learning_rate": 8.522366374236155e-06,
+      "loss": 0.3407,
+      "step": 4352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.375648340144036,
+      "learning_rate": 8.52164345730529e-06,
+      "loss": 0.3235,
+      "step": 4353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.759064297095795,
+      "learning_rate": 8.520920394255183e-06,
+      "loss": 0.3283,
+      "step": 4354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.989751854975891,
+      "learning_rate": 8.520197185115836e-06,
+      "loss": 0.3171,
+      "step": 4355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5708895020208051,
+      "learning_rate": 8.519473829917254e-06,
+      "loss": 0.3305,
+      "step": 4356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4383389833553923,
+      "learning_rate": 8.518750328689452e-06,
+      "loss": 0.3189,
+      "step": 4357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.702322153472481,
+      "learning_rate": 8.518026681462448e-06,
+      "loss": 0.3633,
+      "step": 4358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1957765031845529,
+      "learning_rate": 8.517302888266268e-06,
+      "loss": 0.3427,
+      "step": 4359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7601932345222768,
+      "learning_rate": 8.516578949130945e-06,
+      "loss": 0.3347,
+      "step": 4360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.01378365039657,
+      "learning_rate": 8.515854864086514e-06,
+      "loss": 0.346,
+      "step": 4361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3238848278812276,
+      "learning_rate": 8.515130633163023e-06,
+      "loss": 0.3405,
+      "step": 4362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4612797367053143,
+      "learning_rate": 8.514406256390514e-06,
+      "loss": 0.365,
+      "step": 4363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3676256160402094,
+      "learning_rate": 8.513681733799051e-06,
+      "loss": 0.3427,
+      "step": 4364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7017726229022991,
+      "learning_rate": 8.51295706541869e-06,
+      "loss": 0.3312,
+      "step": 4365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8504650548271053,
+      "learning_rate": 8.5122322512795e-06,
+      "loss": 0.3181,
+      "step": 4366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6175004487616216,
+      "learning_rate": 8.511507291411559e-06,
+      "loss": 0.3372,
+      "step": 4367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6717481403154533,
+      "learning_rate": 8.51078218584494e-06,
+      "loss": 0.3453,
+      "step": 4368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6367362300887454,
+      "learning_rate": 8.510056934609731e-06,
+      "loss": 0.3267,
+      "step": 4369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6889658034505874,
+      "learning_rate": 8.509331537736027e-06,
+      "loss": 0.3549,
+      "step": 4370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.563582268903867,
+      "learning_rate": 8.508605995253925e-06,
+      "loss": 0.3461,
+      "step": 4371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6291706586706887,
+      "learning_rate": 8.507880307193528e-06,
+      "loss": 0.3227,
+      "step": 4372
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.3164096190998,
+      "learning_rate": 8.507154473584947e-06,
+      "loss": 0.3685,
+      "step": 4373
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4672278234014753,
+      "learning_rate": 8.506428494458295e-06,
+      "loss": 0.3194,
+      "step": 4374
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4037526339260669,
+      "learning_rate": 8.505702369843698e-06,
+      "loss": 0.3379,
+      "step": 4375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6108254330532286,
+      "learning_rate": 8.504976099771282e-06,
+      "loss": 0.3407,
+      "step": 4376
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0522815552892464,
+      "learning_rate": 8.504249684271184e-06,
+      "loss": 0.3404,
+      "step": 4377
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5535046442865097,
+      "learning_rate": 8.503523123373542e-06,
+      "loss": 0.3355,
+      "step": 4378
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3438725864520216,
+      "learning_rate": 8.502796417108502e-06,
+      "loss": 0.329,
+      "step": 4379
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2135373884053635,
+      "learning_rate": 8.502069565506217e-06,
+      "loss": 0.3317,
+      "step": 4380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7996234124505242,
+      "learning_rate": 8.501342568596845e-06,
+      "loss": 0.3336,
+      "step": 4381
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8120541851369305,
+      "learning_rate": 8.500615426410552e-06,
+      "loss": 0.379,
+      "step": 4382
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9061961758448398,
+      "learning_rate": 8.499888138977505e-06,
+      "loss": 0.3278,
+      "step": 4383
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.535295646925185,
+      "learning_rate": 8.499160706327885e-06,
+      "loss": 0.3321,
+      "step": 4384
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6314075826180614,
+      "learning_rate": 8.498433128491871e-06,
+      "loss": 0.3036,
+      "step": 4385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3899585689416163,
+      "learning_rate": 8.497705405499653e-06,
+      "loss": 0.3316,
+      "step": 4386
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8082256460455335,
+      "learning_rate": 8.496977537381423e-06,
+      "loss": 0.3213,
+      "step": 4387
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6111633913066594,
+      "learning_rate": 8.496249524167385e-06,
+      "loss": 0.3156,
+      "step": 4388
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.178875953257289,
+      "learning_rate": 8.495521365887746e-06,
+      "loss": 0.3249,
+      "step": 4389
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5600057353396926,
+      "learning_rate": 8.494793062572715e-06,
+      "loss": 0.3251,
+      "step": 4390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9363086132080793,
+      "learning_rate": 8.494064614252514e-06,
+      "loss": 0.3483,
+      "step": 4391
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3219061346519227,
+      "learning_rate": 8.493336020957363e-06,
+      "loss": 0.3385,
+      "step": 4392
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0627756771022723,
+      "learning_rate": 8.492607282717498e-06,
+      "loss": 0.3138,
+      "step": 4393
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.969644771274343,
+      "learning_rate": 8.491878399563152e-06,
+      "loss": 0.3509,
+      "step": 4394
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7271320488528492,
+      "learning_rate": 8.49114937152457e-06,
+      "loss": 0.3435,
+      "step": 4395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.002131941452337,
+      "learning_rate": 8.490420198631998e-06,
+      "loss": 0.3385,
+      "step": 4396
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.077766615899552,
+      "learning_rate": 8.489690880915695e-06,
+      "loss": 0.3174,
+      "step": 4397
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4175821939526254,
+      "learning_rate": 8.488961418405917e-06,
+      "loss": 0.3143,
+      "step": 4398
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.373658890823445,
+      "learning_rate": 8.488231811132932e-06,
+      "loss": 0.357,
+      "step": 4399
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7707450262173765,
+      "learning_rate": 8.487502059127015e-06,
+      "loss": 0.3439,
+      "step": 4400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3140756892535392,
+      "learning_rate": 8.486772162418442e-06,
+      "loss": 0.3364,
+      "step": 4401
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2274575926710707,
+      "learning_rate": 8.4860421210375e-06,
+      "loss": 0.3485,
+      "step": 4402
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1946148108466015,
+      "learning_rate": 8.485311935014478e-06,
+      "loss": 0.3448,
+      "step": 4403
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2386776861371314,
+      "learning_rate": 8.484581604379673e-06,
+      "loss": 0.3226,
+      "step": 4404
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1940307130584027,
+      "learning_rate": 8.483851129163388e-06,
+      "loss": 0.3419,
+      "step": 4405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4218911382822264,
+      "learning_rate": 8.48312050939593e-06,
+      "loss": 0.3458,
+      "step": 4406
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4026977993571956,
+      "learning_rate": 8.482389745107618e-06,
+      "loss": 0.3425,
+      "step": 4407
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7591858574730552,
+      "learning_rate": 8.481658836328767e-06,
+      "loss": 0.333,
+      "step": 4408
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3768237191404413,
+      "learning_rate": 8.48092778308971e-06,
+      "loss": 0.3452,
+      "step": 4409
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1607715160729817,
+      "learning_rate": 8.480196585420775e-06,
+      "loss": 0.3401,
+      "step": 4410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7690587304651568,
+      "learning_rate": 8.479465243352303e-06,
+      "loss": 0.3353,
+      "step": 4411
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.166355849091709,
+      "learning_rate": 8.478733756914636e-06,
+      "loss": 0.3216,
+      "step": 4412
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2161506734271486,
+      "learning_rate": 8.478002126138127e-06,
+      "loss": 0.3247,
+      "step": 4413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3162179304770918,
+      "learning_rate": 8.477270351053133e-06,
+      "loss": 0.3306,
+      "step": 4414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4929854712675306,
+      "learning_rate": 8.476538431690016e-06,
+      "loss": 0.3497,
+      "step": 4415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.624411338106835,
+      "learning_rate": 8.475806368079144e-06,
+      "loss": 0.3609,
+      "step": 4416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0073075734938985,
+      "learning_rate": 8.475074160250892e-06,
+      "loss": 0.3313,
+      "step": 4417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.807099711074169,
+      "learning_rate": 8.47434180823564e-06,
+      "loss": 0.3398,
+      "step": 4418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.677998276150886,
+      "learning_rate": 8.473609312063778e-06,
+      "loss": 0.3462,
+      "step": 4419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6659747390409879,
+      "learning_rate": 8.472876671765692e-06,
+      "loss": 0.489,
+      "step": 4420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.037584135778137,
+      "learning_rate": 8.472143887371786e-06,
+      "loss": 0.3379,
+      "step": 4421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3806910811268274,
+      "learning_rate": 8.471410958912463e-06,
+      "loss": 0.3233,
+      "step": 4422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0217306989118162,
+      "learning_rate": 8.470677886418135e-06,
+      "loss": 0.3474,
+      "step": 4423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.211865032210044,
+      "learning_rate": 8.469944669919214e-06,
+      "loss": 0.3365,
+      "step": 4424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0957457196553895,
+      "learning_rate": 8.469211309446127e-06,
+      "loss": 0.3423,
+      "step": 4425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9589516548469754,
+      "learning_rate": 8.4684778050293e-06,
+      "loss": 0.3364,
+      "step": 4426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0751073588242526,
+      "learning_rate": 8.467744156699168e-06,
+      "loss": 0.3377,
+      "step": 4427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.164582665735496,
+      "learning_rate": 8.467010364486173e-06,
+      "loss": 0.3701,
+      "step": 4428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.773615112194289,
+      "learning_rate": 8.466276428420759e-06,
+      "loss": 0.3499,
+      "step": 4429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7825839864747932,
+      "learning_rate": 8.465542348533379e-06,
+      "loss": 0.3574,
+      "step": 4430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2084379741261175,
+      "learning_rate": 8.46480812485449e-06,
+      "loss": 0.35,
+      "step": 4431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4643439280425021,
+      "learning_rate": 8.464073757414561e-06,
+      "loss": 0.3522,
+      "step": 4432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6935562458211901,
+      "learning_rate": 8.463339246244058e-06,
+      "loss": 0.3346,
+      "step": 4433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6989172639998948,
+      "learning_rate": 8.462604591373457e-06,
+      "loss": 0.3388,
+      "step": 4434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6921131120928505,
+      "learning_rate": 8.46186979283324e-06,
+      "loss": 0.3521,
+      "step": 4435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.552014840242945,
+      "learning_rate": 8.461134850653899e-06,
+      "loss": 0.3172,
+      "step": 4436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5128879381537257,
+      "learning_rate": 8.460399764865923e-06,
+      "loss": 0.353,
+      "step": 4437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2140103876396258,
+      "learning_rate": 8.459664535499816e-06,
+      "loss": 0.3549,
+      "step": 4438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9802234604045736,
+      "learning_rate": 8.458929162586083e-06,
+      "loss": 0.3498,
+      "step": 4439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.301930046058223,
+      "learning_rate": 8.458193646155234e-06,
+      "loss": 0.329,
+      "step": 4440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6136415581442636,
+      "learning_rate": 8.457457986237788e-06,
+      "loss": 0.3502,
+      "step": 4441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7302452323933113,
+      "learning_rate": 8.45672218286427e-06,
+      "loss": 0.3241,
+      "step": 4442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1475670132876745,
+      "learning_rate": 8.455986236065208e-06,
+      "loss": 0.3252,
+      "step": 4443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6465254320379765,
+      "learning_rate": 8.455250145871139e-06,
+      "loss": 0.3287,
+      "step": 4444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2861135255516976,
+      "learning_rate": 8.454513912312605e-06,
+      "loss": 0.3169,
+      "step": 4445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.697120502963351,
+      "learning_rate": 8.45377753542015e-06,
+      "loss": 0.5319,
+      "step": 4446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.733200793446351,
+      "learning_rate": 8.453041015224332e-06,
+      "loss": 0.3337,
+      "step": 4447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.060444362366062,
+      "learning_rate": 8.452304351755712e-06,
+      "loss": 0.3254,
+      "step": 4448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8920764037027373,
+      "learning_rate": 8.451567545044848e-06,
+      "loss": 0.3392,
+      "step": 4449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7831363366263777,
+      "learning_rate": 8.450830595122317e-06,
+      "loss": 0.3381,
+      "step": 4450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7005389347084314,
+      "learning_rate": 8.450093502018698e-06,
+      "loss": 0.3388,
+      "step": 4451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8978674480258144,
+      "learning_rate": 8.449356265764568e-06,
+      "loss": 0.3497,
+      "step": 4452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6128982561238903,
+      "learning_rate": 8.448618886390523e-06,
+      "loss": 0.4854,
+      "step": 4453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.225443077703998,
+      "learning_rate": 8.447881363927152e-06,
+      "loss": 0.3379,
+      "step": 4454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.170752349368643,
+      "learning_rate": 8.44714369840506e-06,
+      "loss": 0.3485,
+      "step": 4455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2547032767916018,
+      "learning_rate": 8.446405889854853e-06,
+      "loss": 0.3316,
+      "step": 4456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.396362027733872,
+      "learning_rate": 8.445667938307145e-06,
+      "loss": 0.355,
+      "step": 4457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.902622725375012,
+      "learning_rate": 8.444929843792554e-06,
+      "loss": 0.3245,
+      "step": 4458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0892517371336243,
+      "learning_rate": 8.444191606341705e-06,
+      "loss": 0.3295,
+      "step": 4459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5573086843510329,
+      "learning_rate": 8.443453225985227e-06,
+      "loss": 0.3189,
+      "step": 4460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.786192565503805,
+      "learning_rate": 8.442714702753761e-06,
+      "loss": 0.3289,
+      "step": 4461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9314382992937276,
+      "learning_rate": 8.441976036677945e-06,
+      "loss": 0.3311,
+      "step": 4462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4805191730640384,
+      "learning_rate": 8.44123722778843e-06,
+      "loss": 0.3388,
+      "step": 4463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4171308586400473,
+      "learning_rate": 8.440498276115872e-06,
+      "loss": 0.3672,
+      "step": 4464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.083731638138706,
+      "learning_rate": 8.439759181690927e-06,
+      "loss": 0.3624,
+      "step": 4465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6896070284168334,
+      "learning_rate": 8.439019944544265e-06,
+      "loss": 0.338,
+      "step": 4466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5516054145198155,
+      "learning_rate": 8.438280564706556e-06,
+      "loss": 0.3368,
+      "step": 4467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3380291500414585,
+      "learning_rate": 8.43754104220848e-06,
+      "loss": 0.3423,
+      "step": 4468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5818858495087924,
+      "learning_rate": 8.436801377080721e-06,
+      "loss": 0.3434,
+      "step": 4469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4457063362733797,
+      "learning_rate": 8.436061569353969e-06,
+      "loss": 0.3159,
+      "step": 4470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4635786151827817,
+      "learning_rate": 8.435321619058918e-06,
+      "loss": 0.3433,
+      "step": 4471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6632148533614326,
+      "learning_rate": 8.434581526226271e-06,
+      "loss": 0.3433,
+      "step": 4472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4451986013557694,
+      "learning_rate": 8.433841290886736e-06,
+      "loss": 0.3227,
+      "step": 4473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4790957748402054,
+      "learning_rate": 8.433100913071028e-06,
+      "loss": 0.3395,
+      "step": 4474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3330617974772014,
+      "learning_rate": 8.432360392809864e-06,
+      "loss": 0.3293,
+      "step": 4475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.413761591379426,
+      "learning_rate": 8.43161973013397e-06,
+      "loss": 0.3207,
+      "step": 4476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3253103720450454,
+      "learning_rate": 8.43087892507408e-06,
+      "loss": 0.3223,
+      "step": 4477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.140728020517338,
+      "learning_rate": 8.430137977660929e-06,
+      "loss": 0.3436,
+      "step": 4478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7925152385170382,
+      "learning_rate": 8.42939688792526e-06,
+      "loss": 0.3324,
+      "step": 4479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8727619116462806,
+      "learning_rate": 8.428655655897824e-06,
+      "loss": 0.3391,
+      "step": 4480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9297848501336063,
+      "learning_rate": 8.427914281609374e-06,
+      "loss": 0.3289,
+      "step": 4481
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4249350042801914,
+      "learning_rate": 8.427172765090671e-06,
+      "loss": 0.3563,
+      "step": 4482
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4973898360340145,
+      "learning_rate": 8.426431106372484e-06,
+      "loss": 0.3346,
+      "step": 4483
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7919917592138357,
+      "learning_rate": 8.425689305485583e-06,
+      "loss": 0.3486,
+      "step": 4484
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.976785259701706,
+      "learning_rate": 8.42494736246075e-06,
+      "loss": 0.3702,
+      "step": 4485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6446864876745837,
+      "learning_rate": 8.424205277328767e-06,
+      "loss": 0.3395,
+      "step": 4486
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5223096088696004,
+      "learning_rate": 8.423463050120424e-06,
+      "loss": 0.338,
+      "step": 4487
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6865376594309599,
+      "learning_rate": 8.42272068086652e-06,
+      "loss": 0.4876,
+      "step": 4488
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3421517139266632,
+      "learning_rate": 8.421978169597853e-06,
+      "loss": 0.3248,
+      "step": 4489
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.213398892885125,
+      "learning_rate": 8.421235516345236e-06,
+      "loss": 0.3262,
+      "step": 4490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.924517519833465,
+      "learning_rate": 8.420492721139479e-06,
+      "loss": 0.3204,
+      "step": 4491
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6575116442764894,
+      "learning_rate": 8.419749784011405e-06,
+      "loss": 0.3299,
+      "step": 4492
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7710530905246775,
+      "learning_rate": 8.419006704991837e-06,
+      "loss": 0.3431,
+      "step": 4493
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0894328764514416,
+      "learning_rate": 8.418263484111607e-06,
+      "loss": 0.325,
+      "step": 4494
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4823357952923764,
+      "learning_rate": 8.417520121401555e-06,
+      "loss": 0.3408,
+      "step": 4495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6770118194333938,
+      "learning_rate": 8.416776616892523e-06,
+      "loss": 0.529,
+      "step": 4496
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.061292293556901,
+      "learning_rate": 8.41603297061536e-06,
+      "loss": 0.3535,
+      "step": 4497
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3675176565008405,
+      "learning_rate": 8.41528918260092e-06,
+      "loss": 0.3406,
+      "step": 4498
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8532972918923218,
+      "learning_rate": 8.414545252880069e-06,
+      "loss": 0.3355,
+      "step": 4499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5945330744361292,
+      "learning_rate": 8.413801181483668e-06,
+      "loss": 0.3247,
+      "step": 4500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7164676116388484,
+      "learning_rate": 8.413056968442593e-06,
+      "loss": 0.3302,
+      "step": 4501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5914714367032117,
+      "learning_rate": 8.412312613787723e-06,
+      "loss": 0.5142,
+      "step": 4502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4003420472063788,
+      "learning_rate": 8.41156811754994e-06,
+      "loss": 0.3173,
+      "step": 4503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.867307296426739,
+      "learning_rate": 8.410823479760138e-06,
+      "loss": 0.3393,
+      "step": 4504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0443264810913835,
+      "learning_rate": 8.41007870044921e-06,
+      "loss": 0.3329,
+      "step": 4505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9652394315038801,
+      "learning_rate": 8.40933377964806e-06,
+      "loss": 0.3539,
+      "step": 4506
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.792371914748029,
+      "learning_rate": 8.408588717387595e-06,
+      "loss": 0.3289,
+      "step": 4507
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5403269130661204,
+      "learning_rate": 8.407843513698732e-06,
+      "loss": 0.3216,
+      "step": 4508
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8494122686288157,
+      "learning_rate": 8.407098168612388e-06,
+      "loss": 0.3396,
+      "step": 4509
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0371687097009716,
+      "learning_rate": 8.406352682159488e-06,
+      "loss": 0.3519,
+      "step": 4510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.395283305635252,
+      "learning_rate": 8.405607054370967e-06,
+      "loss": 0.3355,
+      "step": 4511
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.936582424310372,
+      "learning_rate": 8.404861285277759e-06,
+      "loss": 0.3329,
+      "step": 4512
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1771448147845263,
+      "learning_rate": 8.404115374910808e-06,
+      "loss": 0.3288,
+      "step": 4513
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.045448180983775,
+      "learning_rate": 8.403369323301065e-06,
+      "loss": 0.3564,
+      "step": 4514
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7095027800576776,
+      "learning_rate": 8.402623130479482e-06,
+      "loss": 0.3402,
+      "step": 4515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.582270828514155,
+      "learning_rate": 8.401876796477023e-06,
+      "loss": 0.348,
+      "step": 4516
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1942599791844697,
+      "learning_rate": 8.401130321324653e-06,
+      "loss": 0.3301,
+      "step": 4517
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6912184946799662,
+      "learning_rate": 8.400383705053344e-06,
+      "loss": 0.3518,
+      "step": 4518
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1691049921985823,
+      "learning_rate": 8.399636947694077e-06,
+      "loss": 0.3298,
+      "step": 4519
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4106364768480293,
+      "learning_rate": 8.398890049277834e-06,
+      "loss": 0.35,
+      "step": 4520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.3155101896571524,
+      "learning_rate": 8.398143009835604e-06,
+      "loss": 0.3349,
+      "step": 4521
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6146989377182868,
+      "learning_rate": 8.397395829398387e-06,
+      "loss": 0.3234,
+      "step": 4522
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8812126832136733,
+      "learning_rate": 8.396648507997181e-06,
+      "loss": 0.3219,
+      "step": 4523
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6515107372210474,
+      "learning_rate": 8.395901045662995e-06,
+      "loss": 0.334,
+      "step": 4524
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.862319841649697,
+      "learning_rate": 8.395153442426844e-06,
+      "loss": 0.3393,
+      "step": 4525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.00465234726739,
+      "learning_rate": 8.394405698319748e-06,
+      "loss": 0.3186,
+      "step": 4526
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.657619312354189,
+      "learning_rate": 8.393657813372728e-06,
+      "loss": 0.336,
+      "step": 4527
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7031161656919498,
+      "learning_rate": 8.392909787616817e-06,
+      "loss": 0.5135,
+      "step": 4528
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9786773654926382,
+      "learning_rate": 8.392161621083054e-06,
+      "loss": 0.3445,
+      "step": 4529
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.810524317888919,
+      "learning_rate": 8.39141331380248e-06,
+      "loss": 0.3348,
+      "step": 4530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2980793145349176,
+      "learning_rate": 8.390664865806142e-06,
+      "loss": 0.3429,
+      "step": 4531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.032304141132441,
+      "learning_rate": 8.389916277125099e-06,
+      "loss": 0.3271,
+      "step": 4532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.605666291927062,
+      "learning_rate": 8.389167547790407e-06,
+      "loss": 0.3182,
+      "step": 4533
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7435515785285134,
+      "learning_rate": 8.388418677833132e-06,
+      "loss": 0.3269,
+      "step": 4534
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3544372758044134,
+      "learning_rate": 8.387669667284351e-06,
+      "loss": 0.3291,
+      "step": 4535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.319882159195161,
+      "learning_rate": 8.386920516175135e-06,
+      "loss": 0.32,
+      "step": 4536
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8330191414718793,
+      "learning_rate": 8.386171224536573e-06,
+      "loss": 0.3397,
+      "step": 4537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6620619363299594,
+      "learning_rate": 8.385421792399753e-06,
+      "loss": 0.3347,
+      "step": 4538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.28094788853945,
+      "learning_rate": 8.384672219795766e-06,
+      "loss": 0.3278,
+      "step": 4539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.745254652276401,
+      "learning_rate": 8.38392250675572e-06,
+      "loss": 0.3362,
+      "step": 4540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7561920695282103,
+      "learning_rate": 8.383172653310718e-06,
+      "loss": 0.3304,
+      "step": 4541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8161784410716515,
+      "learning_rate": 8.382422659491873e-06,
+      "loss": 0.3289,
+      "step": 4542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7828874551969827,
+      "learning_rate": 8.381672525330305e-06,
+      "loss": 0.3089,
+      "step": 4543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8769878242003888,
+      "learning_rate": 8.380922250857137e-06,
+      "loss": 0.3293,
+      "step": 4544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4945991479535552,
+      "learning_rate": 8.380171836103499e-06,
+      "loss": 0.3183,
+      "step": 4545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.5309737210634635,
+      "learning_rate": 8.379421281100529e-06,
+      "loss": 0.3357,
+      "step": 4546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9862667447555324,
+      "learning_rate": 8.378670585879365e-06,
+      "loss": 0.314,
+      "step": 4547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1606467474212208,
+      "learning_rate": 8.37791975047116e-06,
+      "loss": 0.3195,
+      "step": 4548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.4590132781243565,
+      "learning_rate": 8.377168774907063e-06,
+      "loss": 0.3186,
+      "step": 4549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1345334959308437,
+      "learning_rate": 8.376417659218234e-06,
+      "loss": 0.3297,
+      "step": 4550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2319609286801514,
+      "learning_rate": 8.375666403435842e-06,
+      "loss": 0.3329,
+      "step": 4551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.181362994384634,
+      "learning_rate": 8.374915007591053e-06,
+      "loss": 0.3396,
+      "step": 4552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.959476969109861,
+      "learning_rate": 8.374163471715046e-06,
+      "loss": 0.3473,
+      "step": 4553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.787575127099224,
+      "learning_rate": 8.373411795839003e-06,
+      "loss": 0.3293,
+      "step": 4554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6641799563675863,
+      "learning_rate": 8.372659979994116e-06,
+      "loss": 0.3301,
+      "step": 4555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.621177104530586,
+      "learning_rate": 8.371908024211572e-06,
+      "loss": 0.3157,
+      "step": 4556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2845599080276258,
+      "learning_rate": 8.371155928522578e-06,
+      "loss": 0.3256,
+      "step": 4557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9115502875416628,
+      "learning_rate": 8.370403692958334e-06,
+      "loss": 0.3152,
+      "step": 4558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9359034905984105,
+      "learning_rate": 8.369651317550055e-06,
+      "loss": 0.326,
+      "step": 4559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.47395122757619,
+      "learning_rate": 8.368898802328957e-06,
+      "loss": 0.3353,
+      "step": 4560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.112484592043131,
+      "learning_rate": 8.368146147326266e-06,
+      "loss": 0.3227,
+      "step": 4561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.837742348531059,
+      "learning_rate": 8.367393352573207e-06,
+      "loss": 0.3228,
+      "step": 4562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8198363011071297,
+      "learning_rate": 8.366640418101018e-06,
+      "loss": 0.3044,
+      "step": 4563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4792177581157653,
+      "learning_rate": 8.365887343940937e-06,
+      "loss": 0.3083,
+      "step": 4564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0749427789794592,
+      "learning_rate": 8.365134130124214e-06,
+      "loss": 0.3141,
+      "step": 4565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4680932736556904,
+      "learning_rate": 8.364380776682095e-06,
+      "loss": 0.3257,
+      "step": 4566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5283672458531308,
+      "learning_rate": 8.363627283645843e-06,
+      "loss": 0.3326,
+      "step": 4567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.393840568899809,
+      "learning_rate": 8.362873651046722e-06,
+      "loss": 0.3201,
+      "step": 4568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5101834231715774,
+      "learning_rate": 8.362119878916e-06,
+      "loss": 0.3444,
+      "step": 4569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8470073253909227,
+      "learning_rate": 8.361365967284951e-06,
+      "loss": 0.326,
+      "step": 4570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.531618393204744,
+      "learning_rate": 8.360611916184859e-06,
+      "loss": 0.3273,
+      "step": 4571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8945311703301655,
+      "learning_rate": 8.359857725647009e-06,
+      "loss": 0.3266,
+      "step": 4572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.212260774091573,
+      "learning_rate": 8.359103395702692e-06,
+      "loss": 0.3195,
+      "step": 4573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.352783258966287,
+      "learning_rate": 8.358348926383211e-06,
+      "loss": 0.3288,
+      "step": 4574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5333192301589995,
+      "learning_rate": 8.357594317719868e-06,
+      "loss": 0.322,
+      "step": 4575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4610462430917213,
+      "learning_rate": 8.356839569743974e-06,
+      "loss": 0.3299,
+      "step": 4576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.887917428489183,
+      "learning_rate": 8.35608468248684e-06,
+      "loss": 0.3276,
+      "step": 4577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.56190176490763,
+      "learning_rate": 8.355329655979794e-06,
+      "loss": 0.3346,
+      "step": 4578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1840627550859883,
+      "learning_rate": 8.35457449025416e-06,
+      "loss": 0.3176,
+      "step": 4579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8388061367703301,
+      "learning_rate": 8.353819185341275e-06,
+      "loss": 0.3181,
+      "step": 4580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8328163528404189,
+      "learning_rate": 8.353063741272471e-06,
+      "loss": 0.3113,
+      "step": 4581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4526576376010834,
+      "learning_rate": 8.352308158079099e-06,
+      "loss": 0.3248,
+      "step": 4582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7219056871357408,
+      "learning_rate": 8.351552435792507e-06,
+      "loss": 0.316,
+      "step": 4583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6905315900768709,
+      "learning_rate": 8.35079657444405e-06,
+      "loss": 0.5059,
+      "step": 4584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.641832623872991,
+      "learning_rate": 8.350040574065093e-06,
+      "loss": 0.4916,
+      "step": 4585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6668764451013662,
+      "learning_rate": 8.349284434687001e-06,
+      "loss": 0.3373,
+      "step": 4586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5941483496192381,
+      "learning_rate": 8.348528156341148e-06,
+      "loss": 0.4646,
+      "step": 4587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.479937255342335,
+      "learning_rate": 8.347771739058917e-06,
+      "loss": 0.3096,
+      "step": 4588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.488931151716721,
+      "learning_rate": 8.34701518287169e-06,
+      "loss": 0.3407,
+      "step": 4589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.957149558452516,
+      "learning_rate": 8.346258487810855e-06,
+      "loss": 0.3442,
+      "step": 4590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.175477160163789,
+      "learning_rate": 8.345501653907813e-06,
+      "loss": 0.3293,
+      "step": 4591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7340860703640896,
+      "learning_rate": 8.344744681193966e-06,
+      "loss": 0.5371,
+      "step": 4592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4367219093086712,
+      "learning_rate": 8.343987569700723e-06,
+      "loss": 0.3157,
+      "step": 4593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.348886619276724,
+      "learning_rate": 8.343230319459493e-06,
+      "loss": 0.3172,
+      "step": 4594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.85393782325142,
+      "learning_rate": 8.3424729305017e-06,
+      "loss": 0.3538,
+      "step": 4595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5583309786518522,
+      "learning_rate": 8.341715402858768e-06,
+      "loss": 0.3281,
+      "step": 4596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.494209692042095,
+      "learning_rate": 8.340957736562127e-06,
+      "loss": 0.3219,
+      "step": 4597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.722372494117495,
+      "learning_rate": 8.340199931643218e-06,
+      "loss": 0.3217,
+      "step": 4598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2868989733023135,
+      "learning_rate": 8.339441988133478e-06,
+      "loss": 0.3336,
+      "step": 4599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7866223372150567,
+      "learning_rate": 8.338683906064361e-06,
+      "loss": 0.318,
+      "step": 4600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2134149476295644,
+      "learning_rate": 8.337925685467318e-06,
+      "loss": 0.3155,
+      "step": 4601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5640175712540683,
+      "learning_rate": 8.337167326373809e-06,
+      "loss": 0.3295,
+      "step": 4602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7998947475940565,
+      "learning_rate": 8.3364088288153e-06,
+      "loss": 0.3357,
+      "step": 4603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6051205017039099,
+      "learning_rate": 8.335650192823263e-06,
+      "loss": 0.3163,
+      "step": 4604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.161571352742066,
+      "learning_rate": 8.334891418429174e-06,
+      "loss": 0.3105,
+      "step": 4605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.988979064688914,
+      "learning_rate": 8.334132505664519e-06,
+      "loss": 0.3223,
+      "step": 4606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.736417459213877,
+      "learning_rate": 8.333373454560782e-06,
+      "loss": 0.3307,
+      "step": 4607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6014179630032326,
+      "learning_rate": 8.33261426514946e-06,
+      "loss": 0.3223,
+      "step": 4608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.657381201666818,
+      "learning_rate": 8.331854937462052e-06,
+      "loss": 0.5172,
+      "step": 4609
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3674659918524494,
+      "learning_rate": 8.331095471530066e-06,
+      "loss": 0.3325,
+      "step": 4610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3468979278569857,
+      "learning_rate": 8.33033586738501e-06,
+      "loss": 0.3156,
+      "step": 4611
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7585464608941155,
+      "learning_rate": 8.329576125058406e-06,
+      "loss": 0.3245,
+      "step": 4612
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.77020996096185,
+      "learning_rate": 8.328816244581774e-06,
+      "loss": 0.3179,
+      "step": 4613
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.35598120381295,
+      "learning_rate": 8.328056225986642e-06,
+      "loss": 0.3384,
+      "step": 4614
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6658406559141374,
+      "learning_rate": 8.327296069304547e-06,
+      "loss": 0.3289,
+      "step": 4615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.467570226639747,
+      "learning_rate": 8.326535774567029e-06,
+      "loss": 0.3234,
+      "step": 4616
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6365120975622698,
+      "learning_rate": 8.325775341805634e-06,
+      "loss": 0.3279,
+      "step": 4617
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7478469575446611,
+      "learning_rate": 8.325014771051909e-06,
+      "loss": 0.3288,
+      "step": 4618
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.997977303998103,
+      "learning_rate": 8.324254062337419e-06,
+      "loss": 0.3227,
+      "step": 4619
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9471147861575473,
+      "learning_rate": 8.323493215693721e-06,
+      "loss": 0.3211,
+      "step": 4620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7943934292722723,
+      "learning_rate": 8.322732231152386e-06,
+      "loss": 0.3436,
+      "step": 4621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.04573953156783,
+      "learning_rate": 8.32197110874499e-06,
+      "loss": 0.3142,
+      "step": 4622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8827940454540946,
+      "learning_rate": 8.321209848503112e-06,
+      "loss": 0.3174,
+      "step": 4623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3611606771710143,
+      "learning_rate": 8.320448450458339e-06,
+      "loss": 0.3154,
+      "step": 4624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3159068448606388,
+      "learning_rate": 8.319686914642261e-06,
+      "loss": 0.3214,
+      "step": 4625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6221485082863762,
+      "learning_rate": 8.318925241086477e-06,
+      "loss": 0.5184,
+      "step": 4626
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.030321585008596,
+      "learning_rate": 8.31816342982259e-06,
+      "loss": 0.3225,
+      "step": 4627
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3398635464810131,
+      "learning_rate": 8.317401480882207e-06,
+      "loss": 0.3279,
+      "step": 4628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.528329638237705,
+      "learning_rate": 8.316639394296947e-06,
+      "loss": 0.3319,
+      "step": 4629
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.3018221375463925,
+      "learning_rate": 8.315877170098425e-06,
+      "loss": 0.3286,
+      "step": 4630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8875170625119404,
+      "learning_rate": 8.315114808318269e-06,
+      "loss": 0.3098,
+      "step": 4631
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.28899150095528,
+      "learning_rate": 8.314352308988114e-06,
+      "loss": 0.3377,
+      "step": 4632
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8717690001076601,
+      "learning_rate": 8.313589672139593e-06,
+      "loss": 0.3251,
+      "step": 4633
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5992361475426525,
+      "learning_rate": 8.312826897804352e-06,
+      "loss": 0.495,
+      "step": 4634
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.815845661611426,
+      "learning_rate": 8.312063986014039e-06,
+      "loss": 0.3403,
+      "step": 4635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9168548744208262,
+      "learning_rate": 8.311300936800306e-06,
+      "loss": 0.3238,
+      "step": 4636
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.235787112400163,
+      "learning_rate": 8.310537750194819e-06,
+      "loss": 0.3287,
+      "step": 4637
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.6813044200360205,
+      "learning_rate": 8.309774426229237e-06,
+      "loss": 0.3224,
+      "step": 4638
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4011164727267102,
+      "learning_rate": 8.309010964935236e-06,
+      "loss": 0.3136,
+      "step": 4639
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6158207209231936,
+      "learning_rate": 8.308247366344494e-06,
+      "loss": 0.3341,
+      "step": 4640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.733443758580431,
+      "learning_rate": 8.307483630488693e-06,
+      "loss": 0.3332,
+      "step": 4641
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.319928529755119,
+      "learning_rate": 8.30671975739952e-06,
+      "loss": 0.3152,
+      "step": 4642
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8969761869791246,
+      "learning_rate": 8.305955747108672e-06,
+      "loss": 0.329,
+      "step": 4643
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.934293984723378,
+      "learning_rate": 8.305191599647846e-06,
+      "loss": 0.3244,
+      "step": 4644
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1444513735312634,
+      "learning_rate": 8.304427315048752e-06,
+      "loss": 0.328,
+      "step": 4645
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0981589553165523,
+      "learning_rate": 8.303662893343099e-06,
+      "loss": 0.3279,
+      "step": 4646
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4599554538301995,
+      "learning_rate": 8.302898334562604e-06,
+      "loss": 0.3039,
+      "step": 4647
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.818042291414395,
+      "learning_rate": 8.30213363873899e-06,
+      "loss": 0.3419,
+      "step": 4648
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6640616622599838,
+      "learning_rate": 8.301368805903988e-06,
+      "loss": 0.322,
+      "step": 4649
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3387610727095836,
+      "learning_rate": 8.300603836089329e-06,
+      "loss": 0.3032,
+      "step": 4650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0141298870195135,
+      "learning_rate": 8.299838729326754e-06,
+      "loss": 0.3531,
+      "step": 4651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9261902144520084,
+      "learning_rate": 8.29907348564801e-06,
+      "loss": 0.3197,
+      "step": 4652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8634593622131974,
+      "learning_rate": 8.298308105084845e-06,
+      "loss": 0.3114,
+      "step": 4653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5803961812105665,
+      "learning_rate": 8.29754258766902e-06,
+      "loss": 0.3529,
+      "step": 4654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7374020598179145,
+      "learning_rate": 8.296776933432296e-06,
+      "loss": 0.348,
+      "step": 4655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7131753006160757,
+      "learning_rate": 8.296011142406442e-06,
+      "loss": 0.331,
+      "step": 4656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3262733550127155,
+      "learning_rate": 8.29524521462323e-06,
+      "loss": 0.3108,
+      "step": 4657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.006111614964186,
+      "learning_rate": 8.294479150114443e-06,
+      "loss": 0.346,
+      "step": 4658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5931057728483875,
+      "learning_rate": 8.293712948911862e-06,
+      "loss": 0.3248,
+      "step": 4659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5811684522172256,
+      "learning_rate": 8.292946611047282e-06,
+      "loss": 0.3283,
+      "step": 4660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5822400594471997,
+      "learning_rate": 8.292180136552498e-06,
+      "loss": 0.3131,
+      "step": 4661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3242294608492324,
+      "learning_rate": 8.29141352545931e-06,
+      "loss": 0.3421,
+      "step": 4662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2683721125412544,
+      "learning_rate": 8.290646777799531e-06,
+      "loss": 0.3164,
+      "step": 4663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7722620972330796,
+      "learning_rate": 8.289879893604973e-06,
+      "loss": 0.3224,
+      "step": 4664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.155502333050992,
+      "learning_rate": 8.289112872907454e-06,
+      "loss": 0.3285,
+      "step": 4665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9188493048321282,
+      "learning_rate": 8.2883457157388e-06,
+      "loss": 0.3478,
+      "step": 4666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.03795725118682,
+      "learning_rate": 8.28757842213084e-06,
+      "loss": 0.339,
+      "step": 4667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.084669198820233,
+      "learning_rate": 8.286810992115413e-06,
+      "loss": 0.3402,
+      "step": 4668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9010436454227169,
+      "learning_rate": 8.28604342572436e-06,
+      "loss": 0.2943,
+      "step": 4669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8805846955981191,
+      "learning_rate": 8.28527572298953e-06,
+      "loss": 0.3494,
+      "step": 4670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5105951866836607,
+      "learning_rate": 8.284507883942771e-06,
+      "loss": 0.3298,
+      "step": 4671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8701312657424063,
+      "learning_rate": 8.283739908615949e-06,
+      "loss": 0.3401,
+      "step": 4672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9356588574831817,
+      "learning_rate": 8.282971797040925e-06,
+      "loss": 0.326,
+      "step": 4673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5187844902542402,
+      "learning_rate": 8.28220354924957e-06,
+      "loss": 0.3107,
+      "step": 4674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1035198049693054,
+      "learning_rate": 8.28143516527376e-06,
+      "loss": 0.3319,
+      "step": 4675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 22.636282393178313,
+      "learning_rate": 8.280666645145377e-06,
+      "loss": 0.3254,
+      "step": 4676
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.341425307095218,
+      "learning_rate": 8.279897988896306e-06,
+      "loss": 0.3503,
+      "step": 4677
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2542353033770426,
+      "learning_rate": 8.279129196558443e-06,
+      "loss": 0.3463,
+      "step": 4678
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.656661930080295,
+      "learning_rate": 8.278360268163686e-06,
+      "loss": 0.3345,
+      "step": 4679
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2763264469002285,
+      "learning_rate": 8.277591203743937e-06,
+      "loss": 0.3364,
+      "step": 4680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5888613647386105,
+      "learning_rate": 8.276822003331108e-06,
+      "loss": 0.3112,
+      "step": 4681
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.301783583669015,
+      "learning_rate": 8.276052666957113e-06,
+      "loss": 0.3297,
+      "step": 4682
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6590518295557342,
+      "learning_rate": 8.275283194653876e-06,
+      "loss": 0.3207,
+      "step": 4683
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.573784037298755,
+      "learning_rate": 8.27451358645332e-06,
+      "loss": 0.3242,
+      "step": 4684
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6889412945575218,
+      "learning_rate": 8.27374384238738e-06,
+      "loss": 0.5102,
+      "step": 4685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0184875092933416,
+      "learning_rate": 8.272973962487991e-06,
+      "loss": 0.3179,
+      "step": 4686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7308184830125093,
+      "learning_rate": 8.272203946787101e-06,
+      "loss": 0.3371,
+      "step": 4687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.621088868982925,
+      "learning_rate": 8.271433795316655e-06,
+      "loss": 0.319,
+      "step": 4688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0747084076055846,
+      "learning_rate": 8.270663508108611e-06,
+      "loss": 0.3242,
+      "step": 4689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6603367027911864,
+      "learning_rate": 8.26989308519493e-06,
+      "loss": 0.5008,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.85065133230469,
+      "learning_rate": 8.269122526607577e-06,
+      "loss": 0.3447,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8972568965185854,
+      "learning_rate": 8.268351832378521e-06,
+      "loss": 0.3276,
+      "step": 4692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.379889960167494,
+      "learning_rate": 8.267581002539745e-06,
+      "loss": 0.3097,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.813434996376796,
+      "learning_rate": 8.266810037123229e-06,
+      "loss": 0.331,
+      "step": 4694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.022182380703507,
+      "learning_rate": 8.26603893616096e-06,
+      "loss": 0.3046,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.313427973205733,
+      "learning_rate": 8.265267699684937e-06,
+      "loss": 0.3517,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2638290308411952,
+      "learning_rate": 8.264496327727157e-06,
+      "loss": 0.3138,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1232531724630594,
+      "learning_rate": 8.263724820319624e-06,
+      "loss": 0.3038,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.553147675273588,
+      "learning_rate": 8.262953177494353e-06,
+      "loss": 0.314,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.351131033381673,
+      "learning_rate": 8.262181399283359e-06,
+      "loss": 0.3085,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.609495727538814,
+      "learning_rate": 8.261409485718663e-06,
+      "loss": 0.3145,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9284030865348525,
+      "learning_rate": 8.260637436832295e-06,
+      "loss": 0.3384,
+      "step": 4702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7319450065480584,
+      "learning_rate": 8.259865252656289e-06,
+      "loss": 0.319,
+      "step": 4703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6200426032426879,
+      "learning_rate": 8.259092933222683e-06,
+      "loss": 0.4603,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.693101313184172,
+      "learning_rate": 8.258320478563524e-06,
+      "loss": 0.3158,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.889716943238856,
+      "learning_rate": 8.25754788871086e-06,
+      "loss": 0.3037,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4358539033489492,
+      "learning_rate": 8.256775163696746e-06,
+      "loss": 0.3106,
+      "step": 4707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.238543110556,
+      "learning_rate": 8.256002303553248e-06,
+      "loss": 0.337,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5476751159653763,
+      "learning_rate": 8.25522930831243e-06,
+      "loss": 0.3175,
+      "step": 4709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3005169638419916,
+      "learning_rate": 8.254456178006368e-06,
+      "loss": 0.337,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.599553246533363,
+      "learning_rate": 8.253682912667136e-06,
+      "loss": 0.4831,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5740293443919963,
+      "learning_rate": 8.252909512326821e-06,
+      "loss": 0.4913,
+      "step": 4712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.546963579802148,
+      "learning_rate": 8.252135977017513e-06,
+      "loss": 0.3189,
+      "step": 4713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.429125795642956,
+      "learning_rate": 8.251362306771306e-06,
+      "loss": 0.3568,
+      "step": 4714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.189305238838425,
+      "learning_rate": 8.250588501620305e-06,
+      "loss": 0.3221,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5843130933913105,
+      "learning_rate": 8.249814561596612e-06,
+      "loss": 0.3119,
+      "step": 4716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.071138163688954,
+      "learning_rate": 8.24904048673234e-06,
+      "loss": 0.3593,
+      "step": 4717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5238599435112223,
+      "learning_rate": 8.248266277059607e-06,
+      "loss": 0.3147,
+      "step": 4718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1740048627002473,
+      "learning_rate": 8.247491932610536e-06,
+      "loss": 0.3211,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.140818851619252,
+      "learning_rate": 8.246717453417257e-06,
+      "loss": 0.3097,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4118365828492045,
+      "learning_rate": 8.245942839511905e-06,
+      "loss": 0.3084,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8780104654279364,
+      "learning_rate": 8.245168090926618e-06,
+      "loss": 0.3362,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.509275582536811,
+      "learning_rate": 8.244393207693544e-06,
+      "loss": 0.3217,
+      "step": 4723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.307156329842178,
+      "learning_rate": 8.243618189844831e-06,
+      "loss": 0.3173,
+      "step": 4724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6175065624426581,
+      "learning_rate": 8.24284303741264e-06,
+      "loss": 0.3281,
+      "step": 4725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.091261383505988,
+      "learning_rate": 8.242067750429131e-06,
+      "loss": 0.3256,
+      "step": 4726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6042560866381548,
+      "learning_rate": 8.241292328926473e-06,
+      "loss": 0.3124,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4559835196693374,
+      "learning_rate": 8.240516772936837e-06,
+      "loss": 0.3455,
+      "step": 4728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4597397892968809,
+      "learning_rate": 8.239741082492405e-06,
+      "loss": 0.329,
+      "step": 4729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.714973562946189,
+      "learning_rate": 8.238965257625363e-06,
+      "loss": 0.5296,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8917801488245396,
+      "learning_rate": 8.238189298367898e-06,
+      "loss": 0.3334,
+      "step": 4731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.188231882240646,
+      "learning_rate": 8.237413204752208e-06,
+      "loss": 0.3409,
+      "step": 4732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7461597182270046,
+      "learning_rate": 8.236636976810492e-06,
+      "loss": 0.3443,
+      "step": 4733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.224238394081794,
+      "learning_rate": 8.235860614574962e-06,
+      "loss": 0.3288,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.68912360675444,
+      "learning_rate": 8.235084118077826e-06,
+      "loss": 0.2999,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6644920691114202,
+      "learning_rate": 8.234307487351303e-06,
+      "loss": 0.3414,
+      "step": 4736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9254952544282533,
+      "learning_rate": 8.233530722427618e-06,
+      "loss": 0.3194,
+      "step": 4737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7836344216234346,
+      "learning_rate": 8.232753823339001e-06,
+      "loss": 0.3346,
+      "step": 4738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9185557546294354,
+      "learning_rate": 8.231976790117684e-06,
+      "loss": 0.336,
+      "step": 4739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7885959894956402,
+      "learning_rate": 8.231199622795912e-06,
+      "loss": 0.3007,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.401987209489957,
+      "learning_rate": 8.230422321405925e-06,
+      "loss": 0.3291,
+      "step": 4741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.816715072691335,
+      "learning_rate": 8.22964488597998e-06,
+      "loss": 0.3445,
+      "step": 4742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8542396643594432,
+      "learning_rate": 8.228867316550332e-06,
+      "loss": 0.3322,
+      "step": 4743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5081081724294076,
+      "learning_rate": 8.228089613149244e-06,
+      "loss": 0.3352,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6356742044834944,
+      "learning_rate": 8.227311775808984e-06,
+      "loss": 0.3042,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.520840142695837,
+      "learning_rate": 8.226533804561828e-06,
+      "loss": 0.3259,
+      "step": 4746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.768885534148715,
+      "learning_rate": 8.22575569944005e-06,
+      "loss": 0.3233,
+      "step": 4747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3583230985499926,
+      "learning_rate": 8.22497746047594e-06,
+      "loss": 0.3386,
+      "step": 4748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5577941188261932,
+      "learning_rate": 8.224199087701787e-06,
+      "loss": 0.3178,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7525710060671866,
+      "learning_rate": 8.223420581149887e-06,
+      "loss": 0.3219,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.948686355608059,
+      "learning_rate": 8.222641940852543e-06,
+      "loss": 0.3244,
+      "step": 4751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5969601204197872,
+      "learning_rate": 8.221863166842057e-06,
+      "loss": 0.3273,
+      "step": 4752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6072918833551268,
+      "learning_rate": 8.22108425915075e-06,
+      "loss": 0.32,
+      "step": 4753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7374397474057224,
+      "learning_rate": 8.220305217810933e-06,
+      "loss": 0.3058,
+      "step": 4754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5285658154607755,
+      "learning_rate": 8.219526042854933e-06,
+      "loss": 0.314,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6978959663770667,
+      "learning_rate": 8.21874673431508e-06,
+      "loss": 0.3296,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.685246099049766,
+      "learning_rate": 8.217967292223707e-06,
+      "loss": 0.3229,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.516889116784165,
+      "learning_rate": 8.217187716613156e-06,
+      "loss": 0.3214,
+      "step": 4758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5307710719855767,
+      "learning_rate": 8.21640800751577e-06,
+      "loss": 0.3346,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.339994045329839,
+      "learning_rate": 8.215628164963906e-06,
+      "loss": 0.3115,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.643101351187922,
+      "learning_rate": 8.214848188989916e-06,
+      "loss": 0.3225,
+      "step": 4761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7033291289125447,
+      "learning_rate": 8.214068079626166e-06,
+      "loss": 0.491,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1133177261523097,
+      "learning_rate": 8.213287836905021e-06,
+      "loss": 0.3404,
+      "step": 4763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.45630998250199,
+      "learning_rate": 8.212507460858857e-06,
+      "loss": 0.3314,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4375229160064453,
+      "learning_rate": 8.21172695152005e-06,
+      "loss": 0.3505,
+      "step": 4765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0368390595621215,
+      "learning_rate": 8.21094630892099e-06,
+      "loss": 0.3092,
+      "step": 4766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8435608026026553,
+      "learning_rate": 8.210165533094064e-06,
+      "loss": 0.3264,
+      "step": 4767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.623385415676598,
+      "learning_rate": 8.209384624071667e-06,
+      "loss": 0.3433,
+      "step": 4768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2817442670181367,
+      "learning_rate": 8.208603581886203e-06,
+      "loss": 0.3367,
+      "step": 4769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6621890877153036,
+      "learning_rate": 8.207822406570077e-06,
+      "loss": 0.3169,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3938859499120446,
+      "learning_rate": 8.207041098155701e-06,
+      "loss": 0.3193,
+      "step": 4771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8851538537508348,
+      "learning_rate": 8.206259656675493e-06,
+      "loss": 0.3333,
+      "step": 4772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.531857536210539,
+      "learning_rate": 8.205478082161877e-06,
+      "loss": 0.3247,
+      "step": 4773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.44634320497755,
+      "learning_rate": 8.204696374647282e-06,
+      "loss": 0.3224,
+      "step": 4774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.280888599599943,
+      "learning_rate": 8.203914534164143e-06,
+      "loss": 0.3238,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6805330915645307,
+      "learning_rate": 8.203132560744898e-06,
+      "loss": 0.3106,
+      "step": 4776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.865548206504745,
+      "learning_rate": 8.202350454421995e-06,
+      "loss": 0.3333,
+      "step": 4777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.318842918556991,
+      "learning_rate": 8.201568215227882e-06,
+      "loss": 0.3232,
+      "step": 4778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.388261525992749,
+      "learning_rate": 8.200785843195019e-06,
+      "loss": 0.3273,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3297509743118145,
+      "learning_rate": 8.200003338355863e-06,
+      "loss": 0.3306,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.393080028369493,
+      "learning_rate": 8.199220700742889e-06,
+      "loss": 0.3222,
+      "step": 4781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6313111032338217,
+      "learning_rate": 8.198437930388562e-06,
+      "loss": 0.3494,
+      "step": 4782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4352270482371632,
+      "learning_rate": 8.197655027325366e-06,
+      "loss": 0.2982,
+      "step": 4783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.206024029455773,
+      "learning_rate": 8.196871991585784e-06,
+      "loss": 0.3443,
+      "step": 4784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.653110086631432,
+      "learning_rate": 8.196088823202302e-06,
+      "loss": 0.3143,
+      "step": 4785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2011751161782405,
+      "learning_rate": 8.19530552220742e-06,
+      "loss": 0.3529,
+      "step": 4786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.924828819173896,
+      "learning_rate": 8.194522088633637e-06,
+      "loss": 0.3324,
+      "step": 4787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1398642044622593,
+      "learning_rate": 8.193738522513458e-06,
+      "loss": 0.3341,
+      "step": 4788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7157670836447547,
+      "learning_rate": 8.192954823879392e-06,
+      "loss": 0.3054,
+      "step": 4789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8737679459565855,
+      "learning_rate": 8.192170992763962e-06,
+      "loss": 0.2988,
+      "step": 4790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.999504728006175,
+      "learning_rate": 8.191387029199687e-06,
+      "loss": 0.3295,
+      "step": 4791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7098577395923357,
+      "learning_rate": 8.190602933219096e-06,
+      "loss": 0.5288,
+      "step": 4792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.072568427445285,
+      "learning_rate": 8.189818704854721e-06,
+      "loss": 0.3392,
+      "step": 4793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9416515731766864,
+      "learning_rate": 8.189034344139102e-06,
+      "loss": 0.3369,
+      "step": 4794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.5445706462693876,
+      "learning_rate": 8.188249851104785e-06,
+      "loss": 0.3238,
+      "step": 4795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6009115325262264,
+      "learning_rate": 8.18746522578432e-06,
+      "loss": 0.3071,
+      "step": 4796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8634447593031986,
+      "learning_rate": 8.186680468210258e-06,
+      "loss": 0.3042,
+      "step": 4797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7643107661326325,
+      "learning_rate": 8.185895578415164e-06,
+      "loss": 0.3352,
+      "step": 4798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7129930040941912,
+      "learning_rate": 8.185110556431606e-06,
+      "loss": 0.3261,
+      "step": 4799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8987423140500093,
+      "learning_rate": 8.184325402292151e-06,
+      "loss": 0.3098,
+      "step": 4800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5328024607127184,
+      "learning_rate": 8.18354011602938e-06,
+      "loss": 0.3157,
+      "step": 4801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.333752570392361,
+      "learning_rate": 8.182754697675875e-06,
+      "loss": 0.3052,
+      "step": 4802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3306054321306156,
+      "learning_rate": 8.181969147264224e-06,
+      "loss": 0.3358,
+      "step": 4803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1146384907583307,
+      "learning_rate": 8.181183464827022e-06,
+      "loss": 0.306,
+      "step": 4804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0758202525059914,
+      "learning_rate": 8.180397650396868e-06,
+      "loss": 0.3044,
+      "step": 4805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6975350539339977,
+      "learning_rate": 8.179611704006364e-06,
+      "loss": 0.3127,
+      "step": 4806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4776622404553157,
+      "learning_rate": 8.178825625688126e-06,
+      "loss": 0.3298,
+      "step": 4807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8269808749535046,
+      "learning_rate": 8.178039415474765e-06,
+      "loss": 0.3403,
+      "step": 4808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.145335388209343,
+      "learning_rate": 8.177253073398903e-06,
+      "loss": 0.3016,
+      "step": 4809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8458166161902985,
+      "learning_rate": 8.17646659949317e-06,
+      "loss": 0.3338,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.144466171744352,
+      "learning_rate": 8.175679993790194e-06,
+      "loss": 0.3207,
+      "step": 4811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6336326139927528,
+      "learning_rate": 8.174893256322613e-06,
+      "loss": 0.3513,
+      "step": 4812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.51376091258254,
+      "learning_rate": 8.174106387123073e-06,
+      "loss": 0.3496,
+      "step": 4813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7210806637465388,
+      "learning_rate": 8.173319386224222e-06,
+      "loss": 0.3484,
+      "step": 4814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.65241854735433,
+      "learning_rate": 8.172532253658711e-06,
+      "loss": 0.3292,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9905391000063817,
+      "learning_rate": 8.171744989459205e-06,
+      "loss": 0.3308,
+      "step": 4816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7518605995680492,
+      "learning_rate": 8.170957593658365e-06,
+      "loss": 0.3068,
+      "step": 4817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.311984135292733,
+      "learning_rate": 8.170170066288861e-06,
+      "loss": 0.3356,
+      "step": 4818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.122085113950779,
+      "learning_rate": 8.169382407383368e-06,
+      "loss": 0.3172,
+      "step": 4819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.995065735215276,
+      "learning_rate": 8.168594616974572e-06,
+      "loss": 0.319,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2246262626361033,
+      "learning_rate": 8.167806695095157e-06,
+      "loss": 0.2997,
+      "step": 4821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7612564023005002,
+      "learning_rate": 8.167018641777817e-06,
+      "loss": 0.3288,
+      "step": 4822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8208704590702947,
+      "learning_rate": 8.166230457055247e-06,
+      "loss": 0.3219,
+      "step": 4823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8920000445494063,
+      "learning_rate": 8.16544214096015e-06,
+      "loss": 0.3308,
+      "step": 4824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.997619539738235,
+      "learning_rate": 8.164653693525239e-06,
+      "loss": 0.3073,
+      "step": 4825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3368032693286307,
+      "learning_rate": 8.163865114783223e-06,
+      "loss": 0.3083,
+      "step": 4826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.614647051404713,
+      "learning_rate": 8.163076404766823e-06,
+      "loss": 0.3022,
+      "step": 4827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7694390217696063,
+      "learning_rate": 8.162287563508767e-06,
+      "loss": 0.3218,
+      "step": 4828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3419304443359232,
+      "learning_rate": 8.161498591041783e-06,
+      "loss": 0.3246,
+      "step": 4829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.935574256862518,
+      "learning_rate": 8.160709487398605e-06,
+      "loss": 0.3258,
+      "step": 4830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.4843662681417458,
+      "learning_rate": 8.159920252611979e-06,
+      "loss": 0.3159,
+      "step": 4831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.47709063905116,
+      "learning_rate": 8.159130886714647e-06,
+      "loss": 0.3029,
+      "step": 4832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.059276053569302,
+      "learning_rate": 8.158341389739364e-06,
+      "loss": 0.3372,
+      "step": 4833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7453200918215377,
+      "learning_rate": 8.157551761718886e-06,
+      "loss": 0.4911,
+      "step": 4834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.658487968733307,
+      "learning_rate": 8.156762002685978e-06,
+      "loss": 0.316,
+      "step": 4835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.804281861595954,
+      "learning_rate": 8.155972112673406e-06,
+      "loss": 0.3176,
+      "step": 4836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.20719583393348,
+      "learning_rate": 8.155182091713946e-06,
+      "loss": 0.3131,
+      "step": 4837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3365859821620707,
+      "learning_rate": 8.154391939840379e-06,
+      "loss": 0.3064,
+      "step": 4838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7261160213559534,
+      "learning_rate": 8.153601657085486e-06,
+      "loss": 0.3408,
+      "step": 4839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.563921809444925,
+      "learning_rate": 8.15281124348206e-06,
+      "loss": 0.3439,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.779037035691217,
+      "learning_rate": 8.152020699062892e-06,
+      "loss": 0.3169,
+      "step": 4841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6113109661634013,
+      "learning_rate": 8.15123002386079e-06,
+      "loss": 0.3282,
+      "step": 4842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6786433123007816,
+      "learning_rate": 8.150439217908557e-06,
+      "loss": 0.2942,
+      "step": 4843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3948528641322693,
+      "learning_rate": 8.149648281239005e-06,
+      "loss": 0.3266,
+      "step": 4844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.406817990679698,
+      "learning_rate": 8.148857213884948e-06,
+      "loss": 0.3394,
+      "step": 4845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.435589378225438,
+      "learning_rate": 8.148066015879214e-06,
+      "loss": 0.338,
+      "step": 4846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 73.95252745390995,
+      "learning_rate": 8.147274687254632e-06,
+      "loss": 0.3143,
+      "step": 4847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.933090324683608,
+      "learning_rate": 8.146483228044032e-06,
+      "loss": 0.3226,
+      "step": 4848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.083182320648684,
+      "learning_rate": 8.14569163828025e-06,
+      "loss": 0.3297,
+      "step": 4849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.81068733173486,
+      "learning_rate": 8.144899917996139e-06,
+      "loss": 0.3232,
+      "step": 4850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5115646565506933,
+      "learning_rate": 8.144108067224544e-06,
+      "loss": 0.3594,
+      "step": 4851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3704651812605,
+      "learning_rate": 8.14331608599832e-06,
+      "loss": 0.3486,
+      "step": 4852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.750737390975413,
+      "learning_rate": 8.142523974350327e-06,
+      "loss": 0.3344,
+      "step": 4853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.641300177158323,
+      "learning_rate": 8.141731732313434e-06,
+      "loss": 0.321,
+      "step": 4854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3377095274666442,
+      "learning_rate": 8.140939359920512e-06,
+      "loss": 0.3266,
+      "step": 4855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2454839829427358,
+      "learning_rate": 8.140146857204433e-06,
+      "loss": 0.345,
+      "step": 4856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6052923028643362,
+      "learning_rate": 8.139354224198087e-06,
+      "loss": 0.3172,
+      "step": 4857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2004381869875216,
+      "learning_rate": 8.13856146093436e-06,
+      "loss": 0.3211,
+      "step": 4858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8407810825110196,
+      "learning_rate": 8.137768567446141e-06,
+      "loss": 0.3172,
+      "step": 4859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8974771862862094,
+      "learning_rate": 8.136975543766328e-06,
+      "loss": 0.3347,
+      "step": 4860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.834938564558332,
+      "learning_rate": 8.136182389927832e-06,
+      "loss": 0.3347,
+      "step": 4861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9249037333511576,
+      "learning_rate": 8.135389105963556e-06,
+      "loss": 0.3306,
+      "step": 4862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.869441650437946,
+      "learning_rate": 8.134595691906416e-06,
+      "loss": 0.3262,
+      "step": 4863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4694686351112236,
+      "learning_rate": 8.133802147789337e-06,
+      "loss": 0.3013,
+      "step": 4864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6259746653299456,
+      "learning_rate": 8.133008473645237e-06,
+      "loss": 0.3365,
+      "step": 4865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.679726853444609,
+      "learning_rate": 8.132214669507052e-06,
+      "loss": 0.3034,
+      "step": 4866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8052295131154517,
+      "learning_rate": 8.131420735407716e-06,
+      "loss": 0.3467,
+      "step": 4867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5368211521749648,
+      "learning_rate": 8.130626671380172e-06,
+      "loss": 0.3316,
+      "step": 4868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1732339031379313,
+      "learning_rate": 8.129832477457367e-06,
+      "loss": 0.3287,
+      "step": 4869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.860292595498335,
+      "learning_rate": 8.129038153672255e-06,
+      "loss": 0.3396,
+      "step": 4870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.155789115355333,
+      "learning_rate": 8.128243700057788e-06,
+      "loss": 0.3324,
+      "step": 4871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6100435575597394,
+      "learning_rate": 8.127449116646937e-06,
+      "loss": 0.311,
+      "step": 4872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7138489497901312,
+      "learning_rate": 8.126654403472668e-06,
+      "loss": 0.3104,
+      "step": 4873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7018536243420075,
+      "learning_rate": 8.125859560567952e-06,
+      "loss": 0.3157,
+      "step": 4874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.75272407707653,
+      "learning_rate": 8.12506458796577e-06,
+      "loss": 0.3296,
+      "step": 4875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.877593274176602,
+      "learning_rate": 8.124269485699108e-06,
+      "loss": 0.3159,
+      "step": 4876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.099054102035657,
+      "learning_rate": 8.123474253800958e-06,
+      "loss": 0.305,
+      "step": 4877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6060449639620602,
+      "learning_rate": 8.12267889230431e-06,
+      "loss": 0.3247,
+      "step": 4878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.09337815632032,
+      "learning_rate": 8.12188340124217e-06,
+      "loss": 0.3221,
+      "step": 4879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4440770016397826,
+      "learning_rate": 8.121087780647543e-06,
+      "loss": 0.3189,
+      "step": 4880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.290636335528498,
+      "learning_rate": 8.120292030553441e-06,
+      "loss": 0.3163,
+      "step": 4881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.955720730453499,
+      "learning_rate": 8.119496150992879e-06,
+      "loss": 0.3247,
+      "step": 4882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.585402146371944,
+      "learning_rate": 8.118700141998879e-06,
+      "loss": 0.328,
+      "step": 4883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8193494686149803,
+      "learning_rate": 8.117904003604472e-06,
+      "loss": 0.3118,
+      "step": 4884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5046889722588577,
+      "learning_rate": 8.117107735842693e-06,
+      "loss": 0.3476,
+      "step": 4885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0278179379997585,
+      "learning_rate": 8.116311338746574e-06,
+      "loss": 0.3097,
+      "step": 4886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3797793022230813,
+      "learning_rate": 8.115514812349163e-06,
+      "loss": 0.3331,
+      "step": 4887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9909270540906427,
+      "learning_rate": 8.11471815668351e-06,
+      "loss": 0.3227,
+      "step": 4888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5316997453624013,
+      "learning_rate": 8.113921371782667e-06,
+      "loss": 0.3279,
+      "step": 4889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2260613296304994,
+      "learning_rate": 8.113124457679695e-06,
+      "loss": 0.2957,
+      "step": 4890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1211395293278983,
+      "learning_rate": 8.112327414407662e-06,
+      "loss": 0.3133,
+      "step": 4891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7679163209400715,
+      "learning_rate": 8.111530241999634e-06,
+      "loss": 0.3431,
+      "step": 4892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.741665979087409,
+      "learning_rate": 8.11073294048869e-06,
+      "loss": 0.3054,
+      "step": 4893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2713386087154355,
+      "learning_rate": 8.109935509907911e-06,
+      "loss": 0.3154,
+      "step": 4894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5499479964956717,
+      "learning_rate": 8.109137950290385e-06,
+      "loss": 0.2943,
+      "step": 4895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6241753185389316,
+      "learning_rate": 8.108340261669204e-06,
+      "loss": 0.3314,
+      "step": 4896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.065475597908749,
+      "learning_rate": 8.107542444077461e-06,
+      "loss": 0.3514,
+      "step": 4897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3892492590291545,
+      "learning_rate": 8.106744497548265e-06,
+      "loss": 0.3229,
+      "step": 4898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7329539118650756,
+      "learning_rate": 8.105946422114721e-06,
+      "loss": 0.2981,
+      "step": 4899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7567142614783802,
+      "learning_rate": 8.105148217809941e-06,
+      "loss": 0.3045,
+      "step": 4900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6199260900646555,
+      "learning_rate": 8.104349884667048e-06,
+      "loss": 0.3203,
+      "step": 4901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.185619067488261,
+      "learning_rate": 8.103551422719164e-06,
+      "loss": 0.3026,
+      "step": 4902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0196664867388665,
+      "learning_rate": 8.10275283199942e-06,
+      "loss": 0.3277,
+      "step": 4903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4596505677728213,
+      "learning_rate": 8.10195411254095e-06,
+      "loss": 0.3148,
+      "step": 4904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0219263492276407,
+      "learning_rate": 8.101155264376892e-06,
+      "loss": 0.3206,
+      "step": 4905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.787627867209791,
+      "learning_rate": 8.100356287540397e-06,
+      "loss": 0.3254,
+      "step": 4906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.410642521194272,
+      "learning_rate": 8.09955718206461e-06,
+      "loss": 0.3093,
+      "step": 4907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0444840723299285,
+      "learning_rate": 8.098757947982692e-06,
+      "loss": 0.321,
+      "step": 4908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4640748718609533,
+      "learning_rate": 8.097958585327801e-06,
+      "loss": 0.3154,
+      "step": 4909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.42997998563266,
+      "learning_rate": 8.097159094133106e-06,
+      "loss": 0.2988,
+      "step": 4910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8019002532735567,
+      "learning_rate": 8.09635947443178e-06,
+      "loss": 0.5156,
+      "step": 4911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2354178381952123,
+      "learning_rate": 8.095559726256998e-06,
+      "loss": 0.3281,
+      "step": 4912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.041393490542275,
+      "learning_rate": 8.094759849641946e-06,
+      "loss": 0.3238,
+      "step": 4913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7435732778585893,
+      "learning_rate": 8.093959844619812e-06,
+      "loss": 0.3223,
+      "step": 4914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9563665748240093,
+      "learning_rate": 8.093159711223788e-06,
+      "loss": 0.3196,
+      "step": 4915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 16.896494403766674,
+      "learning_rate": 8.09235944948707e-06,
+      "loss": 0.3184,
+      "step": 4916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5403427987104072,
+      "learning_rate": 8.09155905944287e-06,
+      "loss": 0.3232,
+      "step": 4917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.278523899803821,
+      "learning_rate": 8.090758541124394e-06,
+      "loss": 0.313,
+      "step": 4918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0731620283807266,
+      "learning_rate": 8.089957894564851e-06,
+      "loss": 0.3244,
+      "step": 4919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9387298702383564,
+      "learning_rate": 8.08915711979747e-06,
+      "loss": 0.3258,
+      "step": 4920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0861885799415876,
+      "learning_rate": 8.088356216855474e-06,
+      "loss": 0.3176,
+      "step": 4921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3201997461856845,
+      "learning_rate": 8.087555185772093e-06,
+      "loss": 0.3651,
+      "step": 4922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.122575679762488,
+      "learning_rate": 8.08675402658056e-06,
+      "loss": 0.3112,
+      "step": 4923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7670045849754261,
+      "learning_rate": 8.085952739314123e-06,
+      "loss": 0.3215,
+      "step": 4924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9158837604562065,
+      "learning_rate": 8.085151324006024e-06,
+      "loss": 0.3022,
+      "step": 4925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.144228081612862,
+      "learning_rate": 8.084349780689517e-06,
+      "loss": 0.3358,
+      "step": 4926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.215449578261158,
+      "learning_rate": 8.08354810939786e-06,
+      "loss": 0.3011,
+      "step": 4927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8327874943896953,
+      "learning_rate": 8.082746310164316e-06,
+      "loss": 0.3094,
+      "step": 4928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.907844742173548,
+      "learning_rate": 8.08194438302215e-06,
+      "loss": 0.3482,
+      "step": 4929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2764122764388346,
+      "learning_rate": 8.081142328004638e-06,
+      "loss": 0.3163,
+      "step": 4930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6559404359038419,
+      "learning_rate": 8.080340145145058e-06,
+      "loss": 0.3305,
+      "step": 4931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8500527172802412,
+      "learning_rate": 8.079537834476692e-06,
+      "loss": 0.3137,
+      "step": 4932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6798642355980546,
+      "learning_rate": 8.078735396032835e-06,
+      "loss": 0.3312,
+      "step": 4933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9532205992175309,
+      "learning_rate": 8.077932829846774e-06,
+      "loss": 0.308,
+      "step": 4934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.587313719105938,
+      "learning_rate": 8.077130135951817e-06,
+      "loss": 0.3021,
+      "step": 4935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7471957552749761,
+      "learning_rate": 8.076327314381262e-06,
+      "loss": 0.5175,
+      "step": 4936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.477394257055449,
+      "learning_rate": 8.075524365168426e-06,
+      "loss": 0.3153,
+      "step": 4937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6740954872672904,
+      "learning_rate": 8.074721288346617e-06,
+      "loss": 0.3046,
+      "step": 4938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0022588763356945,
+      "learning_rate": 8.073918083949164e-06,
+      "loss": 0.3143,
+      "step": 4939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.107769896618522,
+      "learning_rate": 8.073114752009388e-06,
+      "loss": 0.3132,
+      "step": 4940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.767220171014733,
+      "learning_rate": 8.072311292560624e-06,
+      "loss": 0.308,
+      "step": 4941
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.565922059987354,
+      "learning_rate": 8.071507705636204e-06,
+      "loss": 0.3194,
+      "step": 4942
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.574200775674833,
+      "learning_rate": 8.070703991269477e-06,
+      "loss": 0.308,
+      "step": 4943
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9724278178689933,
+      "learning_rate": 8.069900149493786e-06,
+      "loss": 0.3212,
+      "step": 4944
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8255427530008028,
+      "learning_rate": 8.069096180342486e-06,
+      "loss": 0.3095,
+      "step": 4945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.60634504768918,
+      "learning_rate": 8.068292083848932e-06,
+      "loss": 0.3256,
+      "step": 4946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.373028519392421,
+      "learning_rate": 8.067487860046492e-06,
+      "loss": 0.3094,
+      "step": 4947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.664247309678174,
+      "learning_rate": 8.06668350896853e-06,
+      "loss": 0.3132,
+      "step": 4948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.753065592450937,
+      "learning_rate": 8.065879030648423e-06,
+      "loss": 0.3261,
+      "step": 4949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2169032541546447,
+      "learning_rate": 8.06507442511955e-06,
+      "loss": 0.318,
+      "step": 4950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.545690984191229,
+      "learning_rate": 8.064269692415296e-06,
+      "loss": 0.3255,
+      "step": 4951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.970656893769879,
+      "learning_rate": 8.063464832569048e-06,
+      "loss": 0.3155,
+      "step": 4952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2575909231858047,
+      "learning_rate": 8.062659845614202e-06,
+      "loss": 0.3308,
+      "step": 4953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.801862396995543,
+      "learning_rate": 8.06185473158416e-06,
+      "loss": 0.33,
+      "step": 4954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.160705406548701,
+      "learning_rate": 8.061049490512326e-06,
+      "loss": 0.3401,
+      "step": 4955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0949235628577947,
+      "learning_rate": 8.060244122432116e-06,
+      "loss": 0.3147,
+      "step": 4956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.941835161034645,
+      "learning_rate": 8.059438627376936e-06,
+      "loss": 0.3074,
+      "step": 4957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.744498126916404,
+      "learning_rate": 8.058633005380216e-06,
+      "loss": 0.3168,
+      "step": 4958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7371791076787418,
+      "learning_rate": 8.05782725647538e-06,
+      "loss": 0.4851,
+      "step": 4959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.648512559719333,
+      "learning_rate": 8.057021380695858e-06,
+      "loss": 0.3249,
+      "step": 4960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8496177680693084,
+      "learning_rate": 8.05621537807509e-06,
+      "loss": 0.3165,
+      "step": 4961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.874420013697325,
+      "learning_rate": 8.055409248646517e-06,
+      "loss": 0.308,
+      "step": 4962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9362228664429233,
+      "learning_rate": 8.05460299244359e-06,
+      "loss": 0.3011,
+      "step": 4963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8530717830919292,
+      "learning_rate": 8.053796609499755e-06,
+      "loss": 0.3365,
+      "step": 4964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.188405320612547,
+      "learning_rate": 8.052990099848478e-06,
+      "loss": 0.3054,
+      "step": 4965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9289555558700955,
+      "learning_rate": 8.052183463523218e-06,
+      "loss": 0.3065,
+      "step": 4966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.985572830928523,
+      "learning_rate": 8.051376700557445e-06,
+      "loss": 0.3128,
+      "step": 4967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9974222898370826,
+      "learning_rate": 8.050569810984632e-06,
+      "loss": 0.3191,
+      "step": 4968
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7153716759739575,
+      "learning_rate": 8.049762794838258e-06,
+      "loss": 0.3156,
+      "step": 4969
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9904526917215217,
+      "learning_rate": 8.04895565215181e-06,
+      "loss": 0.3171,
+      "step": 4970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.6899332359955075,
+      "learning_rate": 8.048148382958777e-06,
+      "loss": 0.3442,
+      "step": 4971
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.683926435589481,
+      "learning_rate": 8.047340987292653e-06,
+      "loss": 0.3386,
+      "step": 4972
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9911786050649964,
+      "learning_rate": 8.046533465186939e-06,
+      "loss": 0.3041,
+      "step": 4973
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.073778338160897,
+      "learning_rate": 8.04572581667514e-06,
+      "loss": 0.312,
+      "step": 4974
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.57490476913578,
+      "learning_rate": 8.044918041790768e-06,
+      "loss": 0.3215,
+      "step": 4975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.919997197183369,
+      "learning_rate": 8.044110140567337e-06,
+      "loss": 0.3131,
+      "step": 4976
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6494139966985812,
+      "learning_rate": 8.04330211303837e-06,
+      "loss": 0.4924,
+      "step": 4977
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7070700658497352,
+      "learning_rate": 8.042493959237391e-06,
+      "loss": 0.5016,
+      "step": 4978
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0483476573986326,
+      "learning_rate": 8.041685679197936e-06,
+      "loss": 0.3127,
+      "step": 4979
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9444596376193637,
+      "learning_rate": 8.040877272953538e-06,
+      "loss": 0.3319,
+      "step": 4980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.477915831462931,
+      "learning_rate": 8.040068740537741e-06,
+      "loss": 0.3317,
+      "step": 4981
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.230217332345066,
+      "learning_rate": 8.039260081984092e-06,
+      "loss": 0.3055,
+      "step": 4982
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7654025561999291,
+      "learning_rate": 8.038451297326146e-06,
+      "loss": 0.3122,
+      "step": 4983
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9280784523657415,
+      "learning_rate": 8.037642386597456e-06,
+      "loss": 0.308,
+      "step": 4984
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.166152756269824,
+      "learning_rate": 8.03683334983159e-06,
+      "loss": 0.333,
+      "step": 4985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.922670774962704,
+      "learning_rate": 8.036024187062113e-06,
+      "loss": 0.3172,
+      "step": 4986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9193641648410344,
+      "learning_rate": 8.0352148983226e-06,
+      "loss": 0.3169,
+      "step": 4987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.648087946869346,
+      "learning_rate": 8.03440548364663e-06,
+      "loss": 0.3145,
+      "step": 4988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9904013501905227,
+      "learning_rate": 8.033595943067786e-06,
+      "loss": 0.3083,
+      "step": 4989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.056318935392141,
+      "learning_rate": 8.032786276619658e-06,
+      "loss": 0.349,
+      "step": 4990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7745935112318016,
+      "learning_rate": 8.031976484335841e-06,
+      "loss": 0.3216,
+      "step": 4991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2124179343633745,
+      "learning_rate": 8.031166566249935e-06,
+      "loss": 0.3111,
+      "step": 4992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.971833357269287,
+      "learning_rate": 8.030356522395544e-06,
+      "loss": 0.3174,
+      "step": 4993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.053995436223221,
+      "learning_rate": 8.029546352806277e-06,
+      "loss": 0.3173,
+      "step": 4994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.351064506842642,
+      "learning_rate": 8.028736057515753e-06,
+      "loss": 0.3318,
+      "step": 4995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1149409158563297,
+      "learning_rate": 8.02792563655759e-06,
+      "loss": 0.3247,
+      "step": 4996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9858625097368985,
+      "learning_rate": 8.027115089965413e-06,
+      "loss": 0.3302,
+      "step": 4997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.136126573258584,
+      "learning_rate": 8.026304417772854e-06,
+      "loss": 0.326,
+      "step": 4998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0309427426539384,
+      "learning_rate": 8.025493620013549e-06,
+      "loss": 0.3437,
+      "step": 4999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.14193524045625,
+      "learning_rate": 8.02468269672114e-06,
+      "loss": 0.3162,
+      "step": 5000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.995483997129608,
+      "learning_rate": 8.023871647929276e-06,
+      "loss": 0.3087,
+      "step": 5001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3277628667057098,
+      "learning_rate": 8.023060473671605e-06,
+      "loss": 0.3102,
+      "step": 5002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9788008089983757,
+      "learning_rate": 8.022249173981784e-06,
+      "loss": 0.3104,
+      "step": 5003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6996974523522912,
+      "learning_rate": 8.021437748893478e-06,
+      "loss": 0.4972,
+      "step": 5004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2174180211368424,
+      "learning_rate": 8.020626198440352e-06,
+      "loss": 0.3132,
+      "step": 5005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.640476656829514,
+      "learning_rate": 8.01981452265608e-06,
+      "loss": 0.3143,
+      "step": 5006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.334402462138009,
+      "learning_rate": 8.01900272157434e-06,
+      "loss": 0.3069,
+      "step": 5007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.900448913831037,
+      "learning_rate": 8.018190795228815e-06,
+      "loss": 0.3441,
+      "step": 5008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.063654509610975,
+      "learning_rate": 8.017378743653193e-06,
+      "loss": 0.3195,
+      "step": 5009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3909167925522223,
+      "learning_rate": 8.016566566881166e-06,
+      "loss": 0.3042,
+      "step": 5010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.532290570296978,
+      "learning_rate": 8.015754264946435e-06,
+      "loss": 0.3315,
+      "step": 5011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7373760127048294,
+      "learning_rate": 8.014941837882703e-06,
+      "loss": 0.3432,
+      "step": 5012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0618836638509697,
+      "learning_rate": 8.01412928572368e-06,
+      "loss": 0.3232,
+      "step": 5013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.876704991000476,
+      "learning_rate": 8.01331660850308e-06,
+      "loss": 0.3114,
+      "step": 5014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.145002344501987,
+      "learning_rate": 8.01250380625462e-06,
+      "loss": 0.302,
+      "step": 5015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.40383478343123,
+      "learning_rate": 8.011690879012027e-06,
+      "loss": 0.3279,
+      "step": 5016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8666662275168235,
+      "learning_rate": 8.010877826809028e-06,
+      "loss": 0.3141,
+      "step": 5017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.595550835367412,
+      "learning_rate": 8.01006464967936e-06,
+      "loss": 0.3118,
+      "step": 5018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.383487226000707,
+      "learning_rate": 8.009251347656766e-06,
+      "loss": 0.3132,
+      "step": 5019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.675162959329228,
+      "learning_rate": 8.008437920774987e-06,
+      "loss": 0.3088,
+      "step": 5020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6265349900072492,
+      "learning_rate": 8.007624369067776e-06,
+      "loss": 0.2884,
+      "step": 5021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.6831646227136945,
+      "learning_rate": 8.006810692568886e-06,
+      "loss": 0.3141,
+      "step": 5022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.556303551718319,
+      "learning_rate": 8.005996891312082e-06,
+      "loss": 0.3039,
+      "step": 5023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3412784436245517,
+      "learning_rate": 8.005182965331127e-06,
+      "loss": 0.3417,
+      "step": 5024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.284655036838846,
+      "learning_rate": 8.004368914659792e-06,
+      "loss": 0.3232,
+      "step": 5025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.063304621855258,
+      "learning_rate": 8.003554739331856e-06,
+      "loss": 0.3083,
+      "step": 5026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4726216877214298,
+      "learning_rate": 8.002740439381097e-06,
+      "loss": 0.3258,
+      "step": 5027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2303048752751287,
+      "learning_rate": 8.001926014841306e-06,
+      "loss": 0.3162,
+      "step": 5028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.301325872784528,
+      "learning_rate": 8.001111465746269e-06,
+      "loss": 0.3141,
+      "step": 5029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9678152985026538,
+      "learning_rate": 8.00029679212979e-06,
+      "loss": 0.3597,
+      "step": 5030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.746665528896893,
+      "learning_rate": 7.999481994025666e-06,
+      "loss": 0.3292,
+      "step": 5031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8384568544072035,
+      "learning_rate": 7.998667071467709e-06,
+      "loss": 0.3117,
+      "step": 5032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.151200798535191,
+      "learning_rate": 7.997852024489727e-06,
+      "loss": 0.3197,
+      "step": 5033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4186715153542795,
+      "learning_rate": 7.99703685312554e-06,
+      "loss": 0.2909,
+      "step": 5034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.76518638238281,
+      "learning_rate": 7.99622155740897e-06,
+      "loss": 0.3145,
+      "step": 5035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2394895644670703,
+      "learning_rate": 7.995406137373848e-06,
+      "loss": 0.3109,
+      "step": 5036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5883403986052236,
+      "learning_rate": 7.994590593054001e-06,
+      "loss": 0.3132,
+      "step": 5037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0702547423913122,
+      "learning_rate": 7.993774924483275e-06,
+      "loss": 0.3181,
+      "step": 5038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0816594562824275,
+      "learning_rate": 7.992959131695508e-06,
+      "loss": 0.3235,
+      "step": 5039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.63469659261388,
+      "learning_rate": 7.99214321472455e-06,
+      "loss": 0.3176,
+      "step": 5040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0387904745798937,
+      "learning_rate": 7.991327173604257e-06,
+      "loss": 0.3086,
+      "step": 5041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.001197198470353,
+      "learning_rate": 7.990511008368485e-06,
+      "loss": 0.314,
+      "step": 5042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.091889925430689,
+      "learning_rate": 7.989694719051101e-06,
+      "loss": 0.3303,
+      "step": 5043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.46003381239394,
+      "learning_rate": 7.988878305685972e-06,
+      "loss": 0.2925,
+      "step": 5044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.684788091771633,
+      "learning_rate": 7.988061768306973e-06,
+      "loss": 0.3333,
+      "step": 5045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.578808138336421,
+      "learning_rate": 7.987245106947984e-06,
+      "loss": 0.3042,
+      "step": 5046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.158788842711406,
+      "learning_rate": 7.98642832164289e-06,
+      "loss": 0.3235,
+      "step": 5047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.113235636521349,
+      "learning_rate": 7.98561141242558e-06,
+      "loss": 0.3354,
+      "step": 5048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.889934836684406,
+      "learning_rate": 7.98479437932995e-06,
+      "loss": 0.3025,
+      "step": 5049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9492874215275315,
+      "learning_rate": 7.983977222389898e-06,
+      "loss": 0.3066,
+      "step": 5050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1226520046678004,
+      "learning_rate": 7.983159941639334e-06,
+      "loss": 0.3231,
+      "step": 5051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3378657451530405,
+      "learning_rate": 7.982342537112162e-06,
+      "loss": 0.3139,
+      "step": 5052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6246577298481006,
+      "learning_rate": 7.981525008842302e-06,
+      "loss": 0.3338,
+      "step": 5053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.021490134537214,
+      "learning_rate": 7.980707356863673e-06,
+      "loss": 0.2936,
+      "step": 5054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7435643705464416,
+      "learning_rate": 7.979889581210203e-06,
+      "loss": 0.3131,
+      "step": 5055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5048846923382837,
+      "learning_rate": 7.979071681915821e-06,
+      "loss": 0.3009,
+      "step": 5056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3088242899896563,
+      "learning_rate": 7.978253659014463e-06,
+      "loss": 0.3153,
+      "step": 5057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.168719405448897,
+      "learning_rate": 7.977435512540072e-06,
+      "loss": 0.3066,
+      "step": 5058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7285131836876428,
+      "learning_rate": 7.976617242526592e-06,
+      "loss": 0.31,
+      "step": 5059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7048803919940461,
+      "learning_rate": 7.975798849007974e-06,
+      "loss": 0.5259,
+      "step": 5060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7197787822123591,
+      "learning_rate": 7.974980332018178e-06,
+      "loss": 0.3088,
+      "step": 5061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.799530498120901,
+      "learning_rate": 7.974161691591165e-06,
+      "loss": 0.3328,
+      "step": 5062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5861386311184247,
+      "learning_rate": 7.973342927760898e-06,
+      "loss": 0.3019,
+      "step": 5063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2818011562125737,
+      "learning_rate": 7.972524040561353e-06,
+      "loss": 0.3123,
+      "step": 5064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.008332525866915,
+      "learning_rate": 7.971705030026507e-06,
+      "loss": 0.3118,
+      "step": 5065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.256715118907958,
+      "learning_rate": 7.97088589619034e-06,
+      "loss": 0.2927,
+      "step": 5066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.674991938528282,
+      "learning_rate": 7.970066639086839e-06,
+      "loss": 0.301,
+      "step": 5067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1804556456397246,
+      "learning_rate": 7.969247258749999e-06,
+      "loss": 0.302,
+      "step": 5068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9860280400051207,
+      "learning_rate": 7.968427755213814e-06,
+      "loss": 0.3068,
+      "step": 5069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.831230541476977,
+      "learning_rate": 7.967608128512293e-06,
+      "loss": 0.3122,
+      "step": 5070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5164274142461585,
+      "learning_rate": 7.966788378679435e-06,
+      "loss": 0.3592,
+      "step": 5071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.837298725090378,
+      "learning_rate": 7.965968505749262e-06,
+      "loss": 0.291,
+      "step": 5072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5549773023611877,
+      "learning_rate": 7.965148509755785e-06,
+      "loss": 0.2925,
+      "step": 5073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2986968139774238,
+      "learning_rate": 7.96432839073303e-06,
+      "loss": 0.322,
+      "step": 5074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.409482418167394,
+      "learning_rate": 7.963508148715024e-06,
+      "loss": 0.3006,
+      "step": 5075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6148311753650015,
+      "learning_rate": 7.962687783735803e-06,
+      "loss": 0.2995,
+      "step": 5076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.440228470832698,
+      "learning_rate": 7.961867295829403e-06,
+      "loss": 0.31,
+      "step": 5077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.806841277108355,
+      "learning_rate": 7.961046685029868e-06,
+      "loss": 0.3409,
+      "step": 5078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4445096722768909,
+      "learning_rate": 7.960225951371245e-06,
+      "loss": 0.3029,
+      "step": 5079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4989431264792445,
+      "learning_rate": 7.959405094887591e-06,
+      "loss": 0.2938,
+      "step": 5080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6791499788006095,
+      "learning_rate": 7.958584115612963e-06,
+      "loss": 0.5283,
+      "step": 5081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0749415291052133,
+      "learning_rate": 7.957763013581425e-06,
+      "loss": 0.3239,
+      "step": 5082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0466423740691315,
+      "learning_rate": 7.956941788827046e-06,
+      "loss": 0.3171,
+      "step": 5083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 18.875177393278847,
+      "learning_rate": 7.9561204413839e-06,
+      "loss": 0.3002,
+      "step": 5084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7414609372383736,
+      "learning_rate": 7.955298971286066e-06,
+      "loss": 0.3201,
+      "step": 5085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8438288218830583,
+      "learning_rate": 7.954477378567629e-06,
+      "loss": 0.3316,
+      "step": 5086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.979266579438885,
+      "learning_rate": 7.953655663262676e-06,
+      "loss": 0.3338,
+      "step": 5087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.596344116499169,
+      "learning_rate": 7.952833825405306e-06,
+      "loss": 0.3172,
+      "step": 5088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.374949279742272,
+      "learning_rate": 7.952011865029614e-06,
+      "loss": 0.3357,
+      "step": 5089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.76099965096983,
+      "learning_rate": 7.951189782169706e-06,
+      "loss": 0.311,
+      "step": 5090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.791065430818122,
+      "learning_rate": 7.950367576859694e-06,
+      "loss": 0.3222,
+      "step": 5091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.064459872491958,
+      "learning_rate": 7.94954524913369e-06,
+      "loss": 0.3111,
+      "step": 5092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6358258563186658,
+      "learning_rate": 7.948722799025814e-06,
+      "loss": 0.5034,
+      "step": 5093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0212872046446795,
+      "learning_rate": 7.94790022657019e-06,
+      "loss": 0.3342,
+      "step": 5094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9751574512457168,
+      "learning_rate": 7.947077531800953e-06,
+      "loss": 0.3167,
+      "step": 5095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3474367355703467,
+      "learning_rate": 7.946254714752233e-06,
+      "loss": 0.3204,
+      "step": 5096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.2114763334420475,
+      "learning_rate": 7.945431775458172e-06,
+      "loss": 0.3321,
+      "step": 5097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5620017746977082,
+      "learning_rate": 7.944608713952913e-06,
+      "loss": 0.3223,
+      "step": 5098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5993428243811117,
+      "learning_rate": 7.94378553027061e-06,
+      "loss": 0.4901,
+      "step": 5099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7364528052657522,
+      "learning_rate": 7.942962224445416e-06,
+      "loss": 0.3038,
+      "step": 5100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5768580631474247,
+      "learning_rate": 7.942138796511493e-06,
+      "loss": 0.3406,
+      "step": 5101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7684532732243523,
+      "learning_rate": 7.941315246503006e-06,
+      "loss": 0.281,
+      "step": 5102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8710782619801423,
+      "learning_rate": 7.940491574454123e-06,
+      "loss": 0.3424,
+      "step": 5103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3112896392186466,
+      "learning_rate": 7.939667780399023e-06,
+      "loss": 0.309,
+      "step": 5104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7026879528276058,
+      "learning_rate": 7.938843864371886e-06,
+      "loss": 0.3511,
+      "step": 5105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9839296373220543,
+      "learning_rate": 7.938019826406895e-06,
+      "loss": 0.308,
+      "step": 5106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2861315904284585,
+      "learning_rate": 7.937195666538245e-06,
+      "loss": 0.2957,
+      "step": 5107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9380431598967383,
+      "learning_rate": 7.936371384800127e-06,
+      "loss": 0.3362,
+      "step": 5108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.770307461234666,
+      "learning_rate": 7.935546981226747e-06,
+      "loss": 0.3473,
+      "step": 5109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6061355860759877,
+      "learning_rate": 7.934722455852308e-06,
+      "loss": 0.4995,
+      "step": 5110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6253222845390156,
+      "learning_rate": 7.933897808711022e-06,
+      "loss": 0.2948,
+      "step": 5111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.468429389333703,
+      "learning_rate": 7.933073039837104e-06,
+      "loss": 0.3258,
+      "step": 5112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.1328610215687336,
+      "learning_rate": 7.932248149264778e-06,
+      "loss": 0.3124,
+      "step": 5113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1885687747250673,
+      "learning_rate": 7.931423137028265e-06,
+      "loss": 0.3138,
+      "step": 5114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6322587392956598,
+      "learning_rate": 7.9305980031618e-06,
+      "loss": 0.3143,
+      "step": 5115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.691617172288375,
+      "learning_rate": 7.929772747699621e-06,
+      "loss": 0.3284,
+      "step": 5116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4915008100558418,
+      "learning_rate": 7.928947370675963e-06,
+      "loss": 0.3271,
+      "step": 5117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.874462099288567,
+      "learning_rate": 7.928121872125079e-06,
+      "loss": 0.328,
+      "step": 5118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8486747584394894,
+      "learning_rate": 7.927296252081218e-06,
+      "loss": 0.3323,
+      "step": 5119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6815775240784545,
+      "learning_rate": 7.926470510578634e-06,
+      "loss": 0.3315,
+      "step": 5120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.587468779319142,
+      "learning_rate": 7.925644647651591e-06,
+      "loss": 0.2997,
+      "step": 5121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.234666949519142,
+      "learning_rate": 7.924818663334356e-06,
+      "loss": 0.3351,
+      "step": 5122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6422725811821304,
+      "learning_rate": 7.9239925576612e-06,
+      "loss": 0.3039,
+      "step": 5123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7219740756730872,
+      "learning_rate": 7.923166330666397e-06,
+      "loss": 0.3198,
+      "step": 5124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5385086878779366,
+      "learning_rate": 7.922339982384232e-06,
+      "loss": 0.3387,
+      "step": 5125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2084383566082457,
+      "learning_rate": 7.92151351284899e-06,
+      "loss": 0.3314,
+      "step": 5126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4743707311509031,
+      "learning_rate": 7.920686922094964e-06,
+      "loss": 0.3374,
+      "step": 5127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.915323215254238,
+      "learning_rate": 7.91986021015645e-06,
+      "loss": 0.3095,
+      "step": 5128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2837458177313508,
+      "learning_rate": 7.919033377067748e-06,
+      "loss": 0.3379,
+      "step": 5129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2373511139087423,
+      "learning_rate": 7.918206422863169e-06,
+      "loss": 0.3304,
+      "step": 5130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5184108331320194,
+      "learning_rate": 7.91737934757702e-06,
+      "loss": 0.3183,
+      "step": 5131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.169044407235978,
+      "learning_rate": 7.91655215124362e-06,
+      "loss": 0.3342,
+      "step": 5132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5965243231458044,
+      "learning_rate": 7.915724833897291e-06,
+      "loss": 0.318,
+      "step": 5133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.310497701992734,
+      "learning_rate": 7.914897395572362e-06,
+      "loss": 0.3191,
+      "step": 5134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5435031686691405,
+      "learning_rate": 7.91406983630316e-06,
+      "loss": 0.3314,
+      "step": 5135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4206004188264725,
+      "learning_rate": 7.913242156124026e-06,
+      "loss": 0.3411,
+      "step": 5136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0573240464480085,
+      "learning_rate": 7.9124143550693e-06,
+      "loss": 0.3216,
+      "step": 5137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2794630248249228,
+      "learning_rate": 7.911586433173328e-06,
+      "loss": 0.305,
+      "step": 5138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3966530422442425,
+      "learning_rate": 7.910758390470465e-06,
+      "loss": 0.3013,
+      "step": 5139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.358532216350067,
+      "learning_rate": 7.909930226995066e-06,
+      "loss": 0.3184,
+      "step": 5140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2842048344007564,
+      "learning_rate": 7.909101942781494e-06,
+      "loss": 0.3089,
+      "step": 5141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3209506365425492,
+      "learning_rate": 7.908273537864114e-06,
+      "loss": 0.3221,
+      "step": 5142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.289693722947336,
+      "learning_rate": 7.9074450122773e-06,
+      "loss": 0.314,
+      "step": 5143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7378343070889373,
+      "learning_rate": 7.906616366055427e-06,
+      "loss": 0.3512,
+      "step": 5144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4920375500741807,
+      "learning_rate": 7.905787599232878e-06,
+      "loss": 0.3159,
+      "step": 5145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.78577601822971,
+      "learning_rate": 7.904958711844042e-06,
+      "loss": 0.2949,
+      "step": 5146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6478832983625598,
+      "learning_rate": 7.90412970392331e-06,
+      "loss": 0.3033,
+      "step": 5147
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7544514373362592,
+      "learning_rate": 7.903300575505077e-06,
+      "loss": 0.357,
+      "step": 5148
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6290573379952231,
+      "learning_rate": 7.902471326623746e-06,
+      "loss": 0.3407,
+      "step": 5149
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0217958670928295,
+      "learning_rate": 7.901641957313724e-06,
+      "loss": 0.3255,
+      "step": 5150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3195554411977217,
+      "learning_rate": 7.900812467609423e-06,
+      "loss": 0.3208,
+      "step": 5151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.635799297520632,
+      "learning_rate": 7.899982857545263e-06,
+      "loss": 0.3306,
+      "step": 5152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8028536636867845,
+      "learning_rate": 7.899153127155661e-06,
+      "loss": 0.3093,
+      "step": 5153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.760933628232171,
+      "learning_rate": 7.898323276475045e-06,
+      "loss": 0.3177,
+      "step": 5154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3091170778311194,
+      "learning_rate": 7.897493305537851e-06,
+      "loss": 0.3024,
+      "step": 5155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4170006007915341,
+      "learning_rate": 7.896663214378512e-06,
+      "loss": 0.3335,
+      "step": 5156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5269484964884819,
+      "learning_rate": 7.895833003031472e-06,
+      "loss": 0.3463,
+      "step": 5157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5275599386026182,
+      "learning_rate": 7.895002671531175e-06,
+      "loss": 0.3159,
+      "step": 5158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3944402239154978,
+      "learning_rate": 7.894172219912077e-06,
+      "loss": 0.3123,
+      "step": 5159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3081950411158063,
+      "learning_rate": 7.893341648208632e-06,
+      "loss": 0.3119,
+      "step": 5160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1249826879700624,
+      "learning_rate": 7.892510956455305e-06,
+      "loss": 0.3248,
+      "step": 5161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.247131020693949,
+      "learning_rate": 7.891680144686558e-06,
+      "loss": 0.3134,
+      "step": 5162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6270633376295832,
+      "learning_rate": 7.890849212936866e-06,
+      "loss": 0.4912,
+      "step": 5163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3888517215084442,
+      "learning_rate": 7.890018161240707e-06,
+      "loss": 0.3212,
+      "step": 5164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4951206362262988,
+      "learning_rate": 7.889186989632559e-06,
+      "loss": 0.32,
+      "step": 5165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4390563164043506,
+      "learning_rate": 7.888355698146913e-06,
+      "loss": 0.3042,
+      "step": 5166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4237042466124175,
+      "learning_rate": 7.887524286818257e-06,
+      "loss": 0.3224,
+      "step": 5167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0496892155503756,
+      "learning_rate": 7.886692755681091e-06,
+      "loss": 0.323,
+      "step": 5168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3230823370618103,
+      "learning_rate": 7.885861104769915e-06,
+      "loss": 0.326,
+      "step": 5169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3049544437398652,
+      "learning_rate": 7.885029334119237e-06,
+      "loss": 0.3466,
+      "step": 5170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1769160680954185,
+      "learning_rate": 7.884197443763566e-06,
+      "loss": 0.3108,
+      "step": 5171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.48672535677508,
+      "learning_rate": 7.88336543373742e-06,
+      "loss": 0.3526,
+      "step": 5172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1648446508690258,
+      "learning_rate": 7.88253330407532e-06,
+      "loss": 0.3276,
+      "step": 5173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.340396240680619,
+      "learning_rate": 7.881701054811796e-06,
+      "loss": 0.3056,
+      "step": 5174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.728428883608017,
+      "learning_rate": 7.880868685981376e-06,
+      "loss": 0.3191,
+      "step": 5175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4732255951886062,
+      "learning_rate": 7.880036197618596e-06,
+      "loss": 0.3213,
+      "step": 5176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.623338077004673,
+      "learning_rate": 7.879203589757998e-06,
+      "loss": 0.3124,
+      "step": 5177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1853307873795726,
+      "learning_rate": 7.878370862434131e-06,
+      "loss": 0.3251,
+      "step": 5178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9052155557771617,
+      "learning_rate": 7.877538015681542e-06,
+      "loss": 0.3197,
+      "step": 5179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.389894072532732,
+      "learning_rate": 7.876705049534792e-06,
+      "loss": 0.317,
+      "step": 5180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5412997149227012,
+      "learning_rate": 7.875871964028439e-06,
+      "loss": 0.3191,
+      "step": 5181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.485679210130768,
+      "learning_rate": 7.87503875919705e-06,
+      "loss": 0.3386,
+      "step": 5182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6576984661492888,
+      "learning_rate": 7.874205435075196e-06,
+      "loss": 0.3207,
+      "step": 5183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8776673260890777,
+      "learning_rate": 7.873371991697454e-06,
+      "loss": 0.3164,
+      "step": 5184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8228351970614167,
+      "learning_rate": 7.872538429098404e-06,
+      "loss": 0.3204,
+      "step": 5185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.01354442212947,
+      "learning_rate": 7.871704747312631e-06,
+      "loss": 0.3079,
+      "step": 5186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6381023841770297,
+      "learning_rate": 7.87087094637473e-06,
+      "loss": 0.3119,
+      "step": 5187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.230613493458902,
+      "learning_rate": 7.870037026319293e-06,
+      "loss": 0.3333,
+      "step": 5188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.271359396118646,
+      "learning_rate": 7.869202987180921e-06,
+      "loss": 0.3418,
+      "step": 5189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.813108543229716,
+      "learning_rate": 7.868368828994222e-06,
+      "loss": 0.3196,
+      "step": 5190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.8761358812895885,
+      "learning_rate": 7.867534551793805e-06,
+      "loss": 0.3143,
+      "step": 5191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.808748874398439,
+      "learning_rate": 7.866700155614286e-06,
+      "loss": 0.3301,
+      "step": 5192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6232777439998736,
+      "learning_rate": 7.865865640490287e-06,
+      "loss": 0.3238,
+      "step": 5193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1706212707037715,
+      "learning_rate": 7.865031006456432e-06,
+      "loss": 0.3351,
+      "step": 5194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 53.29319560873858,
+      "learning_rate": 7.86419625354735e-06,
+      "loss": 0.3162,
+      "step": 5195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7152615452368334,
+      "learning_rate": 7.863361381797678e-06,
+      "loss": 0.3194,
+      "step": 5196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.723064779092635,
+      "learning_rate": 7.862526391242058e-06,
+      "loss": 0.3283,
+      "step": 5197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5969489563372483,
+      "learning_rate": 7.861691281915133e-06,
+      "loss": 0.3139,
+      "step": 5198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4729317597380558,
+      "learning_rate": 7.860856053851554e-06,
+      "loss": 0.3064,
+      "step": 5199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7380473371823648,
+      "learning_rate": 7.860020707085976e-06,
+      "loss": 0.3284,
+      "step": 5200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 32.738389296582035,
+      "learning_rate": 7.859185241653058e-06,
+      "loss": 0.3313,
+      "step": 5201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8081041996036473,
+      "learning_rate": 7.858349657587465e-06,
+      "loss": 0.312,
+      "step": 5202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7493535284366597,
+      "learning_rate": 7.85751395492387e-06,
+      "loss": 0.3087,
+      "step": 5203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7270254734427284,
+      "learning_rate": 7.856678133696945e-06,
+      "loss": 0.3228,
+      "step": 5204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.577619806778296,
+      "learning_rate": 7.85584219394137e-06,
+      "loss": 0.3132,
+      "step": 5205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7439505639616169,
+      "learning_rate": 7.855006135691829e-06,
+      "loss": 0.3037,
+      "step": 5206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7274519998258857,
+      "learning_rate": 7.854169958983014e-06,
+      "loss": 0.326,
+      "step": 5207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.939987603120739,
+      "learning_rate": 7.853333663849616e-06,
+      "loss": 0.3271,
+      "step": 5208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.389205281536424,
+      "learning_rate": 7.852497250326337e-06,
+      "loss": 0.3267,
+      "step": 5209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7981692340837032,
+      "learning_rate": 7.85166071844788e-06,
+      "loss": 0.3054,
+      "step": 5210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.775791113021543,
+      "learning_rate": 7.850824068248955e-06,
+      "loss": 0.3166,
+      "step": 5211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.622538023807633,
+      "learning_rate": 7.849987299764276e-06,
+      "loss": 0.3206,
+      "step": 5212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.576171702100024,
+      "learning_rate": 7.849150413028562e-06,
+      "loss": 0.3217,
+      "step": 5213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2095791500536244,
+      "learning_rate": 7.848313408076535e-06,
+      "loss": 0.3331,
+      "step": 5214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4865642336849416,
+      "learning_rate": 7.847476284942927e-06,
+      "loss": 0.3322,
+      "step": 5215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.634545356371704,
+      "learning_rate": 7.846639043662472e-06,
+      "loss": 0.3243,
+      "step": 5216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2618720493918716,
+      "learning_rate": 7.845801684269905e-06,
+      "loss": 0.3018,
+      "step": 5217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.498254707575457,
+      "learning_rate": 7.84496420679997e-06,
+      "loss": 0.343,
+      "step": 5218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0083962241665834,
+      "learning_rate": 7.84412661128742e-06,
+      "loss": 0.3385,
+      "step": 5219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.171225818129115,
+      "learning_rate": 7.843288897767003e-06,
+      "loss": 0.3316,
+      "step": 5220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.813160418605828,
+      "learning_rate": 7.842451066273478e-06,
+      "loss": 0.3266,
+      "step": 5221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4333856933473443,
+      "learning_rate": 7.841613116841612e-06,
+      "loss": 0.3052,
+      "step": 5222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5155183948816697,
+      "learning_rate": 7.84077504950617e-06,
+      "loss": 0.3201,
+      "step": 5223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9748036782521914,
+      "learning_rate": 7.839936864301924e-06,
+      "loss": 0.3265,
+      "step": 5224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9203394613494438,
+      "learning_rate": 7.839098561263655e-06,
+      "loss": 0.3036,
+      "step": 5225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8496166679425865,
+      "learning_rate": 7.838260140426145e-06,
+      "loss": 0.325,
+      "step": 5226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.574634293112357,
+      "learning_rate": 7.837421601824178e-06,
+      "loss": 0.3214,
+      "step": 5227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.281911113871297,
+      "learning_rate": 7.83658294549255e-06,
+      "loss": 0.3317,
+      "step": 5228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.792772421014619,
+      "learning_rate": 7.835744171466056e-06,
+      "loss": 0.3063,
+      "step": 5229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6417754999989007,
+      "learning_rate": 7.834905279779501e-06,
+      "loss": 0.3132,
+      "step": 5230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.9212116571629325,
+      "learning_rate": 7.83406627046769e-06,
+      "loss": 0.3224,
+      "step": 5231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6016643983710616,
+      "learning_rate": 7.833227143565436e-06,
+      "loss": 0.5195,
+      "step": 5232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9190803509802752,
+      "learning_rate": 7.832387899107555e-06,
+      "loss": 0.3076,
+      "step": 5233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6528497273908247,
+      "learning_rate": 7.831548537128868e-06,
+      "loss": 0.327,
+      "step": 5234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.203993813077144,
+      "learning_rate": 7.830709057664205e-06,
+      "loss": 0.3281,
+      "step": 5235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.885793144761983,
+      "learning_rate": 7.829869460748394e-06,
+      "loss": 0.3215,
+      "step": 5236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5274124736436414,
+      "learning_rate": 7.829029746416272e-06,
+      "loss": 0.3208,
+      "step": 5237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.570056472955061,
+      "learning_rate": 7.828189914702681e-06,
+      "loss": 0.4846,
+      "step": 5238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3544901211877676,
+      "learning_rate": 7.82734996564247e-06,
+      "loss": 0.3348,
+      "step": 5239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4647389721696236,
+      "learning_rate": 7.826509899270484e-06,
+      "loss": 0.3369,
+      "step": 5240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.73010096681408,
+      "learning_rate": 7.825669715621581e-06,
+      "loss": 0.3229,
+      "step": 5241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9056460904486618,
+      "learning_rate": 7.824829414730625e-06,
+      "loss": 0.3263,
+      "step": 5242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3501727826015766,
+      "learning_rate": 7.82398899663248e-06,
+      "loss": 0.3186,
+      "step": 5243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.88366306271441,
+      "learning_rate": 7.823148461362013e-06,
+      "loss": 0.3225,
+      "step": 5244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6255590264741315,
+      "learning_rate": 7.822307808954102e-06,
+      "loss": 0.3296,
+      "step": 5245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0484755965364156,
+      "learning_rate": 7.821467039443628e-06,
+      "loss": 0.3136,
+      "step": 5246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.480892802698961,
+      "learning_rate": 7.820626152865476e-06,
+      "loss": 0.3027,
+      "step": 5247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1877077552477995,
+      "learning_rate": 7.819785149254534e-06,
+      "loss": 0.3071,
+      "step": 5248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.65753819552042,
+      "learning_rate": 7.818944028645696e-06,
+      "loss": 0.3295,
+      "step": 5249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7443418408508877,
+      "learning_rate": 7.818102791073866e-06,
+      "loss": 0.3104,
+      "step": 5250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3554998717394255,
+      "learning_rate": 7.817261436573944e-06,
+      "loss": 0.3153,
+      "step": 5251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6195504060972558,
+      "learning_rate": 7.816419965180842e-06,
+      "loss": 0.4923,
+      "step": 5252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6823505897589057,
+      "learning_rate": 7.815578376929472e-06,
+      "loss": 0.3216,
+      "step": 5253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8260645558341941,
+      "learning_rate": 7.814736671854755e-06,
+      "loss": 0.3281,
+      "step": 5254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.262671085875684,
+      "learning_rate": 7.813894849991613e-06,
+      "loss": 0.3362,
+      "step": 5255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.189242209242595,
+      "learning_rate": 7.813052911374976e-06,
+      "loss": 0.3016,
+      "step": 5256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.031791925024394,
+      "learning_rate": 7.812210856039777e-06,
+      "loss": 0.3191,
+      "step": 5257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8736459076904737,
+      "learning_rate": 7.811368684020955e-06,
+      "loss": 0.3345,
+      "step": 5258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.377015528104474,
+      "learning_rate": 7.810526395353453e-06,
+      "loss": 0.3238,
+      "step": 5259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.798601382094193,
+      "learning_rate": 7.809683990072217e-06,
+      "loss": 0.2939,
+      "step": 5260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.363927610327008,
+      "learning_rate": 7.808841468212204e-06,
+      "loss": 0.3274,
+      "step": 5261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6227323400475644,
+      "learning_rate": 7.807998829808368e-06,
+      "loss": 0.3111,
+      "step": 5262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5637446224382276,
+      "learning_rate": 7.807156074895673e-06,
+      "loss": 0.328,
+      "step": 5263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.846290688888176,
+      "learning_rate": 7.806313203509087e-06,
+      "loss": 0.3353,
+      "step": 5264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.078383612207726,
+      "learning_rate": 7.805470215683582e-06,
+      "loss": 0.3266,
+      "step": 5265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.979762433478405,
+      "learning_rate": 7.804627111454132e-06,
+      "loss": 0.3076,
+      "step": 5266
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.660159241950354,
+      "learning_rate": 7.803783890855725e-06,
+      "loss": 0.3431,
+      "step": 5267
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6204503421514473,
+      "learning_rate": 7.802940553923344e-06,
+      "loss": 0.2973,
+      "step": 5268
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6614170714011247,
+      "learning_rate": 7.80209710069198e-06,
+      "loss": 0.3181,
+      "step": 5269
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.763459680820304,
+      "learning_rate": 7.801253531196629e-06,
+      "loss": 0.3141,
+      "step": 5270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4819628461307792,
+      "learning_rate": 7.800409845472294e-06,
+      "loss": 0.3238,
+      "step": 5271
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.747752388557921,
+      "learning_rate": 7.799566043553983e-06,
+      "loss": 0.3047,
+      "step": 5272
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1211767852187933,
+      "learning_rate": 7.798722125476702e-06,
+      "loss": 0.3111,
+      "step": 5273
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5505624779238907,
+      "learning_rate": 7.797878091275468e-06,
+      "loss": 0.3108,
+      "step": 5274
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5776271307834984,
+      "learning_rate": 7.797033940985303e-06,
+      "loss": 0.3152,
+      "step": 5275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7226888476530333,
+      "learning_rate": 7.796189674641232e-06,
+      "loss": 0.3272,
+      "step": 5276
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.308974819580043,
+      "learning_rate": 7.795345292278284e-06,
+      "loss": 0.299,
+      "step": 5277
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3214760572985496,
+      "learning_rate": 7.794500793931496e-06,
+      "loss": 0.3226,
+      "step": 5278
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.811168133254081,
+      "learning_rate": 7.793656179635907e-06,
+      "loss": 0.3115,
+      "step": 5279
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4014193914832929,
+      "learning_rate": 7.79281144942656e-06,
+      "loss": 0.3258,
+      "step": 5280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3525619311131196,
+      "learning_rate": 7.791966603338505e-06,
+      "loss": 0.2961,
+      "step": 5281
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.386191233696376,
+      "learning_rate": 7.791121641406798e-06,
+      "loss": 0.3316,
+      "step": 5282
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7465533301761018,
+      "learning_rate": 7.790276563666496e-06,
+      "loss": 0.3113,
+      "step": 5283
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.848497159369416,
+      "learning_rate": 7.789431370152663e-06,
+      "loss": 0.3033,
+      "step": 5284
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1902861996520175,
+      "learning_rate": 7.788586060900366e-06,
+      "loss": 0.3257,
+      "step": 5285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6632439520845537,
+      "learning_rate": 7.787740635944682e-06,
+      "loss": 0.3112,
+      "step": 5286
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.109090444914576,
+      "learning_rate": 7.78689509532069e-06,
+      "loss": 0.2934,
+      "step": 5287
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5185527042333267,
+      "learning_rate": 7.786049439063466e-06,
+      "loss": 0.2903,
+      "step": 5288
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3430976374311951,
+      "learning_rate": 7.785203667208104e-06,
+      "loss": 0.3141,
+      "step": 5289
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.615437410967679,
+      "learning_rate": 7.784357779789695e-06,
+      "loss": 0.3289,
+      "step": 5290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.73339053657673,
+      "learning_rate": 7.783511776843336e-06,
+      "loss": 0.3611,
+      "step": 5291
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.078945198035065,
+      "learning_rate": 7.78266565840413e-06,
+      "loss": 0.325,
+      "step": 5292
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8732262119614633,
+      "learning_rate": 7.781819424507183e-06,
+      "loss": 0.3265,
+      "step": 5293
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7033490086311325,
+      "learning_rate": 7.780973075187606e-06,
+      "loss": 0.3422,
+      "step": 5294
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8274348365490694,
+      "learning_rate": 7.780126610480516e-06,
+      "loss": 0.3221,
+      "step": 5295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.921197079915513,
+      "learning_rate": 7.779280030421036e-06,
+      "loss": 0.3124,
+      "step": 5296
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3620246030895387,
+      "learning_rate": 7.77843333504429e-06,
+      "loss": 0.3154,
+      "step": 5297
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.135086271585444,
+      "learning_rate": 7.777586524385412e-06,
+      "loss": 0.324,
+      "step": 5298
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0187237282935357,
+      "learning_rate": 7.776739598479534e-06,
+      "loss": 0.3084,
+      "step": 5299
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.132250643424089,
+      "learning_rate": 7.775892557361799e-06,
+      "loss": 0.3285,
+      "step": 5300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2241173953747528,
+      "learning_rate": 7.77504540106735e-06,
+      "loss": 0.329,
+      "step": 5301
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4780742936978613,
+      "learning_rate": 7.774198129631339e-06,
+      "loss": 0.32,
+      "step": 5302
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6916449616547573,
+      "learning_rate": 7.77335074308892e-06,
+      "loss": 0.3068,
+      "step": 5303
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1138490450859275,
+      "learning_rate": 7.772503241475253e-06,
+      "loss": 0.3271,
+      "step": 5304
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5593199988123474,
+      "learning_rate": 7.771655624825501e-06,
+      "loss": 0.3124,
+      "step": 5305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1304137145738844,
+      "learning_rate": 7.770807893174836e-06,
+      "loss": 0.3112,
+      "step": 5306
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.545415392976424,
+      "learning_rate": 7.769960046558427e-06,
+      "loss": 0.3149,
+      "step": 5307
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6742429253967805,
+      "learning_rate": 7.769112085011458e-06,
+      "loss": 0.3154,
+      "step": 5308
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8038429385142214,
+      "learning_rate": 7.76826400856911e-06,
+      "loss": 0.3309,
+      "step": 5309
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 27.067237088615375,
+      "learning_rate": 7.76741581726657e-06,
+      "loss": 0.3173,
+      "step": 5310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9502668050879342,
+      "learning_rate": 7.766567511139034e-06,
+      "loss": 0.305,
+      "step": 5311
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4164528935586955,
+      "learning_rate": 7.765719090221697e-06,
+      "loss": 0.3214,
+      "step": 5312
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6561867214571613,
+      "learning_rate": 7.764870554549762e-06,
+      "loss": 0.3193,
+      "step": 5313
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.10729431809092,
+      "learning_rate": 7.764021904158436e-06,
+      "loss": 0.3334,
+      "step": 5314
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4436624384800332,
+      "learning_rate": 7.763173139082934e-06,
+      "loss": 0.3215,
+      "step": 5315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4107756570196157,
+      "learning_rate": 7.762324259358469e-06,
+      "loss": 0.3056,
+      "step": 5316
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.486026851265413,
+      "learning_rate": 7.761475265020265e-06,
+      "loss": 0.3283,
+      "step": 5317
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8007758072365392,
+      "learning_rate": 7.760626156103545e-06,
+      "loss": 0.3022,
+      "step": 5318
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3263991040429182,
+      "learning_rate": 7.759776932643546e-06,
+      "loss": 0.3285,
+      "step": 5319
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.808546364182805,
+      "learning_rate": 7.758927594675498e-06,
+      "loss": 0.305,
+      "step": 5320
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7624843777258097,
+      "learning_rate": 7.758078142234644e-06,
+      "loss": 0.3198,
+      "step": 5321
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3564778264776514,
+      "learning_rate": 7.757228575356227e-06,
+      "loss": 0.3052,
+      "step": 5322
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4638086532198022,
+      "learning_rate": 7.756378894075503e-06,
+      "loss": 0.3362,
+      "step": 5323
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2779272416213177,
+      "learning_rate": 7.75552909842772e-06,
+      "loss": 0.2917,
+      "step": 5324
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3558217933412742,
+      "learning_rate": 7.754679188448142e-06,
+      "loss": 0.3017,
+      "step": 5325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.537657372578936,
+      "learning_rate": 7.75382916417203e-06,
+      "loss": 0.326,
+      "step": 5326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2663583805097514,
+      "learning_rate": 7.752979025634657e-06,
+      "loss": 0.3155,
+      "step": 5327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2802139646917126,
+      "learning_rate": 7.752128772871292e-06,
+      "loss": 0.3323,
+      "step": 5328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.355152512161311,
+      "learning_rate": 7.751278405917216e-06,
+      "loss": 0.3211,
+      "step": 5329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.245421577558753,
+      "learning_rate": 7.750427924807712e-06,
+      "loss": 0.3127,
+      "step": 5330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4510547679293457,
+      "learning_rate": 7.74957732957807e-06,
+      "loss": 0.3106,
+      "step": 5331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0675998325860987,
+      "learning_rate": 7.748726620263581e-06,
+      "loss": 0.3312,
+      "step": 5332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.368644564235117,
+      "learning_rate": 7.74787579689954e-06,
+      "loss": 0.3126,
+      "step": 5333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8347779240013087,
+      "learning_rate": 7.747024859521253e-06,
+      "loss": 0.3666,
+      "step": 5334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.782028464072049,
+      "learning_rate": 7.746173808164023e-06,
+      "loss": 0.3212,
+      "step": 5335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6486331249127584,
+      "learning_rate": 7.745322642863167e-06,
+      "loss": 0.3083,
+      "step": 5336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7510337222465016,
+      "learning_rate": 7.744471363653996e-06,
+      "loss": 0.3236,
+      "step": 5337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.808508462901859,
+      "learning_rate": 7.743619970571834e-06,
+      "loss": 0.3005,
+      "step": 5338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.957146390700833,
+      "learning_rate": 7.742768463652007e-06,
+      "loss": 0.3306,
+      "step": 5339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8095262232863547,
+      "learning_rate": 7.741916842929843e-06,
+      "loss": 0.3229,
+      "step": 5340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8734776294503417,
+      "learning_rate": 7.741065108440682e-06,
+      "loss": 0.3192,
+      "step": 5341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9952200392603823,
+      "learning_rate": 7.74021326021986e-06,
+      "loss": 0.3217,
+      "step": 5342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8703840335891198,
+      "learning_rate": 7.739361298302721e-06,
+      "loss": 0.3231,
+      "step": 5343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.960761681104746,
+      "learning_rate": 7.738509222724617e-06,
+      "loss": 0.3142,
+      "step": 5344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.620839815612159,
+      "learning_rate": 7.7376570335209e-06,
+      "loss": 0.3136,
+      "step": 5345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8388352318752739,
+      "learning_rate": 7.736804730726932e-06,
+      "loss": 0.3101,
+      "step": 5346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5383323686602937,
+      "learning_rate": 7.735952314378075e-06,
+      "loss": 0.3097,
+      "step": 5347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.649249721364612,
+      "learning_rate": 7.735099784509694e-06,
+      "loss": 0.329,
+      "step": 5348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7810964237810616,
+      "learning_rate": 7.734247141157168e-06,
+      "loss": 0.3559,
+      "step": 5349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3182104693968337,
+      "learning_rate": 7.733394384355869e-06,
+      "loss": 0.3044,
+      "step": 5350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5859054146228193,
+      "learning_rate": 7.732541514141182e-06,
+      "loss": 0.297,
+      "step": 5351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.34680833618101,
+      "learning_rate": 7.731688530548495e-06,
+      "loss": 0.2968,
+      "step": 5352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4483123754731415,
+      "learning_rate": 7.7308354336132e-06,
+      "loss": 0.3031,
+      "step": 5353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0288664027543004,
+      "learning_rate": 7.72998222337069e-06,
+      "loss": 0.3147,
+      "step": 5354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8317005133311832,
+      "learning_rate": 7.72912889985637e-06,
+      "loss": 0.306,
+      "step": 5355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4893061078897964,
+      "learning_rate": 7.728275463105645e-06,
+      "loss": 0.304,
+      "step": 5356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4565648018620156,
+      "learning_rate": 7.727421913153925e-06,
+      "loss": 0.3262,
+      "step": 5357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.115236869711557,
+      "learning_rate": 7.726568250036625e-06,
+      "loss": 0.3133,
+      "step": 5358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6771999997855551,
+      "learning_rate": 7.725714473789166e-06,
+      "loss": 0.336,
+      "step": 5359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5117515210362769,
+      "learning_rate": 7.724860584446969e-06,
+      "loss": 0.2993,
+      "step": 5360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5205162452653131,
+      "learning_rate": 7.724006582045468e-06,
+      "loss": 0.3185,
+      "step": 5361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3642739193931954,
+      "learning_rate": 7.723152466620098e-06,
+      "loss": 0.3335,
+      "step": 5362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9031473059944004,
+      "learning_rate": 7.722298238206294e-06,
+      "loss": 0.3343,
+      "step": 5363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.611505317540388,
+      "learning_rate": 7.721443896839499e-06,
+      "loss": 0.321,
+      "step": 5364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1382766240312923,
+      "learning_rate": 7.720589442555164e-06,
+      "loss": 0.3062,
+      "step": 5365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6297875963547095,
+      "learning_rate": 7.719734875388742e-06,
+      "loss": 0.3134,
+      "step": 5366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3373581391367027,
+      "learning_rate": 7.718880195375686e-06,
+      "loss": 0.3046,
+      "step": 5367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.488188711680331,
+      "learning_rate": 7.718025402551464e-06,
+      "loss": 0.3134,
+      "step": 5368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.349829263295176,
+      "learning_rate": 7.71717049695154e-06,
+      "loss": 0.3255,
+      "step": 5369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4576402250034397,
+      "learning_rate": 7.716315478611386e-06,
+      "loss": 0.3396,
+      "step": 5370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3413342808191373,
+      "learning_rate": 7.715460347566476e-06,
+      "loss": 0.3198,
+      "step": 5371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1504449548549363,
+      "learning_rate": 7.714605103852297e-06,
+      "loss": 0.335,
+      "step": 5372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3610529802936822,
+      "learning_rate": 7.713749747504327e-06,
+      "loss": 0.3352,
+      "step": 5373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7288968360147436,
+      "learning_rate": 7.712894278558063e-06,
+      "loss": 0.3354,
+      "step": 5374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5781156822755555,
+      "learning_rate": 7.712038697048993e-06,
+      "loss": 0.332,
+      "step": 5375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7727885194311908,
+      "learning_rate": 7.711183003012624e-06,
+      "loss": 0.3212,
+      "step": 5376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8413907164823824,
+      "learning_rate": 7.710327196484453e-06,
+      "loss": 0.3069,
+      "step": 5377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6852659524346563,
+      "learning_rate": 7.709471277499995e-06,
+      "loss": 0.317,
+      "step": 5378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.089573112753548,
+      "learning_rate": 7.708615246094759e-06,
+      "loss": 0.2965,
+      "step": 5379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.573012052285078,
+      "learning_rate": 7.707759102304267e-06,
+      "loss": 0.328,
+      "step": 5380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1998696507379023,
+      "learning_rate": 7.706902846164042e-06,
+      "loss": 0.312,
+      "step": 5381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3072517004354953,
+      "learning_rate": 7.706046477709607e-06,
+      "loss": 0.3219,
+      "step": 5382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.829271017570389,
+      "learning_rate": 7.705189996976497e-06,
+      "loss": 0.3117,
+      "step": 5383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7827362078826985,
+      "learning_rate": 7.704333404000252e-06,
+      "loss": 0.3062,
+      "step": 5384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.907364631886873,
+      "learning_rate": 7.703476698816408e-06,
+      "loss": 0.3182,
+      "step": 5385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.68795382812034,
+      "learning_rate": 7.702619881460515e-06,
+      "loss": 0.3042,
+      "step": 5386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6616793374426807,
+      "learning_rate": 7.701762951968121e-06,
+      "loss": 0.4993,
+      "step": 5387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.252617426188242,
+      "learning_rate": 7.700905910374786e-06,
+      "loss": 0.3106,
+      "step": 5388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.460615221403972,
+      "learning_rate": 7.700048756716066e-06,
+      "loss": 0.3081,
+      "step": 5389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.79653436274685,
+      "learning_rate": 7.699191491027527e-06,
+      "loss": 0.3311,
+      "step": 5390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7241779184814705,
+      "learning_rate": 7.69833411334474e-06,
+      "loss": 0.301,
+      "step": 5391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4955979687830763,
+      "learning_rate": 7.697476623703278e-06,
+      "loss": 0.3124,
+      "step": 5392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6196195767461121,
+      "learning_rate": 7.696619022138718e-06,
+      "loss": 0.3263,
+      "step": 5393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.594175641596578,
+      "learning_rate": 7.695761308686645e-06,
+      "loss": 0.318,
+      "step": 5394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.267398055036615,
+      "learning_rate": 7.694903483382648e-06,
+      "loss": 0.3158,
+      "step": 5395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.85803065073527,
+      "learning_rate": 7.694045546262318e-06,
+      "loss": 0.3264,
+      "step": 5396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8948503634483607,
+      "learning_rate": 7.693187497361254e-06,
+      "loss": 0.3358,
+      "step": 5397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.251156620044382,
+      "learning_rate": 7.692329336715056e-06,
+      "loss": 0.3063,
+      "step": 5398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7417624228202535,
+      "learning_rate": 7.691471064359333e-06,
+      "loss": 0.3052,
+      "step": 5399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1760982421251651,
+      "learning_rate": 7.690612680329694e-06,
+      "loss": 0.3022,
+      "step": 5400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.883751960610199,
+      "learning_rate": 7.689754184661757e-06,
+      "loss": 0.3308,
+      "step": 5401
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.4214720740442326,
+      "learning_rate": 7.68889557739114e-06,
+      "loss": 0.3177,
+      "step": 5402
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.165026328108929,
+      "learning_rate": 7.68803685855347e-06,
+      "loss": 0.3196,
+      "step": 5403
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4245136398666445,
+      "learning_rate": 7.687178028184376e-06,
+      "loss": 0.3203,
+      "step": 5404
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8519914167079905,
+      "learning_rate": 7.686319086319494e-06,
+      "loss": 0.3122,
+      "step": 5405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.359881063028313,
+      "learning_rate": 7.68546003299446e-06,
+      "loss": 0.32,
+      "step": 5406
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.026191006986111,
+      "learning_rate": 7.68460086824492e-06,
+      "loss": 0.3046,
+      "step": 5407
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.10177207340718,
+      "learning_rate": 7.683741592106521e-06,
+      "loss": 0.3249,
+      "step": 5408
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.159585352314094,
+      "learning_rate": 7.682882204614918e-06,
+      "loss": 0.3246,
+      "step": 5409
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4964786417229705,
+      "learning_rate": 7.682022705805765e-06,
+      "loss": 0.3102,
+      "step": 5410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.013546214111861,
+      "learning_rate": 7.681163095714727e-06,
+      "loss": 0.309,
+      "step": 5411
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3057323594457988,
+      "learning_rate": 7.68030337437747e-06,
+      "loss": 0.3052,
+      "step": 5412
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7469025350698477,
+      "learning_rate": 7.679443541829665e-06,
+      "loss": 0.3161,
+      "step": 5413
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5623748611092583,
+      "learning_rate": 7.67858359810699e-06,
+      "loss": 0.3116,
+      "step": 5414
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.571531505164328,
+      "learning_rate": 7.677723543245122e-06,
+      "loss": 0.3091,
+      "step": 5415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9415793968538617,
+      "learning_rate": 7.676863377279747e-06,
+      "loss": 0.316,
+      "step": 5416
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6608659713305773,
+      "learning_rate": 7.676003100246558e-06,
+      "loss": 0.3273,
+      "step": 5417
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.873022642099225,
+      "learning_rate": 7.675142712181247e-06,
+      "loss": 0.315,
+      "step": 5418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.545331336651048,
+      "learning_rate": 7.674282213119511e-06,
+      "loss": 0.327,
+      "step": 5419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8345110872299504,
+      "learning_rate": 7.673421603097058e-06,
+      "loss": 0.3159,
+      "step": 5420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6698593550026084,
+      "learning_rate": 7.672560882149594e-06,
+      "loss": 0.3215,
+      "step": 5421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.270607186862521,
+      "learning_rate": 7.671700050312832e-06,
+      "loss": 0.2988,
+      "step": 5422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.852817023142275,
+      "learning_rate": 7.670839107622488e-06,
+      "loss": 0.309,
+      "step": 5423
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.437957536653853,
+      "learning_rate": 7.669978054114286e-06,
+      "loss": 0.3238,
+      "step": 5424
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6842640113997158,
+      "learning_rate": 7.669116889823955e-06,
+      "loss": 0.2997,
+      "step": 5425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.206636580338994,
+      "learning_rate": 7.668255614787219e-06,
+      "loss": 0.3039,
+      "step": 5426
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9246387885589153,
+      "learning_rate": 7.667394229039822e-06,
+      "loss": 0.3201,
+      "step": 5427
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.708216086376897,
+      "learning_rate": 7.666532732617498e-06,
+      "loss": 0.3288,
+      "step": 5428
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.0998039251897485,
+      "learning_rate": 7.665671125555996e-06,
+      "loss": 0.3177,
+      "step": 5429
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8879890695701977,
+      "learning_rate": 7.664809407891063e-06,
+      "loss": 0.3019,
+      "step": 5430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5072264638777915,
+      "learning_rate": 7.663947579658456e-06,
+      "loss": 0.3139,
+      "step": 5431
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8296488345571333,
+      "learning_rate": 7.66308564089393e-06,
+      "loss": 0.3299,
+      "step": 5432
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0746004340729414,
+      "learning_rate": 7.662223591633254e-06,
+      "loss": 0.3015,
+      "step": 5433
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5580206769405773,
+      "learning_rate": 7.66136143191219e-06,
+      "loss": 0.308,
+      "step": 5434
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4659077118525965,
+      "learning_rate": 7.660499161766514e-06,
+      "loss": 0.3092,
+      "step": 5435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6012483303825125,
+      "learning_rate": 7.659636781232001e-06,
+      "loss": 0.3118,
+      "step": 5436
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9973094192329948,
+      "learning_rate": 7.658774290344435e-06,
+      "loss": 0.316,
+      "step": 5437
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.555775572637292,
+      "learning_rate": 7.6579116891396e-06,
+      "loss": 0.3129,
+      "step": 5438
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.538207398097279,
+      "learning_rate": 7.65704897765329e-06,
+      "loss": 0.3021,
+      "step": 5439
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.673180852549415,
+      "learning_rate": 7.656186155921296e-06,
+      "loss": 0.5251,
+      "step": 5440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.902886639712222,
+      "learning_rate": 7.655323223979424e-06,
+      "loss": 0.3129,
+      "step": 5441
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4532476838787907,
+      "learning_rate": 7.654460181863473e-06,
+      "loss": 0.2983,
+      "step": 5442
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1444827784271028,
+      "learning_rate": 7.653597029609254e-06,
+      "loss": 0.3406,
+      "step": 5443
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5583675835258517,
+      "learning_rate": 7.65273376725258e-06,
+      "loss": 0.3173,
+      "step": 5444
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3527596301442983,
+      "learning_rate": 7.651870394829272e-06,
+      "loss": 0.3132,
+      "step": 5445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2149606730870763,
+      "learning_rate": 7.651006912375149e-06,
+      "loss": 0.3034,
+      "step": 5446
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5382464713059398,
+      "learning_rate": 7.650143319926044e-06,
+      "loss": 0.3045,
+      "step": 5447
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.665601896616514,
+      "learning_rate": 7.649279617517782e-06,
+      "loss": 0.3081,
+      "step": 5448
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5132442279557612,
+      "learning_rate": 7.648415805186205e-06,
+      "loss": 0.3163,
+      "step": 5449
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.677185939394891,
+      "learning_rate": 7.64755188296715e-06,
+      "loss": 0.3287,
+      "step": 5450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.492380227783077,
+      "learning_rate": 7.646687850896468e-06,
+      "loss": 0.3185,
+      "step": 5451
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8895882032413054,
+      "learning_rate": 7.645823709010003e-06,
+      "loss": 0.3086,
+      "step": 5452
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4730331216348589,
+      "learning_rate": 7.644959457343615e-06,
+      "loss": 0.3091,
+      "step": 5453
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6305290041659515,
+      "learning_rate": 7.644095095933159e-06,
+      "loss": 0.3231,
+      "step": 5454
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9441251975660379,
+      "learning_rate": 7.643230624814504e-06,
+      "loss": 0.2992,
+      "step": 5455
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8346439660799356,
+      "learning_rate": 7.642366044023513e-06,
+      "loss": 0.3424,
+      "step": 5456
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5807228655220897,
+      "learning_rate": 7.641501353596063e-06,
+      "loss": 0.3012,
+      "step": 5457
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.646570700792608,
+      "learning_rate": 7.640636553568028e-06,
+      "loss": 0.4817,
+      "step": 5458
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4828639952335128,
+      "learning_rate": 7.639771643975294e-06,
+      "loss": 0.2872,
+      "step": 5459
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.608629996755941,
+      "learning_rate": 7.638906624853744e-06,
+      "loss": 0.3362,
+      "step": 5460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3528149154625222,
+      "learning_rate": 7.638041496239273e-06,
+      "loss": 0.3083,
+      "step": 5461
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3575733824670244,
+      "learning_rate": 7.637176258167773e-06,
+      "loss": 0.3014,
+      "step": 5462
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8014531298328489,
+      "learning_rate": 7.636310910675146e-06,
+      "loss": 0.3332,
+      "step": 5463
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.9861518497379675,
+      "learning_rate": 7.635445453797296e-06,
+      "loss": 0.3078,
+      "step": 5464
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6201889026965566,
+      "learning_rate": 7.634579887570135e-06,
+      "loss": 0.3286,
+      "step": 5465
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8175134249410336,
+      "learning_rate": 7.633714212029575e-06,
+      "loss": 0.308,
+      "step": 5466
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5878357263015503,
+      "learning_rate": 7.632848427211533e-06,
+      "loss": 0.3027,
+      "step": 5467
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.030875584750693,
+      "learning_rate": 7.631982533151934e-06,
+      "loss": 0.3293,
+      "step": 5468
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.93191629823843,
+      "learning_rate": 7.631116529886704e-06,
+      "loss": 0.3043,
+      "step": 5469
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.165738276276089,
+      "learning_rate": 7.630250417451777e-06,
+      "loss": 0.3533,
+      "step": 5470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0068217250358273,
+      "learning_rate": 7.6293841958830874e-06,
+      "loss": 0.3192,
+      "step": 5471
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 55.71888785457274,
+      "learning_rate": 7.6285178652165785e-06,
+      "loss": 0.2905,
+      "step": 5472
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2722623000258324,
+      "learning_rate": 7.627651425488193e-06,
+      "loss": 0.3256,
+      "step": 5473
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.124205261884713,
+      "learning_rate": 7.626784876733884e-06,
+      "loss": 0.307,
+      "step": 5474
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2024345777044216,
+      "learning_rate": 7.625918218989603e-06,
+      "loss": 0.3101,
+      "step": 5475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5904398430538602,
+      "learning_rate": 7.625051452291312e-06,
+      "loss": 0.2959,
+      "step": 5476
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9110644815770113,
+      "learning_rate": 7.624184576674973e-06,
+      "loss": 0.304,
+      "step": 5477
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.974369373027941,
+      "learning_rate": 7.6233175921765545e-06,
+      "loss": 0.3172,
+      "step": 5478
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5451315928813225,
+      "learning_rate": 7.622450498832031e-06,
+      "loss": 0.3024,
+      "step": 5479
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.039741423737045,
+      "learning_rate": 7.621583296677377e-06,
+      "loss": 0.3058,
+      "step": 5480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4226758284193703,
+      "learning_rate": 7.620715985748578e-06,
+      "loss": 0.3276,
+      "step": 5481
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6398475391365492,
+      "learning_rate": 7.619848566081615e-06,
+      "loss": 0.3309,
+      "step": 5482
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.706618561917322,
+      "learning_rate": 7.6189810377124826e-06,
+      "loss": 0.3284,
+      "step": 5483
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5754415149307848,
+      "learning_rate": 7.618113400677176e-06,
+      "loss": 0.3229,
+      "step": 5484
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.194172050701368,
+      "learning_rate": 7.617245655011694e-06,
+      "loss": 0.3234,
+      "step": 5485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.050477147220965,
+      "learning_rate": 7.6163778007520395e-06,
+      "loss": 0.3344,
+      "step": 5486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6299172795638901,
+      "learning_rate": 7.615509837934224e-06,
+      "loss": 0.5109,
+      "step": 5487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.237847129493219,
+      "learning_rate": 7.61464176659426e-06,
+      "loss": 0.318,
+      "step": 5488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.729576606015852,
+      "learning_rate": 7.6137735867681635e-06,
+      "loss": 0.3284,
+      "step": 5489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9396008089136665,
+      "learning_rate": 7.612905298491958e-06,
+      "loss": 0.308,
+      "step": 5490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0854074033250587,
+      "learning_rate": 7.6120369018016735e-06,
+      "loss": 0.3206,
+      "step": 5491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.697844250934911,
+      "learning_rate": 7.611168396733336e-06,
+      "loss": 0.3042,
+      "step": 5492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2836817591059266,
+      "learning_rate": 7.610299783322984e-06,
+      "loss": 0.3369,
+      "step": 5493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9248168667966188,
+      "learning_rate": 7.609431061606658e-06,
+      "loss": 0.3085,
+      "step": 5494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5950846935985723,
+      "learning_rate": 7.6085622316204036e-06,
+      "loss": 0.3099,
+      "step": 5495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7311641935530067,
+      "learning_rate": 7.607693293400266e-06,
+      "loss": 0.3181,
+      "step": 5496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.14941056256365,
+      "learning_rate": 7.606824246982305e-06,
+      "loss": 0.3021,
+      "step": 5497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4293199970961823,
+      "learning_rate": 7.6059550924025726e-06,
+      "loss": 0.3105,
+      "step": 5498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.303225757039219,
+      "learning_rate": 7.605085829697139e-06,
+      "loss": 0.2994,
+      "step": 5499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6735844612911548,
+      "learning_rate": 7.604216458902063e-06,
+      "loss": 0.3008,
+      "step": 5500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.401747720583864,
+      "learning_rate": 7.6033469800534235e-06,
+      "loss": 0.3203,
+      "step": 5501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0383037151347123,
+      "learning_rate": 7.602477393187291e-06,
+      "loss": 0.3148,
+      "step": 5502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.922985791964369,
+      "learning_rate": 7.601607698339752e-06,
+      "loss": 0.3433,
+      "step": 5503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9108365082688668,
+      "learning_rate": 7.600737895546886e-06,
+      "loss": 0.3222,
+      "step": 5504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1092126191259695,
+      "learning_rate": 7.599867984844786e-06,
+      "loss": 0.3178,
+      "step": 5505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5546278203591173,
+      "learning_rate": 7.598997966269545e-06,
+      "loss": 0.316,
+      "step": 5506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8597113984338176,
+      "learning_rate": 7.5981278398572634e-06,
+      "loss": 0.3008,
+      "step": 5507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3142683648737583,
+      "learning_rate": 7.597257605644042e-06,
+      "loss": 0.3095,
+      "step": 5508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.85963895739769,
+      "learning_rate": 7.59638726366599e-06,
+      "loss": 0.3422,
+      "step": 5509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.724941260177425,
+      "learning_rate": 7.595516813959218e-06,
+      "loss": 0.4937,
+      "step": 5510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7242903690271727,
+      "learning_rate": 7.594646256559843e-06,
+      "loss": 0.3016,
+      "step": 5511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0813544568681692,
+      "learning_rate": 7.593775591503986e-06,
+      "loss": 0.3284,
+      "step": 5512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8868391929230388,
+      "learning_rate": 7.592904818827774e-06,
+      "loss": 0.3198,
+      "step": 5513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.37117999579112,
+      "learning_rate": 7.592033938567335e-06,
+      "loss": 0.3097,
+      "step": 5514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5172549935797623,
+      "learning_rate": 7.591162950758803e-06,
+      "loss": 0.304,
+      "step": 5515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.1622814448229715,
+      "learning_rate": 7.590291855438319e-06,
+      "loss": 0.3181,
+      "step": 5516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.976591741928131,
+      "learning_rate": 7.589420652642025e-06,
+      "loss": 0.2925,
+      "step": 5517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8019794737536576,
+      "learning_rate": 7.588549342406068e-06,
+      "loss": 0.3089,
+      "step": 5518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9974672229337505,
+      "learning_rate": 7.587677924766601e-06,
+      "loss": 0.313,
+      "step": 5519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6339012014344183,
+      "learning_rate": 7.586806399759781e-06,
+      "loss": 0.3142,
+      "step": 5520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2350736606527395,
+      "learning_rate": 7.58593476742177e-06,
+      "loss": 0.3176,
+      "step": 5521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.936643097261422,
+      "learning_rate": 7.58506302778873e-06,
+      "loss": 0.3138,
+      "step": 5522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3658792722133504,
+      "learning_rate": 7.584191180896835e-06,
+      "loss": 0.3223,
+      "step": 5523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8727397687703295,
+      "learning_rate": 7.583319226782258e-06,
+      "loss": 0.3138,
+      "step": 5524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7414055833013713,
+      "learning_rate": 7.582447165481177e-06,
+      "loss": 0.2938,
+      "step": 5525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6510634094253187,
+      "learning_rate": 7.581574997029777e-06,
+      "loss": 0.3133,
+      "step": 5526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4435014538369866,
+      "learning_rate": 7.580702721464244e-06,
+      "loss": 0.321,
+      "step": 5527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.090746458260443,
+      "learning_rate": 7.57983033882077e-06,
+      "loss": 0.3228,
+      "step": 5528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.309977355524843,
+      "learning_rate": 7.578957849135555e-06,
+      "loss": 0.3054,
+      "step": 5529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6793297113210444,
+      "learning_rate": 7.578085252444798e-06,
+      "loss": 0.5371,
+      "step": 5530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.109829442603735,
+      "learning_rate": 7.577212548784705e-06,
+      "loss": 0.3139,
+      "step": 5531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.731572911362363,
+      "learning_rate": 7.576339738191485e-06,
+      "loss": 0.3142,
+      "step": 5532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9886255444932028,
+      "learning_rate": 7.5754668207013535e-06,
+      "loss": 0.3152,
+      "step": 5533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0076796278792806,
+      "learning_rate": 7.574593796350527e-06,
+      "loss": 0.2962,
+      "step": 5534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.511482315572893,
+      "learning_rate": 7.573720665175234e-06,
+      "loss": 0.3052,
+      "step": 5535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.816292089019836,
+      "learning_rate": 7.572847427211698e-06,
+      "loss": 0.335,
+      "step": 5536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9239117791983056,
+      "learning_rate": 7.5719740824961516e-06,
+      "loss": 0.3081,
+      "step": 5537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.679967083566104,
+      "learning_rate": 7.5711006310648316e-06,
+      "loss": 0.3028,
+      "step": 5538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.576149825720993,
+      "learning_rate": 7.570227072953982e-06,
+      "loss": 0.306,
+      "step": 5539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.592116405588844,
+      "learning_rate": 7.569353408199844e-06,
+      "loss": 0.3108,
+      "step": 5540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5437412534820862,
+      "learning_rate": 7.568479636838672e-06,
+      "loss": 0.2998,
+      "step": 5541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7713488575365337,
+      "learning_rate": 7.567605758906714e-06,
+      "loss": 0.3167,
+      "step": 5542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2016171579570147,
+      "learning_rate": 7.5667317744402345e-06,
+      "loss": 0.3175,
+      "step": 5543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0686350989662605,
+      "learning_rate": 7.565857683475495e-06,
+      "loss": 0.3285,
+      "step": 5544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7541404676501888,
+      "learning_rate": 7.564983486048763e-06,
+      "loss": 0.3258,
+      "step": 5545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.637809120824183,
+      "learning_rate": 7.564109182196309e-06,
+      "loss": 0.304,
+      "step": 5546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.065564160982038,
+      "learning_rate": 7.563234771954411e-06,
+      "loss": 0.2936,
+      "step": 5547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.124177000960691,
+      "learning_rate": 7.562360255359349e-06,
+      "loss": 0.3253,
+      "step": 5548
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.836370184152582,
+      "learning_rate": 7.5614856324474095e-06,
+      "loss": 0.3187,
+      "step": 5549
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2304473261043967,
+      "learning_rate": 7.560610903254881e-06,
+      "loss": 0.3165,
+      "step": 5550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.211099580150378,
+      "learning_rate": 7.55973606781806e-06,
+      "loss": 0.2984,
+      "step": 5551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.106253574896783,
+      "learning_rate": 7.558861126173241e-06,
+      "loss": 0.3222,
+      "step": 5552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.43029548593907,
+      "learning_rate": 7.557986078356727e-06,
+      "loss": 0.3309,
+      "step": 5553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4655876812264395,
+      "learning_rate": 7.557110924404829e-06,
+      "loss": 0.3112,
+      "step": 5554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5698564984498824,
+      "learning_rate": 7.5562356643538586e-06,
+      "loss": 0.3449,
+      "step": 5555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8906463936232092,
+      "learning_rate": 7.555360298240128e-06,
+      "loss": 0.3077,
+      "step": 5556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5490763519009287,
+      "learning_rate": 7.554484826099959e-06,
+      "loss": 0.3296,
+      "step": 5557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6689562673367422,
+      "learning_rate": 7.55360924796968e-06,
+      "loss": 0.5084,
+      "step": 5558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4674652561581336,
+      "learning_rate": 7.552733563885616e-06,
+      "loss": 0.3063,
+      "step": 5559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1978635073512747,
+      "learning_rate": 7.551857773884101e-06,
+      "loss": 0.3116,
+      "step": 5560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4731960314675723,
+      "learning_rate": 7.550981878001476e-06,
+      "loss": 0.3118,
+      "step": 5561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4934639668928678,
+      "learning_rate": 7.550105876274082e-06,
+      "loss": 0.299,
+      "step": 5562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4704119077333533,
+      "learning_rate": 7.549229768738266e-06,
+      "loss": 0.3229,
+      "step": 5563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.678290582969975,
+      "learning_rate": 7.548353555430378e-06,
+      "loss": 0.3035,
+      "step": 5564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.050749009636137,
+      "learning_rate": 7.547477236386774e-06,
+      "loss": 0.3078,
+      "step": 5565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.146993060143408,
+      "learning_rate": 7.546600811643816e-06,
+      "loss": 0.3065,
+      "step": 5566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5933902958493147,
+      "learning_rate": 7.545724281237867e-06,
+      "loss": 0.5215,
+      "step": 5567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.452429749055142,
+      "learning_rate": 7.5448476452052955e-06,
+      "loss": 0.3335,
+      "step": 5568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1746540608976783,
+      "learning_rate": 7.543970903582473e-06,
+      "loss": 0.3232,
+      "step": 5569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5632182117070235,
+      "learning_rate": 7.543094056405782e-06,
+      "loss": 0.3209,
+      "step": 5570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9638717159951005,
+      "learning_rate": 7.542217103711601e-06,
+      "loss": 0.3121,
+      "step": 5571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.144578477203159,
+      "learning_rate": 7.5413400455363154e-06,
+      "loss": 0.3242,
+      "step": 5572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.298585420131417,
+      "learning_rate": 7.54046288191632e-06,
+      "loss": 0.3155,
+      "step": 5573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6620275982754027,
+      "learning_rate": 7.5395856128880045e-06,
+      "loss": 0.3326,
+      "step": 5574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.614520449225384,
+      "learning_rate": 7.5387082384877715e-06,
+      "loss": 0.3193,
+      "step": 5575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6048965548854468,
+      "learning_rate": 7.537830758752025e-06,
+      "loss": 0.5238,
+      "step": 5576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.22001852217234,
+      "learning_rate": 7.536953173717174e-06,
+      "loss": 0.3176,
+      "step": 5577
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3787820806039615,
+      "learning_rate": 7.5360754834196284e-06,
+      "loss": 0.301,
+      "step": 5578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.234464955645964,
+      "learning_rate": 7.535197687895805e-06,
+      "loss": 0.3451,
+      "step": 5579
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5321716541892436,
+      "learning_rate": 7.53431978718213e-06,
+      "loss": 0.3164,
+      "step": 5580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.646921846462847,
+      "learning_rate": 7.533441781315023e-06,
+      "loss": 0.3161,
+      "step": 5581
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8979577692014298,
+      "learning_rate": 7.5325636703309165e-06,
+      "loss": 0.3135,
+      "step": 5582
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1137255471128475,
+      "learning_rate": 7.531685454266246e-06,
+      "loss": 0.3035,
+      "step": 5583
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5792122713511922,
+      "learning_rate": 7.5308071331574485e-06,
+      "loss": 0.2999,
+      "step": 5584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5997296727761863,
+      "learning_rate": 7.529928707040969e-06,
+      "loss": 0.3255,
+      "step": 5585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.950600940282619,
+      "learning_rate": 7.529050175953253e-06,
+      "loss": 0.3175,
+      "step": 5586
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6416121319325168,
+      "learning_rate": 7.528171539930753e-06,
+      "loss": 0.3453,
+      "step": 5587
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6169929566759427,
+      "learning_rate": 7.527292799009924e-06,
+      "loss": 0.5057,
+      "step": 5588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5772616025881065,
+      "learning_rate": 7.526413953227231e-06,
+      "loss": 0.3108,
+      "step": 5589
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3937024867886079,
+      "learning_rate": 7.5255350026191336e-06,
+      "loss": 0.3059,
+      "step": 5590
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7036393897636453,
+      "learning_rate": 7.524655947222104e-06,
+      "loss": 0.2912,
+      "step": 5591
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.303718523195396,
+      "learning_rate": 7.523776787072613e-06,
+      "loss": 0.3108,
+      "step": 5592
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4651604061668437,
+      "learning_rate": 7.5228975222071425e-06,
+      "loss": 0.2991,
+      "step": 5593
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.744966095349305,
+      "learning_rate": 7.522018152662171e-06,
+      "loss": 0.3381,
+      "step": 5594
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9847177841081725,
+      "learning_rate": 7.521138678474187e-06,
+      "loss": 0.3055,
+      "step": 5595
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1188786865678044,
+      "learning_rate": 7.520259099679681e-06,
+      "loss": 0.3004,
+      "step": 5596
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8088717725026118,
+      "learning_rate": 7.5193794163151516e-06,
+      "loss": 0.3149,
+      "step": 5597
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0462386040246234,
+      "learning_rate": 7.518499628417091e-06,
+      "loss": 0.2974,
+      "step": 5598
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5040063755350557,
+      "learning_rate": 7.5176197360220096e-06,
+      "loss": 0.3011,
+      "step": 5599
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3183981886848737,
+      "learning_rate": 7.516739739166412e-06,
+      "loss": 0.3122,
+      "step": 5600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9674896556573818,
+      "learning_rate": 7.515859637886814e-06,
+      "loss": 0.3014,
+      "step": 5601
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7152203946403568,
+      "learning_rate": 7.514979432219731e-06,
+      "loss": 0.3145,
+      "step": 5602
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.403720004356182,
+      "learning_rate": 7.514099122201683e-06,
+      "loss": 0.3127,
+      "step": 5603
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7211492577570522,
+      "learning_rate": 7.513218707869199e-06,
+      "loss": 0.3112,
+      "step": 5604
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6104140208393277,
+      "learning_rate": 7.512338189258805e-06,
+      "loss": 0.4887,
+      "step": 5605
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5343344536735042,
+      "learning_rate": 7.511457566407038e-06,
+      "loss": 0.3113,
+      "step": 5606
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8576184789734569,
+      "learning_rate": 7.510576839350435e-06,
+      "loss": 0.3208,
+      "step": 5607
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7988619285075296,
+      "learning_rate": 7.50969600812554e-06,
+      "loss": 0.312,
+      "step": 5608
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1524713125801425,
+      "learning_rate": 7.5088150727689006e-06,
+      "loss": 0.2999,
+      "step": 5609
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8278447817509023,
+      "learning_rate": 7.5079340333170654e-06,
+      "loss": 0.3255,
+      "step": 5610
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4991055701569922,
+      "learning_rate": 7.507052889806595e-06,
+      "loss": 0.3143,
+      "step": 5611
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6456788524314352,
+      "learning_rate": 7.506171642274046e-06,
+      "loss": 0.3322,
+      "step": 5612
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9083932346937371,
+      "learning_rate": 7.505290290755985e-06,
+      "loss": 0.2952,
+      "step": 5613
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6959065724865343,
+      "learning_rate": 7.5044088352889785e-06,
+      "loss": 0.3051,
+      "step": 5614
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5812591366602096,
+      "learning_rate": 7.503527275909603e-06,
+      "loss": 0.3112,
+      "step": 5615
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4145626855077336,
+      "learning_rate": 7.502645612654432e-06,
+      "loss": 0.2984,
+      "step": 5616
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5959014876694964,
+      "learning_rate": 7.5017638455600505e-06,
+      "loss": 0.3104,
+      "step": 5617
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9324527414534696,
+      "learning_rate": 7.5008819746630435e-06,
+      "loss": 0.2996,
+      "step": 5618
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.672622208377859,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2966,
+      "step": 5619
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5514313095791672,
+      "learning_rate": 7.499117921607518e-06,
+      "loss": 0.3099,
+      "step": 5620
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8434493314240692,
+      "learning_rate": 7.498235739522193e-06,
+      "loss": 0.3271,
+      "step": 5621
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.674892198816176,
+      "learning_rate": 7.497353453780631e-06,
+      "loss": 0.2908,
+      "step": 5622
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.470732541528881,
+      "learning_rate": 7.496471064419437e-06,
+      "loss": 0.3165,
+      "step": 5623
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.469813071897201,
+      "learning_rate": 7.495588571475225e-06,
+      "loss": 0.3097,
+      "step": 5624
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.559372444624539,
+      "learning_rate": 7.494705974984611e-06,
+      "loss": 0.3126,
+      "step": 5625
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4119806716099852,
+      "learning_rate": 7.493823274984215e-06,
+      "loss": 0.3063,
+      "step": 5626
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.228598150709018,
+      "learning_rate": 7.492940471510662e-06,
+      "loss": 0.2916,
+      "step": 5627
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.987139959495272,
+      "learning_rate": 7.492057564600579e-06,
+      "loss": 0.3042,
+      "step": 5628
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9055052025674284,
+      "learning_rate": 7.491174554290602e-06,
+      "loss": 0.3166,
+      "step": 5629
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3521912386910904,
+      "learning_rate": 7.490291440617369e-06,
+      "loss": 0.3072,
+      "step": 5630
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9339713899719875,
+      "learning_rate": 7.489408223617521e-06,
+      "loss": 0.3212,
+      "step": 5631
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2991312285805725,
+      "learning_rate": 7.488524903327703e-06,
+      "loss": 0.3222,
+      "step": 5632
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5746372787736225,
+      "learning_rate": 7.487641479784566e-06,
+      "loss": 0.3023,
+      "step": 5633
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.046394299527198,
+      "learning_rate": 7.486757953024767e-06,
+      "loss": 0.3154,
+      "step": 5634
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5720029238898356,
+      "learning_rate": 7.485874323084963e-06,
+      "loss": 0.3192,
+      "step": 5635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1660435220907193,
+      "learning_rate": 7.484990590001818e-06,
+      "loss": 0.3327,
+      "step": 5636
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6374788207877375,
+      "learning_rate": 7.484106753811999e-06,
+      "loss": 0.3035,
+      "step": 5637
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6121513654484185,
+      "learning_rate": 7.4832228145521805e-06,
+      "loss": 0.3076,
+      "step": 5638
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.380066003424248,
+      "learning_rate": 7.4823387722590345e-06,
+      "loss": 0.3108,
+      "step": 5639
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.200349463868485,
+      "learning_rate": 7.481454626969244e-06,
+      "loss": 0.3084,
+      "step": 5640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.807919801223866,
+      "learning_rate": 7.480570378719494e-06,
+      "loss": 0.3304,
+      "step": 5641
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7027573751488092,
+      "learning_rate": 7.4796860275464734e-06,
+      "loss": 0.3271,
+      "step": 5642
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9935426564123864,
+      "learning_rate": 7.4788015734868745e-06,
+      "loss": 0.3291,
+      "step": 5643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1901492399519804,
+      "learning_rate": 7.477917016577396e-06,
+      "loss": 0.3183,
+      "step": 5644
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5598227356088434,
+      "learning_rate": 7.477032356854739e-06,
+      "loss": 0.3284,
+      "step": 5645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5607753249772587,
+      "learning_rate": 7.47614759435561e-06,
+      "loss": 0.3168,
+      "step": 5646
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2946975217902708,
+      "learning_rate": 7.475262729116718e-06,
+      "loss": 0.3079,
+      "step": 5647
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1161331653257256,
+      "learning_rate": 7.47437776117478e-06,
+      "loss": 0.3147,
+      "step": 5648
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.567136292200706,
+      "learning_rate": 7.473492690566513e-06,
+      "loss": 0.3097,
+      "step": 5649
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6040788197915585,
+      "learning_rate": 7.472607517328641e-06,
+      "loss": 0.3063,
+      "step": 5650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.251851878767218,
+      "learning_rate": 7.4717222414978915e-06,
+      "loss": 0.319,
+      "step": 5651
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1348626804101776,
+      "learning_rate": 7.470836863110996e-06,
+      "loss": 0.3091,
+      "step": 5652
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.685702944543895,
+      "learning_rate": 7.46995138220469e-06,
+      "loss": 0.3122,
+      "step": 5653
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3611225750638158,
+      "learning_rate": 7.469065798815715e-06,
+      "loss": 0.3128,
+      "step": 5654
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 58.515560938211244,
+      "learning_rate": 7.468180112980813e-06,
+      "loss": 0.2983,
+      "step": 5655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.780049182137719,
+      "learning_rate": 7.467294324736735e-06,
+      "loss": 0.3055,
+      "step": 5656
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.674422368242577,
+      "learning_rate": 7.466408434120231e-06,
+      "loss": 0.33,
+      "step": 5657
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.670010019365439,
+      "learning_rate": 7.4655224411680635e-06,
+      "loss": 0.3054,
+      "step": 5658
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3692648213304666,
+      "learning_rate": 7.464636345916989e-06,
+      "loss": 0.3146,
+      "step": 5659
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.428873933394962,
+      "learning_rate": 7.463750148403776e-06,
+      "loss": 0.3024,
+      "step": 5660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.018528477444935,
+      "learning_rate": 7.462863848665191e-06,
+      "loss": 0.3379,
+      "step": 5661
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.79207294349736,
+      "learning_rate": 7.4619774467380135e-06,
+      "loss": 0.3152,
+      "step": 5662
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8969671134218307,
+      "learning_rate": 7.461090942659016e-06,
+      "loss": 0.305,
+      "step": 5663
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1867364046379683,
+      "learning_rate": 7.4602043364649845e-06,
+      "loss": 0.3102,
+      "step": 5664
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9064886275259811,
+      "learning_rate": 7.459317628192706e-06,
+      "loss": 0.3092,
+      "step": 5665
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.51862220711062,
+      "learning_rate": 7.458430817878971e-06,
+      "loss": 0.3202,
+      "step": 5666
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3127277383490619,
+      "learning_rate": 7.457543905560574e-06,
+      "loss": 0.3002,
+      "step": 5667
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4085002021476818,
+      "learning_rate": 7.4566568912743156e-06,
+      "loss": 0.2915,
+      "step": 5668
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.169750998987912,
+      "learning_rate": 7.455769775056998e-06,
+      "loss": 0.3137,
+      "step": 5669
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4067814726613208,
+      "learning_rate": 7.454882556945433e-06,
+      "loss": 0.2997,
+      "step": 5670
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7267074013942263,
+      "learning_rate": 7.453995236976428e-06,
+      "loss": 0.3116,
+      "step": 5671
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.585783007967785,
+      "learning_rate": 7.453107815186803e-06,
+      "loss": 0.3141,
+      "step": 5672
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4987161190589928,
+      "learning_rate": 7.452220291613377e-06,
+      "loss": 0.307,
+      "step": 5673
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2510466920102021,
+      "learning_rate": 7.451332666292977e-06,
+      "loss": 0.3022,
+      "step": 5674
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0184887848923205,
+      "learning_rate": 7.450444939262429e-06,
+      "loss": 0.3041,
+      "step": 5675
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6660699120502802,
+      "learning_rate": 7.4495571105585685e-06,
+      "loss": 0.3086,
+      "step": 5676
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8567088182384004,
+      "learning_rate": 7.448669180218232e-06,
+      "loss": 0.3147,
+      "step": 5677
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0114870151083903,
+      "learning_rate": 7.447781148278264e-06,
+      "loss": 0.306,
+      "step": 5678
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.152110807995157,
+      "learning_rate": 7.446893014775506e-06,
+      "loss": 0.3235,
+      "step": 5679
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3991897600041783,
+      "learning_rate": 7.446004779746811e-06,
+      "loss": 0.3234,
+      "step": 5680
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6360703079205299,
+      "learning_rate": 7.445116443229033e-06,
+      "loss": 0.3147,
+      "step": 5681
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6753473981753595,
+      "learning_rate": 7.4442280052590325e-06,
+      "loss": 0.4836,
+      "step": 5682
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3987344153282701,
+      "learning_rate": 7.443339465873668e-06,
+      "loss": 0.3206,
+      "step": 5683
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1578346647493634,
+      "learning_rate": 7.442450825109811e-06,
+      "loss": 0.3032,
+      "step": 5684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3961765871672758,
+      "learning_rate": 7.4415620830043315e-06,
+      "loss": 0.3422,
+      "step": 5685
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3999347252216263,
+      "learning_rate": 7.440673239594104e-06,
+      "loss": 0.338,
+      "step": 5686
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.458042249081322,
+      "learning_rate": 7.439784294916006e-06,
+      "loss": 0.3071,
+      "step": 5687
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.442999583186028,
+      "learning_rate": 7.438895249006928e-06,
+      "loss": 0.3237,
+      "step": 5688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2220124555028837,
+      "learning_rate": 7.438006101903752e-06,
+      "loss": 0.3352,
+      "step": 5689
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5793718312701053,
+      "learning_rate": 7.4371168536433736e-06,
+      "loss": 0.3303,
+      "step": 5690
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.402348641953203,
+      "learning_rate": 7.436227504262686e-06,
+      "loss": 0.3095,
+      "step": 5691
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4817249184171337,
+      "learning_rate": 7.435338053798594e-06,
+      "loss": 0.3038,
+      "step": 5692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3749261366728374,
+      "learning_rate": 7.434448502288e-06,
+      "loss": 0.3013,
+      "step": 5693
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.353155418341018,
+      "learning_rate": 7.433558849767814e-06,
+      "loss": 0.3008,
+      "step": 5694
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7818034013591229,
+      "learning_rate": 7.4326690962749475e-06,
+      "loss": 0.3158,
+      "step": 5695
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7948200058142285,
+      "learning_rate": 7.431779241846321e-06,
+      "loss": 0.3059,
+      "step": 5696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.411385096486676,
+      "learning_rate": 7.430889286518853e-06,
+      "loss": 0.35,
+      "step": 5697
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6324917356146929,
+      "learning_rate": 7.429999230329472e-06,
+      "loss": 0.3016,
+      "step": 5698
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.281234103921979,
+      "learning_rate": 7.429109073315105e-06,
+      "loss": 0.3344,
+      "step": 5699
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4392219907285286,
+      "learning_rate": 7.42821881551269e-06,
+      "loss": 0.298,
+      "step": 5700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.384668267321661,
+      "learning_rate": 7.427328456959162e-06,
+      "loss": 0.3195,
+      "step": 5701
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1959807559594506,
+      "learning_rate": 7.4264379976914654e-06,
+      "loss": 0.3035,
+      "step": 5702
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.690632201664417,
+      "learning_rate": 7.425547437746546e-06,
+      "loss": 0.3098,
+      "step": 5703
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6763078633531703,
+      "learning_rate": 7.424656777161357e-06,
+      "loss": 0.3167,
+      "step": 5704
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3469438471630495,
+      "learning_rate": 7.4237660159728496e-06,
+      "loss": 0.3122,
+      "step": 5705
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6327036984516017,
+      "learning_rate": 7.422875154217986e-06,
+      "loss": 0.3062,
+      "step": 5706
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3215204754316474,
+      "learning_rate": 7.421984191933728e-06,
+      "loss": 0.3152,
+      "step": 5707
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6332467955778492,
+      "learning_rate": 7.421093129157044e-06,
+      "loss": 0.5212,
+      "step": 5708
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6653563696691982,
+      "learning_rate": 7.4202019659249066e-06,
+      "loss": 0.3062,
+      "step": 5709
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4541565492272985,
+      "learning_rate": 7.419310702274289e-06,
+      "loss": 0.3029,
+      "step": 5710
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8504941946587374,
+      "learning_rate": 7.418419338242176e-06,
+      "loss": 0.2965,
+      "step": 5711
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7463594995990253,
+      "learning_rate": 7.417527873865548e-06,
+      "loss": 0.3294,
+      "step": 5712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.557788398969939,
+      "learning_rate": 7.416636309181393e-06,
+      "loss": 0.3411,
+      "step": 5713
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6168204406296904,
+      "learning_rate": 7.415744644226706e-06,
+      "loss": 0.5094,
+      "step": 5714
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6182301560704233,
+      "learning_rate": 7.414852879038483e-06,
+      "loss": 0.4927,
+      "step": 5715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4139939106748585,
+      "learning_rate": 7.413961013653725e-06,
+      "loss": 0.3065,
+      "step": 5716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8082851084173472,
+      "learning_rate": 7.4130690481094356e-06,
+      "loss": 0.3026,
+      "step": 5717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6075580421704037,
+      "learning_rate": 7.412176982442629e-06,
+      "loss": 0.3098,
+      "step": 5718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6559692571334896,
+      "learning_rate": 7.41128481669031e-06,
+      "loss": 0.3189,
+      "step": 5719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.604115556433833,
+      "learning_rate": 7.4103925508895046e-06,
+      "loss": 0.2997,
+      "step": 5720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2416715029221943,
+      "learning_rate": 7.40950018507723e-06,
+      "loss": 0.3014,
+      "step": 5721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3092676267296375,
+      "learning_rate": 7.408607719290512e-06,
+      "loss": 0.3169,
+      "step": 5722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.535729355769897,
+      "learning_rate": 7.407715153566383e-06,
+      "loss": 0.3111,
+      "step": 5723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9132874418537968,
+      "learning_rate": 7.4068224879418734e-06,
+      "loss": 0.3238,
+      "step": 5724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.296794143617355,
+      "learning_rate": 7.405929722454026e-06,
+      "loss": 0.2883,
+      "step": 5725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6492150870470486,
+      "learning_rate": 7.40503685713988e-06,
+      "loss": 0.4716,
+      "step": 5726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3639429959010156,
+      "learning_rate": 7.404143892036484e-06,
+      "loss": 0.3275,
+      "step": 5727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.2386137948775255,
+      "learning_rate": 7.403250827180887e-06,
+      "loss": 0.3349,
+      "step": 5728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2563994628426658,
+      "learning_rate": 7.402357662610144e-06,
+      "loss": 0.3175,
+      "step": 5729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7851120512806526,
+      "learning_rate": 7.4014643983613155e-06,
+      "loss": 0.3157,
+      "step": 5730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7893059924225179,
+      "learning_rate": 7.4005710344714624e-06,
+      "loss": 0.3052,
+      "step": 5731
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4542712392411767,
+      "learning_rate": 7.399677570977653e-06,
+      "loss": 0.3406,
+      "step": 5732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9076157660029271,
+      "learning_rate": 7.39878400791696e-06,
+      "loss": 0.3087,
+      "step": 5733
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3486840979132322,
+      "learning_rate": 7.397890345326458e-06,
+      "loss": 0.3052,
+      "step": 5734
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6639528761915583,
+      "learning_rate": 7.396996583243227e-06,
+      "loss": 0.3196,
+      "step": 5735
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3565853131676544,
+      "learning_rate": 7.396102721704348e-06,
+      "loss": 0.3005,
+      "step": 5736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2639933629837592,
+      "learning_rate": 7.395208760746912e-06,
+      "loss": 0.3085,
+      "step": 5737
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.657557845945615,
+      "learning_rate": 7.394314700408012e-06,
+      "loss": 0.3254,
+      "step": 5738
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6632622764957616,
+      "learning_rate": 7.39342054072474e-06,
+      "loss": 0.3337,
+      "step": 5739
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.487677377720476,
+      "learning_rate": 7.3925262817341996e-06,
+      "loss": 0.2991,
+      "step": 5740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.701000162151653,
+      "learning_rate": 7.3916319234734935e-06,
+      "loss": 0.3186,
+      "step": 5741
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6058788602510203,
+      "learning_rate": 7.390737465979732e-06,
+      "loss": 0.287,
+      "step": 5742
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.710631298606624,
+      "learning_rate": 7.389842909290025e-06,
+      "loss": 0.3284,
+      "step": 5743
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7821217348464067,
+      "learning_rate": 7.388948253441492e-06,
+      "loss": 0.3274,
+      "step": 5744
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5799089572036988,
+      "learning_rate": 7.388053498471253e-06,
+      "loss": 0.3133,
+      "step": 5745
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6500341324063666,
+      "learning_rate": 7.387158644416432e-06,
+      "loss": 0.4969,
+      "step": 5746
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2714644253683998,
+      "learning_rate": 7.386263691314157e-06,
+      "loss": 0.3199,
+      "step": 5747
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9249387926907506,
+      "learning_rate": 7.385368639201567e-06,
+      "loss": 0.3239,
+      "step": 5748
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0455498736956876,
+      "learning_rate": 7.384473488115792e-06,
+      "loss": 0.3217,
+      "step": 5749
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6326238233539525,
+      "learning_rate": 7.383578238093979e-06,
+      "loss": 0.3071,
+      "step": 5750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0793840173506237,
+      "learning_rate": 7.38268288917327e-06,
+      "loss": 0.3343,
+      "step": 5751
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3578410524260112,
+      "learning_rate": 7.381787441390815e-06,
+      "loss": 0.311,
+      "step": 5752
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.504546431557493,
+      "learning_rate": 7.380891894783769e-06,
+      "loss": 0.3332,
+      "step": 5753
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.478990359257859,
+      "learning_rate": 7.3799962493892895e-06,
+      "loss": 0.3078,
+      "step": 5754
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5863958876970328,
+      "learning_rate": 7.379100505244538e-06,
+      "loss": 0.3193,
+      "step": 5755
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7685270421751085,
+      "learning_rate": 7.378204662386683e-06,
+      "loss": 0.2973,
+      "step": 5756
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5372639700328596,
+      "learning_rate": 7.37730872085289e-06,
+      "loss": 0.3192,
+      "step": 5757
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6127723564068863,
+      "learning_rate": 7.376412680680336e-06,
+      "loss": 0.3131,
+      "step": 5758
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4735496079413415,
+      "learning_rate": 7.375516541906199e-06,
+      "loss": 0.2981,
+      "step": 5759
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.877987572363265,
+      "learning_rate": 7.3746203045676625e-06,
+      "loss": 0.3228,
+      "step": 5760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6400665714471598,
+      "learning_rate": 7.37372396870191e-06,
+      "loss": 0.3026,
+      "step": 5761
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5285887762229906,
+      "learning_rate": 7.372827534346134e-06,
+      "loss": 0.322,
+      "step": 5762
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5689585017596384,
+      "learning_rate": 7.371931001537529e-06,
+      "loss": 0.3102,
+      "step": 5763
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.342024338730037,
+      "learning_rate": 7.371034370313296e-06,
+      "loss": 0.3575,
+      "step": 5764
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.468124222734872,
+      "learning_rate": 7.370137640710632e-06,
+      "loss": 0.3145,
+      "step": 5765
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.320101177618037,
+      "learning_rate": 7.36924081276675e-06,
+      "loss": 0.3196,
+      "step": 5766
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4339543507967045,
+      "learning_rate": 7.368343886518857e-06,
+      "loss": 0.2987,
+      "step": 5767
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8696148075773147,
+      "learning_rate": 7.3674468620041716e-06,
+      "loss": 0.313,
+      "step": 5768
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.5474706499442066,
+      "learning_rate": 7.366549739259908e-06,
+      "loss": 0.3221,
+      "step": 5769
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2964753833558096,
+      "learning_rate": 7.365652518323294e-06,
+      "loss": 0.3021,
+      "step": 5770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4044127931489798,
+      "learning_rate": 7.364755199231555e-06,
+      "loss": 0.3229,
+      "step": 5771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7601533043645516,
+      "learning_rate": 7.363857782021922e-06,
+      "loss": 0.3362,
+      "step": 5772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6087853838590078,
+      "learning_rate": 7.362960266731632e-06,
+      "loss": 0.4977,
+      "step": 5773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3781158828767315,
+      "learning_rate": 7.362062653397923e-06,
+      "loss": 0.32,
+      "step": 5774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.658366266497918,
+      "learning_rate": 7.361164942058038e-06,
+      "loss": 0.3192,
+      "step": 5775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8875124548609907,
+      "learning_rate": 7.360267132749227e-06,
+      "loss": 0.3087,
+      "step": 5776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2727850664024363,
+      "learning_rate": 7.359369225508738e-06,
+      "loss": 0.3185,
+      "step": 5777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.450642338466432,
+      "learning_rate": 7.358471220373831e-06,
+      "loss": 0.3204,
+      "step": 5778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.6278554362256004,
+      "learning_rate": 7.357573117381764e-06,
+      "loss": 0.3008,
+      "step": 5779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9078310390802198,
+      "learning_rate": 7.3566749165698e-06,
+      "loss": 0.3023,
+      "step": 5780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4518003132283595,
+      "learning_rate": 7.355776617975209e-06,
+      "loss": 0.3324,
+      "step": 5781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9831437966496162,
+      "learning_rate": 7.354878221635262e-06,
+      "loss": 0.3067,
+      "step": 5782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9558412798166955,
+      "learning_rate": 7.353979727587234e-06,
+      "loss": 0.284,
+      "step": 5783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.75995278297336,
+      "learning_rate": 7.353081135868405e-06,
+      "loss": 0.3316,
+      "step": 5784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4438940009092813,
+      "learning_rate": 7.3521824465160605e-06,
+      "loss": 0.3071,
+      "step": 5785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4964626549739581,
+      "learning_rate": 7.3512836595674896e-06,
+      "loss": 0.3031,
+      "step": 5786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.195188164153965,
+      "learning_rate": 7.350384775059983e-06,
+      "loss": 0.3179,
+      "step": 5787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4793362289035,
+      "learning_rate": 7.349485793030837e-06,
+      "loss": 0.3084,
+      "step": 5788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7286359867807102,
+      "learning_rate": 7.3485867135173514e-06,
+      "loss": 0.3196,
+      "step": 5789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5745322911322044,
+      "learning_rate": 7.347687536556833e-06,
+      "loss": 0.3144,
+      "step": 5790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4459368475648984,
+      "learning_rate": 7.346788262186588e-06,
+      "loss": 0.3082,
+      "step": 5791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5842852315498037,
+      "learning_rate": 7.34588889044393e-06,
+      "loss": 0.3118,
+      "step": 5792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6431567181335914,
+      "learning_rate": 7.344989421366175e-06,
+      "loss": 0.5115,
+      "step": 5793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7974130873253658,
+      "learning_rate": 7.3440898549906435e-06,
+      "loss": 0.308,
+      "step": 5794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.149415807747741,
+      "learning_rate": 7.34319019135466e-06,
+      "loss": 0.3145,
+      "step": 5795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0781212669866096,
+      "learning_rate": 7.342290430495554e-06,
+      "loss": 0.3185,
+      "step": 5796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6616416544173185,
+      "learning_rate": 7.341390572450659e-06,
+      "loss": 0.3071,
+      "step": 5797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.433830528273421,
+      "learning_rate": 7.340490617257309e-06,
+      "loss": 0.3169,
+      "step": 5798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9030889016357984,
+      "learning_rate": 7.339590564952845e-06,
+      "loss": 0.3178,
+      "step": 5799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.646943345805651,
+      "learning_rate": 7.338690415574614e-06,
+      "loss": 0.3067,
+      "step": 5800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6316014584446898,
+      "learning_rate": 7.337790169159964e-06,
+      "loss": 0.3009,
+      "step": 5801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.144307940388868,
+      "learning_rate": 7.3368898257462486e-06,
+      "loss": 0.3031,
+      "step": 5802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4620217959917476,
+      "learning_rate": 7.3359893853708205e-06,
+      "loss": 0.3002,
+      "step": 5803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9917308700716931,
+      "learning_rate": 7.335088848071046e-06,
+      "loss": 0.3085,
+      "step": 5804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.515370739560277,
+      "learning_rate": 7.334188213884287e-06,
+      "loss": 0.3089,
+      "step": 5805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8956378841874428,
+      "learning_rate": 7.333287482847913e-06,
+      "loss": 0.2921,
+      "step": 5806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0689430657937513,
+      "learning_rate": 7.332386654999296e-06,
+      "loss": 0.3124,
+      "step": 5807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1196299524360107,
+      "learning_rate": 7.3314857303758155e-06,
+      "loss": 0.3207,
+      "step": 5808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8664441193295482,
+      "learning_rate": 7.330584709014849e-06,
+      "loss": 0.3125,
+      "step": 5809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5672815518063077,
+      "learning_rate": 7.329683590953785e-06,
+      "loss": 0.3214,
+      "step": 5810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9459149264097486,
+      "learning_rate": 7.328782376230011e-06,
+      "loss": 0.3032,
+      "step": 5811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8956083387799723,
+      "learning_rate": 7.32788106488092e-06,
+      "loss": 0.3189,
+      "step": 5812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.724382629263787,
+      "learning_rate": 7.326979656943907e-06,
+      "loss": 0.301,
+      "step": 5813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.040882141972441,
+      "learning_rate": 7.326078152456375e-06,
+      "loss": 0.3041,
+      "step": 5814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9496157807918788,
+      "learning_rate": 7.325176551455729e-06,
+      "loss": 0.334,
+      "step": 5815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.224184541491048,
+      "learning_rate": 7.324274853979381e-06,
+      "loss": 0.3067,
+      "step": 5816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.00410639397539,
+      "learning_rate": 7.323373060064738e-06,
+      "loss": 0.2864,
+      "step": 5817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3888165581844634,
+      "learning_rate": 7.322471169749219e-06,
+      "loss": 0.3062,
+      "step": 5818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.829558357402964,
+      "learning_rate": 7.321569183070247e-06,
+      "loss": 0.2967,
+      "step": 5819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2188317970095683,
+      "learning_rate": 7.320667100065248e-06,
+      "loss": 0.309,
+      "step": 5820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7218714069460432,
+      "learning_rate": 7.319764920771646e-06,
+      "loss": 0.3075,
+      "step": 5821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.48354112910123,
+      "learning_rate": 7.31886264522688e-06,
+      "loss": 0.289,
+      "step": 5822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.687544450349418,
+      "learning_rate": 7.3179602734683815e-06,
+      "loss": 0.5144,
+      "step": 5823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6247972565050295,
+      "learning_rate": 7.317057805533596e-06,
+      "loss": 0.305,
+      "step": 5824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.133622568335391,
+      "learning_rate": 7.316155241459966e-06,
+      "loss": 0.3133,
+      "step": 5825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.28950572443818,
+      "learning_rate": 7.315252581284942e-06,
+      "loss": 0.3122,
+      "step": 5826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.472810479886116,
+      "learning_rate": 7.314349825045975e-06,
+      "loss": 0.3088,
+      "step": 5827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6783569563645876,
+      "learning_rate": 7.313446972780522e-06,
+      "loss": 0.5448,
+      "step": 5828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.821842629114185,
+      "learning_rate": 7.312544024526045e-06,
+      "loss": 0.3261,
+      "step": 5829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.204742629300778,
+      "learning_rate": 7.311640980320012e-06,
+      "loss": 0.3178,
+      "step": 5830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2984830118945174,
+      "learning_rate": 7.310737840199886e-06,
+      "loss": 0.312,
+      "step": 5831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2660732407209636,
+      "learning_rate": 7.3098346042031435e-06,
+      "loss": 0.3022,
+      "step": 5832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0585727921304535,
+      "learning_rate": 7.308931272367259e-06,
+      "loss": 0.3282,
+      "step": 5833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.814486531397894,
+      "learning_rate": 7.308027844729717e-06,
+      "loss": 0.3001,
+      "step": 5834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.139194598575215,
+      "learning_rate": 7.3071243213279994e-06,
+      "loss": 0.2996,
+      "step": 5835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.854070756875088,
+      "learning_rate": 7.306220702199596e-06,
+      "loss": 0.3038,
+      "step": 5836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2854665347947902,
+      "learning_rate": 7.305316987381998e-06,
+      "loss": 0.3136,
+      "step": 5837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.451365980106976,
+      "learning_rate": 7.304413176912706e-06,
+      "loss": 0.2991,
+      "step": 5838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5527165313661646,
+      "learning_rate": 7.303509270829217e-06,
+      "loss": 0.2992,
+      "step": 5839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.184094753030994,
+      "learning_rate": 7.302605269169036e-06,
+      "loss": 0.3067,
+      "step": 5840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.009560153174841,
+      "learning_rate": 7.301701171969673e-06,
+      "loss": 0.3107,
+      "step": 5841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8787207694283758,
+      "learning_rate": 7.300796979268641e-06,
+      "loss": 0.3032,
+      "step": 5842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.867082892121984,
+      "learning_rate": 7.299892691103455e-06,
+      "loss": 0.3259,
+      "step": 5843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9656140868400405,
+      "learning_rate": 7.298988307511637e-06,
+      "loss": 0.3353,
+      "step": 5844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3475489149999587,
+      "learning_rate": 7.298083828530708e-06,
+      "loss": 0.3119,
+      "step": 5845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.625330661974828,
+      "learning_rate": 7.297179254198202e-06,
+      "loss": 0.3035,
+      "step": 5846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5264940611890767,
+      "learning_rate": 7.296274584551647e-06,
+      "loss": 0.3146,
+      "step": 5847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.45465879872521,
+      "learning_rate": 7.295369819628582e-06,
+      "loss": 0.317,
+      "step": 5848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.891544266720139,
+      "learning_rate": 7.294464959466545e-06,
+      "loss": 0.289,
+      "step": 5849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0262458195071824,
+      "learning_rate": 7.2935600041030815e-06,
+      "loss": 0.3021,
+      "step": 5850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5013778140899698,
+      "learning_rate": 7.292654953575739e-06,
+      "loss": 0.2926,
+      "step": 5851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0427202461352336,
+      "learning_rate": 7.291749807922072e-06,
+      "loss": 0.308,
+      "step": 5852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7416220802159856,
+      "learning_rate": 7.290844567179635e-06,
+      "loss": 0.2939,
+      "step": 5853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8268191816210502,
+      "learning_rate": 7.289939231385987e-06,
+      "loss": 0.3055,
+      "step": 5854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5882337922854566,
+      "learning_rate": 7.289033800578692e-06,
+      "loss": 0.3254,
+      "step": 5855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.622749871598468,
+      "learning_rate": 7.2881282747953195e-06,
+      "loss": 0.3143,
+      "step": 5856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.408186670504864,
+      "learning_rate": 7.287222654073442e-06,
+      "loss": 0.2995,
+      "step": 5857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.733951039006919,
+      "learning_rate": 7.286316938450633e-06,
+      "loss": 0.3032,
+      "step": 5858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.96746684889688,
+      "learning_rate": 7.285411127964472e-06,
+      "loss": 0.2968,
+      "step": 5859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7309144654104038,
+      "learning_rate": 7.284505222652546e-06,
+      "loss": 0.2935,
+      "step": 5860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3396226499138895,
+      "learning_rate": 7.28359922255244e-06,
+      "loss": 0.2992,
+      "step": 5861
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3488703823264145,
+      "learning_rate": 7.2826931277017455e-06,
+      "loss": 0.3092,
+      "step": 5862
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7107288739922248,
+      "learning_rate": 7.281786938138058e-06,
+      "loss": 0.2875,
+      "step": 5863
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.714898213298103,
+      "learning_rate": 7.28088065389898e-06,
+      "loss": 0.311,
+      "step": 5864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.156603177681332,
+      "learning_rate": 7.279974275022111e-06,
+      "loss": 0.3169,
+      "step": 5865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0778148806959416,
+      "learning_rate": 7.279067801545059e-06,
+      "loss": 0.292,
+      "step": 5866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1673113087277693,
+      "learning_rate": 7.278161233505435e-06,
+      "loss": 0.2949,
+      "step": 5867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5382648247895252,
+      "learning_rate": 7.277254570940857e-06,
+      "loss": 0.3342,
+      "step": 5868
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.723274882034758,
+      "learning_rate": 7.276347813888942e-06,
+      "loss": 0.3053,
+      "step": 5869
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4908694377216396,
+      "learning_rate": 7.275440962387311e-06,
+      "loss": 0.294,
+      "step": 5870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.122961585526543,
+      "learning_rate": 7.274534016473595e-06,
+      "loss": 0.2908,
+      "step": 5871
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5943014249013667,
+      "learning_rate": 7.273626976185422e-06,
+      "loss": 0.3053,
+      "step": 5872
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8921446662808121,
+      "learning_rate": 7.272719841560426e-06,
+      "loss": 0.3093,
+      "step": 5873
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.442207526439503,
+      "learning_rate": 7.271812612636249e-06,
+      "loss": 0.3018,
+      "step": 5874
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.687574785421414,
+      "learning_rate": 7.270905289450529e-06,
+      "loss": 0.3031,
+      "step": 5875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3049954404147388,
+      "learning_rate": 7.269997872040918e-06,
+      "loss": 0.2979,
+      "step": 5876
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.494379088008921,
+      "learning_rate": 7.26909036044506e-06,
+      "loss": 0.3144,
+      "step": 5877
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4272037481827549,
+      "learning_rate": 7.268182754700616e-06,
+      "loss": 0.2785,
+      "step": 5878
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5092362448586174,
+      "learning_rate": 7.267275054845238e-06,
+      "loss": 0.3021,
+      "step": 5879
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.204995359379711,
+      "learning_rate": 7.266367260916594e-06,
+      "loss": 0.3205,
+      "step": 5880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.67255212483285,
+      "learning_rate": 7.265459372952343e-06,
+      "loss": 0.3152,
+      "step": 5881
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5525485891146222,
+      "learning_rate": 7.264551390990161e-06,
+      "loss": 0.2957,
+      "step": 5882
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7766000510995803,
+      "learning_rate": 7.2636433150677185e-06,
+      "loss": 0.3206,
+      "step": 5883
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3475064172629985,
+      "learning_rate": 7.262735145222696e-06,
+      "loss": 0.316,
+      "step": 5884
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.233693133714042,
+      "learning_rate": 7.261826881492771e-06,
+      "loss": 0.3069,
+      "step": 5885
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.422949658024172,
+      "learning_rate": 7.260918523915632e-06,
+      "loss": 0.3156,
+      "step": 5886
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.485762226091767,
+      "learning_rate": 7.260010072528968e-06,
+      "loss": 0.2902,
+      "step": 5887
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.882394089851754,
+      "learning_rate": 7.259101527370471e-06,
+      "loss": 0.3042,
+      "step": 5888
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5545202236983964,
+      "learning_rate": 7.25819288847784e-06,
+      "loss": 0.3175,
+      "step": 5889
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1366015587414267,
+      "learning_rate": 7.257284155888775e-06,
+      "loss": 0.328,
+      "step": 5890
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1565696032774317,
+      "learning_rate": 7.25637532964098e-06,
+      "loss": 0.3377,
+      "step": 5891
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.005323489987182,
+      "learning_rate": 7.255466409772165e-06,
+      "loss": 0.3136,
+      "step": 5892
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.50084749243097,
+      "learning_rate": 7.254557396320043e-06,
+      "loss": 0.3229,
+      "step": 5893
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0641706062751233,
+      "learning_rate": 7.253648289322331e-06,
+      "loss": 0.3279,
+      "step": 5894
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.578493390770236,
+      "learning_rate": 7.252739088816747e-06,
+      "loss": 0.2983,
+      "step": 5895
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0699252430865474,
+      "learning_rate": 7.251829794841017e-06,
+      "loss": 0.2977,
+      "step": 5896
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4448905027082404,
+      "learning_rate": 7.25092040743287e-06,
+      "loss": 0.3252,
+      "step": 5897
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6582162666588753,
+      "learning_rate": 7.250010926630038e-06,
+      "loss": 0.3047,
+      "step": 5898
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.029785768214059,
+      "learning_rate": 7.2491013524702545e-06,
+      "loss": 0.3072,
+      "step": 5899
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9210438445093962,
+      "learning_rate": 7.248191684991262e-06,
+      "loss": 0.3059,
+      "step": 5900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5119819916754642,
+      "learning_rate": 7.247281924230802e-06,
+      "loss": 0.3119,
+      "step": 5901
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.348791307836979,
+      "learning_rate": 7.246372070226625e-06,
+      "loss": 0.3045,
+      "step": 5902
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.462685294050548,
+      "learning_rate": 7.245462123016478e-06,
+      "loss": 0.3037,
+      "step": 5903
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.657400086378503,
+      "learning_rate": 7.244552082638122e-06,
+      "loss": 0.3045,
+      "step": 5904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3256327447015273,
+      "learning_rate": 7.243641949129312e-06,
+      "loss": 0.3082,
+      "step": 5905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6509889104987481,
+      "learning_rate": 7.242731722527814e-06,
+      "loss": 0.4944,
+      "step": 5906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5261226749791494,
+      "learning_rate": 7.24182140287139e-06,
+      "loss": 0.3184,
+      "step": 5907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2445305914520484,
+      "learning_rate": 7.2409109901978185e-06,
+      "loss": 0.2872,
+      "step": 5908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7015555458004097,
+      "learning_rate": 7.240000484544866e-06,
+      "loss": 0.3241,
+      "step": 5909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3164805378662296,
+      "learning_rate": 7.239089885950317e-06,
+      "loss": 0.3069,
+      "step": 5910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7882030170435197,
+      "learning_rate": 7.238179194451949e-06,
+      "loss": 0.3103,
+      "step": 5911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2328517241237735,
+      "learning_rate": 7.237268410087553e-06,
+      "loss": 0.3195,
+      "step": 5912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5985679519054314,
+      "learning_rate": 7.236357532894916e-06,
+      "loss": 0.5348,
+      "step": 5913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7307987927378585,
+      "learning_rate": 7.235446562911834e-06,
+      "loss": 0.3005,
+      "step": 5914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8246946174040426,
+      "learning_rate": 7.234535500176101e-06,
+      "loss": 0.3063,
+      "step": 5915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6524543979794832,
+      "learning_rate": 7.233624344725524e-06,
+      "loss": 0.3209,
+      "step": 5916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3908195391034117,
+      "learning_rate": 7.232713096597903e-06,
+      "loss": 0.2883,
+      "step": 5917
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2579837209121814,
+      "learning_rate": 7.231801755831052e-06,
+      "loss": 0.3053,
+      "step": 5918
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.698560050406029,
+      "learning_rate": 7.230890322462781e-06,
+      "loss": 0.2831,
+      "step": 5919
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1326315798504343,
+      "learning_rate": 7.2299787965309075e-06,
+      "loss": 0.3413,
+      "step": 5920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6627139754391318,
+      "learning_rate": 7.229067178073253e-06,
+      "loss": 0.2878,
+      "step": 5921
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7135080849189828,
+      "learning_rate": 7.228155467127642e-06,
+      "loss": 0.3277,
+      "step": 5922
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.844747474669683,
+      "learning_rate": 7.227243663731904e-06,
+      "loss": 0.3259,
+      "step": 5923
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8893996693073825,
+      "learning_rate": 7.226331767923871e-06,
+      "loss": 0.3204,
+      "step": 5924
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9262178045103964,
+      "learning_rate": 7.225419779741376e-06,
+      "loss": 0.3036,
+      "step": 5925
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0445039946056185,
+      "learning_rate": 7.224507699222263e-06,
+      "loss": 0.3105,
+      "step": 5926
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6887161344967128,
+      "learning_rate": 7.223595526404374e-06,
+      "loss": 0.3332,
+      "step": 5927
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.096963681172321,
+      "learning_rate": 7.2226832613255584e-06,
+      "loss": 0.3196,
+      "step": 5928
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.771536720348542,
+      "learning_rate": 7.221770904023664e-06,
+      "loss": 0.3128,
+      "step": 5929
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.399124860506791,
+      "learning_rate": 7.2208584545365505e-06,
+      "loss": 0.3207,
+      "step": 5930
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.545844740184038,
+      "learning_rate": 7.219945912902073e-06,
+      "loss": 0.2978,
+      "step": 5931
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4043635595207227,
+      "learning_rate": 7.2190332791580995e-06,
+      "loss": 0.325,
+      "step": 5932
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.540370324210669,
+      "learning_rate": 7.218120553342492e-06,
+      "loss": 0.3149,
+      "step": 5933
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.080938864836028,
+      "learning_rate": 7.217207735493122e-06,
+      "loss": 0.2886,
+      "step": 5934
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3301992556399873,
+      "learning_rate": 7.216294825647866e-06,
+      "loss": 0.3128,
+      "step": 5935
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4832123564266622,
+      "learning_rate": 7.215381823844601e-06,
+      "loss": 0.2998,
+      "step": 5936
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.960706285117124,
+      "learning_rate": 7.214468730121209e-06,
+      "loss": 0.301,
+      "step": 5937
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.255026371704293,
+      "learning_rate": 7.213555544515577e-06,
+      "loss": 0.3036,
+      "step": 5938
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4455784418932116,
+      "learning_rate": 7.212642267065593e-06,
+      "loss": 0.3153,
+      "step": 5939
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6126928210379949,
+      "learning_rate": 7.211728897809151e-06,
+      "loss": 0.4656,
+      "step": 5940
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9074410032665503,
+      "learning_rate": 7.210815436784148e-06,
+      "loss": 0.2978,
+      "step": 5941
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1222972203089223,
+      "learning_rate": 7.209901884028487e-06,
+      "loss": 0.2926,
+      "step": 5942
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0324798593508495,
+      "learning_rate": 7.20898823958007e-06,
+      "loss": 0.2851,
+      "step": 5943
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8127984813845401,
+      "learning_rate": 7.208074503476808e-06,
+      "loss": 0.3103,
+      "step": 5944
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6075085901736177,
+      "learning_rate": 7.207160675756614e-06,
+      "loss": 0.2974,
+      "step": 5945
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2821636272598704,
+      "learning_rate": 7.206246756457402e-06,
+      "loss": 0.3086,
+      "step": 5946
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.726983264962258,
+      "learning_rate": 7.205332745617095e-06,
+      "loss": 0.3128,
+      "step": 5947
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9714385902606477,
+      "learning_rate": 7.204418643273613e-06,
+      "loss": 0.3057,
+      "step": 5948
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.710025784138471,
+      "learning_rate": 7.2035044494648865e-06,
+      "loss": 0.3012,
+      "step": 5949
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5061638534445656,
+      "learning_rate": 7.202590164228849e-06,
+      "loss": 0.3223,
+      "step": 5950
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.99989521694326,
+      "learning_rate": 7.20167578760343e-06,
+      "loss": 0.3166,
+      "step": 5951
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.156413550352374,
+      "learning_rate": 7.200761319626574e-06,
+      "loss": 0.3094,
+      "step": 5952
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4174844838680327,
+      "learning_rate": 7.199846760336221e-06,
+      "loss": 0.2987,
+      "step": 5953
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6831474095576104,
+      "learning_rate": 7.198932109770319e-06,
+      "loss": 0.2978,
+      "step": 5954
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.494215493871055,
+      "learning_rate": 7.198017367966817e-06,
+      "loss": 0.3126,
+      "step": 5955
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.3713637925407065,
+      "learning_rate": 7.197102534963671e-06,
+      "loss": 0.3164,
+      "step": 5956
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1745249570043237,
+      "learning_rate": 7.19618761079884e-06,
+      "loss": 0.3085,
+      "step": 5957
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6721736852571504,
+      "learning_rate": 7.195272595510282e-06,
+      "loss": 0.5019,
+      "step": 5958
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.759645629800595,
+      "learning_rate": 7.194357489135966e-06,
+      "loss": 0.3258,
+      "step": 5959
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1010763393118905,
+      "learning_rate": 7.193442291713858e-06,
+      "loss": 0.3065,
+      "step": 5960
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.523171481090262,
+      "learning_rate": 7.192527003281935e-06,
+      "loss": 0.3185,
+      "step": 5961
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8455391431495372,
+      "learning_rate": 7.191611623878173e-06,
+      "loss": 0.2899,
+      "step": 5962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.526983148103827,
+      "learning_rate": 7.1906961535405505e-06,
+      "loss": 0.3059,
+      "step": 5963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.225402680218895,
+      "learning_rate": 7.189780592307054e-06,
+      "loss": 0.3093,
+      "step": 5964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5167138065359553,
+      "learning_rate": 7.188864940215671e-06,
+      "loss": 0.2865,
+      "step": 5965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.356918859670861,
+      "learning_rate": 7.187949197304395e-06,
+      "loss": 0.3218,
+      "step": 5966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.89189404408381,
+      "learning_rate": 7.187033363611219e-06,
+      "loss": 0.318,
+      "step": 5967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.11700703467767,
+      "learning_rate": 7.186117439174145e-06,
+      "loss": 0.3133,
+      "step": 5968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.005995948321859,
+      "learning_rate": 7.185201424031174e-06,
+      "loss": 0.3221,
+      "step": 5969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.470314521155802,
+      "learning_rate": 7.184285318220316e-06,
+      "loss": 0.2872,
+      "step": 5970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.751373484900817,
+      "learning_rate": 7.18336912177958e-06,
+      "loss": 0.3091,
+      "step": 5971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0781543745306235,
+      "learning_rate": 7.182452834746982e-06,
+      "loss": 0.3037,
+      "step": 5972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7377097311702716,
+      "learning_rate": 7.181536457160538e-06,
+      "loss": 0.2901,
+      "step": 5973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.736701016666469,
+      "learning_rate": 7.180619989058273e-06,
+      "loss": 0.3156,
+      "step": 5974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3994772603359749,
+      "learning_rate": 7.17970343047821e-06,
+      "loss": 0.2894,
+      "step": 5975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8318057658099611,
+      "learning_rate": 7.178786781458381e-06,
+      "loss": 0.3041,
+      "step": 5976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.227806643026452,
+      "learning_rate": 7.177870042036819e-06,
+      "loss": 0.2914,
+      "step": 5977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.697526425726963,
+      "learning_rate": 7.176953212251559e-06,
+      "loss": 0.3074,
+      "step": 5978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.038781239013182,
+      "learning_rate": 7.176036292140644e-06,
+      "loss": 0.3121,
+      "step": 5979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1796377758250554,
+      "learning_rate": 7.175119281742119e-06,
+      "loss": 0.3029,
+      "step": 5980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.080976565060239,
+      "learning_rate": 7.174202181094031e-06,
+      "loss": 0.3286,
+      "step": 5981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.309657195172605,
+      "learning_rate": 7.173284990234433e-06,
+      "loss": 0.2914,
+      "step": 5982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6046106125210732,
+      "learning_rate": 7.17236770920138e-06,
+      "loss": 0.4895,
+      "step": 5983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.87831532572056,
+      "learning_rate": 7.1714503380329326e-06,
+      "loss": 0.3074,
+      "step": 5984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.794917558533384,
+      "learning_rate": 7.170532876767153e-06,
+      "loss": 0.3168,
+      "step": 5985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.461320850332969,
+      "learning_rate": 7.16961532544211e-06,
+      "loss": 0.3167,
+      "step": 5986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3543175351221206,
+      "learning_rate": 7.168697684095873e-06,
+      "loss": 0.3126,
+      "step": 5987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4638785596026485,
+      "learning_rate": 7.1677799527665186e-06,
+      "loss": 0.3172,
+      "step": 5988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.10295359690129,
+      "learning_rate": 7.166862131492122e-06,
+      "loss": 0.3179,
+      "step": 5989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.904067089162369,
+      "learning_rate": 7.165944220310766e-06,
+      "loss": 0.2811,
+      "step": 5990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2090260665197143,
+      "learning_rate": 7.165026219260538e-06,
+      "loss": 0.3006,
+      "step": 5991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5909880695299142,
+      "learning_rate": 7.164108128379528e-06,
+      "loss": 0.5293,
+      "step": 5992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5514547504538685,
+      "learning_rate": 7.163189947705826e-06,
+      "loss": 0.3085,
+      "step": 5993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5225524480726405,
+      "learning_rate": 7.162271677277532e-06,
+      "loss": 0.2955,
+      "step": 5994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2985759849943497,
+      "learning_rate": 7.161353317132744e-06,
+      "loss": 0.3193,
+      "step": 5995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9162294349107698,
+      "learning_rate": 7.160434867309569e-06,
+      "loss": 0.3037,
+      "step": 5996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4891738981892237,
+      "learning_rate": 7.159516327846114e-06,
+      "loss": 0.3105,
+      "step": 5997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0724693220140935,
+      "learning_rate": 7.1585976987804895e-06,
+      "loss": 0.3097,
+      "step": 5998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.259566667810883,
+      "learning_rate": 7.157678980150814e-06,
+      "loss": 0.3068,
+      "step": 5999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6444077881578327,
+      "learning_rate": 7.156760171995204e-06,
+      "loss": 0.5023,
+      "step": 6000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7271831934927013,
+      "learning_rate": 7.155841274351784e-06,
+      "loss": 0.3108,
+      "step": 6001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.6636631154283315,
+      "learning_rate": 7.154922287258681e-06,
+      "loss": 0.2844,
+      "step": 6002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8343864224279804,
+      "learning_rate": 7.1540032107540245e-06,
+      "loss": 0.3197,
+      "step": 6003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.848427973087907,
+      "learning_rate": 7.1530840448759484e-06,
+      "loss": 0.3049,
+      "step": 6004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8418480083532276,
+      "learning_rate": 7.152164789662592e-06,
+      "loss": 0.3111,
+      "step": 6005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7151396141888278,
+      "learning_rate": 7.151245445152096e-06,
+      "loss": 0.2947,
+      "step": 6006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6177711619734679,
+      "learning_rate": 7.1503260113826035e-06,
+      "loss": 0.484,
+      "step": 6007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.707515386143622,
+      "learning_rate": 7.1494064883922655e-06,
+      "loss": 0.2996,
+      "step": 6008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4864084445145718,
+      "learning_rate": 7.148486876219235e-06,
+      "loss": 0.2861,
+      "step": 6009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7944702242957387,
+      "learning_rate": 7.14756717490167e-06,
+      "loss": 0.3191,
+      "step": 6010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9664301090511955,
+      "learning_rate": 7.146647384477725e-06,
+      "loss": 0.3012,
+      "step": 6011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.66737973962642,
+      "learning_rate": 7.145727504985569e-06,
+      "loss": 0.2933,
+      "step": 6012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.647342014756224,
+      "learning_rate": 7.144807536463368e-06,
+      "loss": 0.2938,
+      "step": 6013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.518720756555653,
+      "learning_rate": 7.143887478949292e-06,
+      "loss": 0.3153,
+      "step": 6014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7140269567116875,
+      "learning_rate": 7.142967332481516e-06,
+      "loss": 0.3074,
+      "step": 6015
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9129138897222344,
+      "learning_rate": 7.142047097098219e-06,
+      "loss": 0.3017,
+      "step": 6016
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.826972517723574,
+      "learning_rate": 7.1411267728375845e-06,
+      "loss": 0.3117,
+      "step": 6017
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.980931301259122,
+      "learning_rate": 7.140206359737797e-06,
+      "loss": 0.2969,
+      "step": 6018
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2998712903592073,
+      "learning_rate": 7.139285857837046e-06,
+      "loss": 0.3169,
+      "step": 6019
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.756500055755841,
+      "learning_rate": 7.138365267173524e-06,
+      "loss": 0.31,
+      "step": 6020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2435415782378745,
+      "learning_rate": 7.13744458778543e-06,
+      "loss": 0.2979,
+      "step": 6021
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.228875636301696,
+      "learning_rate": 7.136523819710963e-06,
+      "loss": 0.3237,
+      "step": 6022
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9203626706075438,
+      "learning_rate": 7.135602962988327e-06,
+      "loss": 0.3089,
+      "step": 6023
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2807773580502912,
+      "learning_rate": 7.134682017655732e-06,
+      "loss": 0.2801,
+      "step": 6024
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.883383447407521,
+      "learning_rate": 7.1337609837513875e-06,
+      "loss": 0.3036,
+      "step": 6025
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9989008456799338,
+      "learning_rate": 7.132839861313511e-06,
+      "loss": 0.2943,
+      "step": 6026
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7139691608818124,
+      "learning_rate": 7.13191865038032e-06,
+      "loss": 0.3214,
+      "step": 6027
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.205294008880047,
+      "learning_rate": 7.130997350990037e-06,
+      "loss": 0.3241,
+      "step": 6028
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8248584968970416,
+      "learning_rate": 7.130075963180889e-06,
+      "loss": 0.3089,
+      "step": 6029
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7235284044945178,
+      "learning_rate": 7.129154486991105e-06,
+      "loss": 0.2949,
+      "step": 6030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.491676776637185,
+      "learning_rate": 7.128232922458922e-06,
+      "loss": 0.2964,
+      "step": 6031
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.141666058168301,
+      "learning_rate": 7.127311269622573e-06,
+      "loss": 0.2879,
+      "step": 6032
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.352111095265622,
+      "learning_rate": 7.126389528520301e-06,
+      "loss": 0.3072,
+      "step": 6033
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7030797976818755,
+      "learning_rate": 7.125467699190351e-06,
+      "loss": 0.3125,
+      "step": 6034
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6068189899876875,
+      "learning_rate": 7.1245457816709705e-06,
+      "loss": 0.2837,
+      "step": 6035
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6235180671792958,
+      "learning_rate": 7.123623776000412e-06,
+      "loss": 0.4872,
+      "step": 6036
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5349126519788254,
+      "learning_rate": 7.1227016822169315e-06,
+      "loss": 0.3251,
+      "step": 6037
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.72208190723723,
+      "learning_rate": 7.121779500358788e-06,
+      "loss": 0.2987,
+      "step": 6038
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.946054516048329,
+      "learning_rate": 7.120857230464244e-06,
+      "loss": 0.3218,
+      "step": 6039
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9674156888459624,
+      "learning_rate": 7.119934872571566e-06,
+      "loss": 0.2882,
+      "step": 6040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4279407274139766,
+      "learning_rate": 7.119012426719024e-06,
+      "loss": 0.3161,
+      "step": 6041
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.351650021155563,
+      "learning_rate": 7.118089892944894e-06,
+      "loss": 0.3239,
+      "step": 6042
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0681820473331536,
+      "learning_rate": 7.117167271287453e-06,
+      "loss": 0.2923,
+      "step": 6043
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.118400686194755,
+      "learning_rate": 7.116244561784979e-06,
+      "loss": 0.3165,
+      "step": 6044
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6274338441549956,
+      "learning_rate": 7.11532176447576e-06,
+      "loss": 0.3317,
+      "step": 6045
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8604762163536679,
+      "learning_rate": 7.114398879398084e-06,
+      "loss": 0.296,
+      "step": 6046
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4985839502161586,
+      "learning_rate": 7.113475906590243e-06,
+      "loss": 0.3202,
+      "step": 6047
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.256378121651349,
+      "learning_rate": 7.112552846090533e-06,
+      "loss": 0.3179,
+      "step": 6048
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.719310780296274,
+      "learning_rate": 7.111629697937253e-06,
+      "loss": 0.3108,
+      "step": 6049
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0849549900381477,
+      "learning_rate": 7.110706462168706e-06,
+      "loss": 0.3093,
+      "step": 6050
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0654583039100354,
+      "learning_rate": 7.109783138823199e-06,
+      "loss": 0.3036,
+      "step": 6051
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3209794508592054,
+      "learning_rate": 7.108859727939042e-06,
+      "loss": 0.3176,
+      "step": 6052
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.340267151334955,
+      "learning_rate": 7.107936229554549e-06,
+      "loss": 0.297,
+      "step": 6053
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9390822601459914,
+      "learning_rate": 7.107012643708039e-06,
+      "loss": 0.2924,
+      "step": 6054
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.77185725744881,
+      "learning_rate": 7.10608897043783e-06,
+      "loss": 0.2911,
+      "step": 6055
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7767442966723028,
+      "learning_rate": 7.10516520978225e-06,
+      "loss": 0.3039,
+      "step": 6056
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.794659391202409,
+      "learning_rate": 7.104241361779627e-06,
+      "loss": 0.2888,
+      "step": 6057
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.4730653608893003,
+      "learning_rate": 7.10331742646829e-06,
+      "loss": 0.3009,
+      "step": 6058
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.28545990046778,
+      "learning_rate": 7.102393403886578e-06,
+      "loss": 0.3212,
+      "step": 6059
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.170668734452948,
+      "learning_rate": 7.101469294072829e-06,
+      "loss": 0.3019,
+      "step": 6060
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.303197692180006,
+      "learning_rate": 7.100545097065389e-06,
+      "loss": 0.2969,
+      "step": 6061
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.766199437062035,
+      "learning_rate": 7.099620812902599e-06,
+      "loss": 0.2956,
+      "step": 6062
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.234164688795876,
+      "learning_rate": 7.098696441622814e-06,
+      "loss": 0.2914,
+      "step": 6063
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.962299086533189,
+      "learning_rate": 7.097771983264384e-06,
+      "loss": 0.3295,
+      "step": 6064
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.501685559116226,
+      "learning_rate": 7.096847437865671e-06,
+      "loss": 0.2999,
+      "step": 6065
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0932316992331916,
+      "learning_rate": 7.095922805465031e-06,
+      "loss": 0.3018,
+      "step": 6066
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.123320078935363,
+      "learning_rate": 7.0949980861008315e-06,
+      "loss": 0.2807,
+      "step": 6067
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0151895169381566,
+      "learning_rate": 7.0940732798114395e-06,
+      "loss": 0.3129,
+      "step": 6068
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5562491480090626,
+      "learning_rate": 7.0931483866352305e-06,
+      "loss": 0.3018,
+      "step": 6069
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2349881269372105,
+      "learning_rate": 7.092223406610574e-06,
+      "loss": 0.3057,
+      "step": 6070
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8705404189940347,
+      "learning_rate": 7.091298339775854e-06,
+      "loss": 0.3209,
+      "step": 6071
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9862349323270718,
+      "learning_rate": 7.0903731861694505e-06,
+      "loss": 0.304,
+      "step": 6072
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.685636385013489,
+      "learning_rate": 7.089447945829752e-06,
+      "loss": 0.2902,
+      "step": 6073
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.142857306447189,
+      "learning_rate": 7.088522618795145e-06,
+      "loss": 0.2931,
+      "step": 6074
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3857036789661261,
+      "learning_rate": 7.087597205104026e-06,
+      "loss": 0.3033,
+      "step": 6075
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.160089716025319,
+      "learning_rate": 7.08667170479479e-06,
+      "loss": 0.3069,
+      "step": 6076
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1547547172595234,
+      "learning_rate": 7.085746117905841e-06,
+      "loss": 0.2862,
+      "step": 6077
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.707743010491154,
+      "learning_rate": 7.084820444475579e-06,
+      "loss": 0.3067,
+      "step": 6078
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.786143828091818,
+      "learning_rate": 7.083894684542413e-06,
+      "loss": 0.3,
+      "step": 6079
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.907577231597122,
+      "learning_rate": 7.082968838144756e-06,
+      "loss": 0.3341,
+      "step": 6080
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2263893472766974,
+      "learning_rate": 7.082042905321022e-06,
+      "loss": 0.3041,
+      "step": 6081
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.33554888014138,
+      "learning_rate": 7.081116886109629e-06,
+      "loss": 0.3103,
+      "step": 6082
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8114787057696915,
+      "learning_rate": 7.080190780549002e-06,
+      "loss": 0.3229,
+      "step": 6083
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9649870410684962,
+      "learning_rate": 7.079264588677564e-06,
+      "loss": 0.3382,
+      "step": 6084
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.799504074440034,
+      "learning_rate": 7.078338310533744e-06,
+      "loss": 0.3073,
+      "step": 6085
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6390799041427644,
+      "learning_rate": 7.077411946155975e-06,
+      "loss": 0.3015,
+      "step": 6086
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.308849381739574,
+      "learning_rate": 7.076485495582696e-06,
+      "loss": 0.2826,
+      "step": 6087
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7807209331700014,
+      "learning_rate": 7.0755589588523464e-06,
+      "loss": 0.3127,
+      "step": 6088
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.062526210664507,
+      "learning_rate": 7.074632336003368e-06,
+      "loss": 0.3222,
+      "step": 6089
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8211249363935218,
+      "learning_rate": 7.0737056270742085e-06,
+      "loss": 0.2989,
+      "step": 6090
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8539341258880238,
+      "learning_rate": 7.072778832103321e-06,
+      "loss": 0.3089,
+      "step": 6091
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.903834652848273,
+      "learning_rate": 7.071851951129156e-06,
+      "loss": 0.3173,
+      "step": 6092
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2189344812108156,
+      "learning_rate": 7.070924984190175e-06,
+      "loss": 0.2947,
+      "step": 6093
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6056192596870574,
+      "learning_rate": 7.069997931324837e-06,
+      "loss": 0.5081,
+      "step": 6094
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.701728321342455,
+      "learning_rate": 7.069070792571608e-06,
+      "loss": 0.3021,
+      "step": 6095
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3949983094981673,
+      "learning_rate": 7.068143567968958e-06,
+      "loss": 0.3015,
+      "step": 6096
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2379630527545697,
+      "learning_rate": 7.067216257555357e-06,
+      "loss": 0.3115,
+      "step": 6097
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5775514592063432,
+      "learning_rate": 7.0662888613692815e-06,
+      "loss": 0.4892,
+      "step": 6098
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.39311315109939,
+      "learning_rate": 7.065361379449213e-06,
+      "loss": 0.3138,
+      "step": 6099
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2324433257890504,
+      "learning_rate": 7.06443381183363e-06,
+      "loss": 0.2861,
+      "step": 6100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8595551351518653,
+      "learning_rate": 7.063506158561022e-06,
+      "loss": 0.2958,
+      "step": 6101
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.989934325758585,
+      "learning_rate": 7.062578419669877e-06,
+      "loss": 0.2935,
+      "step": 6102
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4691404106150727,
+      "learning_rate": 7.061650595198692e-06,
+      "loss": 0.3079,
+      "step": 6103
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7616358926720772,
+      "learning_rate": 7.060722685185961e-06,
+      "loss": 0.2921,
+      "step": 6104
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.8235942004169186,
+      "learning_rate": 7.0597946896701854e-06,
+      "loss": 0.2923,
+      "step": 6105
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6410184610210388,
+      "learning_rate": 7.05886660868987e-06,
+      "loss": 0.2969,
+      "step": 6106
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6815154076557426,
+      "learning_rate": 7.057938442283523e-06,
+      "loss": 0.2984,
+      "step": 6107
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.720821468920267,
+      "learning_rate": 7.057010190489651e-06,
+      "loss": 0.289,
+      "step": 6108
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6209093699482704,
+      "learning_rate": 7.056081853346776e-06,
+      "loss": 0.2944,
+      "step": 6109
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6758360820245468,
+      "learning_rate": 7.055153430893412e-06,
+      "loss": 0.4849,
+      "step": 6110
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8280460061261037,
+      "learning_rate": 7.054224923168083e-06,
+      "loss": 0.3002,
+      "step": 6111
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9719266559597206,
+      "learning_rate": 7.053296330209309e-06,
+      "loss": 0.3061,
+      "step": 6112
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6649698942428275,
+      "learning_rate": 7.052367652055628e-06,
+      "loss": 0.3133,
+      "step": 6113
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.329141100603012,
+      "learning_rate": 7.051438888745566e-06,
+      "loss": 0.322,
+      "step": 6114
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6134970792257957,
+      "learning_rate": 7.050510040317661e-06,
+      "loss": 0.3054,
+      "step": 6115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.600146444124667,
+      "learning_rate": 7.0495811068104505e-06,
+      "loss": 0.3033,
+      "step": 6116
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.484537435259375,
+      "learning_rate": 7.048652088262481e-06,
+      "loss": 0.3137,
+      "step": 6117
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.037218912035509,
+      "learning_rate": 7.047722984712298e-06,
+      "loss": 0.3306,
+      "step": 6118
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.391179915946461,
+      "learning_rate": 7.0467937961984505e-06,
+      "loss": 0.2994,
+      "step": 6119
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.731809271949276,
+      "learning_rate": 7.045864522759492e-06,
+      "loss": 0.3172,
+      "step": 6120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8965117256482238,
+      "learning_rate": 7.044935164433982e-06,
+      "loss": 0.2978,
+      "step": 6121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.130368883395997,
+      "learning_rate": 7.0440057212604784e-06,
+      "loss": 0.3091,
+      "step": 6122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6828950671125487,
+      "learning_rate": 7.043076193277548e-06,
+      "loss": 0.2963,
+      "step": 6123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.812860545122394,
+      "learning_rate": 7.042146580523757e-06,
+      "loss": 0.2947,
+      "step": 6124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2461511729264543,
+      "learning_rate": 7.0412168830376785e-06,
+      "loss": 0.2904,
+      "step": 6125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.780148370156523,
+      "learning_rate": 7.040287100857885e-06,
+      "loss": 0.3078,
+      "step": 6126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.443994902112538,
+      "learning_rate": 7.039357234022954e-06,
+      "loss": 0.2968,
+      "step": 6127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.738888871812275,
+      "learning_rate": 7.03842728257147e-06,
+      "loss": 0.2973,
+      "step": 6128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7900939277729429,
+      "learning_rate": 7.037497246542018e-06,
+      "loss": 0.3074,
+      "step": 6129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.45456775526933,
+      "learning_rate": 7.036567125973187e-06,
+      "loss": 0.2964,
+      "step": 6130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7554449753870167,
+      "learning_rate": 7.035636920903568e-06,
+      "loss": 0.304,
+      "step": 6131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5027604018520393,
+      "learning_rate": 7.034706631371756e-06,
+      "loss": 0.3056,
+      "step": 6132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6697743466379795,
+      "learning_rate": 7.033776257416354e-06,
+      "loss": 0.3144,
+      "step": 6133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.565734714733023,
+      "learning_rate": 7.03284579907596e-06,
+      "loss": 0.2951,
+      "step": 6134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1532350004877565,
+      "learning_rate": 7.031915256389186e-06,
+      "loss": 0.3155,
+      "step": 6135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8870110224742196,
+      "learning_rate": 7.030984629394637e-06,
+      "loss": 0.3057,
+      "step": 6136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.501300650823364,
+      "learning_rate": 7.03005391813093e-06,
+      "loss": 0.3081,
+      "step": 6137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4477855092138006,
+      "learning_rate": 7.029123122636678e-06,
+      "loss": 0.3133,
+      "step": 6138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3984298105624784,
+      "learning_rate": 7.028192242950506e-06,
+      "loss": 0.3016,
+      "step": 6139
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.2803001292669896,
+      "learning_rate": 7.027261279111033e-06,
+      "loss": 0.3061,
+      "step": 6140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 42.62180019802184,
+      "learning_rate": 7.02633023115689e-06,
+      "loss": 0.3123,
+      "step": 6141
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8786333964794473,
+      "learning_rate": 7.025399099126705e-06,
+      "loss": 0.2947,
+      "step": 6142
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6413535434928973,
+      "learning_rate": 7.024467883059116e-06,
+      "loss": 0.5041,
+      "step": 6143
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.591790878851951,
+      "learning_rate": 7.023536582992758e-06,
+      "loss": 0.2987,
+      "step": 6144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.416299671075495,
+      "learning_rate": 7.022605198966274e-06,
+      "loss": 0.2956,
+      "step": 6145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.183675660620761,
+      "learning_rate": 7.021673731018305e-06,
+      "loss": 0.2978,
+      "step": 6146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2303023644855005,
+      "learning_rate": 7.0207421791875045e-06,
+      "loss": 0.3042,
+      "step": 6147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1889082267322033,
+      "learning_rate": 7.01981054351252e-06,
+      "loss": 0.2879,
+      "step": 6148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.339385146404882,
+      "learning_rate": 7.0188788240320095e-06,
+      "loss": 0.3128,
+      "step": 6149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9486207992351334,
+      "learning_rate": 7.017947020784629e-06,
+      "loss": 0.307,
+      "step": 6150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.696794029026286,
+      "learning_rate": 7.017015133809044e-06,
+      "loss": 0.2968,
+      "step": 6151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0364165147095457,
+      "learning_rate": 7.016083163143918e-06,
+      "loss": 0.2984,
+      "step": 6152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7201399926681409,
+      "learning_rate": 7.015151108827921e-06,
+      "loss": 0.2891,
+      "step": 6153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2876844560550476,
+      "learning_rate": 7.014218970899724e-06,
+      "loss": 0.3287,
+      "step": 6154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5222707925705408,
+      "learning_rate": 7.013286749398008e-06,
+      "loss": 0.3477,
+      "step": 6155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.754050839924,
+      "learning_rate": 7.0123544443614445e-06,
+      "loss": 0.2759,
+      "step": 6156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2849969363113685,
+      "learning_rate": 7.011422055828721e-06,
+      "loss": 0.3232,
+      "step": 6157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9947993971964004,
+      "learning_rate": 7.010489583838525e-06,
+      "loss": 0.3348,
+      "step": 6158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0603287475019276,
+      "learning_rate": 7.009557028429547e-06,
+      "loss": 0.2989,
+      "step": 6159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.601370430986458,
+      "learning_rate": 7.008624389640476e-06,
+      "loss": 0.3173,
+      "step": 6160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2302915464410913,
+      "learning_rate": 7.0076916675100115e-06,
+      "loss": 0.3141,
+      "step": 6161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.503792068663737,
+      "learning_rate": 7.0067588620768535e-06,
+      "loss": 0.3128,
+      "step": 6162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.704506418702534,
+      "learning_rate": 7.005825973379707e-06,
+      "loss": 0.2945,
+      "step": 6163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.332277329306159,
+      "learning_rate": 7.004893001457277e-06,
+      "loss": 0.3014,
+      "step": 6164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6437128335933913,
+      "learning_rate": 7.003959946348277e-06,
+      "loss": 0.3038,
+      "step": 6165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.875263815796361,
+      "learning_rate": 7.003026808091417e-06,
+      "loss": 0.308,
+      "step": 6166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3694084605402639,
+      "learning_rate": 7.002093586725419e-06,
+      "loss": 0.3067,
+      "step": 6167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8694930266490688,
+      "learning_rate": 7.001160282289e-06,
+      "loss": 0.305,
+      "step": 6168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7609396688890446,
+      "learning_rate": 7.000226894820888e-06,
+      "loss": 0.2942,
+      "step": 6169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0930306386116806,
+      "learning_rate": 6.99929342435981e-06,
+      "loss": 0.323,
+      "step": 6170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8113095138643645,
+      "learning_rate": 6.998359870944495e-06,
+      "loss": 0.2783,
+      "step": 6171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.371474334736473,
+      "learning_rate": 6.99742623461368e-06,
+      "loss": 0.3029,
+      "step": 6172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.530787928579816,
+      "learning_rate": 6.996492515406104e-06,
+      "loss": 0.3019,
+      "step": 6173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1724544346209704,
+      "learning_rate": 6.995558713360505e-06,
+      "loss": 0.3042,
+      "step": 6174
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.301487508566316,
+      "learning_rate": 6.994624828515632e-06,
+      "loss": 0.2969,
+      "step": 6175
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9765221901474412,
+      "learning_rate": 6.993690860910232e-06,
+      "loss": 0.3137,
+      "step": 6176
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.297429021572898,
+      "learning_rate": 6.992756810583057e-06,
+      "loss": 0.3203,
+      "step": 6177
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4405690815449252,
+      "learning_rate": 6.991822677572862e-06,
+      "loss": 0.2862,
+      "step": 6178
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9009653441134091,
+      "learning_rate": 6.9908884619184054e-06,
+      "loss": 0.2977,
+      "step": 6179
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.049451385108946,
+      "learning_rate": 6.98995416365845e-06,
+      "loss": 0.302,
+      "step": 6180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8342916199501595,
+      "learning_rate": 6.989019782831764e-06,
+      "loss": 0.3101,
+      "step": 6181
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0547641381279673,
+      "learning_rate": 6.988085319477114e-06,
+      "loss": 0.3038,
+      "step": 6182
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3825054175343645,
+      "learning_rate": 6.987150773633271e-06,
+      "loss": 0.3111,
+      "step": 6183
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8135322516377184,
+      "learning_rate": 6.9862161453390145e-06,
+      "loss": 0.2849,
+      "step": 6184
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4229262153991158,
+      "learning_rate": 6.9852814346331225e-06,
+      "loss": 0.3119,
+      "step": 6185
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.675703339366878,
+      "learning_rate": 6.984346641554376e-06,
+      "loss": 0.2941,
+      "step": 6186
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.055667374737458,
+      "learning_rate": 6.983411766141563e-06,
+      "loss": 0.3274,
+      "step": 6187
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6502831156998723,
+      "learning_rate": 6.9824768084334736e-06,
+      "loss": 0.2932,
+      "step": 6188
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7299002277457307,
+      "learning_rate": 6.9815417684689e-06,
+      "loss": 0.3117,
+      "step": 6189
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6634212748601143,
+      "learning_rate": 6.980606646286637e-06,
+      "loss": 0.3262,
+      "step": 6190
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.774921914155989,
+      "learning_rate": 6.97967144192549e-06,
+      "loss": 0.3047,
+      "step": 6191
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1742148632581366,
+      "learning_rate": 6.978736155424255e-06,
+      "loss": 0.2885,
+      "step": 6192
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.024499807094335,
+      "learning_rate": 6.977800786821744e-06,
+      "loss": 0.2962,
+      "step": 6193
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.601096611957067,
+      "learning_rate": 6.976865336156765e-06,
+      "loss": 0.3191,
+      "step": 6194
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.990699948167919,
+      "learning_rate": 6.975929803468133e-06,
+      "loss": 0.3044,
+      "step": 6195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4728847852061975,
+      "learning_rate": 6.974994188794662e-06,
+      "loss": 0.2947,
+      "step": 6196
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4848283441348078,
+      "learning_rate": 6.974058492175176e-06,
+      "loss": 0.2987,
+      "step": 6197
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5987230287646015,
+      "learning_rate": 6.973122713648495e-06,
+      "loss": 0.2955,
+      "step": 6198
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8083417251027383,
+      "learning_rate": 6.97218685325345e-06,
+      "loss": 0.293,
+      "step": 6199
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.970791728847927,
+      "learning_rate": 6.9712509110288686e-06,
+      "loss": 0.2979,
+      "step": 6200
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2241409000408114,
+      "learning_rate": 6.970314887013585e-06,
+      "loss": 0.3084,
+      "step": 6201
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.776637136199845,
+      "learning_rate": 6.969378781246436e-06,
+      "loss": 0.3005,
+      "step": 6202
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3347987153060936,
+      "learning_rate": 6.968442593766266e-06,
+      "loss": 0.319,
+      "step": 6203
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2445610065185355,
+      "learning_rate": 6.967506324611915e-06,
+      "loss": 0.314,
+      "step": 6204
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6001916367689244,
+      "learning_rate": 6.9665699738222316e-06,
+      "loss": 0.2999,
+      "step": 6205
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.2452831174311605,
+      "learning_rate": 6.965633541436066e-06,
+      "loss": 0.3092,
+      "step": 6206
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9719513079696007,
+      "learning_rate": 6.964697027492277e-06,
+      "loss": 0.3064,
+      "step": 6207
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3324282619334165,
+      "learning_rate": 6.963760432029716e-06,
+      "loss": 0.3328,
+      "step": 6208
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3561641089060426,
+      "learning_rate": 6.9628237550872465e-06,
+      "loss": 0.3092,
+      "step": 6209
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5942604909686575,
+      "learning_rate": 6.961886996703733e-06,
+      "loss": 0.3166,
+      "step": 6210
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 152.38072679946387,
+      "learning_rate": 6.960950156918045e-06,
+      "loss": 0.2966,
+      "step": 6211
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6104351817313225,
+      "learning_rate": 6.960013235769051e-06,
+      "loss": 0.287,
+      "step": 6212
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9535967986263134,
+      "learning_rate": 6.959076233295625e-06,
+      "loss": 0.3308,
+      "step": 6213
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 27.233761279908748,
+      "learning_rate": 6.958139149536648e-06,
+      "loss": 0.3323,
+      "step": 6214
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0368906440057692,
+      "learning_rate": 6.957201984531e-06,
+      "loss": 0.3064,
+      "step": 6215
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6859032984801705,
+      "learning_rate": 6.956264738317564e-06,
+      "loss": 0.5166,
+      "step": 6216
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.352088624571709,
+      "learning_rate": 6.9553274109352305e-06,
+      "loss": 0.3143,
+      "step": 6217
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8277867727762485,
+      "learning_rate": 6.954390002422889e-06,
+      "loss": 0.2988,
+      "step": 6218
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.602359366787512,
+      "learning_rate": 6.953452512819435e-06,
+      "loss": 0.3167,
+      "step": 6219
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5034782055592268,
+      "learning_rate": 6.952514942163766e-06,
+      "loss": 0.2939,
+      "step": 6220
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8499237300295603,
+      "learning_rate": 6.951577290494784e-06,
+      "loss": 0.3092,
+      "step": 6221
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5586430421078536,
+      "learning_rate": 6.950639557851395e-06,
+      "loss": 0.3071,
+      "step": 6222
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 31.27574456336476,
+      "learning_rate": 6.949701744272506e-06,
+      "loss": 0.3065,
+      "step": 6223
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0730910107267553,
+      "learning_rate": 6.9487638497970266e-06,
+      "loss": 0.3084,
+      "step": 6224
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.949287808216297,
+      "learning_rate": 6.947825874463876e-06,
+      "loss": 0.3057,
+      "step": 6225
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9460132840550055,
+      "learning_rate": 6.946887818311969e-06,
+      "loss": 0.292,
+      "step": 6226
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5116987370461523,
+      "learning_rate": 6.945949681380229e-06,
+      "loss": 0.3014,
+      "step": 6227
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.727836623745217,
+      "learning_rate": 6.9450114637075785e-06,
+      "loss": 0.3037,
+      "step": 6228
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5963086534531157,
+      "learning_rate": 6.944073165332949e-06,
+      "loss": 0.303,
+      "step": 6229
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5931571148043875,
+      "learning_rate": 6.943134786295272e-06,
+      "loss": 0.3196,
+      "step": 6230
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.749088870863373,
+      "learning_rate": 6.942196326633479e-06,
+      "loss": 0.298,
+      "step": 6231
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0658775691579407,
+      "learning_rate": 6.941257786386511e-06,
+      "loss": 0.3008,
+      "step": 6232
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.504533425052485,
+      "learning_rate": 6.940319165593312e-06,
+      "loss": 0.2871,
+      "step": 6233
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.921881588484143,
+      "learning_rate": 6.939380464292822e-06,
+      "loss": 0.3153,
+      "step": 6234
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8339638466264867,
+      "learning_rate": 6.938441682523992e-06,
+      "loss": 0.3271,
+      "step": 6235
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.858992607009354,
+      "learning_rate": 6.9375028203257745e-06,
+      "loss": 0.3066,
+      "step": 6236
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5885599558565247,
+      "learning_rate": 6.936563877737124e-06,
+      "loss": 0.3032,
+      "step": 6237
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4347952306799847,
+      "learning_rate": 6.935624854796996e-06,
+      "loss": 0.2824,
+      "step": 6238
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0816583249155474,
+      "learning_rate": 6.934685751544356e-06,
+      "loss": 0.2944,
+      "step": 6239
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.806656978109954,
+      "learning_rate": 6.933746568018168e-06,
+      "loss": 0.2914,
+      "step": 6240
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.12612682284159,
+      "learning_rate": 6.932807304257401e-06,
+      "loss": 0.2904,
+      "step": 6241
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2979071889701084,
+      "learning_rate": 6.931867960301024e-06,
+      "loss": 0.3017,
+      "step": 6242
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.028160544617963,
+      "learning_rate": 6.930928536188015e-06,
+      "loss": 0.3065,
+      "step": 6243
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7094229070482596,
+      "learning_rate": 6.929989031957352e-06,
+      "loss": 0.4919,
+      "step": 6244
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.496110455846103,
+      "learning_rate": 6.929049447648015e-06,
+      "loss": 0.3175,
+      "step": 6245
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.453174483302288,
+      "learning_rate": 6.928109783298989e-06,
+      "loss": 0.3055,
+      "step": 6246
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.99618496581895,
+      "learning_rate": 6.927170038949267e-06,
+      "loss": 0.3059,
+      "step": 6247
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.6402182504686644,
+      "learning_rate": 6.926230214637833e-06,
+      "loss": 0.3034,
+      "step": 6248
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5233747917849692,
+      "learning_rate": 6.925290310403689e-06,
+      "loss": 0.3063,
+      "step": 6249
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6409800461397064,
+      "learning_rate": 6.9243503262858285e-06,
+      "loss": 0.3051,
+      "step": 6250
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.977322549412839,
+      "learning_rate": 6.9234102623232555e-06,
+      "loss": 0.2981,
+      "step": 6251
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.656496815216476,
+      "learning_rate": 6.922470118554975e-06,
+      "loss": 0.5009,
+      "step": 6252
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.431708885245252,
+      "learning_rate": 6.921529895019995e-06,
+      "loss": 0.2936,
+      "step": 6253
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.556348654702341,
+      "learning_rate": 6.920589591757324e-06,
+      "loss": 0.3245,
+      "step": 6254
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.667316568236721,
+      "learning_rate": 6.919649208805982e-06,
+      "loss": 0.3048,
+      "step": 6255
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.935196027183069,
+      "learning_rate": 6.9187087462049825e-06,
+      "loss": 0.2968,
+      "step": 6256
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.454611931272937,
+      "learning_rate": 6.917768203993351e-06,
+      "loss": 0.3067,
+      "step": 6257
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5713638831818426,
+      "learning_rate": 6.91682758221011e-06,
+      "loss": 0.3024,
+      "step": 6258
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5883838395154235,
+      "learning_rate": 6.915886880894288e-06,
+      "loss": 0.3081,
+      "step": 6259
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.262431695447782,
+      "learning_rate": 6.914946100084916e-06,
+      "loss": 0.3142,
+      "step": 6260
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8432977159515525,
+      "learning_rate": 6.914005239821029e-06,
+      "loss": 0.295,
+      "step": 6261
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4358082624150301,
+      "learning_rate": 6.913064300141664e-06,
+      "loss": 0.3194,
+      "step": 6262
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5763974867524773,
+      "learning_rate": 6.912123281085865e-06,
+      "loss": 0.2939,
+      "step": 6263
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9178292226801221,
+      "learning_rate": 6.911182182692674e-06,
+      "loss": 0.3076,
+      "step": 6264
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1300446005851024,
+      "learning_rate": 6.910241005001139e-06,
+      "loss": 0.3028,
+      "step": 6265
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.342598594435115,
+      "learning_rate": 6.9092997480503125e-06,
+      "loss": 0.2987,
+      "step": 6266
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8683884499375756,
+      "learning_rate": 6.908358411879249e-06,
+      "loss": 0.3218,
+      "step": 6267
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8350418262789883,
+      "learning_rate": 6.907416996527003e-06,
+      "loss": 0.2927,
+      "step": 6268
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4211472264303675,
+      "learning_rate": 6.906475502032639e-06,
+      "loss": 0.2996,
+      "step": 6269
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7646794548493567,
+      "learning_rate": 6.90553392843522e-06,
+      "loss": 0.2933,
+      "step": 6270
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2783824298141386,
+      "learning_rate": 6.904592275773816e-06,
+      "loss": 0.3227,
+      "step": 6271
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.576439189656656,
+      "learning_rate": 6.9036505440874915e-06,
+      "loss": 0.3012,
+      "step": 6272
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6992834889994706,
+      "learning_rate": 6.902708733415328e-06,
+      "loss": 0.3077,
+      "step": 6273
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.317652696958572,
+      "learning_rate": 6.901766843796398e-06,
+      "loss": 0.3153,
+      "step": 6274
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.432923743401709,
+      "learning_rate": 6.900824875269785e-06,
+      "loss": 0.3336,
+      "step": 6275
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4839330873472332,
+      "learning_rate": 6.8998828278745686e-06,
+      "loss": 0.3213,
+      "step": 6276
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5807956176225497,
+      "learning_rate": 6.898940701649842e-06,
+      "loss": 0.2924,
+      "step": 6277
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3568996076431077,
+      "learning_rate": 6.8979984966346914e-06,
+      "loss": 0.3095,
+      "step": 6278
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.43490768163794,
+      "learning_rate": 6.897056212868214e-06,
+      "loss": 0.3039,
+      "step": 6279
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9956618792777565,
+      "learning_rate": 6.8961138503895005e-06,
+      "loss": 0.3061,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9596725519419582,
+      "learning_rate": 6.89517140923766e-06,
+      "loss": 0.3204,
+      "step": 6281
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7387812431055278,
+      "learning_rate": 6.89422888945179e-06,
+      "loss": 0.2911,
+      "step": 6282
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4441244823517567,
+      "learning_rate": 6.893286291071e-06,
+      "loss": 0.3176,
+      "step": 6283
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0913105776521217,
+      "learning_rate": 6.892343614134395e-06,
+      "loss": 0.2994,
+      "step": 6284
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.550371241432521,
+      "learning_rate": 6.891400858681097e-06,
+      "loss": 0.3072,
+      "step": 6285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4329921081678139,
+      "learning_rate": 6.890458024750214e-06,
+      "loss": 0.3028,
+      "step": 6286
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6129288678403952,
+      "learning_rate": 6.889515112380871e-06,
+      "loss": 0.3189,
+      "step": 6287
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6080109905841038,
+      "learning_rate": 6.88857212161219e-06,
+      "loss": 0.3228,
+      "step": 6288
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3992892619014397,
+      "learning_rate": 6.887629052483299e-06,
+      "loss": 0.2798,
+      "step": 6289
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.384952561702454,
+      "learning_rate": 6.886685905033324e-06,
+      "loss": 0.3005,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.456135948922505,
+      "learning_rate": 6.885742679301399e-06,
+      "loss": 0.3127,
+      "step": 6291
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.424520455355706,
+      "learning_rate": 6.884799375326662e-06,
+      "loss": 0.3069,
+      "step": 6292
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9529667834212303,
+      "learning_rate": 6.883855993148252e-06,
+      "loss": 0.3136,
+      "step": 6293
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.515609439982961,
+      "learning_rate": 6.882912532805308e-06,
+      "loss": 0.3064,
+      "step": 6294
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9749881320005336,
+      "learning_rate": 6.8819689943369805e-06,
+      "loss": 0.3119,
+      "step": 6295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.952786377641116,
+      "learning_rate": 6.881025377782415e-06,
+      "loss": 0.318,
+      "step": 6296
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8379560886548978,
+      "learning_rate": 6.880081683180768e-06,
+      "loss": 0.313,
+      "step": 6297
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1068557721472867,
+      "learning_rate": 6.879137910571191e-06,
+      "loss": 0.2853,
+      "step": 6298
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9633950582282589,
+      "learning_rate": 6.878194059992846e-06,
+      "loss": 0.2957,
+      "step": 6299
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4827048174594124,
+      "learning_rate": 6.8772501314848915e-06,
+      "loss": 0.3068,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4277887708746932,
+      "learning_rate": 6.876306125086496e-06,
+      "loss": 0.2938,
+      "step": 6301
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.445439710481941,
+      "learning_rate": 6.8753620408368235e-06,
+      "loss": 0.2964,
+      "step": 6302
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.037116657515197,
+      "learning_rate": 6.8744178787750526e-06,
+      "loss": 0.3028,
+      "step": 6303
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.177952274362125,
+      "learning_rate": 6.873473638940354e-06,
+      "loss": 0.2987,
+      "step": 6304
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9791992705023835,
+      "learning_rate": 6.872529321371906e-06,
+      "loss": 0.3175,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6868036208965926,
+      "learning_rate": 6.87158492610889e-06,
+      "loss": 0.3118,
+      "step": 6306
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.198675008942469,
+      "learning_rate": 6.870640453190491e-06,
+      "loss": 0.3157,
+      "step": 6307
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.433621451491889,
+      "learning_rate": 6.869695902655898e-06,
+      "loss": 0.304,
+      "step": 6308
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9460073702454201,
+      "learning_rate": 6.868751274544301e-06,
+      "loss": 0.294,
+      "step": 6309
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9670261253683712,
+      "learning_rate": 6.867806568894893e-06,
+      "loss": 0.311,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6591038286039345,
+      "learning_rate": 6.866861785746873e-06,
+      "loss": 0.5118,
+      "step": 6311
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4779175952239088,
+      "learning_rate": 6.865916925139442e-06,
+      "loss": 0.3067,
+      "step": 6312
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.511401538514108,
+      "learning_rate": 6.864971987111804e-06,
+      "loss": 0.3111,
+      "step": 6313
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.692742950586186,
+      "learning_rate": 6.864026971703166e-06,
+      "loss": 0.3071,
+      "step": 6314
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.481765192599177,
+      "learning_rate": 6.863081878952738e-06,
+      "loss": 0.3295,
+      "step": 6315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8603319326306396,
+      "learning_rate": 6.8621367088997325e-06,
+      "loss": 0.3238,
+      "step": 6316
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.715309388077017,
+      "learning_rate": 6.8611914615833676e-06,
+      "loss": 0.2939,
+      "step": 6317
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9872358204036504,
+      "learning_rate": 6.860246137042863e-06,
+      "loss": 0.2956,
+      "step": 6318
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8065665490318836,
+      "learning_rate": 6.859300735317444e-06,
+      "loss": 0.3002,
+      "step": 6319
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.342606612220326,
+      "learning_rate": 6.858355256446333e-06,
+      "loss": 0.3155,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.590118021556689,
+      "learning_rate": 6.857409700468762e-06,
+      "loss": 0.2907,
+      "step": 6321
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.921239443935161,
+      "learning_rate": 6.856464067423963e-06,
+      "loss": 0.3159,
+      "step": 6322
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.574425643570183,
+      "learning_rate": 6.855518357351174e-06,
+      "loss": 0.3008,
+      "step": 6323
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5398285287025202,
+      "learning_rate": 6.854572570289632e-06,
+      "loss": 0.3065,
+      "step": 6324
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8041753943042287,
+      "learning_rate": 6.853626706278579e-06,
+      "loss": 0.3078,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.057675770228045,
+      "learning_rate": 6.852680765357262e-06,
+      "loss": 0.3064,
+      "step": 6326
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4892062230634326,
+      "learning_rate": 6.85173474756493e-06,
+      "loss": 0.31,
+      "step": 6327
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0511590689152306,
+      "learning_rate": 6.850788652940832e-06,
+      "loss": 0.3012,
+      "step": 6328
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0263552318565505,
+      "learning_rate": 6.849842481524228e-06,
+      "loss": 0.3301,
+      "step": 6329
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8728865182573091,
+      "learning_rate": 6.8488962333543715e-06,
+      "loss": 0.3122,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.517592855766456,
+      "learning_rate": 6.847949908470529e-06,
+      "loss": 0.3074,
+      "step": 6331
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.178893106391159,
+      "learning_rate": 6.84700350691196e-06,
+      "loss": 0.307,
+      "step": 6332
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6021767766807424,
+      "learning_rate": 6.846057028717937e-06,
+      "loss": 0.3197,
+      "step": 6333
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.501002131768941,
+      "learning_rate": 6.845110473927727e-06,
+      "loss": 0.3286,
+      "step": 6334
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2132623633872157,
+      "learning_rate": 6.844163842580608e-06,
+      "loss": 0.3226,
+      "step": 6335
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5350315206393872,
+      "learning_rate": 6.8432171347158535e-06,
+      "loss": 0.3115,
+      "step": 6336
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1798307847448,
+      "learning_rate": 6.842270350372749e-06,
+      "loss": 0.2947,
+      "step": 6337
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6360090486603727,
+      "learning_rate": 6.8413234895905726e-06,
+      "loss": 0.2967,
+      "step": 6338
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9848580915620224,
+      "learning_rate": 6.840376552408614e-06,
+      "loss": 0.3115,
+      "step": 6339
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3405026564912617,
+      "learning_rate": 6.839429538866164e-06,
+      "loss": 0.2822,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5362126550497548,
+      "learning_rate": 6.838482449002517e-06,
+      "loss": 0.286,
+      "step": 6341
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3069866190457384,
+      "learning_rate": 6.837535282856966e-06,
+      "loss": 0.2885,
+      "step": 6342
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4765326992977221,
+      "learning_rate": 6.836588040468812e-06,
+      "loss": 0.3005,
+      "step": 6343
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.565457286700779,
+      "learning_rate": 6.835640721877359e-06,
+      "loss": 0.2854,
+      "step": 6344
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.559378590319128,
+      "learning_rate": 6.834693327121913e-06,
+      "loss": 0.3041,
+      "step": 6345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0677104007652565,
+      "learning_rate": 6.83374585624178e-06,
+      "loss": 0.3031,
+      "step": 6346
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1149379020446104,
+      "learning_rate": 6.832798309276275e-06,
+      "loss": 0.3142,
+      "step": 6347
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.225979158875827,
+      "learning_rate": 6.831850686264712e-06,
+      "loss": 0.2883,
+      "step": 6348
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4380784270508573,
+      "learning_rate": 6.830902987246413e-06,
+      "loss": 0.3088,
+      "step": 6349
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1902049420419636,
+      "learning_rate": 6.8299552122606934e-06,
+      "loss": 0.3209,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.4983139239455925,
+      "learning_rate": 6.829007361346885e-06,
+      "loss": 0.3236,
+      "step": 6351
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.38038349341543,
+      "learning_rate": 6.828059434544309e-06,
+      "loss": 0.3019,
+      "step": 6352
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5095839511716451,
+      "learning_rate": 6.827111431892303e-06,
+      "loss": 0.2996,
+      "step": 6353
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3002850599763964,
+      "learning_rate": 6.826163353430197e-06,
+      "loss": 0.2835,
+      "step": 6354
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.782981876014829,
+      "learning_rate": 6.82521519919733e-06,
+      "loss": 0.3044,
+      "step": 6355
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4652631665027267,
+      "learning_rate": 6.8242669692330424e-06,
+      "loss": 0.3209,
+      "step": 6356
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2667979646321514,
+      "learning_rate": 6.823318663576679e-06,
+      "loss": 0.3175,
+      "step": 6357
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3577067350887904,
+      "learning_rate": 6.822370282267585e-06,
+      "loss": 0.2862,
+      "step": 6358
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5255784700649988,
+      "learning_rate": 6.82142182534511e-06,
+      "loss": 0.2999,
+      "step": 6359
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7601001630013533,
+      "learning_rate": 6.8204732928486096e-06,
+      "loss": 0.2916,
+      "step": 6360
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6091656405623176,
+      "learning_rate": 6.819524684817439e-06,
+      "loss": 0.312,
+      "step": 6361
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7177599903981735,
+      "learning_rate": 6.8185760012909566e-06,
+      "loss": 0.2972,
+      "step": 6362
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9267907314481778,
+      "learning_rate": 6.817627242308525e-06,
+      "loss": 0.3185,
+      "step": 6363
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8570406736156893,
+      "learning_rate": 6.816678407909511e-06,
+      "loss": 0.316,
+      "step": 6364
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0796908376535903,
+      "learning_rate": 6.815729498133286e-06,
+      "loss": 0.302,
+      "step": 6365
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4157855196807505,
+      "learning_rate": 6.814780513019214e-06,
+      "loss": 0.3037,
+      "step": 6366
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6743395572870021,
+      "learning_rate": 6.813831452606678e-06,
+      "loss": 0.2874,
+      "step": 6367
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.692394679105457,
+      "learning_rate": 6.8128823169350535e-06,
+      "loss": 0.3051,
+      "step": 6368
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.763263681427266,
+      "learning_rate": 6.811933106043721e-06,
+      "loss": 0.2882,
+      "step": 6369
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9985345857072736,
+      "learning_rate": 6.8109838199720655e-06,
+      "loss": 0.3012,
+      "step": 6370
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2706514808811527,
+      "learning_rate": 6.8100344587594754e-06,
+      "loss": 0.307,
+      "step": 6371
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3051438527903696,
+      "learning_rate": 6.809085022445341e-06,
+      "loss": 0.3037,
+      "step": 6372
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.297934797149001,
+      "learning_rate": 6.808135511069054e-06,
+      "loss": 0.3042,
+      "step": 6373
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.325539927654011,
+      "learning_rate": 6.807185924670013e-06,
+      "loss": 0.2952,
+      "step": 6374
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4080357034994866,
+      "learning_rate": 6.80623626328762e-06,
+      "loss": 0.2911,
+      "step": 6375
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9720138672113132,
+      "learning_rate": 6.805286526961274e-06,
+      "loss": 0.3278,
+      "step": 6376
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8250325174334556,
+      "learning_rate": 6.804336715730385e-06,
+      "loss": 0.3113,
+      "step": 6377
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6273334819042335,
+      "learning_rate": 6.803386829634361e-06,
+      "loss": 0.3016,
+      "step": 6378
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9125793157768267,
+      "learning_rate": 6.8024368687126145e-06,
+      "loss": 0.3044,
+      "step": 6379
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.037940233798932,
+      "learning_rate": 6.801486833004559e-06,
+      "loss": 0.3044,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7520585156058943,
+      "learning_rate": 6.8005367225496155e-06,
+      "loss": 0.2997,
+      "step": 6381
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6297250564109693,
+      "learning_rate": 6.799586537387206e-06,
+      "loss": 0.3251,
+      "step": 6382
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6972051848994267,
+      "learning_rate": 6.7986362775567545e-06,
+      "loss": 0.3171,
+      "step": 6383
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5832104903379132,
+      "learning_rate": 6.79768594309769e-06,
+      "loss": 0.4883,
+      "step": 6384
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5790471983614522,
+      "learning_rate": 6.796735534049441e-06,
+      "loss": 0.3225,
+      "step": 6385
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.330783457874948,
+      "learning_rate": 6.795785050451443e-06,
+      "loss": 0.3118,
+      "step": 6386
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.835327429306243,
+      "learning_rate": 6.7948344923431355e-06,
+      "loss": 0.3374,
+      "step": 6387
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.617774730322946,
+      "learning_rate": 6.793883859763955e-06,
+      "loss": 0.3133,
+      "step": 6388
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7923144480516422,
+      "learning_rate": 6.792933152753348e-06,
+      "loss": 0.3036,
+      "step": 6389
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0125838015849853,
+      "learning_rate": 6.791982371350761e-06,
+      "loss": 0.2968,
+      "step": 6390
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5522094024516724,
+      "learning_rate": 6.791031515595641e-06,
+      "loss": 0.3029,
+      "step": 6391
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.270122226573881,
+      "learning_rate": 6.790080585527442e-06,
+      "loss": 0.3262,
+      "step": 6392
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4299924927621477,
+      "learning_rate": 6.789129581185621e-06,
+      "loss": 0.2933,
+      "step": 6393
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.428656152912421,
+      "learning_rate": 6.788178502609635e-06,
+      "loss": 0.2936,
+      "step": 6394
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5712739983277206,
+      "learning_rate": 6.787227349838946e-06,
+      "loss": 0.3108,
+      "step": 6395
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7508718567894106,
+      "learning_rate": 6.786276122913021e-06,
+      "loss": 0.3109,
+      "step": 6396
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6948998904039196,
+      "learning_rate": 6.785324821871326e-06,
+      "loss": 0.2911,
+      "step": 6397
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8042237657566123,
+      "learning_rate": 6.784373446753334e-06,
+      "loss": 0.3153,
+      "step": 6398
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.203912760157126,
+      "learning_rate": 6.783421997598518e-06,
+      "loss": 0.3216,
+      "step": 6399
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.128446226595074,
+      "learning_rate": 6.782470474446357e-06,
+      "loss": 0.2887,
+      "step": 6400
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2689239331071493,
+      "learning_rate": 6.781518877336328e-06,
+      "loss": 0.3201,
+      "step": 6401
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9469366874019725,
+      "learning_rate": 6.7805672063079166e-06,
+      "loss": 0.2989,
+      "step": 6402
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8764163133376295,
+      "learning_rate": 6.77961546140061e-06,
+      "loss": 0.3321,
+      "step": 6403
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9326197487964842,
+      "learning_rate": 6.778663642653897e-06,
+      "loss": 0.2928,
+      "step": 6404
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.683167891228867,
+      "learning_rate": 6.77771175010727e-06,
+      "loss": 0.2963,
+      "step": 6405
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.184407438462736,
+      "learning_rate": 6.776759783800224e-06,
+      "loss": 0.3145,
+      "step": 6406
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2446580362794384,
+      "learning_rate": 6.775807743772258e-06,
+      "loss": 0.3081,
+      "step": 6407
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.926659571692815,
+      "learning_rate": 6.7748556300628764e-06,
+      "loss": 0.3009,
+      "step": 6408
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6655192578806834,
+      "learning_rate": 6.773903442711582e-06,
+      "loss": 0.3182,
+      "step": 6409
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3488201062460332,
+      "learning_rate": 6.772951181757883e-06,
+      "loss": 0.3062,
+      "step": 6410
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2345830430784606,
+      "learning_rate": 6.77199884724129e-06,
+      "loss": 0.302,
+      "step": 6411
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.440144329245238,
+      "learning_rate": 6.7710464392013165e-06,
+      "loss": 0.3013,
+      "step": 6412
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4499405252676987,
+      "learning_rate": 6.770093957677483e-06,
+      "loss": 0.2919,
+      "step": 6413
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4967558730981563,
+      "learning_rate": 6.769141402709305e-06,
+      "loss": 0.2891,
+      "step": 6414
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5492769404609563,
+      "learning_rate": 6.7681887743363085e-06,
+      "loss": 0.3092,
+      "step": 6415
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.085038937768041,
+      "learning_rate": 6.767236072598018e-06,
+      "loss": 0.3176,
+      "step": 6416
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.315609889684815,
+      "learning_rate": 6.766283297533965e-06,
+      "loss": 0.2823,
+      "step": 6417
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0800677507978316,
+      "learning_rate": 6.765330449183682e-06,
+      "loss": 0.3076,
+      "step": 6418
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8613492189867349,
+      "learning_rate": 6.764377527586701e-06,
+      "loss": 0.3399,
+      "step": 6419
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.716488026652963,
+      "learning_rate": 6.763424532782562e-06,
+      "loss": 0.3122,
+      "step": 6420
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.537163245755776,
+      "learning_rate": 6.762471464810808e-06,
+      "loss": 0.2919,
+      "step": 6421
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.444778544636523,
+      "learning_rate": 6.761518323710983e-06,
+      "loss": 0.289,
+      "step": 6422
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4808424561732993,
+      "learning_rate": 6.760565109522634e-06,
+      "loss": 0.3142,
+      "step": 6423
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5107185897637505,
+      "learning_rate": 6.75961182228531e-06,
+      "loss": 0.2986,
+      "step": 6424
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6168280353607318,
+      "learning_rate": 6.758658462038568e-06,
+      "loss": 0.2971,
+      "step": 6425
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0215373404663097,
+      "learning_rate": 6.757705028821961e-06,
+      "loss": 0.3061,
+      "step": 6426
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0190968985247033,
+      "learning_rate": 6.756751522675051e-06,
+      "loss": 0.3095,
+      "step": 6427
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5502530079717356,
+      "learning_rate": 6.755797943637401e-06,
+      "loss": 0.2924,
+      "step": 6428
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7631284176320388,
+      "learning_rate": 6.754844291748575e-06,
+      "loss": 0.2864,
+      "step": 6429
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1932181990125468,
+      "learning_rate": 6.753890567048141e-06,
+      "loss": 0.2986,
+      "step": 6430
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0542045306519885,
+      "learning_rate": 6.752936769575673e-06,
+      "loss": 0.3014,
+      "step": 6431
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8273598593855904,
+      "learning_rate": 6.751982899370746e-06,
+      "loss": 0.2978,
+      "step": 6432
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.733642977777779,
+      "learning_rate": 6.751028956472935e-06,
+      "loss": 0.314,
+      "step": 6433
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.397989501728536,
+      "learning_rate": 6.7500749409218235e-06,
+      "loss": 0.3007,
+      "step": 6434
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7413082411923697,
+      "learning_rate": 6.749120852756994e-06,
+      "loss": 0.2918,
+      "step": 6435
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7194939410932075,
+      "learning_rate": 6.748166692018033e-06,
+      "loss": 0.2819,
+      "step": 6436
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.772032781328613,
+      "learning_rate": 6.7472124587445306e-06,
+      "loss": 0.3071,
+      "step": 6437
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.731602382722578,
+      "learning_rate": 6.746258152976082e-06,
+      "loss": 0.2945,
+      "step": 6438
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0333223022746494,
+      "learning_rate": 6.745303774752279e-06,
+      "loss": 0.3064,
+      "step": 6439
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.76952663531311,
+      "learning_rate": 6.744349324112722e-06,
+      "loss": 0.3132,
+      "step": 6440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6825494996694594,
+      "learning_rate": 6.743394801097014e-06,
+      "loss": 0.3132,
+      "step": 6441
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.51728955276763,
+      "learning_rate": 6.7424402057447606e-06,
+      "loss": 0.2873,
+      "step": 6442
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7091262720438556,
+      "learning_rate": 6.741485538095566e-06,
+      "loss": 0.3023,
+      "step": 6443
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8394371015675834,
+      "learning_rate": 6.7405307981890436e-06,
+      "loss": 0.3047,
+      "step": 6444
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0800861005997704,
+      "learning_rate": 6.739575986064807e-06,
+      "loss": 0.3005,
+      "step": 6445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.598482420410477,
+      "learning_rate": 6.738621101762472e-06,
+      "loss": 0.2811,
+      "step": 6446
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7975966513433286,
+      "learning_rate": 6.737666145321662e-06,
+      "loss": 0.2843,
+      "step": 6447
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9658994276713817,
+      "learning_rate": 6.7367111167819955e-06,
+      "loss": 0.2969,
+      "step": 6448
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4202457333872267,
+      "learning_rate": 6.735756016183099e-06,
+      "loss": 0.2972,
+      "step": 6449
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4356628148175767,
+      "learning_rate": 6.734800843564604e-06,
+      "loss": 0.2804,
+      "step": 6450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.434828003590157,
+      "learning_rate": 6.73384559896614e-06,
+      "loss": 0.3023,
+      "step": 6451
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9753955136574015,
+      "learning_rate": 6.732890282427342e-06,
+      "loss": 0.2929,
+      "step": 6452
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7334199063473066,
+      "learning_rate": 6.731934893987849e-06,
+      "loss": 0.2801,
+      "step": 6453
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.458364112155257,
+      "learning_rate": 6.7309794336873e-06,
+      "loss": 0.2966,
+      "step": 6454
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3031052151800613,
+      "learning_rate": 6.730023901565341e-06,
+      "loss": 0.304,
+      "step": 6455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6402178108076182,
+      "learning_rate": 6.729068297661618e-06,
+      "loss": 0.3019,
+      "step": 6456
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6320293107294643,
+      "learning_rate": 6.728112622015779e-06,
+      "loss": 0.2945,
+      "step": 6457
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5309669224890703,
+      "learning_rate": 6.727156874667478e-06,
+      "loss": 0.3075,
+      "step": 6458
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.432981301693717,
+      "learning_rate": 6.726201055656369e-06,
+      "loss": 0.2818,
+      "step": 6459
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6973018453399739,
+      "learning_rate": 6.725245165022114e-06,
+      "loss": 0.4881,
+      "step": 6460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.446417171142823,
+      "learning_rate": 6.724289202804373e-06,
+      "loss": 0.3192,
+      "step": 6461
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3604008127576046,
+      "learning_rate": 6.723333169042808e-06,
+      "loss": 0.3047,
+      "step": 6462
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.807463817597169,
+      "learning_rate": 6.722377063777091e-06,
+      "loss": 0.3137,
+      "step": 6463
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8294947866474718,
+      "learning_rate": 6.72142088704689e-06,
+      "loss": 0.3068,
+      "step": 6464
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3060169917182747,
+      "learning_rate": 6.720464638891878e-06,
+      "loss": 0.2916,
+      "step": 6465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.008175265254663,
+      "learning_rate": 6.719508319351733e-06,
+      "loss": 0.3022,
+      "step": 6466
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.124354686119315,
+      "learning_rate": 6.718551928466133e-06,
+      "loss": 0.3146,
+      "step": 6467
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6295233948044467,
+      "learning_rate": 6.717595466274762e-06,
+      "loss": 0.2968,
+      "step": 6468
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4702849046667679,
+      "learning_rate": 6.716638932817303e-06,
+      "loss": 0.2925,
+      "step": 6469
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6444903445591657,
+      "learning_rate": 6.715682328133447e-06,
+      "loss": 0.2863,
+      "step": 6470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8408699849276253,
+      "learning_rate": 6.714725652262882e-06,
+      "loss": 0.2907,
+      "step": 6471
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6301294973434417,
+      "learning_rate": 6.713768905245306e-06,
+      "loss": 0.2855,
+      "step": 6472
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.939913293202927,
+      "learning_rate": 6.712812087120413e-06,
+      "loss": 0.3038,
+      "step": 6473
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3121552370294105,
+      "learning_rate": 6.711855197927904e-06,
+      "loss": 0.2905,
+      "step": 6474
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5014649916385596,
+      "learning_rate": 6.710898237707482e-06,
+      "loss": 0.2995,
+      "step": 6475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4460069249305667,
+      "learning_rate": 6.7099412064988555e-06,
+      "loss": 0.3094,
+      "step": 6476
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9799930823188674,
+      "learning_rate": 6.708984104341728e-06,
+      "loss": 0.3129,
+      "step": 6477
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6505256145852305,
+      "learning_rate": 6.708026931275817e-06,
+      "loss": 0.3025,
+      "step": 6478
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0605434419766584,
+      "learning_rate": 6.707069687340834e-06,
+      "loss": 0.3033,
+      "step": 6479
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0207596390285953,
+      "learning_rate": 6.706112372576499e-06,
+      "loss": 0.2911,
+      "step": 6480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.144713949978419,
+      "learning_rate": 6.705154987022528e-06,
+      "loss": 0.3052,
+      "step": 6481
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6216085321664697,
+      "learning_rate": 6.70419753071865e-06,
+      "loss": 0.3001,
+      "step": 6482
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7893907759835832,
+      "learning_rate": 6.703240003704588e-06,
+      "loss": 0.3005,
+      "step": 6483
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.026551985538254,
+      "learning_rate": 6.702282406020076e-06,
+      "loss": 0.3295,
+      "step": 6484
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.656580054248685,
+      "learning_rate": 6.70132473770484e-06,
+      "loss": 0.3237,
+      "step": 6485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3342468805975165,
+      "learning_rate": 6.700366998798621e-06,
+      "loss": 0.2965,
+      "step": 6486
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3114027899184495,
+      "learning_rate": 6.699409189341153e-06,
+      "loss": 0.2865,
+      "step": 6487
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.660419345588762,
+      "learning_rate": 6.69845130937218e-06,
+      "loss": 0.2929,
+      "step": 6488
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3563661499613435,
+      "learning_rate": 6.697493358931446e-06,
+      "loss": 0.3178,
+      "step": 6489
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5526025441779743,
+      "learning_rate": 6.696535338058699e-06,
+      "loss": 0.2829,
+      "step": 6490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6259505361326623,
+      "learning_rate": 6.695577246793684e-06,
+      "loss": 0.2969,
+      "step": 6491
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6633457265011925,
+      "learning_rate": 6.694619085176159e-06,
+      "loss": 0.5003,
+      "step": 6492
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.104362074544626,
+      "learning_rate": 6.693660853245878e-06,
+      "loss": 0.2913,
+      "step": 6493
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0840101305702516,
+      "learning_rate": 6.6927025510426015e-06,
+      "loss": 0.2877,
+      "step": 6494
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5817337343184006,
+      "learning_rate": 6.691744178606087e-06,
+      "loss": 0.3022,
+      "step": 6495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5247565946653607,
+      "learning_rate": 6.690785735976103e-06,
+      "loss": 0.2988,
+      "step": 6496
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5498495605021445,
+      "learning_rate": 6.6898272231924155e-06,
+      "loss": 0.3035,
+      "step": 6497
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.986600088902626,
+      "learning_rate": 6.688868640294796e-06,
+      "loss": 0.2984,
+      "step": 6498
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0870828973579,
+      "learning_rate": 6.687909987323016e-06,
+      "loss": 0.3022,
+      "step": 6499
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8103354335933153,
+      "learning_rate": 6.686951264316852e-06,
+      "loss": 0.3161,
+      "step": 6500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5621955735177566,
+      "learning_rate": 6.6859924713160825e-06,
+      "loss": 0.2957,
+      "step": 6501
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0167450198214447,
+      "learning_rate": 6.685033608360494e-06,
+      "loss": 0.306,
+      "step": 6502
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.373936432373496,
+      "learning_rate": 6.684074675489864e-06,
+      "loss": 0.3407,
+      "step": 6503
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8784137649234551,
+      "learning_rate": 6.683115672743989e-06,
+      "loss": 0.2972,
+      "step": 6504
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.643761595008346,
+      "learning_rate": 6.682156600162653e-06,
+      "loss": 0.3029,
+      "step": 6505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7342194894575267,
+      "learning_rate": 6.681197457785652e-06,
+      "loss": 0.2768,
+      "step": 6506
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.846046235908457,
+      "learning_rate": 6.680238245652782e-06,
+      "loss": 0.2983,
+      "step": 6507
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.248225598354093,
+      "learning_rate": 6.679278963803843e-06,
+      "loss": 0.2868,
+      "step": 6508
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6656411086377948,
+      "learning_rate": 6.678319612278636e-06,
+      "loss": 0.4823,
+      "step": 6509
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.085584357878793,
+      "learning_rate": 6.677360191116971e-06,
+      "loss": 0.2954,
+      "step": 6510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.875879156190651,
+      "learning_rate": 6.676400700358647e-06,
+      "loss": 0.3085,
+      "step": 6511
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1069616976492127,
+      "learning_rate": 6.6754411400434835e-06,
+      "loss": 0.2905,
+      "step": 6512
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6032407445338612,
+      "learning_rate": 6.674481510211292e-06,
+      "loss": 0.2784,
+      "step": 6513
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0661652484314277,
+      "learning_rate": 6.673521810901886e-06,
+      "loss": 0.3457,
+      "step": 6514
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.013500104714371,
+      "learning_rate": 6.672562042155089e-06,
+      "loss": 0.2974,
+      "step": 6515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.027355388246828,
+      "learning_rate": 6.671602204010722e-06,
+      "loss": 0.28,
+      "step": 6516
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.221326735148984,
+      "learning_rate": 6.67064229650861e-06,
+      "loss": 0.3294,
+      "step": 6517
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5814221531969004,
+      "learning_rate": 6.669682319688582e-06,
+      "loss": 0.2906,
+      "step": 6518
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 26.916348187608442,
+      "learning_rate": 6.6687222735904675e-06,
+      "loss": 0.3017,
+      "step": 6519
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0790376071172463,
+      "learning_rate": 6.667762158254104e-06,
+      "loss": 0.2981,
+      "step": 6520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9109341389632886,
+      "learning_rate": 6.6668019737193255e-06,
+      "loss": 0.2845,
+      "step": 6521
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.032732941934832,
+      "learning_rate": 6.665841720025972e-06,
+      "loss": 0.2976,
+      "step": 6522
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5256747676019773,
+      "learning_rate": 6.664881397213887e-06,
+      "loss": 0.3113,
+      "step": 6523
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0470466405784453,
+      "learning_rate": 6.663921005322917e-06,
+      "loss": 0.3078,
+      "step": 6524
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4551442549662454,
+      "learning_rate": 6.662960544392907e-06,
+      "loss": 0.3108,
+      "step": 6525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.492917944019753,
+      "learning_rate": 6.662000014463711e-06,
+      "loss": 0.3122,
+      "step": 6526
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.261631175411658,
+      "learning_rate": 6.661039415575183e-06,
+      "loss": 0.2973,
+      "step": 6527
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.866713432003546,
+      "learning_rate": 6.660078747767178e-06,
+      "loss": 0.309,
+      "step": 6528
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8077549956648529,
+      "learning_rate": 6.659118011079558e-06,
+      "loss": 0.3034,
+      "step": 6529
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8211784951489682,
+      "learning_rate": 6.658157205552185e-06,
+      "loss": 0.2911,
+      "step": 6530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.25057674642045,
+      "learning_rate": 6.6571963312249236e-06,
+      "loss": 0.3295,
+      "step": 6531
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.04367286593556,
+      "learning_rate": 6.656235388137644e-06,
+      "loss": 0.322,
+      "step": 6532
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2583043092157906,
+      "learning_rate": 6.655274376330214e-06,
+      "loss": 0.2728,
+      "step": 6533
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.162039432449415,
+      "learning_rate": 6.654313295842513e-06,
+      "loss": 0.3119,
+      "step": 6534
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1090665833326745,
+      "learning_rate": 6.653352146714413e-06,
+      "loss": 0.3204,
+      "step": 6535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2228168014288647,
+      "learning_rate": 6.652390928985797e-06,
+      "loss": 0.2973,
+      "step": 6536
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.771665543505001,
+      "learning_rate": 6.651429642696545e-06,
+      "loss": 0.2982,
+      "step": 6537
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1660760953101317,
+      "learning_rate": 6.6504682878865444e-06,
+      "loss": 0.2934,
+      "step": 6538
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.0469822850552575,
+      "learning_rate": 6.649506864595683e-06,
+      "loss": 0.2928,
+      "step": 6539
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6919902299540963,
+      "learning_rate": 6.648545372863853e-06,
+      "loss": 0.3067,
+      "step": 6540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.131151185345146,
+      "learning_rate": 6.647583812730945e-06,
+      "loss": 0.3182,
+      "step": 6541
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5973916018871677,
+      "learning_rate": 6.646622184236861e-06,
+      "loss": 0.3322,
+      "step": 6542
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8035752923220416,
+      "learning_rate": 6.6456604874214955e-06,
+      "loss": 0.3168,
+      "step": 6543
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.097977463715777,
+      "learning_rate": 6.644698722324755e-06,
+      "loss": 0.2925,
+      "step": 6544
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.890742047703468,
+      "learning_rate": 6.643736888986541e-06,
+      "loss": 0.2945,
+      "step": 6545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8278897102287393,
+      "learning_rate": 6.642774987446768e-06,
+      "loss": 0.3043,
+      "step": 6546
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8485565712722765,
+      "learning_rate": 6.641813017745339e-06,
+      "loss": 0.293,
+      "step": 6547
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.766376519928722,
+      "learning_rate": 6.640850979922173e-06,
+      "loss": 0.2878,
+      "step": 6548
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8180337634210066,
+      "learning_rate": 6.639888874017185e-06,
+      "loss": 0.3108,
+      "step": 6549
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8359271171744111,
+      "learning_rate": 6.638926700070296e-06,
+      "loss": 0.3123,
+      "step": 6550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1252229166077155,
+      "learning_rate": 6.637964458121427e-06,
+      "loss": 0.2957,
+      "step": 6551
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7650068696064065,
+      "learning_rate": 6.637002148210502e-06,
+      "loss": 0.2968,
+      "step": 6552
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.545614146749676,
+      "learning_rate": 6.63603977037745e-06,
+      "loss": 0.2915,
+      "step": 6553
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0902969731352163,
+      "learning_rate": 6.635077324662203e-06,
+      "loss": 0.281,
+      "step": 6554
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.898235766333002,
+      "learning_rate": 6.6341148111046935e-06,
+      "loss": 0.2912,
+      "step": 6555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.0526073815626145,
+      "learning_rate": 6.6331522297448584e-06,
+      "loss": 0.3282,
+      "step": 6556
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.580336371799474,
+      "learning_rate": 6.632189580622636e-06,
+      "loss": 0.32,
+      "step": 6557
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3217855246364456,
+      "learning_rate": 6.631226863777968e-06,
+      "loss": 0.3033,
+      "step": 6558
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3520595471580505,
+      "learning_rate": 6.6302640792508e-06,
+      "loss": 0.3188,
+      "step": 6559
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4813151010945194,
+      "learning_rate": 6.629301227081082e-06,
+      "loss": 0.2948,
+      "step": 6560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.940935134432427,
+      "learning_rate": 6.6283383073087595e-06,
+      "loss": 0.2935,
+      "step": 6561
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5880193534118219,
+      "learning_rate": 6.62737531997379e-06,
+      "loss": 0.2863,
+      "step": 6562
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.841185018704397,
+      "learning_rate": 6.626412265116127e-06,
+      "loss": 0.2871,
+      "step": 6563
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.447388579476549,
+      "learning_rate": 6.625449142775731e-06,
+      "loss": 0.2895,
+      "step": 6564
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6735093007773516,
+      "learning_rate": 6.624485952992563e-06,
+      "loss": 0.3117,
+      "step": 6565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.537175043517064,
+      "learning_rate": 6.623522695806588e-06,
+      "loss": 0.3177,
+      "step": 6566
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6200215922676457,
+      "learning_rate": 6.62255937125777e-06,
+      "loss": 0.3127,
+      "step": 6567
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.564596987242483,
+      "learning_rate": 6.621595979386084e-06,
+      "loss": 0.295,
+      "step": 6568
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.986314193108306,
+      "learning_rate": 6.6206325202315e-06,
+      "loss": 0.3048,
+      "step": 6569
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.622307023271134,
+      "learning_rate": 6.6196689938339946e-06,
+      "loss": 0.5279,
+      "step": 6570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.342715204296712,
+      "learning_rate": 6.618705400233544e-06,
+      "loss": 0.2875,
+      "step": 6571
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.057952809297888,
+      "learning_rate": 6.617741739470134e-06,
+      "loss": 0.3058,
+      "step": 6572
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.340990691708877,
+      "learning_rate": 6.616778011583744e-06,
+      "loss": 0.315,
+      "step": 6573
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.1280708954523515,
+      "learning_rate": 6.6158142166143625e-06,
+      "loss": 0.3019,
+      "step": 6574
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9229894990239464,
+      "learning_rate": 6.614850354601978e-06,
+      "loss": 0.2781,
+      "step": 6575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5369966196890847,
+      "learning_rate": 6.613886425586586e-06,
+      "loss": 0.288,
+      "step": 6576
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.641657513587774,
+      "learning_rate": 6.61292242960818e-06,
+      "loss": 0.2982,
+      "step": 6577
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7241730992220534,
+      "learning_rate": 6.611958366706757e-06,
+      "loss": 0.3175,
+      "step": 6578
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8285734215480198,
+      "learning_rate": 6.610994236922317e-06,
+      "loss": 0.296,
+      "step": 6579
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.3209263340897497,
+      "learning_rate": 6.610030040294866e-06,
+      "loss": 0.3147,
+      "step": 6580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.068601570984096,
+      "learning_rate": 6.6090657768644085e-06,
+      "loss": 0.2886,
+      "step": 6581
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3594994164369227,
+      "learning_rate": 6.608101446670953e-06,
+      "loss": 0.312,
+      "step": 6582
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6020994895289615,
+      "learning_rate": 6.607137049754513e-06,
+      "loss": 0.293,
+      "step": 6583
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.850261649027184,
+      "learning_rate": 6.6061725861551026e-06,
+      "loss": 0.2782,
+      "step": 6584
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5573929837511624,
+      "learning_rate": 6.605208055912737e-06,
+      "loss": 0.3199,
+      "step": 6585
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7566236394019696,
+      "learning_rate": 6.604243459067441e-06,
+      "loss": 0.2998,
+      "step": 6586
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8706880225366673,
+      "learning_rate": 6.6032787956592316e-06,
+      "loss": 0.2885,
+      "step": 6587
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8610202506770097,
+      "learning_rate": 6.602314065728139e-06,
+      "loss": 0.3132,
+      "step": 6588
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7072450253287421,
+      "learning_rate": 6.601349269314188e-06,
+      "loss": 0.2845,
+      "step": 6589
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6233152007016336,
+      "learning_rate": 6.600384406457414e-06,
+      "loss": 0.2859,
+      "step": 6590
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.702646182713347,
+      "learning_rate": 6.599419477197846e-06,
+      "loss": 0.3026,
+      "step": 6591
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8484181919293035,
+      "learning_rate": 6.598454481575525e-06,
+      "loss": 0.3033,
+      "step": 6592
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4374414106581042,
+      "learning_rate": 6.597489419630486e-06,
+      "loss": 0.3008,
+      "step": 6593
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7344994247032326,
+      "learning_rate": 6.5965242914027764e-06,
+      "loss": 0.3069,
+      "step": 6594
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.349672451682584,
+      "learning_rate": 6.595559096932436e-06,
+      "loss": 0.3086,
+      "step": 6595
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0954466858856042,
+      "learning_rate": 6.594593836259516e-06,
+      "loss": 0.2854,
+      "step": 6596
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4831756558014324,
+      "learning_rate": 6.5936285094240635e-06,
+      "loss": 0.3074,
+      "step": 6597
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5382543642269466,
+      "learning_rate": 6.592663116466136e-06,
+      "loss": 0.2879,
+      "step": 6598
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0291490477998524,
+      "learning_rate": 6.591697657425785e-06,
+      "loss": 0.3152,
+      "step": 6599
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.254477060601201,
+      "learning_rate": 6.590732132343072e-06,
+      "loss": 0.2717,
+      "step": 6600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8654577208286884,
+      "learning_rate": 6.589766541258056e-06,
+      "loss": 0.3088,
+      "step": 6601
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7547277008358781,
+      "learning_rate": 6.588800884210804e-06,
+      "loss": 0.2805,
+      "step": 6602
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2004128043591042,
+      "learning_rate": 6.587835161241381e-06,
+      "loss": 0.3029,
+      "step": 6603
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.76608811992652,
+      "learning_rate": 6.586869372389857e-06,
+      "loss": 0.2944,
+      "step": 6604
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8336780119686658,
+      "learning_rate": 6.585903517696304e-06,
+      "loss": 0.3146,
+      "step": 6605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3608484381333867,
+      "learning_rate": 6.584937597200797e-06,
+      "loss": 0.2952,
+      "step": 6606
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7509240104914126,
+      "learning_rate": 6.5839716109434136e-06,
+      "loss": 0.3201,
+      "step": 6607
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9604685934523443,
+      "learning_rate": 6.583005558964235e-06,
+      "loss": 0.2982,
+      "step": 6608
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3077049783515002,
+      "learning_rate": 6.582039441303344e-06,
+      "loss": 0.2804,
+      "step": 6609
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6072040591235521,
+      "learning_rate": 6.581073258000827e-06,
+      "loss": 0.3106,
+      "step": 6610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1569202949236703,
+      "learning_rate": 6.580107009096771e-06,
+      "loss": 0.2857,
+      "step": 6611
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7634977147567996,
+      "learning_rate": 6.57914069463127e-06,
+      "loss": 0.3002,
+      "step": 6612
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9597734943808822,
+      "learning_rate": 6.578174314644416e-06,
+      "loss": 0.3044,
+      "step": 6613
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.544044361929901,
+      "learning_rate": 6.5772078691763065e-06,
+      "loss": 0.2972,
+      "step": 6614
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3856537063237093,
+      "learning_rate": 6.57624135826704e-06,
+      "loss": 0.29,
+      "step": 6615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7474925133654111,
+      "learning_rate": 6.575274781956722e-06,
+      "loss": 0.2909,
+      "step": 6616
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7235652080990869,
+      "learning_rate": 6.574308140285454e-06,
+      "loss": 0.2926,
+      "step": 6617
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6267696018741113,
+      "learning_rate": 6.573341433293345e-06,
+      "loss": 0.3091,
+      "step": 6618
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.279834635014297,
+      "learning_rate": 6.572374661020505e-06,
+      "loss": 0.3002,
+      "step": 6619
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7558010712909244,
+      "learning_rate": 6.571407823507049e-06,
+      "loss": 0.2795,
+      "step": 6620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6294631332275649,
+      "learning_rate": 6.5704409207930905e-06,
+      "loss": 0.2777,
+      "step": 6621
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4934477254417544,
+      "learning_rate": 6.569473952918749e-06,
+      "loss": 0.3045,
+      "step": 6622
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.202872314621412,
+      "learning_rate": 6.5685069199241435e-06,
+      "loss": 0.3073,
+      "step": 6623
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6364301520786926,
+      "learning_rate": 6.567539821849403e-06,
+      "loss": 0.302,
+      "step": 6624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4578145608951738,
+      "learning_rate": 6.566572658734649e-06,
+      "loss": 0.2833,
+      "step": 6625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5653620306932607,
+      "learning_rate": 6.565605430620014e-06,
+      "loss": 0.3066,
+      "step": 6626
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7116355398366587,
+      "learning_rate": 6.564638137545627e-06,
+      "loss": 0.3007,
+      "step": 6627
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3005990082073562,
+      "learning_rate": 6.563670779551627e-06,
+      "loss": 0.2968,
+      "step": 6628
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.512059306827473,
+      "learning_rate": 6.562703356678147e-06,
+      "loss": 0.2852,
+      "step": 6629
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8029879121206913,
+      "learning_rate": 6.56173586896533e-06,
+      "loss": 0.5057,
+      "step": 6630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.51892830915327,
+      "learning_rate": 6.560768316453317e-06,
+      "loss": 0.3069,
+      "step": 6631
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4714159838244183,
+      "learning_rate": 6.5598006991822565e-06,
+      "loss": 0.3107,
+      "step": 6632
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.17130969302371,
+      "learning_rate": 6.558833017192292e-06,
+      "loss": 0.3247,
+      "step": 6633
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0429954347034123,
+      "learning_rate": 6.557865270523578e-06,
+      "loss": 0.2966,
+      "step": 6634
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9615465707278945,
+      "learning_rate": 6.556897459216266e-06,
+      "loss": 0.2923,
+      "step": 6635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0565315431508737,
+      "learning_rate": 6.555929583310515e-06,
+      "loss": 0.2982,
+      "step": 6636
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.637913546854928,
+      "learning_rate": 6.55496164284648e-06,
+      "loss": 0.3019,
+      "step": 6637
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1177319613133587,
+      "learning_rate": 6.553993637864325e-06,
+      "loss": 0.2935,
+      "step": 6638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5943474259786545,
+      "learning_rate": 6.5530255684042145e-06,
+      "loss": 0.2905,
+      "step": 6639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5647836227289802,
+      "learning_rate": 6.552057434506314e-06,
+      "loss": 0.2964,
+      "step": 6640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.300130802169015,
+      "learning_rate": 6.551089236210793e-06,
+      "loss": 0.305,
+      "step": 6641
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.525592303022024,
+      "learning_rate": 6.550120973557825e-06,
+      "loss": 0.3034,
+      "step": 6642
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5047122486921707,
+      "learning_rate": 6.549152646587585e-06,
+      "loss": 0.2963,
+      "step": 6643
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5618472647096326,
+      "learning_rate": 6.548184255340251e-06,
+      "loss": 0.3237,
+      "step": 6644
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5165527372393564,
+      "learning_rate": 6.5472157998560005e-06,
+      "loss": 0.3144,
+      "step": 6645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8126356620836233,
+      "learning_rate": 6.5462472801750195e-06,
+      "loss": 0.2912,
+      "step": 6646
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.288393285526579,
+      "learning_rate": 6.545278696337492e-06,
+      "loss": 0.2845,
+      "step": 6647
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.232613800542668,
+      "learning_rate": 6.544310048383608e-06,
+      "loss": 0.3276,
+      "step": 6648
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9751667329786748,
+      "learning_rate": 6.543341336353555e-06,
+      "loss": 0.31,
+      "step": 6649
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6580555248410977,
+      "learning_rate": 6.54237256028753e-06,
+      "loss": 0.2896,
+      "step": 6650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.140158115245598,
+      "learning_rate": 6.541403720225726e-06,
+      "loss": 0.2798,
+      "step": 6651
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.087865297320279,
+      "learning_rate": 6.5404348162083466e-06,
+      "loss": 0.3037,
+      "step": 6652
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5568908169910936,
+      "learning_rate": 6.539465848275588e-06,
+      "loss": 0.2964,
+      "step": 6653
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7141778926779454,
+      "learning_rate": 6.538496816467658e-06,
+      "loss": 0.501,
+      "step": 6654
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0369483023198427,
+      "learning_rate": 6.537527720824763e-06,
+      "loss": 0.3049,
+      "step": 6655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7767126454898903,
+      "learning_rate": 6.536558561387112e-06,
+      "loss": 0.3242,
+      "step": 6656
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.387913680174168,
+      "learning_rate": 6.535589338194916e-06,
+      "loss": 0.2958,
+      "step": 6657
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7439653011474678,
+      "learning_rate": 6.534620051288392e-06,
+      "loss": 0.2962,
+      "step": 6658
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.1073700321500155,
+      "learning_rate": 6.533650700707755e-06,
+      "loss": 0.2867,
+      "step": 6659
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.72888325269297,
+      "learning_rate": 6.532681286493227e-06,
+      "loss": 0.3015,
+      "step": 6660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.754324138536733,
+      "learning_rate": 6.5317118086850296e-06,
+      "loss": 0.3389,
+      "step": 6661
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5798089948703729,
+      "learning_rate": 6.53074226732339e-06,
+      "loss": 0.2954,
+      "step": 6662
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.357269872813877,
+      "learning_rate": 6.5297726624485335e-06,
+      "loss": 0.2935,
+      "step": 6663
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4157827770860054,
+      "learning_rate": 6.5288029941006924e-06,
+      "loss": 0.2849,
+      "step": 6664
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.483880541105889,
+      "learning_rate": 6.527833262320099e-06,
+      "loss": 0.3013,
+      "step": 6665
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8280625928331715,
+      "learning_rate": 6.526863467146991e-06,
+      "loss": 0.311,
+      "step": 6666
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.137689554284446,
+      "learning_rate": 6.525893608621604e-06,
+      "loss": 0.3009,
+      "step": 6667
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4052662791986477,
+      "learning_rate": 6.524923686784184e-06,
+      "loss": 0.3071,
+      "step": 6668
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9190988175049173,
+      "learning_rate": 6.523953701674969e-06,
+      "loss": 0.3225,
+      "step": 6669
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.4271075268165996,
+      "learning_rate": 6.5229836533342095e-06,
+      "loss": 0.3028,
+      "step": 6670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9156249575803732,
+      "learning_rate": 6.5220135418021515e-06,
+      "loss": 0.3241,
+      "step": 6671
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1031247527236507,
+      "learning_rate": 6.521043367119049e-06,
+      "loss": 0.3017,
+      "step": 6672
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4463256394660986,
+      "learning_rate": 6.520073129325156e-06,
+      "loss": 0.2923,
+      "step": 6673
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0932528750837225,
+      "learning_rate": 6.51910282846073e-06,
+      "loss": 0.2976,
+      "step": 6674
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.325389754089071,
+      "learning_rate": 6.518132464566026e-06,
+      "loss": 0.2991,
+      "step": 6675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4153017818677673,
+      "learning_rate": 6.517162037681313e-06,
+      "loss": 0.2982,
+      "step": 6676
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4919394863188316,
+      "learning_rate": 6.5161915478468495e-06,
+      "loss": 0.2834,
+      "step": 6677
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7836137050152663,
+      "learning_rate": 6.5152209951029085e-06,
+      "loss": 0.2939,
+      "step": 6678
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0392726353543766,
+      "learning_rate": 6.514250379489754e-06,
+      "loss": 0.317,
+      "step": 6679
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.336572065414183,
+      "learning_rate": 6.513279701047663e-06,
+      "loss": 0.2825,
+      "step": 6680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5357636682880287,
+      "learning_rate": 6.51230895981691e-06,
+      "loss": 0.3017,
+      "step": 6681
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.056914598784362,
+      "learning_rate": 6.511338155837772e-06,
+      "loss": 0.3285,
+      "step": 6682
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9321872914816856,
+      "learning_rate": 6.510367289150528e-06,
+      "loss": 0.2908,
+      "step": 6683
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9659965305423102,
+      "learning_rate": 6.509396359795465e-06,
+      "loss": 0.3044,
+      "step": 6684
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.701003316316974,
+      "learning_rate": 6.508425367812864e-06,
+      "loss": 0.3084,
+      "step": 6685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3292562347262862,
+      "learning_rate": 6.507454313243016e-06,
+      "loss": 0.2932,
+      "step": 6686
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8085632941667336,
+      "learning_rate": 6.5064831961262095e-06,
+      "loss": 0.2924,
+      "step": 6687
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.099391928724341,
+      "learning_rate": 6.505512016502742e-06,
+      "loss": 0.293,
+      "step": 6688
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8982831618294798,
+      "learning_rate": 6.504540774412905e-06,
+      "loss": 0.3195,
+      "step": 6689
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8581115161281003,
+      "learning_rate": 6.503569469896999e-06,
+      "loss": 0.2985,
+      "step": 6690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0070324488975997,
+      "learning_rate": 6.502598102995326e-06,
+      "loss": 0.3109,
+      "step": 6691
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5611289647158277,
+      "learning_rate": 6.5016266737481895e-06,
+      "loss": 0.3013,
+      "step": 6692
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1415445046374235,
+      "learning_rate": 6.500655182195893e-06,
+      "loss": 0.2832,
+      "step": 6693
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6257866022093999,
+      "learning_rate": 6.4996836283787475e-06,
+      "loss": 0.2883,
+      "step": 6694
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4598524460289037,
+      "learning_rate": 6.498712012337065e-06,
+      "loss": 0.2936,
+      "step": 6695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7563371977421394,
+      "learning_rate": 6.497740334111161e-06,
+      "loss": 0.3083,
+      "step": 6696
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.528684558497098,
+      "learning_rate": 6.496768593741347e-06,
+      "loss": 0.3136,
+      "step": 6697
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0870146974393315,
+      "learning_rate": 6.495796791267948e-06,
+      "loss": 0.3032,
+      "step": 6698
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1498320571843594,
+      "learning_rate": 6.4948249267312826e-06,
+      "loss": 0.3065,
+      "step": 6699
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7312396090469386,
+      "learning_rate": 6.493853000171677e-06,
+      "loss": 0.3167,
+      "step": 6700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.953251234131047,
+      "learning_rate": 6.492881011629455e-06,
+      "loss": 0.2992,
+      "step": 6701
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4939549599591375,
+      "learning_rate": 6.49190896114495e-06,
+      "loss": 0.31,
+      "step": 6702
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.727483209643662,
+      "learning_rate": 6.4909368487584916e-06,
+      "loss": 0.3,
+      "step": 6703
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.580324751770327,
+      "learning_rate": 6.489964674510415e-06,
+      "loss": 0.3069,
+      "step": 6704
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.682575206260421,
+      "learning_rate": 6.488992438441055e-06,
+      "loss": 0.2941,
+      "step": 6705
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6084755212680026,
+      "learning_rate": 6.488020140590758e-06,
+      "loss": 0.333,
+      "step": 6706
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.414128282407443,
+      "learning_rate": 6.48704778099986e-06,
+      "loss": 0.3224,
+      "step": 6707
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3166081613666436,
+      "learning_rate": 6.486075359708709e-06,
+      "loss": 0.319,
+      "step": 6708
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6840612137367519,
+      "learning_rate": 6.485102876757652e-06,
+      "loss": 0.3168,
+      "step": 6709
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9088392499215119,
+      "learning_rate": 6.48413033218704e-06,
+      "loss": 0.2967,
+      "step": 6710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4688670212016326,
+      "learning_rate": 6.483157726037222e-06,
+      "loss": 0.2856,
+      "step": 6711
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.407240901376062,
+      "learning_rate": 6.482185058348556e-06,
+      "loss": 0.3038,
+      "step": 6712
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7259059175476872,
+      "learning_rate": 6.4812123291614005e-06,
+      "loss": 0.2901,
+      "step": 6713
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.816169264353281,
+      "learning_rate": 6.480239538516114e-06,
+      "loss": 0.2976,
+      "step": 6714
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8517276184680693,
+      "learning_rate": 6.47926668645306e-06,
+      "loss": 0.2997,
+      "step": 6715
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.155593367474794,
+      "learning_rate": 6.478293773012603e-06,
+      "loss": 0.305,
+      "step": 6716
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7211834035333056,
+      "learning_rate": 6.477320798235112e-06,
+      "loss": 0.2883,
+      "step": 6717
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5938584569789214,
+      "learning_rate": 6.476347762160957e-06,
+      "loss": 0.3108,
+      "step": 6718
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8052742745153534,
+      "learning_rate": 6.475374664830512e-06,
+      "loss": 0.2925,
+      "step": 6719
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.2814627037738258,
+      "learning_rate": 6.47440150628415e-06,
+      "loss": 0.297,
+      "step": 6720
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5331285415764486,
+      "learning_rate": 6.473428286562251e-06,
+      "loss": 0.2936,
+      "step": 6721
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.916705578993107,
+      "learning_rate": 6.472455005705197e-06,
+      "loss": 0.2896,
+      "step": 6722
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.634617436509288,
+      "learning_rate": 6.471481663753367e-06,
+      "loss": 0.288,
+      "step": 6723
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1405479001930976,
+      "learning_rate": 6.47050826074715e-06,
+      "loss": 0.3104,
+      "step": 6724
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.829690751521165,
+      "learning_rate": 6.469534796726934e-06,
+      "loss": 0.2899,
+      "step": 6725
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.493429523614867,
+      "learning_rate": 6.4685612717331096e-06,
+      "loss": 0.2882,
+      "step": 6726
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.476047469509268,
+      "learning_rate": 6.467587685806067e-06,
+      "loss": 0.2847,
+      "step": 6727
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.330666995233044,
+      "learning_rate": 6.466614038986208e-06,
+      "loss": 0.2914,
+      "step": 6728
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.179439580181806,
+      "learning_rate": 6.465640331313925e-06,
+      "loss": 0.3025,
+      "step": 6729
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7500517706162204,
+      "learning_rate": 6.464666562829624e-06,
+      "loss": 0.2953,
+      "step": 6730
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.3990291736926075,
+      "learning_rate": 6.4636927335737025e-06,
+      "loss": 0.2996,
+      "step": 6731
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.036953456502084,
+      "learning_rate": 6.462718843586572e-06,
+      "loss": 0.3033,
+      "step": 6732
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6268336755904357,
+      "learning_rate": 6.461744892908637e-06,
+      "loss": 0.4836,
+      "step": 6733
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7624383530294316,
+      "learning_rate": 6.460770881580311e-06,
+      "loss": 0.2899,
+      "step": 6734
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6027756428985964,
+      "learning_rate": 6.4597968096420045e-06,
+      "loss": 0.2991,
+      "step": 6735
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1245425071428894,
+      "learning_rate": 6.4588226771341386e-06,
+      "loss": 0.3101,
+      "step": 6736
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5774516435729113,
+      "learning_rate": 6.457848484097128e-06,
+      "loss": 0.2938,
+      "step": 6737
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.307916791625838,
+      "learning_rate": 6.456874230571393e-06,
+      "loss": 0.3357,
+      "step": 6738
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5779917268889785,
+      "learning_rate": 6.455899916597359e-06,
+      "loss": 0.2867,
+      "step": 6739
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.688366142488379,
+      "learning_rate": 6.4549255422154525e-06,
+      "loss": 0.2887,
+      "step": 6740
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9617952338691722,
+      "learning_rate": 6.4539511074660995e-06,
+      "loss": 0.3096,
+      "step": 6741
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.260581858861846,
+      "learning_rate": 6.452976612389733e-06,
+      "loss": 0.3243,
+      "step": 6742
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.004827513471801,
+      "learning_rate": 6.452002057026786e-06,
+      "loss": 0.2999,
+      "step": 6743
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6405049139436518,
+      "learning_rate": 6.451027441417696e-06,
+      "loss": 0.2742,
+      "step": 6744
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6131278877084996,
+      "learning_rate": 6.4500527656028985e-06,
+      "loss": 0.2981,
+      "step": 6745
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9106438014988696,
+      "learning_rate": 6.449078029622837e-06,
+      "loss": 0.2896,
+      "step": 6746
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8401447823120518,
+      "learning_rate": 6.448103233517954e-06,
+      "loss": 0.3143,
+      "step": 6747
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.879439296180402,
+      "learning_rate": 6.447128377328695e-06,
+      "loss": 0.2922,
+      "step": 6748
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.279784983777409,
+      "learning_rate": 6.4461534610955104e-06,
+      "loss": 0.2867,
+      "step": 6749
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.746662266839023,
+      "learning_rate": 6.445178484858849e-06,
+      "loss": 0.2913,
+      "step": 6750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.671399715811076,
+      "learning_rate": 6.444203448659165e-06,
+      "loss": 0.3099,
+      "step": 6751
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0312387518092354,
+      "learning_rate": 6.443228352536918e-06,
+      "loss": 0.2873,
+      "step": 6752
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9592013263028545,
+      "learning_rate": 6.4422531965325594e-06,
+      "loss": 0.2914,
+      "step": 6753
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6378974765143877,
+      "learning_rate": 6.441277980686556e-06,
+      "loss": 0.3196,
+      "step": 6754
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.90402098774003,
+      "learning_rate": 6.4403027050393675e-06,
+      "loss": 0.3007,
+      "step": 6755
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.739745636095288,
+      "learning_rate": 6.439327369631464e-06,
+      "loss": 0.2911,
+      "step": 6756
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7499854498849223,
+      "learning_rate": 6.438351974503309e-06,
+      "loss": 0.3105,
+      "step": 6757
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.36591451892812,
+      "learning_rate": 6.437376519695376e-06,
+      "loss": 0.3302,
+      "step": 6758
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7624095080741506,
+      "learning_rate": 6.436401005248139e-06,
+      "loss": 0.2895,
+      "step": 6759
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.714746600818347,
+      "learning_rate": 6.435425431202074e-06,
+      "loss": 0.2996,
+      "step": 6760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.379943254084911,
+      "learning_rate": 6.434449797597657e-06,
+      "loss": 0.2836,
+      "step": 6761
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6241433111758001,
+      "learning_rate": 6.433474104475369e-06,
+      "loss": 0.4965,
+      "step": 6762
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5604552983852458,
+      "learning_rate": 6.432498351875696e-06,
+      "loss": 0.2715,
+      "step": 6763
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.704246886300958,
+      "learning_rate": 6.43152253983912e-06,
+      "loss": 0.2903,
+      "step": 6764
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7847322408734418,
+      "learning_rate": 6.430546668406133e-06,
+      "loss": 0.2985,
+      "step": 6765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.539045795134168,
+      "learning_rate": 6.429570737617223e-06,
+      "loss": 0.2816,
+      "step": 6766
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8569052860052397,
+      "learning_rate": 6.428594747512884e-06,
+      "loss": 0.2994,
+      "step": 6767
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9623127045136122,
+      "learning_rate": 6.427618698133612e-06,
+      "loss": 0.2873,
+      "step": 6768
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.505181360047184,
+      "learning_rate": 6.4266425895199036e-06,
+      "loss": 0.2754,
+      "step": 6769
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.008662784013174,
+      "learning_rate": 6.425666421712263e-06,
+      "loss": 0.3048,
+      "step": 6770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7449333988077806,
+      "learning_rate": 6.424690194751186e-06,
+      "loss": 0.3042,
+      "step": 6771
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.11664783106136,
+      "learning_rate": 6.423713908677185e-06,
+      "loss": 0.309,
+      "step": 6772
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3902479893888047,
+      "learning_rate": 6.422737563530763e-06,
+      "loss": 0.3006,
+      "step": 6773
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6467051612233414,
+      "learning_rate": 6.4217611593524355e-06,
+      "loss": 0.2875,
+      "step": 6774
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8336430361345313,
+      "learning_rate": 6.420784696182709e-06,
+      "loss": 0.3061,
+      "step": 6775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7093595735481975,
+      "learning_rate": 6.419808174062103e-06,
+      "loss": 0.2974,
+      "step": 6776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0017537665334055,
+      "learning_rate": 6.418831593031134e-06,
+      "loss": 0.3129,
+      "step": 6777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0562628003301255,
+      "learning_rate": 6.417854953130323e-06,
+      "loss": 0.2976,
+      "step": 6778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.636640367023912,
+      "learning_rate": 6.416878254400191e-06,
+      "loss": 0.2958,
+      "step": 6779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6510810730970376,
+      "learning_rate": 6.4159014968812634e-06,
+      "loss": 0.2977,
+      "step": 6780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8031167414175482,
+      "learning_rate": 6.4149246806140675e-06,
+      "loss": 0.3017,
+      "step": 6781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.480695841073619,
+      "learning_rate": 6.413947805639136e-06,
+      "loss": 0.2872,
+      "step": 6782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5016659168010187,
+      "learning_rate": 6.412970871996995e-06,
+      "loss": 0.2858,
+      "step": 6783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9446304107640044,
+      "learning_rate": 6.411993879728184e-06,
+      "loss": 0.2937,
+      "step": 6784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.314690855240622,
+      "learning_rate": 6.411016828873239e-06,
+      "loss": 0.2796,
+      "step": 6785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5776542150995976,
+      "learning_rate": 6.4100397194727005e-06,
+      "loss": 0.4739,
+      "step": 6786
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.40466950989219,
+      "learning_rate": 6.409062551567109e-06,
+      "loss": 0.3025,
+      "step": 6787
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6651469506478778,
+      "learning_rate": 6.4080853251970086e-06,
+      "loss": 0.2947,
+      "step": 6788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1833000065185026,
+      "learning_rate": 6.4071080404029475e-06,
+      "loss": 0.2975,
+      "step": 6789
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8783876707876814,
+      "learning_rate": 6.4061306972254745e-06,
+      "loss": 0.3256,
+      "step": 6790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6929911454243527,
+      "learning_rate": 6.405153295705142e-06,
+      "loss": 0.3157,
+      "step": 6791
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0992830830375038,
+      "learning_rate": 6.404175835882503e-06,
+      "loss": 0.3031,
+      "step": 6792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6012424203390276,
+      "learning_rate": 6.403198317798113e-06,
+      "loss": 0.2832,
+      "step": 6793
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1412361975886807,
+      "learning_rate": 6.402220741492533e-06,
+      "loss": 0.3084,
+      "step": 6794
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8586675520090896,
+      "learning_rate": 6.401243107006325e-06,
+      "loss": 0.2929,
+      "step": 6795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5001244380557788,
+      "learning_rate": 6.4002654143800515e-06,
+      "loss": 0.2797,
+      "step": 6796
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8021882489270977,
+      "learning_rate": 6.399287663654279e-06,
+      "loss": 0.3144,
+      "step": 6797
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.561296588065728,
+      "learning_rate": 6.398309854869574e-06,
+      "loss": 0.2905,
+      "step": 6798
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6822116196263686,
+      "learning_rate": 6.397331988066512e-06,
+      "loss": 0.3022,
+      "step": 6799
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.59133152600341,
+      "learning_rate": 6.396354063285662e-06,
+      "loss": 0.321,
+      "step": 6800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7238356491332714,
+      "learning_rate": 6.395376080567602e-06,
+      "loss": 0.2834,
+      "step": 6801
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8237679990774085,
+      "learning_rate": 6.394398039952911e-06,
+      "loss": 0.2856,
+      "step": 6802
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0978506014726537,
+      "learning_rate": 6.3934199414821674e-06,
+      "loss": 0.2922,
+      "step": 6803
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6291710724231745,
+      "learning_rate": 6.392441785195956e-06,
+      "loss": 0.2884,
+      "step": 6804
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6154631905494288,
+      "learning_rate": 6.391463571134862e-06,
+      "loss": 0.3206,
+      "step": 6805
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8472115736003967,
+      "learning_rate": 6.390485299339473e-06,
+      "loss": 0.3124,
+      "step": 6806
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.049227778719141,
+      "learning_rate": 6.389506969850378e-06,
+      "loss": 0.3071,
+      "step": 6807
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.852184541794767,
+      "learning_rate": 6.3885285827081725e-06,
+      "loss": 0.2909,
+      "step": 6808
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.210604253205481,
+      "learning_rate": 6.3875501379534486e-06,
+      "loss": 0.2947,
+      "step": 6809
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.195088755565158,
+      "learning_rate": 6.386571635626804e-06,
+      "loss": 0.3426,
+      "step": 6810
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.254386928581936,
+      "learning_rate": 6.3855930757688415e-06,
+      "loss": 0.2929,
+      "step": 6811
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.427413066891106,
+      "learning_rate": 6.38461445842016e-06,
+      "loss": 0.3237,
+      "step": 6812
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.479301294544089,
+      "learning_rate": 6.383635783621365e-06,
+      "loss": 0.3242,
+      "step": 6813
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.4322326984415525,
+      "learning_rate": 6.382657051413063e-06,
+      "loss": 0.3136,
+      "step": 6814
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.673907078170811,
+      "learning_rate": 6.3816782618358666e-06,
+      "loss": 0.2763,
+      "step": 6815
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5224796805783334,
+      "learning_rate": 6.380699414930385e-06,
+      "loss": 0.2943,
+      "step": 6816
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9402496872929773,
+      "learning_rate": 6.37972051073723e-06,
+      "loss": 0.2899,
+      "step": 6817
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8635711574000435,
+      "learning_rate": 6.378741549297021e-06,
+      "loss": 0.2881,
+      "step": 6818
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8063934042696637,
+      "learning_rate": 6.377762530650375e-06,
+      "loss": 0.3006,
+      "step": 6819
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.094089157008058,
+      "learning_rate": 6.376783454837916e-06,
+      "loss": 0.3159,
+      "step": 6820
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7428949729006693,
+      "learning_rate": 6.375804321900267e-06,
+      "loss": 0.2843,
+      "step": 6821
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7019580540891321,
+      "learning_rate": 6.3748251318780514e-06,
+      "loss": 0.2909,
+      "step": 6822
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6504712380753666,
+      "learning_rate": 6.3738458848119e-06,
+      "loss": 0.2845,
+      "step": 6823
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.885256138973961,
+      "learning_rate": 6.372866580742442e-06,
+      "loss": 0.2956,
+      "step": 6824
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5142776791261525,
+      "learning_rate": 6.371887219710312e-06,
+      "loss": 0.2929,
+      "step": 6825
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5168769328775624,
+      "learning_rate": 6.370907801756143e-06,
+      "loss": 0.2959,
+      "step": 6826
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6316484169195389,
+      "learning_rate": 6.369928326920575e-06,
+      "loss": 0.2942,
+      "step": 6827
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.365074744515247,
+      "learning_rate": 6.368948795244247e-06,
+      "loss": 0.2868,
+      "step": 6828
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8919705280839447,
+      "learning_rate": 6.367969206767803e-06,
+      "loss": 0.3115,
+      "step": 6829
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8577146692649886,
+      "learning_rate": 6.366989561531887e-06,
+      "loss": 0.28,
+      "step": 6830
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.27404396869779,
+      "learning_rate": 6.3660098595771445e-06,
+      "loss": 0.3005,
+      "step": 6831
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7231158169295044,
+      "learning_rate": 6.365030100944227e-06,
+      "loss": 0.2921,
+      "step": 6832
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8452785989556784,
+      "learning_rate": 6.364050285673788e-06,
+      "loss": 0.3021,
+      "step": 6833
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.034796840341878,
+      "learning_rate": 6.363070413806478e-06,
+      "loss": 0.3077,
+      "step": 6834
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0652641545771186,
+      "learning_rate": 6.362090485382956e-06,
+      "loss": 0.29,
+      "step": 6835
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.953695334345325,
+      "learning_rate": 6.361110500443879e-06,
+      "loss": 0.2861,
+      "step": 6836
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.583631036849745,
+      "learning_rate": 6.360130459029912e-06,
+      "loss": 0.3059,
+      "step": 6837
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4901758759164343,
+      "learning_rate": 6.3591503611817155e-06,
+      "loss": 0.3115,
+      "step": 6838
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.829612896761817,
+      "learning_rate": 6.358170206939955e-06,
+      "loss": 0.3155,
+      "step": 6839
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8677532522239744,
+      "learning_rate": 6.357189996345302e-06,
+      "loss": 0.2933,
+      "step": 6840
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.431188282024258,
+      "learning_rate": 6.356209729438425e-06,
+      "loss": 0.3105,
+      "step": 6841
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2546802077530046,
+      "learning_rate": 6.3552294062599975e-06,
+      "loss": 0.3002,
+      "step": 6842
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2200722143856417,
+      "learning_rate": 6.354249026850694e-06,
+      "loss": 0.3055,
+      "step": 6843
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5011573402567864,
+      "learning_rate": 6.3532685912511934e-06,
+      "loss": 0.2855,
+      "step": 6844
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2676207427517316,
+      "learning_rate": 6.352288099502175e-06,
+      "loss": 0.299,
+      "step": 6845
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.963298980139748,
+      "learning_rate": 6.351307551644322e-06,
+      "loss": 0.3154,
+      "step": 6846
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4566492371647595,
+      "learning_rate": 6.350326947718319e-06,
+      "loss": 0.3119,
+      "step": 6847
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.859494361893289,
+      "learning_rate": 6.3493462877648515e-06,
+      "loss": 0.3009,
+      "step": 6848
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.36861430609139,
+      "learning_rate": 6.348365571824611e-06,
+      "loss": 0.2866,
+      "step": 6849
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8902335631050846,
+      "learning_rate": 6.3473847999382855e-06,
+      "loss": 0.3038,
+      "step": 6850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.148905293926362,
+      "learning_rate": 6.346403972146574e-06,
+      "loss": 0.3024,
+      "step": 6851
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6120394598504009,
+      "learning_rate": 6.345423088490169e-06,
+      "loss": 0.5123,
+      "step": 6852
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.936424933189742,
+      "learning_rate": 6.344442149009771e-06,
+      "loss": 0.2978,
+      "step": 6853
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4006229347917698,
+      "learning_rate": 6.343461153746079e-06,
+      "loss": 0.2899,
+      "step": 6854
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.043274006634977,
+      "learning_rate": 6.3424801027397984e-06,
+      "loss": 0.3094,
+      "step": 6855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3576158863718646,
+      "learning_rate": 6.3414989960316345e-06,
+      "loss": 0.2981,
+      "step": 6856
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.41163007617088,
+      "learning_rate": 6.340517833662293e-06,
+      "loss": 0.2937,
+      "step": 6857
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.530216208151628,
+      "learning_rate": 6.339536615672486e-06,
+      "loss": 0.2888,
+      "step": 6858
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.315749199744174,
+      "learning_rate": 6.338555342102927e-06,
+      "loss": 0.2903,
+      "step": 6859
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.2305350401404063,
+      "learning_rate": 6.337574012994327e-06,
+      "loss": 0.3106,
+      "step": 6860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6948606536470603,
+      "learning_rate": 6.336592628387407e-06,
+      "loss": 0.3062,
+      "step": 6861
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6648834760085593,
+      "learning_rate": 6.335611188322883e-06,
+      "loss": 0.3025,
+      "step": 6862
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.526308257063972,
+      "learning_rate": 6.334629692841481e-06,
+      "loss": 0.2867,
+      "step": 6863
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4595612981056614,
+      "learning_rate": 6.333648141983921e-06,
+      "loss": 0.3038,
+      "step": 6864
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.71043145840467,
+      "learning_rate": 6.33266653579093e-06,
+      "loss": 0.2847,
+      "step": 6865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5690542517115196,
+      "learning_rate": 6.3316848743032385e-06,
+      "loss": 0.3124,
+      "step": 6866
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.705659968142546,
+      "learning_rate": 6.3307031575615775e-06,
+      "loss": 0.292,
+      "step": 6867
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.960419911846101,
+      "learning_rate": 6.329721385606676e-06,
+      "loss": 0.2902,
+      "step": 6868
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4939959391311788,
+      "learning_rate": 6.328739558479275e-06,
+      "loss": 0.3122,
+      "step": 6869
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9057049387463765,
+      "learning_rate": 6.3277576762201074e-06,
+      "loss": 0.332,
+      "step": 6870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.010189182878927,
+      "learning_rate": 6.326775738869917e-06,
+      "loss": 0.2978,
+      "step": 6871
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6184330075214888,
+      "learning_rate": 6.325793746469443e-06,
+      "loss": 0.3035,
+      "step": 6872
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5469203734541837,
+      "learning_rate": 6.324811699059433e-06,
+      "loss": 0.2852,
+      "step": 6873
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6770697859282753,
+      "learning_rate": 6.323829596680633e-06,
+      "loss": 0.2777,
+      "step": 6874
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5636152530996292,
+      "learning_rate": 6.322847439373792e-06,
+      "loss": 0.2983,
+      "step": 6875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6705672976398467,
+      "learning_rate": 6.321865227179658e-06,
+      "loss": 0.2815,
+      "step": 6876
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5980499810557056,
+      "learning_rate": 6.3208829601389896e-06,
+      "loss": 0.2957,
+      "step": 6877
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.736204496034844,
+      "learning_rate": 6.319900638292541e-06,
+      "loss": 0.2895,
+      "step": 6878
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7046260857418942,
+      "learning_rate": 6.318918261681072e-06,
+      "loss": 0.2889,
+      "step": 6879
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.848262816668194,
+      "learning_rate": 6.3179358303453386e-06,
+      "loss": 0.2919,
+      "step": 6880
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 28.675346970086203,
+      "learning_rate": 6.3169533443261085e-06,
+      "loss": 0.3188,
+      "step": 6881
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5106934754378571,
+      "learning_rate": 6.315970803664145e-06,
+      "loss": 0.2992,
+      "step": 6882
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4680564361738777,
+      "learning_rate": 6.314988208400215e-06,
+      "loss": 0.2846,
+      "step": 6883
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.740275830738221,
+      "learning_rate": 6.314005558575089e-06,
+      "loss": 0.3061,
+      "step": 6884
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.59199399670399,
+      "learning_rate": 6.313022854229539e-06,
+      "loss": 0.2955,
+      "step": 6885
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6104358404222554,
+      "learning_rate": 6.312040095404337e-06,
+      "loss": 0.4985,
+      "step": 6886
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0916533426754236,
+      "learning_rate": 6.311057282140261e-06,
+      "loss": 0.2856,
+      "step": 6887
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7000841777319762,
+      "learning_rate": 6.310074414478091e-06,
+      "loss": 0.3012,
+      "step": 6888
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6799428434908734,
+      "learning_rate": 6.309091492458608e-06,
+      "loss": 0.2927,
+      "step": 6889
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.172910443727807,
+      "learning_rate": 6.308108516122591e-06,
+      "loss": 0.3175,
+      "step": 6890
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.277465820846939,
+      "learning_rate": 6.307125485510829e-06,
+      "loss": 0.2921,
+      "step": 6891
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.569782192125306,
+      "learning_rate": 6.306142400664108e-06,
+      "loss": 0.2996,
+      "step": 6892
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.142878579832889,
+      "learning_rate": 6.30515926162322e-06,
+      "loss": 0.2881,
+      "step": 6893
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5730768397650798,
+      "learning_rate": 6.304176068428957e-06,
+      "loss": 0.2816,
+      "step": 6894
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7981238689247325,
+      "learning_rate": 6.30319282112211e-06,
+      "loss": 0.2968,
+      "step": 6895
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5272437486330543,
+      "learning_rate": 6.30220951974348e-06,
+      "loss": 0.2975,
+      "step": 6896
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5952759735211863,
+      "learning_rate": 6.3012261643338635e-06,
+      "loss": 0.2968,
+      "step": 6897
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7209101307714316,
+      "learning_rate": 6.3002427549340615e-06,
+      "loss": 0.285,
+      "step": 6898
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5355351233576975,
+      "learning_rate": 6.299259291584879e-06,
+      "loss": 0.3108,
+      "step": 6899
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.328437310208531,
+      "learning_rate": 6.298275774327121e-06,
+      "loss": 0.3472,
+      "step": 6900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6083503246623474,
+      "learning_rate": 6.297292203201595e-06,
+      "loss": 0.2988,
+      "step": 6901
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9392954330231182,
+      "learning_rate": 6.296308578249109e-06,
+      "loss": 0.3219,
+      "step": 6902
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.639787763238278,
+      "learning_rate": 6.29532489951048e-06,
+      "loss": 0.5236,
+      "step": 6903
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.932409822949063,
+      "learning_rate": 6.29434116702652e-06,
+      "loss": 0.3096,
+      "step": 6904
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.038587484729636,
+      "learning_rate": 6.293357380838046e-06,
+      "loss": 0.3124,
+      "step": 6905
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6437779728204444,
+      "learning_rate": 6.292373540985875e-06,
+      "loss": 0.544,
+      "step": 6906
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4183458562866678,
+      "learning_rate": 6.291389647510833e-06,
+      "loss": 0.2946,
+      "step": 6907
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5418600018117234,
+      "learning_rate": 6.2904057004537385e-06,
+      "loss": 0.3101,
+      "step": 6908
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7109504249161431,
+      "learning_rate": 6.2894216998554215e-06,
+      "loss": 0.2858,
+      "step": 6909
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8419552482693984,
+      "learning_rate": 6.288437645756706e-06,
+      "loss": 0.2864,
+      "step": 6910
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.85911169554198,
+      "learning_rate": 6.287453538198426e-06,
+      "loss": 0.311,
+      "step": 6911
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5054704968118577,
+      "learning_rate": 6.28646937722141e-06,
+      "loss": 0.2945,
+      "step": 6912
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.157171929352471,
+      "learning_rate": 6.285485162866496e-06,
+      "loss": 0.2975,
+      "step": 6913
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.093106304030806,
+      "learning_rate": 6.284500895174518e-06,
+      "loss": 0.2969,
+      "step": 6914
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.951128097908572,
+      "learning_rate": 6.283516574186318e-06,
+      "loss": 0.3002,
+      "step": 6915
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7873416091672234,
+      "learning_rate": 6.282532199942735e-06,
+      "loss": 0.308,
+      "step": 6916
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5250801150628495,
+      "learning_rate": 6.281547772484612e-06,
+      "loss": 0.3029,
+      "step": 6917
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3239861226448197,
+      "learning_rate": 6.280563291852796e-06,
+      "loss": 0.2981,
+      "step": 6918
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.896650528267158,
+      "learning_rate": 6.279578758088135e-06,
+      "loss": 0.2865,
+      "step": 6919
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.046720401907298,
+      "learning_rate": 6.278594171231478e-06,
+      "loss": 0.2971,
+      "step": 6920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.107336850805136,
+      "learning_rate": 6.277609531323678e-06,
+      "loss": 0.3215,
+      "step": 6921
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.557701835265859,
+      "learning_rate": 6.27662483840559e-06,
+      "loss": 0.3184,
+      "step": 6922
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8742361321090764,
+      "learning_rate": 6.27564009251807e-06,
+      "loss": 0.2952,
+      "step": 6923
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2743838864681423,
+      "learning_rate": 6.274655293701974e-06,
+      "loss": 0.285,
+      "step": 6924
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9320725595818318,
+      "learning_rate": 6.273670441998169e-06,
+      "loss": 0.2882,
+      "step": 6925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.645404943097612,
+      "learning_rate": 6.272685537447513e-06,
+      "loss": 0.2664,
+      "step": 6926
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2758878763486856,
+      "learning_rate": 6.271700580090876e-06,
+      "loss": 0.311,
+      "step": 6927
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9482575278035108,
+      "learning_rate": 6.270715569969119e-06,
+      "loss": 0.3052,
+      "step": 6928
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.112387212977741,
+      "learning_rate": 6.26973050712312e-06,
+      "loss": 0.3033,
+      "step": 6929
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7241744926684435,
+      "learning_rate": 6.2687453915937445e-06,
+      "loss": 0.3023,
+      "step": 6930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0555602708856138,
+      "learning_rate": 6.267760223421871e-06,
+      "loss": 0.2969,
+      "step": 6931
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6248297912514691,
+      "learning_rate": 6.266775002648373e-06,
+      "loss": 0.5389,
+      "step": 6932
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.037528832629987,
+      "learning_rate": 6.265789729314131e-06,
+      "loss": 0.2989,
+      "step": 6933
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6910461415496076,
+      "learning_rate": 6.264804403460023e-06,
+      "loss": 0.3171,
+      "step": 6934
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1443824498795303,
+      "learning_rate": 6.263819025126936e-06,
+      "loss": 0.2882,
+      "step": 6935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0070005688932357,
+      "learning_rate": 6.2628335943557505e-06,
+      "loss": 0.2792,
+      "step": 6936
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6577178825907559,
+      "learning_rate": 6.261848111187358e-06,
+      "loss": 0.5032,
+      "step": 6937
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2880460372140314,
+      "learning_rate": 6.260862575662645e-06,
+      "loss": 0.311,
+      "step": 6938
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2988816930377083,
+      "learning_rate": 6.259876987822506e-06,
+      "loss": 0.2874,
+      "step": 6939
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.303903898880498,
+      "learning_rate": 6.258891347707829e-06,
+      "loss": 0.2868,
+      "step": 6940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4554178697573767,
+      "learning_rate": 6.257905655359519e-06,
+      "loss": 0.2776,
+      "step": 6941
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3989185717964019,
+      "learning_rate": 6.256919910818466e-06,
+      "loss": 0.3024,
+      "step": 6942
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4713631313250355,
+      "learning_rate": 6.255934114125574e-06,
+      "loss": 0.2867,
+      "step": 6943
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.477810837095802,
+      "learning_rate": 6.254948265321744e-06,
+      "loss": 0.2911,
+      "step": 6944
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9950190859446826,
+      "learning_rate": 6.253962364447883e-06,
+      "loss": 0.2826,
+      "step": 6945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5048232846671787,
+      "learning_rate": 6.252976411544896e-06,
+      "loss": 0.2893,
+      "step": 6946
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.81103356279454,
+      "learning_rate": 6.251990406653691e-06,
+      "loss": 0.3192,
+      "step": 6947
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.050149692739653,
+      "learning_rate": 6.25100434981518e-06,
+      "loss": 0.3046,
+      "step": 6948
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.911724694471854,
+      "learning_rate": 6.250018241070278e-06,
+      "loss": 0.2898,
+      "step": 6949
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8625527564946367,
+      "learning_rate": 6.249032080459898e-06,
+      "loss": 0.296,
+      "step": 6950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.71810994009719,
+      "learning_rate": 6.248045868024958e-06,
+      "loss": 0.2963,
+      "step": 6951
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6184633602532053,
+      "learning_rate": 6.247059603806379e-06,
+      "loss": 0.287,
+      "step": 6952
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.176729576462319,
+      "learning_rate": 6.246073287845083e-06,
+      "loss": 0.2843,
+      "step": 6953
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2516712664615466,
+      "learning_rate": 6.24508692018199e-06,
+      "loss": 0.2998,
+      "step": 6954
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0138997086144292,
+      "learning_rate": 6.2441005008580314e-06,
+      "loss": 0.3118,
+      "step": 6955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3472168870737558,
+      "learning_rate": 6.243114029914133e-06,
+      "loss": 0.284,
+      "step": 6956
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.171887894936276,
+      "learning_rate": 6.242127507391226e-06,
+      "loss": 0.2945,
+      "step": 6957
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.359229503674128,
+      "learning_rate": 6.241140933330241e-06,
+      "loss": 0.2897,
+      "step": 6958
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8023195721854788,
+      "learning_rate": 6.240154307772115e-06,
+      "loss": 0.2934,
+      "step": 6959
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7584205824226065,
+      "learning_rate": 6.2391676307577845e-06,
+      "loss": 0.2928,
+      "step": 6960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3600575699249644,
+      "learning_rate": 6.238180902328188e-06,
+      "loss": 0.2819,
+      "step": 6961
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.928848767948197,
+      "learning_rate": 6.237194122524264e-06,
+      "loss": 0.2941,
+      "step": 6962
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.866835490502667,
+      "learning_rate": 6.236207291386962e-06,
+      "loss": 0.3024,
+      "step": 6963
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4273734466129238,
+      "learning_rate": 6.235220408957221e-06,
+      "loss": 0.3057,
+      "step": 6964
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.862038667793041,
+      "learning_rate": 6.234233475275994e-06,
+      "loss": 0.2852,
+      "step": 6965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8983752565715197,
+      "learning_rate": 6.233246490384224e-06,
+      "loss": 0.2887,
+      "step": 6966
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7021755181041338,
+      "learning_rate": 6.23225945432287e-06,
+      "loss": 0.2842,
+      "step": 6967
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.287035243422365,
+      "learning_rate": 6.231272367132881e-06,
+      "loss": 0.3123,
+      "step": 6968
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.945247621903697,
+      "learning_rate": 6.230285228855215e-06,
+      "loss": 0.2919,
+      "step": 6969
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.020562013384556,
+      "learning_rate": 6.229298039530829e-06,
+      "loss": 0.2861,
+      "step": 6970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0372056156931793,
+      "learning_rate": 6.228310799200685e-06,
+      "loss": 0.2948,
+      "step": 6971
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.912464584260742,
+      "learning_rate": 6.227323507905743e-06,
+      "loss": 0.2911,
+      "step": 6972
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.374210068437158,
+      "learning_rate": 6.226336165686969e-06,
+      "loss": 0.2929,
+      "step": 6973
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.135244812766038,
+      "learning_rate": 6.225348772585329e-06,
+      "loss": 0.3076,
+      "step": 6974
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6406959607573017,
+      "learning_rate": 6.224361328641794e-06,
+      "loss": 0.2881,
+      "step": 6975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1404839011362387,
+      "learning_rate": 6.2233738338973304e-06,
+      "loss": 0.2953,
+      "step": 6976
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.649553312960678,
+      "learning_rate": 6.222386288392914e-06,
+      "loss": 0.307,
+      "step": 6977
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.513140057397445,
+      "learning_rate": 6.2213986921695194e-06,
+      "loss": 0.3009,
+      "step": 6978
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.273120614566552,
+      "learning_rate": 6.220411045268124e-06,
+      "loss": 0.2908,
+      "step": 6979
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.083398440398737,
+      "learning_rate": 6.219423347729707e-06,
+      "loss": 0.2883,
+      "step": 6980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8022818181447064,
+      "learning_rate": 6.218435599595249e-06,
+      "loss": 0.3016,
+      "step": 6981
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.715627016313357,
+      "learning_rate": 6.217447800905733e-06,
+      "loss": 0.2806,
+      "step": 6982
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9147989051437198,
+      "learning_rate": 6.216459951702146e-06,
+      "loss": 0.2956,
+      "step": 6983
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.57260093166668,
+      "learning_rate": 6.215472052025474e-06,
+      "loss": 0.2859,
+      "step": 6984
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6807125697290495,
+      "learning_rate": 6.214484101916709e-06,
+      "loss": 0.2853,
+      "step": 6985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.712395827031748,
+      "learning_rate": 6.21349610141684e-06,
+      "loss": 0.3045,
+      "step": 6986
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 58.18177499566314,
+      "learning_rate": 6.2125080505668645e-06,
+      "loss": 0.3021,
+      "step": 6987
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5855958975202866,
+      "learning_rate": 6.2115199494077735e-06,
+      "loss": 0.2961,
+      "step": 6988
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.754485949036334,
+      "learning_rate": 6.210531797980571e-06,
+      "loss": 0.3101,
+      "step": 6989
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8998780353349038,
+      "learning_rate": 6.2095435963262514e-06,
+      "loss": 0.3012,
+      "step": 6990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6426158708593297,
+      "learning_rate": 6.2085553444858225e-06,
+      "loss": 0.2826,
+      "step": 6991
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.342512165812579,
+      "learning_rate": 6.207567042500282e-06,
+      "loss": 0.2853,
+      "step": 6992
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.042777282932608,
+      "learning_rate": 6.206578690410643e-06,
+      "loss": 0.277,
+      "step": 6993
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6647666484805896,
+      "learning_rate": 6.20559028825791e-06,
+      "loss": 0.2834,
+      "step": 6994
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5568308132031212,
+      "learning_rate": 6.204601836083094e-06,
+      "loss": 0.2961,
+      "step": 6995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0035566784926324,
+      "learning_rate": 6.203613333927209e-06,
+      "loss": 0.2787,
+      "step": 6996
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3794714496850886,
+      "learning_rate": 6.202624781831269e-06,
+      "loss": 0.2754,
+      "step": 6997
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.683218158441574,
+      "learning_rate": 6.20163617983629e-06,
+      "loss": 0.303,
+      "step": 6998
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.316001094462051,
+      "learning_rate": 6.200647527983292e-06,
+      "loss": 0.2835,
+      "step": 6999
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7957678371881527,
+      "learning_rate": 6.199658826313295e-06,
+      "loss": 0.3096,
+      "step": 7000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.139733121461754,
+      "learning_rate": 6.198670074867324e-06,
+      "loss": 0.3103,
+      "step": 7001
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4758325644328265,
+      "learning_rate": 6.197681273686401e-06,
+      "loss": 0.2833,
+      "step": 7002
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.786111703777581,
+      "learning_rate": 6.196692422811554e-06,
+      "loss": 0.3245,
+      "step": 7003
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4195571136952108,
+      "learning_rate": 6.195703522283813e-06,
+      "loss": 0.304,
+      "step": 7004
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.368769143360143,
+      "learning_rate": 6.194714572144212e-06,
+      "loss": 0.3047,
+      "step": 7005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4142774733542174,
+      "learning_rate": 6.193725572433779e-06,
+      "loss": 0.3021,
+      "step": 7006
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3621891793787175,
+      "learning_rate": 6.192736523193551e-06,
+      "loss": 0.3102,
+      "step": 7007
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.34851259692707,
+      "learning_rate": 6.191747424464567e-06,
+      "loss": 0.306,
+      "step": 7008
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2927614161375742,
+      "learning_rate": 6.1907582762878675e-06,
+      "loss": 0.3111,
+      "step": 7009
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2750354088498064,
+      "learning_rate": 6.189769078704489e-06,
+      "loss": 0.3084,
+      "step": 7010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.013965366878603,
+      "learning_rate": 6.188779831755479e-06,
+      "loss": 0.3017,
+      "step": 7011
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.013557669244629,
+      "learning_rate": 6.1877905354818825e-06,
+      "loss": 0.3053,
+      "step": 7012
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3582681305222066,
+      "learning_rate": 6.186801189924748e-06,
+      "loss": 0.2781,
+      "step": 7013
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.750226125509539,
+      "learning_rate": 6.185811795125122e-06,
+      "loss": 0.309,
+      "step": 7014
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.786131507164553,
+      "learning_rate": 6.18482235112406e-06,
+      "loss": 0.2924,
+      "step": 7015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8849568031186652,
+      "learning_rate": 6.183832857962614e-06,
+      "loss": 0.2992,
+      "step": 7016
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.856593453702831,
+      "learning_rate": 6.18284331568184e-06,
+      "loss": 0.3275,
+      "step": 7017
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2274692600465977,
+      "learning_rate": 6.181853724322795e-06,
+      "loss": 0.2941,
+      "step": 7018
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.120222048101711,
+      "learning_rate": 6.180864083926541e-06,
+      "loss": 0.281,
+      "step": 7019
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9364150542022918,
+      "learning_rate": 6.179874394534138e-06,
+      "loss": 0.3085,
+      "step": 7020
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.128182292913789,
+      "learning_rate": 6.178884656186651e-06,
+      "loss": 0.2761,
+      "step": 7021
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6459217754761926,
+      "learning_rate": 6.177894868925144e-06,
+      "loss": 0.2888,
+      "step": 7022
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.379192067164824,
+      "learning_rate": 6.17690503279069e-06,
+      "loss": 0.306,
+      "step": 7023
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3443047043338927,
+      "learning_rate": 6.175915147824353e-06,
+      "loss": 0.2965,
+      "step": 7024
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9122964560256457,
+      "learning_rate": 6.174925214067209e-06,
+      "loss": 0.3075,
+      "step": 7025
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7078695101143218,
+      "learning_rate": 6.17393523156033e-06,
+      "loss": 0.2822,
+      "step": 7026
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4347276812654037,
+      "learning_rate": 6.172945200344794e-06,
+      "loss": 0.2816,
+      "step": 7027
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.052861955587367,
+      "learning_rate": 6.171955120461679e-06,
+      "loss": 0.2809,
+      "step": 7028
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6833519806050177,
+      "learning_rate": 6.170964991952063e-06,
+      "loss": 0.3065,
+      "step": 7029
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4351329321545907,
+      "learning_rate": 6.169974814857029e-06,
+      "loss": 0.2893,
+      "step": 7030
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6541129622906818,
+      "learning_rate": 6.168984589217665e-06,
+      "loss": 0.4677,
+      "step": 7031
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7955176642312072,
+      "learning_rate": 6.167994315075051e-06,
+      "loss": 0.3035,
+      "step": 7032
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.068203637383184,
+      "learning_rate": 6.16700399247028e-06,
+      "loss": 0.3021,
+      "step": 7033
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5691572599151804,
+      "learning_rate": 6.166013621444439e-06,
+      "loss": 0.293,
+      "step": 7034
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7879507597793993,
+      "learning_rate": 6.165023202038623e-06,
+      "loss": 0.2894,
+      "step": 7035
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1276244932199386,
+      "learning_rate": 6.164032734293923e-06,
+      "loss": 0.2785,
+      "step": 7036
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.676875938315613,
+      "learning_rate": 6.163042218251441e-06,
+      "loss": 0.313,
+      "step": 7037
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8636923280359605,
+      "learning_rate": 6.162051653952268e-06,
+      "loss": 0.2848,
+      "step": 7038
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8956087264028727,
+      "learning_rate": 6.16106104143751e-06,
+      "loss": 0.3253,
+      "step": 7039
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.0370163004082045,
+      "learning_rate": 6.160070380748266e-06,
+      "loss": 0.3339,
+      "step": 7040
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3509695033589586,
+      "learning_rate": 6.159079671925643e-06,
+      "loss": 0.2997,
+      "step": 7041
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.296569875515288,
+      "learning_rate": 6.158088915010744e-06,
+      "loss": 0.2879,
+      "step": 7042
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6044115897038609,
+      "learning_rate": 6.15709811004468e-06,
+      "loss": 0.4974,
+      "step": 7043
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8354352805137628,
+      "learning_rate": 6.156107257068558e-06,
+      "loss": 0.2906,
+      "step": 7044
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2308941184507693,
+      "learning_rate": 6.155116356123495e-06,
+      "loss": 0.2823,
+      "step": 7045
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.911403847403431,
+      "learning_rate": 6.154125407250602e-06,
+      "loss": 0.29,
+      "step": 7046
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.47474944203617,
+      "learning_rate": 6.153134410490995e-06,
+      "loss": 0.2883,
+      "step": 7047
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3271239921105271,
+      "learning_rate": 6.152143365885794e-06,
+      "loss": 0.2917,
+      "step": 7048
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9133580788293076,
+      "learning_rate": 6.151152273476118e-06,
+      "loss": 0.2858,
+      "step": 7049
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.410731303828012,
+      "learning_rate": 6.150161133303088e-06,
+      "loss": 0.3041,
+      "step": 7050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.48746589224747,
+      "learning_rate": 6.149169945407832e-06,
+      "loss": 0.2927,
+      "step": 7051
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2059173948238775,
+      "learning_rate": 6.1481787098314725e-06,
+      "loss": 0.3055,
+      "step": 7052
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0858129348336796,
+      "learning_rate": 6.14718742661514e-06,
+      "loss": 0.2933,
+      "step": 7053
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4960797857699237,
+      "learning_rate": 6.146196095799963e-06,
+      "loss": 0.2749,
+      "step": 7054
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.904959603471647,
+      "learning_rate": 6.145204717427073e-06,
+      "loss": 0.291,
+      "step": 7055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2726555804621618,
+      "learning_rate": 6.144213291537606e-06,
+      "loss": 0.2991,
+      "step": 7056
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5150798268392562,
+      "learning_rate": 6.143221818172699e-06,
+      "loss": 0.2961,
+      "step": 7057
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.619053693609467,
+      "learning_rate": 6.142230297373486e-06,
+      "loss": 0.299,
+      "step": 7058
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5399819273112656,
+      "learning_rate": 6.141238729181109e-06,
+      "loss": 0.2867,
+      "step": 7059
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4249270772240146,
+      "learning_rate": 6.1402471136367115e-06,
+      "loss": 0.2939,
+      "step": 7060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.078587070086647,
+      "learning_rate": 6.139255450781436e-06,
+      "loss": 0.2876,
+      "step": 7061
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5371946901276299,
+      "learning_rate": 6.138263740656427e-06,
+      "loss": 0.2646,
+      "step": 7062
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.154862380993804,
+      "learning_rate": 6.137271983302834e-06,
+      "loss": 0.312,
+      "step": 7063
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2056076224871501,
+      "learning_rate": 6.136280178761806e-06,
+      "loss": 0.2954,
+      "step": 7064
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7749927945554256,
+      "learning_rate": 6.135288327074497e-06,
+      "loss": 0.3172,
+      "step": 7065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8188161739814512,
+      "learning_rate": 6.134296428282056e-06,
+      "loss": 0.3134,
+      "step": 7066
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.233217469619812,
+      "learning_rate": 6.1333044824256435e-06,
+      "loss": 0.2885,
+      "step": 7067
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.975546975137299,
+      "learning_rate": 6.132312489546414e-06,
+      "loss": 0.2984,
+      "step": 7068
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8913759033808324,
+      "learning_rate": 6.131320449685529e-06,
+      "loss": 0.2788,
+      "step": 7069
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.477800206370894,
+      "learning_rate": 6.130328362884148e-06,
+      "loss": 0.2956,
+      "step": 7070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6984344336266934,
+      "learning_rate": 6.129336229183437e-06,
+      "loss": 0.518,
+      "step": 7071
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8583193101830235,
+      "learning_rate": 6.128344048624558e-06,
+      "loss": 0.3035,
+      "step": 7072
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2840903028232784,
+      "learning_rate": 6.1273518212486825e-06,
+      "loss": 0.2788,
+      "step": 7073
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.548828807577619,
+      "learning_rate": 6.126359547096975e-06,
+      "loss": 0.2797,
+      "step": 7074
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.519022640767992,
+      "learning_rate": 6.125367226210612e-06,
+      "loss": 0.299,
+      "step": 7075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4147106372018525,
+      "learning_rate": 6.124374858630762e-06,
+      "loss": 0.2934,
+      "step": 7076
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6580370895773484,
+      "learning_rate": 6.123382444398603e-06,
+      "loss": 0.2755,
+      "step": 7077
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3587927852990556,
+      "learning_rate": 6.12238998355531e-06,
+      "loss": 0.2892,
+      "step": 7078
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.578125520614798,
+      "learning_rate": 6.121397476142064e-06,
+      "loss": 0.2821,
+      "step": 7079
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.038925348742282,
+      "learning_rate": 6.1204049222000435e-06,
+      "loss": 0.2789,
+      "step": 7080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3745016399428254,
+      "learning_rate": 6.1194123217704336e-06,
+      "loss": 0.2845,
+      "step": 7081
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7100527980503495,
+      "learning_rate": 6.118419674894418e-06,
+      "loss": 0.2932,
+      "step": 7082
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.379610724171847,
+      "learning_rate": 6.1174269816131845e-06,
+      "loss": 0.3309,
+      "step": 7083
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2657517778028917,
+      "learning_rate": 6.11643424196792e-06,
+      "loss": 0.295,
+      "step": 7084
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.733993237194969,
+      "learning_rate": 6.115441455999816e-06,
+      "loss": 0.2892,
+      "step": 7085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.657008334448093,
+      "learning_rate": 6.114448623750065e-06,
+      "loss": 0.279,
+      "step": 7086
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3213655311496555,
+      "learning_rate": 6.113455745259861e-06,
+      "loss": 0.3175,
+      "step": 7087
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5280282867737696,
+      "learning_rate": 6.112462820570399e-06,
+      "loss": 0.3049,
+      "step": 7088
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.972235837101983,
+      "learning_rate": 6.11146984972288e-06,
+      "loss": 0.3146,
+      "step": 7089
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5848359505232334,
+      "learning_rate": 6.110476832758503e-06,
+      "loss": 0.4673,
+      "step": 7090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6857764963936,
+      "learning_rate": 6.10948376971847e-06,
+      "loss": 0.2994,
+      "step": 7091
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8886980918386227,
+      "learning_rate": 6.108490660643982e-06,
+      "loss": 0.2876,
+      "step": 7092
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.040459490022842,
+      "learning_rate": 6.107497505576251e-06,
+      "loss": 0.2765,
+      "step": 7093
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.509208148055233,
+      "learning_rate": 6.106504304556479e-06,
+      "loss": 0.2864,
+      "step": 7094
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6197656977857953,
+      "learning_rate": 6.10551105762588e-06,
+      "loss": 0.3046,
+      "step": 7095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9027807338901899,
+      "learning_rate": 6.104517764825662e-06,
+      "loss": 0.2777,
+      "step": 7096
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.766224200303339,
+      "learning_rate": 6.103524426197041e-06,
+      "loss": 0.2775,
+      "step": 7097
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.965032226299235,
+      "learning_rate": 6.10253104178123e-06,
+      "loss": 0.2851,
+      "step": 7098
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9785809208109495,
+      "learning_rate": 6.1015376116194506e-06,
+      "loss": 0.3002,
+      "step": 7099
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3618764576495743,
+      "learning_rate": 6.100544135752916e-06,
+      "loss": 0.2931,
+      "step": 7100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8428649776226034,
+      "learning_rate": 6.0995506142228525e-06,
+      "loss": 0.2921,
+      "step": 7101
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0358845575392865,
+      "learning_rate": 6.09855704707048e-06,
+      "loss": 0.2974,
+      "step": 7102
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.832942096883757,
+      "learning_rate": 6.097563434337026e-06,
+      "loss": 0.2927,
+      "step": 7103
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.005828899782704,
+      "learning_rate": 6.096569776063712e-06,
+      "loss": 0.2955,
+      "step": 7104
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1884957411659056,
+      "learning_rate": 6.095576072291774e-06,
+      "loss": 0.3011,
+      "step": 7105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7264328630247696,
+      "learning_rate": 6.094582323062437e-06,
+      "loss": 0.2942,
+      "step": 7106
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.5422946025736,
+      "learning_rate": 6.093588528416936e-06,
+      "loss": 0.2919,
+      "step": 7107
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.149446507980943,
+      "learning_rate": 6.092594688396504e-06,
+      "loss": 0.3032,
+      "step": 7108
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.714015273782947,
+      "learning_rate": 6.091600803042378e-06,
+      "loss": 0.2971,
+      "step": 7109
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6346469554531453,
+      "learning_rate": 6.090606872395796e-06,
+      "loss": 0.2971,
+      "step": 7110
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6839766515658936,
+      "learning_rate": 6.089612896497996e-06,
+      "loss": 0.311,
+      "step": 7111
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.251165125405875,
+      "learning_rate": 6.088618875390223e-06,
+      "loss": 0.3107,
+      "step": 7112
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.228872572673163,
+      "learning_rate": 6.087624809113721e-06,
+      "loss": 0.2842,
+      "step": 7113
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5442661757751215,
+      "learning_rate": 6.086630697709731e-06,
+      "loss": 0.2978,
+      "step": 7114
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2380844371099577,
+      "learning_rate": 6.085636541219504e-06,
+      "loss": 0.2898,
+      "step": 7115
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.403116211780939,
+      "learning_rate": 6.084642339684289e-06,
+      "loss": 0.3015,
+      "step": 7116
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8748136352845692,
+      "learning_rate": 6.083648093145337e-06,
+      "loss": 0.3157,
+      "step": 7117
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.264386049161767,
+      "learning_rate": 6.0826538016439e-06,
+      "loss": 0.3019,
+      "step": 7118
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.424593499904079,
+      "learning_rate": 6.081659465221234e-06,
+      "loss": 0.2966,
+      "step": 7119
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2664871301678624,
+      "learning_rate": 6.080665083918595e-06,
+      "loss": 0.2923,
+      "step": 7120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6016217598693909,
+      "learning_rate": 6.079670657777244e-06,
+      "loss": 0.5003,
+      "step": 7121
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.599686531993162,
+      "learning_rate": 6.078676186838438e-06,
+      "loss": 0.2986,
+      "step": 7122
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0419308709735158,
+      "learning_rate": 6.077681671143443e-06,
+      "loss": 0.3194,
+      "step": 7123
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.3146080493616905,
+      "learning_rate": 6.076687110733519e-06,
+      "loss": 0.2942,
+      "step": 7124
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3781187233316055,
+      "learning_rate": 6.075692505649937e-06,
+      "loss": 0.2602,
+      "step": 7125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3047037457388875,
+      "learning_rate": 6.074697855933959e-06,
+      "loss": 0.2953,
+      "step": 7126
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7933948377616522,
+      "learning_rate": 6.07370316162686e-06,
+      "loss": 0.2808,
+      "step": 7127
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8672118045593933,
+      "learning_rate": 6.0727084227699095e-06,
+      "loss": 0.2843,
+      "step": 7128
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9693002900450496,
+      "learning_rate": 6.071713639404382e-06,
+      "loss": 0.2911,
+      "step": 7129
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8233998369605118,
+      "learning_rate": 6.07071881157155e-06,
+      "loss": 0.2838,
+      "step": 7130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.053406986259955,
+      "learning_rate": 6.069723939312695e-06,
+      "loss": 0.3018,
+      "step": 7131
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7077429588707012,
+      "learning_rate": 6.068729022669092e-06,
+      "loss": 0.2779,
+      "step": 7132
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6434443393824358,
+      "learning_rate": 6.067734061682024e-06,
+      "loss": 0.2968,
+      "step": 7133
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0316381344492025,
+      "learning_rate": 6.066739056392774e-06,
+      "loss": 0.2954,
+      "step": 7134
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5362265329115818,
+      "learning_rate": 6.065744006842626e-06,
+      "loss": 0.2978,
+      "step": 7135
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.415171399932464,
+      "learning_rate": 6.064748913072864e-06,
+      "loss": 0.2853,
+      "step": 7136
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6488797212567636,
+      "learning_rate": 6.063753775124781e-06,
+      "loss": 0.291,
+      "step": 7137
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6571025766959937,
+      "learning_rate": 6.062758593039663e-06,
+      "loss": 0.2965,
+      "step": 7138
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4006475600524397,
+      "learning_rate": 6.061763366858804e-06,
+      "loss": 0.2868,
+      "step": 7139
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4657827633756655,
+      "learning_rate": 6.060768096623496e-06,
+      "loss": 0.3023,
+      "step": 7140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.710688928026125,
+      "learning_rate": 6.059772782375036e-06,
+      "loss": 0.2826,
+      "step": 7141
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5995991529604687,
+      "learning_rate": 6.0587774241547205e-06,
+      "loss": 0.3118,
+      "step": 7142
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.041110035251263,
+      "learning_rate": 6.057782022003851e-06,
+      "loss": 0.2813,
+      "step": 7143
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0509229914424894,
+      "learning_rate": 6.056786575963725e-06,
+      "loss": 0.2976,
+      "step": 7144
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1445565628745213,
+      "learning_rate": 6.0557910860756466e-06,
+      "loss": 0.2924,
+      "step": 7145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.20498732398269,
+      "learning_rate": 6.054795552380921e-06,
+      "loss": 0.2961,
+      "step": 7146
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8094318107666756,
+      "learning_rate": 6.053799974920856e-06,
+      "loss": 0.2857,
+      "step": 7147
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4924062943715422,
+      "learning_rate": 6.052804353736757e-06,
+      "loss": 0.2792,
+      "step": 7148
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.540322723520255,
+      "learning_rate": 6.051808688869934e-06,
+      "loss": 0.3007,
+      "step": 7149
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.777149268579538,
+      "learning_rate": 6.050812980361701e-06,
+      "loss": 0.3074,
+      "step": 7150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3407254052984399,
+      "learning_rate": 6.049817228253373e-06,
+      "loss": 0.277,
+      "step": 7151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.7574284251759975,
+      "learning_rate": 6.048821432586261e-06,
+      "loss": 0.3029,
+      "step": 7152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4172601326843792,
+      "learning_rate": 6.047825593401686e-06,
+      "loss": 0.2824,
+      "step": 7153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7614750685381748,
+      "learning_rate": 6.046829710740966e-06,
+      "loss": 0.302,
+      "step": 7154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.925876730667548,
+      "learning_rate": 6.045833784645422e-06,
+      "loss": 0.304,
+      "step": 7155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8993360446893568,
+      "learning_rate": 6.044837815156377e-06,
+      "loss": 0.2811,
+      "step": 7156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5832044248453947,
+      "learning_rate": 6.043841802315153e-06,
+      "loss": 0.3002,
+      "step": 7157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.597899906254344,
+      "learning_rate": 6.042845746163081e-06,
+      "loss": 0.3098,
+      "step": 7158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2392190828824825,
+      "learning_rate": 6.041849646741485e-06,
+      "loss": 0.3133,
+      "step": 7159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.049087140610685,
+      "learning_rate": 6.040853504091698e-06,
+      "loss": 0.2963,
+      "step": 7160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7356790660117334,
+      "learning_rate": 6.03985731825505e-06,
+      "loss": 0.2759,
+      "step": 7161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5465860338521598,
+      "learning_rate": 6.038861089272875e-06,
+      "loss": 0.2939,
+      "step": 7162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.560534782086251,
+      "learning_rate": 6.037864817186507e-06,
+      "loss": 0.2859,
+      "step": 7163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.841044845563695,
+      "learning_rate": 6.036868502037286e-06,
+      "loss": 0.2788,
+      "step": 7164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0138936133192953,
+      "learning_rate": 6.035872143866549e-06,
+      "loss": 0.296,
+      "step": 7165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4743239005657713,
+      "learning_rate": 6.034875742715636e-06,
+      "loss": 0.296,
+      "step": 7166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.88237633275076,
+      "learning_rate": 6.03387929862589e-06,
+      "loss": 0.3117,
+      "step": 7167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.762908556869072,
+      "learning_rate": 6.032882811638656e-06,
+      "loss": 0.3031,
+      "step": 7168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.416071537999233,
+      "learning_rate": 6.03188628179528e-06,
+      "loss": 0.2716,
+      "step": 7169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3807843710024055,
+      "learning_rate": 6.030889709137109e-06,
+      "loss": 0.2826,
+      "step": 7170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.425970117898785,
+      "learning_rate": 6.029893093705492e-06,
+      "loss": 0.2963,
+      "step": 7171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.606691631289159,
+      "learning_rate": 6.0288964355417825e-06,
+      "loss": 0.317,
+      "step": 7172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7422822195154766,
+      "learning_rate": 6.027899734687332e-06,
+      "loss": 0.2906,
+      "step": 7173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6710875179664204,
+      "learning_rate": 6.026902991183496e-06,
+      "loss": 0.3053,
+      "step": 7174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.757845693594874,
+      "learning_rate": 6.02590620507163e-06,
+      "loss": 0.3094,
+      "step": 7175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.299915257306338,
+      "learning_rate": 6.024909376393093e-06,
+      "loss": 0.2785,
+      "step": 7176
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3634373999122538,
+      "learning_rate": 6.0239125051892475e-06,
+      "loss": 0.2949,
+      "step": 7177
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.150206470673079,
+      "learning_rate": 6.022915591501453e-06,
+      "loss": 0.2912,
+      "step": 7178
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8093004842615352,
+      "learning_rate": 6.021918635371072e-06,
+      "loss": 0.2975,
+      "step": 7179
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.29026863303202,
+      "learning_rate": 6.020921636839473e-06,
+      "loss": 0.2986,
+      "step": 7180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6955111512391063,
+      "learning_rate": 6.019924595948022e-06,
+      "loss": 0.289,
+      "step": 7181
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.773420507313563,
+      "learning_rate": 6.018927512738088e-06,
+      "loss": 0.2971,
+      "step": 7182
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4514447964474035,
+      "learning_rate": 6.017930387251041e-06,
+      "loss": 0.2911,
+      "step": 7183
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4629645084668805,
+      "learning_rate": 6.016933219528255e-06,
+      "loss": 0.3011,
+      "step": 7184
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3663085418713634,
+      "learning_rate": 6.015936009611103e-06,
+      "loss": 0.2928,
+      "step": 7185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1752263871073754,
+      "learning_rate": 6.0149387575409615e-06,
+      "loss": 0.3098,
+      "step": 7186
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6604531010115278,
+      "learning_rate": 6.0139414633592075e-06,
+      "loss": 0.286,
+      "step": 7187
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3047216009249556,
+      "learning_rate": 6.012944127107222e-06,
+      "loss": 0.2945,
+      "step": 7188
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6102446940460937,
+      "learning_rate": 6.0119467488263846e-06,
+      "loss": 0.3039,
+      "step": 7189
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6873860243434702,
+      "learning_rate": 6.010949328558081e-06,
+      "loss": 0.2813,
+      "step": 7190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.65250739275473,
+      "learning_rate": 6.009951866343693e-06,
+      "loss": 0.2974,
+      "step": 7191
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.136638892242125,
+      "learning_rate": 6.008954362224608e-06,
+      "loss": 0.3219,
+      "step": 7192
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.76969378823849,
+      "learning_rate": 6.007956816242214e-06,
+      "loss": 0.3051,
+      "step": 7193
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0477884946942178,
+      "learning_rate": 6.006959228437903e-06,
+      "loss": 0.279,
+      "step": 7194
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.955981559417683,
+      "learning_rate": 6.005961598853064e-06,
+      "loss": 0.289,
+      "step": 7195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8725487337804585,
+      "learning_rate": 6.004963927529092e-06,
+      "loss": 0.2863,
+      "step": 7196
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5115666860324817,
+      "learning_rate": 6.00396621450738e-06,
+      "loss": 0.2904,
+      "step": 7197
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.806800120613164,
+      "learning_rate": 6.002968459829328e-06,
+      "loss": 0.3106,
+      "step": 7198
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.963758732342158,
+      "learning_rate": 6.001970663536333e-06,
+      "loss": 0.2889,
+      "step": 7199
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 194.9734329297264,
+      "learning_rate": 6.000972825669795e-06,
+      "loss": 0.3226,
+      "step": 7200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5076472658835243,
+      "learning_rate": 5.999974946271116e-06,
+      "loss": 0.2848,
+      "step": 7201
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.935928815908976,
+      "learning_rate": 5.9989770253817015e-06,
+      "loss": 0.2931,
+      "step": 7202
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.462734348506588,
+      "learning_rate": 5.9979790630429556e-06,
+      "loss": 0.2916,
+      "step": 7203
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.087409253017042,
+      "learning_rate": 5.9969810592962866e-06,
+      "loss": 0.3014,
+      "step": 7204
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3780574766695843,
+      "learning_rate": 5.995983014183101e-06,
+      "loss": 0.3041,
+      "step": 7205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.939480393315704,
+      "learning_rate": 5.994984927744812e-06,
+      "loss": 0.2984,
+      "step": 7206
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6145280272522786,
+      "learning_rate": 5.993986800022831e-06,
+      "loss": 0.2807,
+      "step": 7207
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8932343800479923,
+      "learning_rate": 5.9929886310585715e-06,
+      "loss": 0.2725,
+      "step": 7208
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6408730271203036,
+      "learning_rate": 5.9919904208934495e-06,
+      "loss": 0.2866,
+      "step": 7209
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.057829410640411,
+      "learning_rate": 5.990992169568884e-06,
+      "loss": 0.3055,
+      "step": 7210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0850932435888914,
+      "learning_rate": 5.989993877126294e-06,
+      "loss": 0.2901,
+      "step": 7211
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.826839017585909,
+      "learning_rate": 5.988995543607099e-06,
+      "loss": 0.2915,
+      "step": 7212
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.125635241380182,
+      "learning_rate": 5.9879971690527205e-06,
+      "loss": 0.3033,
+      "step": 7213
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8525267201779592,
+      "learning_rate": 5.986998753504586e-06,
+      "loss": 0.2862,
+      "step": 7214
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.721278197258601,
+      "learning_rate": 5.98600029700412e-06,
+      "loss": 0.3181,
+      "step": 7215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8176466356005263,
+      "learning_rate": 5.985001799592751e-06,
+      "loss": 0.2982,
+      "step": 7216
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6227108712118903,
+      "learning_rate": 5.984003261311907e-06,
+      "loss": 0.2962,
+      "step": 7217
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6973316147915964,
+      "learning_rate": 5.983004682203019e-06,
+      "loss": 0.2946,
+      "step": 7218
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8593563400881514,
+      "learning_rate": 5.982006062307523e-06,
+      "loss": 0.2914,
+      "step": 7219
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6367525378794925,
+      "learning_rate": 5.9810074016668505e-06,
+      "loss": 0.3104,
+      "step": 7220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2345639766843854,
+      "learning_rate": 5.980008700322437e-06,
+      "loss": 0.2808,
+      "step": 7221
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.1959260099266358,
+      "learning_rate": 5.979009958315724e-06,
+      "loss": 0.2763,
+      "step": 7222
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.869002819082661,
+      "learning_rate": 5.978011175688146e-06,
+      "loss": 0.3071,
+      "step": 7223
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.07459957347487,
+      "learning_rate": 5.977012352481151e-06,
+      "loss": 0.2802,
+      "step": 7224
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8806389119383606,
+      "learning_rate": 5.976013488736176e-06,
+      "loss": 0.3045,
+      "step": 7225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9032865003983352,
+      "learning_rate": 5.975014584494666e-06,
+      "loss": 0.2754,
+      "step": 7226
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6699520016809288,
+      "learning_rate": 5.974015639798071e-06,
+      "loss": 0.3022,
+      "step": 7227
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.76294421370008,
+      "learning_rate": 5.973016654687838e-06,
+      "loss": 0.3242,
+      "step": 7228
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2832268842075085,
+      "learning_rate": 5.972017629205414e-06,
+      "loss": 0.2946,
+      "step": 7229
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.221390776982483,
+      "learning_rate": 5.971018563392252e-06,
+      "loss": 0.2823,
+      "step": 7230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.806604932737318,
+      "learning_rate": 5.970019457289807e-06,
+      "loss": 0.2818,
+      "step": 7231
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.002898366321213,
+      "learning_rate": 5.969020310939531e-06,
+      "loss": 0.2873,
+      "step": 7232
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7389739878325832,
+      "learning_rate": 5.96802112438288e-06,
+      "loss": 0.2862,
+      "step": 7233
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7332056198816865,
+      "learning_rate": 5.967021897661313e-06,
+      "loss": 0.3007,
+      "step": 7234
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8028012010013845,
+      "learning_rate": 5.96602263081629e-06,
+      "loss": 0.3179,
+      "step": 7235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0725242141711324,
+      "learning_rate": 5.965023323889274e-06,
+      "loss": 0.3128,
+      "step": 7236
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6802336999132215,
+      "learning_rate": 5.964023976921726e-06,
+      "loss": 0.2844,
+      "step": 7237
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1904381635886274,
+      "learning_rate": 5.963024589955109e-06,
+      "loss": 0.3195,
+      "step": 7238
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3641856927142584,
+      "learning_rate": 5.962025163030892e-06,
+      "loss": 0.2782,
+      "step": 7239
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.806119631520209,
+      "learning_rate": 5.961025696190542e-06,
+      "loss": 0.293,
+      "step": 7240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.6191839085668445,
+      "learning_rate": 5.96002618947553e-06,
+      "loss": 0.2976,
+      "step": 7241
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.4019028021268545,
+      "learning_rate": 5.959026642927326e-06,
+      "loss": 0.2995,
+      "step": 7242
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.688824631444094,
+      "learning_rate": 5.958027056587402e-06,
+      "loss": 0.2929,
+      "step": 7243
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6488963726063965,
+      "learning_rate": 5.957027430497234e-06,
+      "loss": 0.303,
+      "step": 7244
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7708326852048644,
+      "learning_rate": 5.956027764698299e-06,
+      "loss": 0.2841,
+      "step": 7245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5246818748926652,
+      "learning_rate": 5.955028059232074e-06,
+      "loss": 0.2819,
+      "step": 7246
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.3463251772205735,
+      "learning_rate": 5.9540283141400375e-06,
+      "loss": 0.312,
+      "step": 7247
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.583674965530944,
+      "learning_rate": 5.953028529463671e-06,
+      "loss": 0.2778,
+      "step": 7248
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2052136113771246,
+      "learning_rate": 5.952028705244461e-06,
+      "loss": 0.2943,
+      "step": 7249
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6632695069557133,
+      "learning_rate": 5.951028841523887e-06,
+      "loss": 0.2919,
+      "step": 7250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9484240863350015,
+      "learning_rate": 5.9500289383434375e-06,
+      "loss": 0.2834,
+      "step": 7251
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.637149342076611,
+      "learning_rate": 5.949028995744599e-06,
+      "loss": 0.2985,
+      "step": 7252
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.3914995576076095,
+      "learning_rate": 5.9480290137688626e-06,
+      "loss": 0.287,
+      "step": 7253
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.931397503932855,
+      "learning_rate": 5.94702899245772e-06,
+      "loss": 0.3006,
+      "step": 7254
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8063217585051436,
+      "learning_rate": 5.94602893185266e-06,
+      "loss": 0.2959,
+      "step": 7255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.289267573423351,
+      "learning_rate": 5.94502883199518e-06,
+      "loss": 0.2917,
+      "step": 7256
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8022680327182692,
+      "learning_rate": 5.944028692926776e-06,
+      "loss": 0.3005,
+      "step": 7257
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 29.957254472344218,
+      "learning_rate": 5.943028514688944e-06,
+      "loss": 0.3,
+      "step": 7258
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.832096501221565,
+      "learning_rate": 5.942028297323183e-06,
+      "loss": 0.2954,
+      "step": 7259
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.930825432132346,
+      "learning_rate": 5.941028040870997e-06,
+      "loss": 0.297,
+      "step": 7260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0185155568750943,
+      "learning_rate": 5.940027745373884e-06,
+      "loss": 0.3066,
+      "step": 7261
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.288879995901175,
+      "learning_rate": 5.9390274108733515e-06,
+      "loss": 0.2902,
+      "step": 7262
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.686703416547137,
+      "learning_rate": 5.938027037410903e-06,
+      "loss": 0.2868,
+      "step": 7263
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1451843873584933,
+      "learning_rate": 5.937026625028047e-06,
+      "loss": 0.2829,
+      "step": 7264
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1755340959470746,
+      "learning_rate": 5.936026173766292e-06,
+      "loss": 0.318,
+      "step": 7265
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 55.750355936623194,
+      "learning_rate": 5.9350256836671494e-06,
+      "loss": 0.29,
+      "step": 7266
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.765183249115334,
+      "learning_rate": 5.934025154772128e-06,
+      "loss": 0.2789,
+      "step": 7267
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9681713972256576,
+      "learning_rate": 5.933024587122745e-06,
+      "loss": 0.3115,
+      "step": 7268
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1816787712382673,
+      "learning_rate": 5.932023980760515e-06,
+      "loss": 0.292,
+      "step": 7269
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6475029963947472,
+      "learning_rate": 5.931023335726957e-06,
+      "loss": 0.2901,
+      "step": 7270
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6838641190179063,
+      "learning_rate": 5.930022652063583e-06,
+      "loss": 0.5215,
+      "step": 7271
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5108281733733957,
+      "learning_rate": 5.929021929811919e-06,
+      "loss": 0.2849,
+      "step": 7272
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.712095615158612,
+      "learning_rate": 5.928021169013485e-06,
+      "loss": 0.2846,
+      "step": 7273
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.674838096644239,
+      "learning_rate": 5.927020369709805e-06,
+      "loss": 0.2866,
+      "step": 7274
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.024535918173516,
+      "learning_rate": 5.926019531942401e-06,
+      "loss": 0.2937,
+      "step": 7275
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4882969756059152,
+      "learning_rate": 5.925018655752804e-06,
+      "loss": 0.3078,
+      "step": 7276
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.561645532711883,
+      "learning_rate": 5.924017741182539e-06,
+      "loss": 0.3147,
+      "step": 7277
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6572663485292467,
+      "learning_rate": 5.923016788273138e-06,
+      "loss": 0.2933,
+      "step": 7278
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.308643302273684,
+      "learning_rate": 5.922015797066129e-06,
+      "loss": 0.2912,
+      "step": 7279
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.030645572286068,
+      "learning_rate": 5.921014767603048e-06,
+      "loss": 0.3169,
+      "step": 7280
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.39587674260693,
+      "learning_rate": 5.920013699925429e-06,
+      "loss": 0.295,
+      "step": 7281
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 112.0111833183104,
+      "learning_rate": 5.919012594074807e-06,
+      "loss": 0.2982,
+      "step": 7282
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6589209320407328,
+      "learning_rate": 5.9180114500927185e-06,
+      "loss": 0.2857,
+      "step": 7283
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2727747773501745,
+      "learning_rate": 5.917010268020707e-06,
+      "loss": 0.2886,
+      "step": 7284
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.849798585055369,
+      "learning_rate": 5.91600904790031e-06,
+      "loss": 0.2879,
+      "step": 7285
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9573511836423125,
+      "learning_rate": 5.915007789773069e-06,
+      "loss": 0.3066,
+      "step": 7286
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8373904254532667,
+      "learning_rate": 5.914006493680531e-06,
+      "loss": 0.2846,
+      "step": 7287
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.30294072982048,
+      "learning_rate": 5.9130051596642404e-06,
+      "loss": 0.2938,
+      "step": 7288
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.657875267381038,
+      "learning_rate": 5.912003787765743e-06,
+      "loss": 0.2852,
+      "step": 7289
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.433293948338587,
+      "learning_rate": 5.911002378026588e-06,
+      "loss": 0.2856,
+      "step": 7290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.630769748538676,
+      "learning_rate": 5.910000930488326e-06,
+      "loss": 0.2885,
+      "step": 7291
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8859035035346345,
+      "learning_rate": 5.908999445192513e-06,
+      "loss": 0.2901,
+      "step": 7292
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.250733228068876,
+      "learning_rate": 5.907997922180695e-06,
+      "loss": 0.2854,
+      "step": 7293
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0362654451623787,
+      "learning_rate": 5.90699636149443e-06,
+      "loss": 0.3085,
+      "step": 7294
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9214597344762963,
+      "learning_rate": 5.905994763175275e-06,
+      "loss": 0.2974,
+      "step": 7295
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5592251145918676,
+      "learning_rate": 5.9049931272647905e-06,
+      "loss": 0.2989,
+      "step": 7296
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.145934769889822,
+      "learning_rate": 5.903991453804532e-06,
+      "loss": 0.304,
+      "step": 7297
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7282478379612267,
+      "learning_rate": 5.9029897428360625e-06,
+      "loss": 0.4977,
+      "step": 7298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.846397462762834,
+      "learning_rate": 5.9019879944009454e-06,
+      "loss": 0.2994,
+      "step": 7299
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.584589394237719,
+      "learning_rate": 5.900986208540744e-06,
+      "loss": 0.3091,
+      "step": 7300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9359017420102511,
+      "learning_rate": 5.899984385297023e-06,
+      "loss": 0.2829,
+      "step": 7301
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3744193182835036,
+      "learning_rate": 5.898982524711353e-06,
+      "loss": 0.2976,
+      "step": 7302
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8582866800424578,
+      "learning_rate": 5.897980626825301e-06,
+      "loss": 0.2817,
+      "step": 7303
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2194878079083267,
+      "learning_rate": 5.896978691680438e-06,
+      "loss": 0.2885,
+      "step": 7304
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6776804598703174,
+      "learning_rate": 5.895976719318335e-06,
+      "loss": 0.2838,
+      "step": 7305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5756283658154336,
+      "learning_rate": 5.894974709780568e-06,
+      "loss": 0.4982,
+      "step": 7306
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1569390316349253,
+      "learning_rate": 5.893972663108709e-06,
+      "loss": 0.286,
+      "step": 7307
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.669584606805467,
+      "learning_rate": 5.892970579344337e-06,
+      "loss": 0.2945,
+      "step": 7308
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7325601149234242,
+      "learning_rate": 5.891968458529028e-06,
+      "loss": 0.3276,
+      "step": 7309
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3263148341652418,
+      "learning_rate": 5.890966300704366e-06,
+      "loss": 0.293,
+      "step": 7310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6667154132344058,
+      "learning_rate": 5.8899641059119284e-06,
+      "loss": 0.2831,
+      "step": 7311
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8787594973105606,
+      "learning_rate": 5.8889618741933e-06,
+      "loss": 0.3063,
+      "step": 7312
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9704987007244803,
+      "learning_rate": 5.887959605590062e-06,
+      "loss": 0.274,
+      "step": 7313
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0981803359192868,
+      "learning_rate": 5.886957300143806e-06,
+      "loss": 0.2906,
+      "step": 7314
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.1888955643199,
+      "learning_rate": 5.885954957896115e-06,
+      "loss": 0.2862,
+      "step": 7315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.018121329207774,
+      "learning_rate": 5.884952578888578e-06,
+      "loss": 0.2933,
+      "step": 7316
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7170643080728223,
+      "learning_rate": 5.883950163162788e-06,
+      "loss": 0.3004,
+      "step": 7317
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.44992481760076,
+      "learning_rate": 5.882947710760336e-06,
+      "loss": 0.3015,
+      "step": 7318
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7227972049219455,
+      "learning_rate": 5.881945221722815e-06,
+      "loss": 0.2998,
+      "step": 7319
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4908067710610826,
+      "learning_rate": 5.880942696091818e-06,
+      "loss": 0.2789,
+      "step": 7320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8873695098000953,
+      "learning_rate": 5.879940133908946e-06,
+      "loss": 0.3,
+      "step": 7321
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2268818679339066,
+      "learning_rate": 5.878937535215795e-06,
+      "loss": 0.3084,
+      "step": 7322
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.069477399588808,
+      "learning_rate": 5.877934900053963e-06,
+      "loss": 0.2885,
+      "step": 7323
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.766883386039501,
+      "learning_rate": 5.876932228465054e-06,
+      "loss": 0.2943,
+      "step": 7324
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5011298653926668,
+      "learning_rate": 5.875929520490669e-06,
+      "loss": 0.2968,
+      "step": 7325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8427409423855354,
+      "learning_rate": 5.874926776172413e-06,
+      "loss": 0.2957,
+      "step": 7326
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.78595019446095,
+      "learning_rate": 5.873923995551888e-06,
+      "loss": 0.2875,
+      "step": 7327
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9306178224589718,
+      "learning_rate": 5.8729211786707075e-06,
+      "loss": 0.3156,
+      "step": 7328
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.985264088394038,
+      "learning_rate": 5.871918325570475e-06,
+      "loss": 0.3139,
+      "step": 7329
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2983958776720996,
+      "learning_rate": 5.8709154362928045e-06,
+      "loss": 0.2781,
+      "step": 7330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5459953800347166,
+      "learning_rate": 5.869912510879303e-06,
+      "loss": 0.3131,
+      "step": 7331
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.871828278846024,
+      "learning_rate": 5.868909549371588e-06,
+      "loss": 0.298,
+      "step": 7332
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6170950853305737,
+      "learning_rate": 5.867906551811271e-06,
+      "loss": 0.2908,
+      "step": 7333
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.107788983459267,
+      "learning_rate": 5.866903518239973e-06,
+      "loss": 0.284,
+      "step": 7334
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.4097658497869885,
+      "learning_rate": 5.865900448699304e-06,
+      "loss": 0.299,
+      "step": 7335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8048049716625916,
+      "learning_rate": 5.864897343230889e-06,
+      "loss": 0.2967,
+      "step": 7336
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9191555674538232,
+      "learning_rate": 5.8638942018763466e-06,
+      "loss": 0.2897,
+      "step": 7337
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9404756393540705,
+      "learning_rate": 5.862891024677299e-06,
+      "loss": 0.2794,
+      "step": 7338
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.765853922751242,
+      "learning_rate": 5.861887811675372e-06,
+      "loss": 0.3067,
+      "step": 7339
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.167100232172597,
+      "learning_rate": 5.860884562912188e-06,
+      "loss": 0.3097,
+      "step": 7340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9071164601550794,
+      "learning_rate": 5.859881278429374e-06,
+      "loss": 0.2868,
+      "step": 7341
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6268019484936718,
+      "learning_rate": 5.858877958268558e-06,
+      "loss": 0.2861,
+      "step": 7342
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7383118914966724,
+      "learning_rate": 5.857874602471371e-06,
+      "loss": 0.2839,
+      "step": 7343
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3895124538247345,
+      "learning_rate": 5.856871211079444e-06,
+      "loss": 0.2999,
+      "step": 7344
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9410957812527907,
+      "learning_rate": 5.855867784134406e-06,
+      "loss": 0.3021,
+      "step": 7345
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4242961577547777,
+      "learning_rate": 5.854864321677894e-06,
+      "loss": 0.2935,
+      "step": 7346
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6225016487997318,
+      "learning_rate": 5.853860823751543e-06,
+      "loss": 0.4978,
+      "step": 7347
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6650591672455808,
+      "learning_rate": 5.852857290396992e-06,
+      "loss": 0.2858,
+      "step": 7348
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6397770623242103,
+      "learning_rate": 5.8518537216558745e-06,
+      "loss": 0.2893,
+      "step": 7349
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9201914622817458,
+      "learning_rate": 5.850850117569834e-06,
+      "loss": 0.2778,
+      "step": 7350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6469677312924365,
+      "learning_rate": 5.84984647818051e-06,
+      "loss": 0.3017,
+      "step": 7351
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3510291544324198,
+      "learning_rate": 5.848842803529547e-06,
+      "loss": 0.2821,
+      "step": 7352
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9665345362927438,
+      "learning_rate": 5.847839093658587e-06,
+      "loss": 0.2919,
+      "step": 7353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.954562981846482,
+      "learning_rate": 5.84683534860928e-06,
+      "loss": 0.2845,
+      "step": 7354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8956564872328672,
+      "learning_rate": 5.8458315684232685e-06,
+      "loss": 0.2717,
+      "step": 7355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.248810262332383,
+      "learning_rate": 5.844827753142203e-06,
+      "loss": 0.2868,
+      "step": 7356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.113520115498332,
+      "learning_rate": 5.843823902807733e-06,
+      "loss": 0.2999,
+      "step": 7357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6703145513935096,
+      "learning_rate": 5.84282001746151e-06,
+      "loss": 0.297,
+      "step": 7358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8700202881329098,
+      "learning_rate": 5.841816097145189e-06,
+      "loss": 0.3041,
+      "step": 7359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0327005899982455,
+      "learning_rate": 5.840812141900423e-06,
+      "loss": 0.3033,
+      "step": 7360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9215461324762908,
+      "learning_rate": 5.839808151768865e-06,
+      "loss": 0.2884,
+      "step": 7361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4390662266099223,
+      "learning_rate": 5.838804126792178e-06,
+      "loss": 0.2844,
+      "step": 7362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.799526167333,
+      "learning_rate": 5.837800067012016e-06,
+      "loss": 0.2845,
+      "step": 7363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.447246239952149,
+      "learning_rate": 5.836795972470041e-06,
+      "loss": 0.3081,
+      "step": 7364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.7774009755870805,
+      "learning_rate": 5.835791843207916e-06,
+      "loss": 0.3062,
+      "step": 7365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6124144871826007,
+      "learning_rate": 5.8347876792673044e-06,
+      "loss": 0.2904,
+      "step": 7366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.649169209059497,
+      "learning_rate": 5.833783480689868e-06,
+      "loss": 0.3021,
+      "step": 7367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4987322099032652,
+      "learning_rate": 5.832779247517273e-06,
+      "loss": 0.2963,
+      "step": 7368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.440929233885759,
+      "learning_rate": 5.831774979791188e-06,
+      "loss": 0.3098,
+      "step": 7369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.051398084478484,
+      "learning_rate": 5.830770677553282e-06,
+      "loss": 0.2834,
+      "step": 7370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.708495911664979,
+      "learning_rate": 5.829766340845225e-06,
+      "loss": 0.2849,
+      "step": 7371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.400937708458666,
+      "learning_rate": 5.828761969708689e-06,
+      "loss": 0.3028,
+      "step": 7372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.098191143314644,
+      "learning_rate": 5.827757564185347e-06,
+      "loss": 0.3081,
+      "step": 7373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5746115675118102,
+      "learning_rate": 5.826753124316873e-06,
+      "loss": 0.2972,
+      "step": 7374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5442040364507443,
+      "learning_rate": 5.8257486501449435e-06,
+      "loss": 0.2807,
+      "step": 7375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.991918654063862,
+      "learning_rate": 5.824744141711235e-06,
+      "loss": 0.315,
+      "step": 7376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6218451131737361,
+      "learning_rate": 5.823739599057427e-06,
+      "loss": 0.2823,
+      "step": 7377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.2641020700786445,
+      "learning_rate": 5.822735022225202e-06,
+      "loss": 0.3095,
+      "step": 7378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2021640666825215,
+      "learning_rate": 5.821730411256238e-06,
+      "loss": 0.2959,
+      "step": 7379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7549027559276418,
+      "learning_rate": 5.82072576619222e-06,
+      "loss": 0.3049,
+      "step": 7380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3850711295652784,
+      "learning_rate": 5.819721087074832e-06,
+      "loss": 0.3245,
+      "step": 7381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6691598483890555,
+      "learning_rate": 5.818716373945762e-06,
+      "loss": 0.3157,
+      "step": 7382
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.657245966556539,
+      "learning_rate": 5.817711626846691e-06,
+      "loss": 0.3141,
+      "step": 7383
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.68096296240861,
+      "learning_rate": 5.816706845819316e-06,
+      "loss": 0.2855,
+      "step": 7384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8480071058746694,
+      "learning_rate": 5.815702030905322e-06,
+      "loss": 0.2913,
+      "step": 7385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3705701435846134,
+      "learning_rate": 5.8146971821464034e-06,
+      "loss": 0.2957,
+      "step": 7386
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.711514997249587,
+      "learning_rate": 5.8136922995842495e-06,
+      "loss": 0.2887,
+      "step": 7387
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5839163728817727,
+      "learning_rate": 5.812687383260558e-06,
+      "loss": 0.2914,
+      "step": 7388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7639738555541817,
+      "learning_rate": 5.811682433217023e-06,
+      "loss": 0.3118,
+      "step": 7389
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8906218401729564,
+      "learning_rate": 5.810677449495343e-06,
+      "loss": 0.3015,
+      "step": 7390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.83842369457081,
+      "learning_rate": 5.809672432137215e-06,
+      "loss": 0.2835,
+      "step": 7391
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.47335619090344,
+      "learning_rate": 5.80866738118434e-06,
+      "loss": 0.2795,
+      "step": 7392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.863035727645143,
+      "learning_rate": 5.807662296678418e-06,
+      "loss": 0.3085,
+      "step": 7393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.350936176999657,
+      "learning_rate": 5.806657178661153e-06,
+      "loss": 0.2814,
+      "step": 7394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.674197133155234,
+      "learning_rate": 5.805652027174249e-06,
+      "loss": 0.3206,
+      "step": 7395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9593121527421304,
+      "learning_rate": 5.804646842259413e-06,
+      "loss": 0.3084,
+      "step": 7396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0677949674119387,
+      "learning_rate": 5.803641623958348e-06,
+      "loss": 0.2954,
+      "step": 7397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.52424879535754,
+      "learning_rate": 5.802636372312766e-06,
+      "loss": 0.2749,
+      "step": 7398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1236913832671624,
+      "learning_rate": 5.801631087364374e-06,
+      "loss": 0.3015,
+      "step": 7399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5553322808566252,
+      "learning_rate": 5.8006257691548865e-06,
+      "loss": 0.2987,
+      "step": 7400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5785389856886474,
+      "learning_rate": 5.799620417726012e-06,
+      "loss": 0.463,
+      "step": 7401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4850186335183841,
+      "learning_rate": 5.798615033119466e-06,
+      "loss": 0.2934,
+      "step": 7402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 25.044942457349375,
+      "learning_rate": 5.797609615376964e-06,
+      "loss": 0.2855,
+      "step": 7403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0788518723451155,
+      "learning_rate": 5.796604164540224e-06,
+      "loss": 0.2762,
+      "step": 7404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9639242140463364,
+      "learning_rate": 5.795598680650959e-06,
+      "loss": 0.2846,
+      "step": 7405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.202881432684013,
+      "learning_rate": 5.794593163750894e-06,
+      "loss": 0.2703,
+      "step": 7406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.61043453216115,
+      "learning_rate": 5.793587613881747e-06,
+      "loss": 0.3026,
+      "step": 7407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.821645440211319,
+      "learning_rate": 5.792582031085241e-06,
+      "loss": 0.2941,
+      "step": 7408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1025830781435055,
+      "learning_rate": 5.791576415403097e-06,
+      "loss": 0.3294,
+      "step": 7409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.43317684151675,
+      "learning_rate": 5.790570766877043e-06,
+      "loss": 0.296,
+      "step": 7410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0351764553562695,
+      "learning_rate": 5.789565085548803e-06,
+      "loss": 0.2845,
+      "step": 7411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6329261443402958,
+      "learning_rate": 5.788559371460107e-06,
+      "loss": 0.2876,
+      "step": 7412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1815197784932674,
+      "learning_rate": 5.787553624652678e-06,
+      "loss": 0.2648,
+      "step": 7413
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.672178902243905,
+      "learning_rate": 5.786547845168253e-06,
+      "loss": 0.2965,
+      "step": 7414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.781489413288844,
+      "learning_rate": 5.78554203304856e-06,
+      "loss": 0.2973,
+      "step": 7415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.906575158794279,
+      "learning_rate": 5.784536188335334e-06,
+      "loss": 0.2977,
+      "step": 7416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 27.952144720770548,
+      "learning_rate": 5.783530311070306e-06,
+      "loss": 0.3185,
+      "step": 7417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3791989634924233,
+      "learning_rate": 5.782524401295216e-06,
+      "loss": 0.2817,
+      "step": 7418
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6515979396728617,
+      "learning_rate": 5.781518459051797e-06,
+      "loss": 0.2805,
+      "step": 7419
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8358959595852473,
+      "learning_rate": 5.78051248438179e-06,
+      "loss": 0.2967,
+      "step": 7420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4987211405224334,
+      "learning_rate": 5.779506477326933e-06,
+      "loss": 0.3037,
+      "step": 7421
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9785200233494902,
+      "learning_rate": 5.778500437928969e-06,
+      "loss": 0.3012,
+      "step": 7422
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1999170666941543,
+      "learning_rate": 5.777494366229637e-06,
+      "loss": 0.3118,
+      "step": 7423
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.253242248500906,
+      "learning_rate": 5.7764882622706834e-06,
+      "loss": 0.296,
+      "step": 7424
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1496977476439847,
+      "learning_rate": 5.7754821260938536e-06,
+      "loss": 0.2802,
+      "step": 7425
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.083243302988889,
+      "learning_rate": 5.774475957740892e-06,
+      "loss": 0.293,
+      "step": 7426
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.395467071501667,
+      "learning_rate": 5.7734697572535485e-06,
+      "loss": 0.2744,
+      "step": 7427
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.17167924037797,
+      "learning_rate": 5.7724635246735695e-06,
+      "loss": 0.2966,
+      "step": 7428
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6577735965038014,
+      "learning_rate": 5.771457260042707e-06,
+      "loss": 0.2826,
+      "step": 7429
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0480434049578315,
+      "learning_rate": 5.770450963402714e-06,
+      "loss": 0.2939,
+      "step": 7430
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.797695230901797,
+      "learning_rate": 5.76944463479534e-06,
+      "loss": 0.2842,
+      "step": 7431
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.2435853244238055,
+      "learning_rate": 5.7684382742623425e-06,
+      "loss": 0.3024,
+      "step": 7432
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.480875688589397,
+      "learning_rate": 5.767431881845475e-06,
+      "loss": 0.2863,
+      "step": 7433
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.460237068746604,
+      "learning_rate": 5.766425457586497e-06,
+      "loss": 0.2841,
+      "step": 7434
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.2983250323644855,
+      "learning_rate": 5.765419001527165e-06,
+      "loss": 0.3179,
+      "step": 7435
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2555286612370506,
+      "learning_rate": 5.7644125137092395e-06,
+      "loss": 0.2768,
+      "step": 7436
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7879728690741918,
+      "learning_rate": 5.76340599417448e-06,
+      "loss": 0.3091,
+      "step": 7437
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.794846458366639,
+      "learning_rate": 5.762399442964652e-06,
+      "loss": 0.299,
+      "step": 7438
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.466717185262108,
+      "learning_rate": 5.761392860121514e-06,
+      "loss": 0.3052,
+      "step": 7439
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.508990408891954,
+      "learning_rate": 5.760386245686836e-06,
+      "loss": 0.3105,
+      "step": 7440
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.176518988435943,
+      "learning_rate": 5.759379599702381e-06,
+      "loss": 0.2905,
+      "step": 7441
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.571342908592038,
+      "learning_rate": 5.758372922209918e-06,
+      "loss": 0.2919,
+      "step": 7442
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.12035874006493,
+      "learning_rate": 5.757366213251215e-06,
+      "loss": 0.2881,
+      "step": 7443
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.872116166511122,
+      "learning_rate": 5.756359472868044e-06,
+      "loss": 0.3009,
+      "step": 7444
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7007860292818515,
+      "learning_rate": 5.755352701102174e-06,
+      "loss": 0.2931,
+      "step": 7445
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6324341594720326,
+      "learning_rate": 5.754345897995378e-06,
+      "loss": 0.2915,
+      "step": 7446
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6107260486933718,
+      "learning_rate": 5.753339063589431e-06,
+      "loss": 0.4902,
+      "step": 7447
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7491841544424926,
+      "learning_rate": 5.752332197926109e-06,
+      "loss": 0.2907,
+      "step": 7448
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8626797410832066,
+      "learning_rate": 5.751325301047188e-06,
+      "loss": 0.3146,
+      "step": 7449
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.509480948330613,
+      "learning_rate": 5.7503183729944454e-06,
+      "loss": 0.2746,
+      "step": 7450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.049359590867218,
+      "learning_rate": 5.749311413809661e-06,
+      "loss": 0.2858,
+      "step": 7451
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.04012364496551,
+      "learning_rate": 5.748304423534615e-06,
+      "loss": 0.2673,
+      "step": 7452
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.413054241860235,
+      "learning_rate": 5.74729740221109e-06,
+      "loss": 0.296,
+      "step": 7453
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3595975797556992,
+      "learning_rate": 5.746290349880867e-06,
+      "loss": 0.2916,
+      "step": 7454
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6144008908891876,
+      "learning_rate": 5.745283266585732e-06,
+      "loss": 0.2895,
+      "step": 7455
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.030825353977351,
+      "learning_rate": 5.744276152367472e-06,
+      "loss": 0.2749,
+      "step": 7456
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3108544995994253,
+      "learning_rate": 5.7432690072678696e-06,
+      "loss": 0.2952,
+      "step": 7457
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0277926170135334,
+      "learning_rate": 5.7422618313287184e-06,
+      "loss": 0.3021,
+      "step": 7458
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9805852476289116,
+      "learning_rate": 5.741254624591804e-06,
+      "loss": 0.2806,
+      "step": 7459
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9802090010372773,
+      "learning_rate": 5.7402473870989205e-06,
+      "loss": 0.3016,
+      "step": 7460
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6529213476321996,
+      "learning_rate": 5.7392401188918555e-06,
+      "loss": 0.3039,
+      "step": 7461
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7974486995779502,
+      "learning_rate": 5.738232820012407e-06,
+      "loss": 0.3013,
+      "step": 7462
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.920592923579262,
+      "learning_rate": 5.737225490502366e-06,
+      "loss": 0.3029,
+      "step": 7463
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8315720248690885,
+      "learning_rate": 5.736218130403532e-06,
+      "loss": 0.298,
+      "step": 7464
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.032439390079224,
+      "learning_rate": 5.735210739757697e-06,
+      "loss": 0.2819,
+      "step": 7465
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4071823116303686,
+      "learning_rate": 5.734203318606666e-06,
+      "loss": 0.2905,
+      "step": 7466
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.486510636408435,
+      "learning_rate": 5.7331958669922335e-06,
+      "loss": 0.3097,
+      "step": 7467
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.174635017024112,
+      "learning_rate": 5.732188384956203e-06,
+      "loss": 0.2864,
+      "step": 7468
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0793500591352876,
+      "learning_rate": 5.731180872540374e-06,
+      "loss": 0.2764,
+      "step": 7469
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.267484630972985,
+      "learning_rate": 5.730173329786554e-06,
+      "loss": 0.2977,
+      "step": 7470
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2034813592776934,
+      "learning_rate": 5.729165756736544e-06,
+      "loss": 0.3035,
+      "step": 7471
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4909879137858109,
+      "learning_rate": 5.728158153432153e-06,
+      "loss": 0.2937,
+      "step": 7472
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9755301438766115,
+      "learning_rate": 5.727150519915186e-06,
+      "loss": 0.2899,
+      "step": 7473
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.884894116487073,
+      "learning_rate": 5.726142856227453e-06,
+      "loss": 0.2662,
+      "step": 7474
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4461228179609695,
+      "learning_rate": 5.725135162410762e-06,
+      "loss": 0.2871,
+      "step": 7475
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.404019744402751,
+      "learning_rate": 5.724127438506925e-06,
+      "loss": 0.2991,
+      "step": 7476
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1658715308810628,
+      "learning_rate": 5.723119684557755e-06,
+      "loss": 0.2907,
+      "step": 7477
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5756943612655252,
+      "learning_rate": 5.722111900605066e-06,
+      "loss": 0.2918,
+      "step": 7478
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.070764156037807,
+      "learning_rate": 5.721104086690671e-06,
+      "loss": 0.2856,
+      "step": 7479
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3053552661157948,
+      "learning_rate": 5.720096242856385e-06,
+      "loss": 0.2934,
+      "step": 7480
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.739945420154742,
+      "learning_rate": 5.719088369144028e-06,
+      "loss": 0.2909,
+      "step": 7481
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5469861022930633,
+      "learning_rate": 5.718080465595419e-06,
+      "loss": 0.2824,
+      "step": 7482
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.480193219958061,
+      "learning_rate": 5.717072532252373e-06,
+      "loss": 0.2886,
+      "step": 7483
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8616266062916327,
+      "learning_rate": 5.716064569156716e-06,
+      "loss": 0.2935,
+      "step": 7484
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.953623144323701,
+      "learning_rate": 5.715056576350267e-06,
+      "loss": 0.2921,
+      "step": 7485
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.594419332724903,
+      "learning_rate": 5.714048553874852e-06,
+      "loss": 0.2933,
+      "step": 7486
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2998054298609465,
+      "learning_rate": 5.713040501772292e-06,
+      "loss": 0.3098,
+      "step": 7487
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.61312532244727,
+      "learning_rate": 5.712032420084418e-06,
+      "loss": 0.3033,
+      "step": 7488
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.946469730706228,
+      "learning_rate": 5.7110243088530525e-06,
+      "loss": 0.3049,
+      "step": 7489
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.989729102428159,
+      "learning_rate": 5.710016168120026e-06,
+      "loss": 0.2912,
+      "step": 7490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.582846243896762,
+      "learning_rate": 5.7090079979271665e-06,
+      "loss": 0.291,
+      "step": 7491
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 55.800817128984455,
+      "learning_rate": 5.707999798316307e-06,
+      "loss": 0.284,
+      "step": 7492
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.815348963611848,
+      "learning_rate": 5.706991569329277e-06,
+      "loss": 0.2922,
+      "step": 7493
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4766139781374577,
+      "learning_rate": 5.705983311007913e-06,
+      "loss": 0.2965,
+      "step": 7494
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.246352461340009,
+      "learning_rate": 5.704975023394045e-06,
+      "loss": 0.2933,
+      "step": 7495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.385660925283995,
+      "learning_rate": 5.703966706529513e-06,
+      "loss": 0.2883,
+      "step": 7496
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.116793870423358,
+      "learning_rate": 5.702958360456151e-06,
+      "loss": 0.3052,
+      "step": 7497
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2875175653246824,
+      "learning_rate": 5.7019499852158e-06,
+      "loss": 0.2993,
+      "step": 7498
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3467825549517634,
+      "learning_rate": 5.700941580850294e-06,
+      "loss": 0.2891,
+      "step": 7499
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0220070722338837,
+      "learning_rate": 5.6999331474014795e-06,
+      "loss": 0.2853,
+      "step": 7500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.889623982996533,
+      "learning_rate": 5.698924684911195e-06,
+      "loss": 0.2908,
+      "step": 7501
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.640199975628539,
+      "learning_rate": 5.697916193421283e-06,
+      "loss": 0.2958,
+      "step": 7502
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.216291310154457,
+      "learning_rate": 5.6969076729735885e-06,
+      "loss": 0.2744,
+      "step": 7503
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9070709643064623,
+      "learning_rate": 5.695899123609957e-06,
+      "loss": 0.2845,
+      "step": 7504
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.766277469410023,
+      "learning_rate": 5.694890545372235e-06,
+      "loss": 0.3101,
+      "step": 7505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6244973568012134,
+      "learning_rate": 5.693881938302271e-06,
+      "loss": 0.4962,
+      "step": 7506
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4504037855239074,
+      "learning_rate": 5.69287330244191e-06,
+      "loss": 0.2893,
+      "step": 7507
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.361867015535825,
+      "learning_rate": 5.691864637833009e-06,
+      "loss": 0.3067,
+      "step": 7508
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.129035141662319,
+      "learning_rate": 5.690855944517413e-06,
+      "loss": 0.3131,
+      "step": 7509
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.733406459487125,
+      "learning_rate": 5.689847222536976e-06,
+      "loss": 0.3003,
+      "step": 7510
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0861672973751695,
+      "learning_rate": 5.688838471933553e-06,
+      "loss": 0.291,
+      "step": 7511
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5806383363046554,
+      "learning_rate": 5.687829692749e-06,
+      "loss": 0.3019,
+      "step": 7512
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7720927581792263,
+      "learning_rate": 5.686820885025169e-06,
+      "loss": 0.2696,
+      "step": 7513
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4731309256315446,
+      "learning_rate": 5.68581204880392e-06,
+      "loss": 0.2839,
+      "step": 7514
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3839875728516073,
+      "learning_rate": 5.684803184127112e-06,
+      "loss": 0.2824,
+      "step": 7515
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.729233174779593,
+      "learning_rate": 5.683794291036604e-06,
+      "loss": 0.2875,
+      "step": 7516
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4638741674560807,
+      "learning_rate": 5.682785369574254e-06,
+      "loss": 0.2753,
+      "step": 7517
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.180851821766722,
+      "learning_rate": 5.681776419781928e-06,
+      "loss": 0.2909,
+      "step": 7518
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1562037127667515,
+      "learning_rate": 5.680767441701487e-06,
+      "loss": 0.296,
+      "step": 7519
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.203281795230169,
+      "learning_rate": 5.679758435374797e-06,
+      "loss": 0.2793,
+      "step": 7520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.809742360489713,
+      "learning_rate": 5.678749400843719e-06,
+      "loss": 0.3229,
+      "step": 7521
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5053153872792064,
+      "learning_rate": 5.677740338150126e-06,
+      "loss": 0.3021,
+      "step": 7522
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5869112896278917,
+      "learning_rate": 5.6767312473358805e-06,
+      "loss": 0.4689,
+      "step": 7523
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8921322492543409,
+      "learning_rate": 5.6757221284428554e-06,
+      "loss": 0.2785,
+      "step": 7524
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.511930460075979,
+      "learning_rate": 5.674712981512915e-06,
+      "loss": 0.3081,
+      "step": 7525
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.569747007160839,
+      "learning_rate": 5.673703806587939e-06,
+      "loss": 0.2707,
+      "step": 7526
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4256604965425432,
+      "learning_rate": 5.672694603709794e-06,
+      "loss": 0.2847,
+      "step": 7527
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0290628140660143,
+      "learning_rate": 5.671685372920355e-06,
+      "loss": 0.2814,
+      "step": 7528
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.446966074724786,
+      "learning_rate": 5.670676114261495e-06,
+      "loss": 0.2957,
+      "step": 7529
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6950616372965643,
+      "learning_rate": 5.669666827775095e-06,
+      "loss": 0.2872,
+      "step": 7530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.926014532229951,
+      "learning_rate": 5.668657513503027e-06,
+      "loss": 0.3254,
+      "step": 7531
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.600040825964026,
+      "learning_rate": 5.667648171487171e-06,
+      "loss": 0.5185,
+      "step": 7532
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6331571157681681,
+      "learning_rate": 5.666638801769406e-06,
+      "loss": 0.489,
+      "step": 7533
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9454720619669392,
+      "learning_rate": 5.665629404391615e-06,
+      "loss": 0.2997,
+      "step": 7534
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.650075794061309,
+      "learning_rate": 5.664619979395676e-06,
+      "loss": 0.3163,
+      "step": 7535
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1008226354398274,
+      "learning_rate": 5.663610526823474e-06,
+      "loss": 0.2842,
+      "step": 7536
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0083496967367487,
+      "learning_rate": 5.662601046716893e-06,
+      "loss": 0.3236,
+      "step": 7537
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.540398965207893,
+      "learning_rate": 5.661591539117818e-06,
+      "loss": 0.2709,
+      "step": 7538
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.263167581705944,
+      "learning_rate": 5.660582004068134e-06,
+      "loss": 0.285,
+      "step": 7539
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.470621667087632,
+      "learning_rate": 5.6595724416097285e-06,
+      "loss": 0.2997,
+      "step": 7540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2905877109285977,
+      "learning_rate": 5.658562851784493e-06,
+      "loss": 0.3073,
+      "step": 7541
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.614383038345281,
+      "learning_rate": 5.657553234634314e-06,
+      "loss": 0.2739,
+      "step": 7542
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.544974988609688,
+      "learning_rate": 5.656543590201084e-06,
+      "loss": 0.2823,
+      "step": 7543
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6832459754722957,
+      "learning_rate": 5.655533918526693e-06,
+      "loss": 0.292,
+      "step": 7544
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.17433286086938,
+      "learning_rate": 5.654524219653035e-06,
+      "loss": 0.3132,
+      "step": 7545
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8187781261421623,
+      "learning_rate": 5.653514493622007e-06,
+      "loss": 0.2918,
+      "step": 7546
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.373290493754437,
+      "learning_rate": 5.652504740475499e-06,
+      "loss": 0.2919,
+      "step": 7547
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2056108905279466,
+      "learning_rate": 5.651494960255412e-06,
+      "loss": 0.3059,
+      "step": 7548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6706705669316926,
+      "learning_rate": 5.650485153003642e-06,
+      "loss": 0.2879,
+      "step": 7549
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.029525412874832,
+      "learning_rate": 5.649475318762088e-06,
+      "loss": 0.2935,
+      "step": 7550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2076103327721412,
+      "learning_rate": 5.648465457572648e-06,
+      "loss": 0.3061,
+      "step": 7551
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.077242752112191,
+      "learning_rate": 5.647455569477225e-06,
+      "loss": 0.2989,
+      "step": 7552
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.812558801833626,
+      "learning_rate": 5.646445654517721e-06,
+      "loss": 0.3027,
+      "step": 7553
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2839181061111227,
+      "learning_rate": 5.645435712736038e-06,
+      "loss": 0.2699,
+      "step": 7554
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2677247153083626,
+      "learning_rate": 5.6444257441740804e-06,
+      "loss": 0.2771,
+      "step": 7555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4156651211418265,
+      "learning_rate": 5.643415748873755e-06,
+      "loss": 0.284,
+      "step": 7556
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.546944978902009,
+      "learning_rate": 5.642405726876967e-06,
+      "loss": 0.2665,
+      "step": 7557
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.834067952436137,
+      "learning_rate": 5.641395678225624e-06,
+      "loss": 0.2789,
+      "step": 7558
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7530318367081685,
+      "learning_rate": 5.640385602961634e-06,
+      "loss": 0.2797,
+      "step": 7559
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3080553025611903,
+      "learning_rate": 5.6393755011269115e-06,
+      "loss": 0.2973,
+      "step": 7560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8055236618951296,
+      "learning_rate": 5.63836537276336e-06,
+      "loss": 0.2944,
+      "step": 7561
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7602991910046204,
+      "learning_rate": 5.6373552179128975e-06,
+      "loss": 0.2921,
+      "step": 7562
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8516542757364982,
+      "learning_rate": 5.636345036617433e-06,
+      "loss": 0.2922,
+      "step": 7563
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8027538131341654,
+      "learning_rate": 5.635334828918886e-06,
+      "loss": 0.3032,
+      "step": 7564
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.391280447097233,
+      "learning_rate": 5.6343245948591655e-06,
+      "loss": 0.3,
+      "step": 7565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7098293431121996,
+      "learning_rate": 5.633314334480191e-06,
+      "loss": 0.2855,
+      "step": 7566
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8748959742737594,
+      "learning_rate": 5.632304047823881e-06,
+      "loss": 0.2792,
+      "step": 7567
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.337783856025002,
+      "learning_rate": 5.631293734932153e-06,
+      "loss": 0.3168,
+      "step": 7568
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1377957947763133,
+      "learning_rate": 5.630283395846926e-06,
+      "loss": 0.2996,
+      "step": 7569
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.395129393465031,
+      "learning_rate": 5.6292730306101215e-06,
+      "loss": 0.2916,
+      "step": 7570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5285132336855396,
+      "learning_rate": 5.6282626392636615e-06,
+      "loss": 0.2921,
+      "step": 7571
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 59.335283073129325,
+      "learning_rate": 5.6272522218494695e-06,
+      "loss": 0.2852,
+      "step": 7572
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.699520023762649,
+      "learning_rate": 5.626241778409467e-06,
+      "loss": 0.2873,
+      "step": 7573
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.392732779304917,
+      "learning_rate": 5.625231308985582e-06,
+      "loss": 0.294,
+      "step": 7574
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6432966342313273,
+      "learning_rate": 5.624220813619739e-06,
+      "loss": 0.3014,
+      "step": 7575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.380568531075516,
+      "learning_rate": 5.623210292353865e-06,
+      "loss": 0.3115,
+      "step": 7576
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.951581765742306,
+      "learning_rate": 5.62219974522989e-06,
+      "loss": 0.2775,
+      "step": 7577
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2032960059647033,
+      "learning_rate": 5.621189172289742e-06,
+      "loss": 0.3008,
+      "step": 7578
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.386585173453586,
+      "learning_rate": 5.620178573575352e-06,
+      "loss": 0.31,
+      "step": 7579
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.233562515286271,
+      "learning_rate": 5.6191679491286525e-06,
+      "loss": 0.2862,
+      "step": 7580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9552337487300702,
+      "learning_rate": 5.618157298991574e-06,
+      "loss": 0.2896,
+      "step": 7581
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7236303988944257,
+      "learning_rate": 5.61714662320605e-06,
+      "loss": 0.2865,
+      "step": 7582
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.463293909716265,
+      "learning_rate": 5.616135921814018e-06,
+      "loss": 0.2837,
+      "step": 7583
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9170548306979804,
+      "learning_rate": 5.615125194857411e-06,
+      "loss": 0.2936,
+      "step": 7584
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6618161709336832,
+      "learning_rate": 5.614114442378169e-06,
+      "loss": 0.5309,
+      "step": 7585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.161281708041211,
+      "learning_rate": 5.613103664418227e-06,
+      "loss": 0.2977,
+      "step": 7586
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8520552299589774,
+      "learning_rate": 5.6120928610195235e-06,
+      "loss": 0.2714,
+      "step": 7587
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3206685316379283,
+      "learning_rate": 5.611082032224001e-06,
+      "loss": 0.293,
+      "step": 7588
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8522249948067022,
+      "learning_rate": 5.610071178073601e-06,
+      "loss": 0.2773,
+      "step": 7589
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6560173966684166,
+      "learning_rate": 5.609060298610263e-06,
+      "loss": 0.2934,
+      "step": 7590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.457576734413159,
+      "learning_rate": 5.6080493938759314e-06,
+      "loss": 0.3222,
+      "step": 7591
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7121792610360973,
+      "learning_rate": 5.607038463912551e-06,
+      "loss": 0.292,
+      "step": 7592
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7733976375907252,
+      "learning_rate": 5.606027508762067e-06,
+      "loss": 0.3028,
+      "step": 7593
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6817503215393352,
+      "learning_rate": 5.605016528466424e-06,
+      "loss": 0.2795,
+      "step": 7594
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.016365833923353,
+      "learning_rate": 5.604005523067569e-06,
+      "loss": 0.2804,
+      "step": 7595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2220054674335756,
+      "learning_rate": 5.602994492607454e-06,
+      "loss": 0.2928,
+      "step": 7596
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2291516796281352,
+      "learning_rate": 5.601983437128027e-06,
+      "loss": 0.2753,
+      "step": 7597
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7494213461557602,
+      "learning_rate": 5.600972356671238e-06,
+      "loss": 0.2812,
+      "step": 7598
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3269791290799406,
+      "learning_rate": 5.599961251279037e-06,
+      "loss": 0.2952,
+      "step": 7599
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2304916030670954,
+      "learning_rate": 5.5989501209933795e-06,
+      "loss": 0.2765,
+      "step": 7600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.459507001809702,
+      "learning_rate": 5.5979389658562165e-06,
+      "loss": 0.2799,
+      "step": 7601
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4381696557232515,
+      "learning_rate": 5.596927785909505e-06,
+      "loss": 0.2888,
+      "step": 7602
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7657934821560524,
+      "learning_rate": 5.595916581195198e-06,
+      "loss": 0.2875,
+      "step": 7603
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6843243807157469,
+      "learning_rate": 5.594905351755254e-06,
+      "loss": 0.2724,
+      "step": 7604
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9240401785154129,
+      "learning_rate": 5.59389409763163e-06,
+      "loss": 0.2747,
+      "step": 7605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8506574965474858,
+      "learning_rate": 5.592882818866286e-06,
+      "loss": 0.2846,
+      "step": 7606
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.044141352460662,
+      "learning_rate": 5.591871515501181e-06,
+      "loss": 0.2951,
+      "step": 7607
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4142261519026587,
+      "learning_rate": 5.590860187578274e-06,
+      "loss": 0.2961,
+      "step": 7608
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8414898976085707,
+      "learning_rate": 5.589848835139529e-06,
+      "loss": 0.2907,
+      "step": 7609
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8158250187790252,
+      "learning_rate": 5.588837458226908e-06,
+      "loss": 0.2802,
+      "step": 7610
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1173824203542102,
+      "learning_rate": 5.587826056882376e-06,
+      "loss": 0.2924,
+      "step": 7611
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8089366617244502,
+      "learning_rate": 5.586814631147895e-06,
+      "loss": 0.2819,
+      "step": 7612
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6528462451797643,
+      "learning_rate": 5.5858031810654345e-06,
+      "loss": 0.2996,
+      "step": 7613
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.735345799156136,
+      "learning_rate": 5.5847917066769585e-06,
+      "loss": 0.2984,
+      "step": 7614
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8993270059871377,
+      "learning_rate": 5.583780208024436e-06,
+      "loss": 0.2917,
+      "step": 7615
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3079231039346049,
+      "learning_rate": 5.582768685149837e-06,
+      "loss": 0.2794,
+      "step": 7616
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6259454534568314,
+      "learning_rate": 5.581757138095128e-06,
+      "loss": 0.329,
+      "step": 7617
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8301455628536605,
+      "learning_rate": 5.5807455669022825e-06,
+      "loss": 0.2785,
+      "step": 7618
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.789044571160198,
+      "learning_rate": 5.579733971613274e-06,
+      "loss": 0.3055,
+      "step": 7619
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.880487546744792,
+      "learning_rate": 5.578722352270071e-06,
+      "loss": 0.2963,
+      "step": 7620
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.533953734467629,
+      "learning_rate": 5.57771070891465e-06,
+      "loss": 0.2783,
+      "step": 7621
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.443776325070946,
+      "learning_rate": 5.576699041588986e-06,
+      "loss": 0.293,
+      "step": 7622
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8706229546458053,
+      "learning_rate": 5.575687350335055e-06,
+      "loss": 0.2737,
+      "step": 7623
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.16178471422946,
+      "learning_rate": 5.574675635194833e-06,
+      "loss": 0.3048,
+      "step": 7624
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6636299220898205,
+      "learning_rate": 5.573663896210297e-06,
+      "loss": 0.5273,
+      "step": 7625
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.486550452320844,
+      "learning_rate": 5.5726521334234275e-06,
+      "loss": 0.3054,
+      "step": 7626
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.261091326497108,
+      "learning_rate": 5.5716403468762045e-06,
+      "loss": 0.2917,
+      "step": 7627
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1698329739370568,
+      "learning_rate": 5.570628536610608e-06,
+      "loss": 0.2911,
+      "step": 7628
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6663475686674074,
+      "learning_rate": 5.56961670266862e-06,
+      "loss": 0.2871,
+      "step": 7629
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4871906189983948,
+      "learning_rate": 5.5686048450922224e-06,
+      "loss": 0.2918,
+      "step": 7630
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6193746845400931,
+      "learning_rate": 5.567592963923401e-06,
+      "loss": 0.2884,
+      "step": 7631
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.180697079712196,
+      "learning_rate": 5.566581059204139e-06,
+      "loss": 0.3088,
+      "step": 7632
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8345635015984232,
+      "learning_rate": 5.5655691309764225e-06,
+      "loss": 0.2951,
+      "step": 7633
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6623792226280405,
+      "learning_rate": 5.564557179282237e-06,
+      "loss": 0.2883,
+      "step": 7634
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1105137740109474,
+      "learning_rate": 5.563545204163574e-06,
+      "loss": 0.3031,
+      "step": 7635
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.000333524241662,
+      "learning_rate": 5.562533205662417e-06,
+      "loss": 0.284,
+      "step": 7636
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.499717981356932,
+      "learning_rate": 5.561521183820759e-06,
+      "loss": 0.2777,
+      "step": 7637
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8619320855670536,
+      "learning_rate": 5.5605091386805896e-06,
+      "loss": 0.2966,
+      "step": 7638
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0606967986000067,
+      "learning_rate": 5.5594970702839005e-06,
+      "loss": 0.2829,
+      "step": 7639
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.561519112722402,
+      "learning_rate": 5.558484978672684e-06,
+      "loss": 0.3106,
+      "step": 7640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9140444866929631,
+      "learning_rate": 5.557472863888935e-06,
+      "loss": 0.2895,
+      "step": 7641
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.599444497655152,
+      "learning_rate": 5.556460725974645e-06,
+      "loss": 0.292,
+      "step": 7642
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.613480966642762,
+      "learning_rate": 5.555448564971812e-06,
+      "loss": 0.3196,
+      "step": 7643
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5954900095024396,
+      "learning_rate": 5.554436380922431e-06,
+      "loss": 0.2959,
+      "step": 7644
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1933243969816356,
+      "learning_rate": 5.553424173868501e-06,
+      "loss": 0.2781,
+      "step": 7645
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.497756212293826,
+      "learning_rate": 5.552411943852017e-06,
+      "loss": 0.2954,
+      "step": 7646
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.920059588064577,
+      "learning_rate": 5.551399690914981e-06,
+      "loss": 0.3116,
+      "step": 7647
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.044223232505679,
+      "learning_rate": 5.550387415099393e-06,
+      "loss": 0.2804,
+      "step": 7648
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.742875490724477,
+      "learning_rate": 5.549375116447254e-06,
+      "loss": 0.3274,
+      "step": 7649
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6956356747388157,
+      "learning_rate": 5.548362795000565e-06,
+      "loss": 0.2997,
+      "step": 7650
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.303764207487624,
+      "learning_rate": 5.54735045080133e-06,
+      "loss": 0.2811,
+      "step": 7651
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4929475893767643,
+      "learning_rate": 5.546338083891551e-06,
+      "loss": 0.2798,
+      "step": 7652
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4465382185303735,
+      "learning_rate": 5.545325694313237e-06,
+      "loss": 0.3021,
+      "step": 7653
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9279209253137302,
+      "learning_rate": 5.544313282108389e-06,
+      "loss": 0.2811,
+      "step": 7654
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7286870769978662,
+      "learning_rate": 5.543300847319017e-06,
+      "loss": 0.2946,
+      "step": 7655
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6097230773486723,
+      "learning_rate": 5.542288389987128e-06,
+      "loss": 0.4863,
+      "step": 7656
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.278823836160525,
+      "learning_rate": 5.541275910154731e-06,
+      "loss": 0.293,
+      "step": 7657
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8592303195067155,
+      "learning_rate": 5.540263407863834e-06,
+      "loss": 0.2974,
+      "step": 7658
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7508571877058219,
+      "learning_rate": 5.53925088315645e-06,
+      "loss": 0.2959,
+      "step": 7659
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.384552100276538,
+      "learning_rate": 5.538238336074587e-06,
+      "loss": 0.2918,
+      "step": 7660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0424155998200018,
+      "learning_rate": 5.537225766660261e-06,
+      "loss": 0.2951,
+      "step": 7661
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.887461789212646,
+      "learning_rate": 5.536213174955484e-06,
+      "loss": 0.2815,
+      "step": 7662
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.936148551169702,
+      "learning_rate": 5.535200561002268e-06,
+      "loss": 0.2975,
+      "step": 7663
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8999436939087193,
+      "learning_rate": 5.534187924842631e-06,
+      "loss": 0.2789,
+      "step": 7664
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.618104287259037,
+      "learning_rate": 5.53317526651859e-06,
+      "loss": 0.2986,
+      "step": 7665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.0458312152352685,
+      "learning_rate": 5.532162586072158e-06,
+      "loss": 0.323,
+      "step": 7666
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0072672077035305,
+      "learning_rate": 5.5311498835453555e-06,
+      "loss": 0.2891,
+      "step": 7667
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0729173082436962,
+      "learning_rate": 5.530137158980201e-06,
+      "loss": 0.2811,
+      "step": 7668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6521791172526483,
+      "learning_rate": 5.529124412418715e-06,
+      "loss": 0.3063,
+      "step": 7669
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6462108771835293,
+      "learning_rate": 5.5281116439029156e-06,
+      "loss": 0.2912,
+      "step": 7670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.316829178533151,
+      "learning_rate": 5.527098853474828e-06,
+      "loss": 0.3091,
+      "step": 7671
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5556510477501275,
+      "learning_rate": 5.526086041176472e-06,
+      "loss": 0.2787,
+      "step": 7672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6613192769447411,
+      "learning_rate": 5.525073207049871e-06,
+      "loss": 0.2797,
+      "step": 7673
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6538928837380741,
+      "learning_rate": 5.524060351137049e-06,
+      "loss": 0.5012,
+      "step": 7674
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.299107483738187,
+      "learning_rate": 5.523047473480036e-06,
+      "loss": 0.2783,
+      "step": 7675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3144949143742457,
+      "learning_rate": 5.522034574120851e-06,
+      "loss": 0.2781,
+      "step": 7676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5728670283895205,
+      "learning_rate": 5.521021653101525e-06,
+      "loss": 0.2924,
+      "step": 7677
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.291260118859905,
+      "learning_rate": 5.520008710464085e-06,
+      "loss": 0.2724,
+      "step": 7678
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.790638424702302,
+      "learning_rate": 5.518995746250561e-06,
+      "loss": 0.2688,
+      "step": 7679
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.192349802244243,
+      "learning_rate": 5.517982760502981e-06,
+      "loss": 0.3109,
+      "step": 7680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6548324309082583,
+      "learning_rate": 5.516969753263376e-06,
+      "loss": 0.2932,
+      "step": 7681
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1810413925839023,
+      "learning_rate": 5.515956724573778e-06,
+      "loss": 0.3011,
+      "step": 7682
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.95441875468757,
+      "learning_rate": 5.51494367447622e-06,
+      "loss": 0.2775,
+      "step": 7683
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9777505854560355,
+      "learning_rate": 5.513930603012733e-06,
+      "loss": 0.2919,
+      "step": 7684
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.322860532123437,
+      "learning_rate": 5.512917510225353e-06,
+      "loss": 0.2773,
+      "step": 7685
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8650869664910847,
+      "learning_rate": 5.511904396156113e-06,
+      "loss": 0.2916,
+      "step": 7686
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.095873520856129,
+      "learning_rate": 5.510891260847053e-06,
+      "loss": 0.2906,
+      "step": 7687
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1493326173476626,
+      "learning_rate": 5.509878104340205e-06,
+      "loss": 0.2932,
+      "step": 7688
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.960223046675856,
+      "learning_rate": 5.508864926677609e-06,
+      "loss": 0.3028,
+      "step": 7689
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3690556805584755,
+      "learning_rate": 5.507851727901304e-06,
+      "loss": 0.2919,
+      "step": 7690
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6671958339056985,
+      "learning_rate": 5.506838508053329e-06,
+      "loss": 0.2885,
+      "step": 7691
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9193056688864807,
+      "learning_rate": 5.505825267175723e-06,
+      "loss": 0.2721,
+      "step": 7692
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4266198792138582,
+      "learning_rate": 5.50481200531053e-06,
+      "loss": 0.2947,
+      "step": 7693
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7212335690916,
+      "learning_rate": 5.503798722499789e-06,
+      "loss": 0.2802,
+      "step": 7694
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9068507374041386,
+      "learning_rate": 5.502785418785545e-06,
+      "loss": 0.2757,
+      "step": 7695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.520798957009045,
+      "learning_rate": 5.501772094209838e-06,
+      "loss": 0.303,
+      "step": 7696
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.073194997785977,
+      "learning_rate": 5.500758748814719e-06,
+      "loss": 0.2753,
+      "step": 7697
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5058041779306213,
+      "learning_rate": 5.4997453826422286e-06,
+      "loss": 0.2885,
+      "step": 7698
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8423842920647373,
+      "learning_rate": 5.498731995734416e-06,
+      "loss": 0.2876,
+      "step": 7699
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.992622462410933,
+      "learning_rate": 5.497718588133325e-06,
+      "loss": 0.2942,
+      "step": 7700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.8564163935973115,
+      "learning_rate": 5.4967051598810074e-06,
+      "loss": 0.3164,
+      "step": 7701
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.905517536132355,
+      "learning_rate": 5.4956917110195096e-06,
+      "loss": 0.2868,
+      "step": 7702
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7386476152167354,
+      "learning_rate": 5.494678241590883e-06,
+      "loss": 0.3107,
+      "step": 7703
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7342282734045082,
+      "learning_rate": 5.493664751637176e-06,
+      "loss": 0.3173,
+      "step": 7704
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0065624921530523,
+      "learning_rate": 5.492651241200443e-06,
+      "loss": 0.3109,
+      "step": 7705
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7653686719302852,
+      "learning_rate": 5.491637710322735e-06,
+      "loss": 0.3246,
+      "step": 7706
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.287865288384318,
+      "learning_rate": 5.490624159046104e-06,
+      "loss": 0.2763,
+      "step": 7707
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5601821995705238,
+      "learning_rate": 5.4896105874126045e-06,
+      "loss": 0.2883,
+      "step": 7708
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5696044566668663,
+      "learning_rate": 5.488596995464295e-06,
+      "loss": 0.281,
+      "step": 7709
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.08586468425905,
+      "learning_rate": 5.4875833832432265e-06,
+      "loss": 0.3066,
+      "step": 7710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.815212508735295,
+      "learning_rate": 5.486569750791457e-06,
+      "loss": 0.2832,
+      "step": 7711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.05406033481365,
+      "learning_rate": 5.485556098151045e-06,
+      "loss": 0.2888,
+      "step": 7712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7899025505151624,
+      "learning_rate": 5.484542425364049e-06,
+      "loss": 0.2945,
+      "step": 7713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4809174048149893,
+      "learning_rate": 5.483528732472525e-06,
+      "loss": 0.2767,
+      "step": 7714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6367459434759566,
+      "learning_rate": 5.482515019518536e-06,
+      "loss": 0.5068,
+      "step": 7715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6513782255389672,
+      "learning_rate": 5.481501286544142e-06,
+      "loss": 0.4778,
+      "step": 7716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.362815848020362,
+      "learning_rate": 5.480487533591405e-06,
+      "loss": 0.319,
+      "step": 7717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.171551057908031,
+      "learning_rate": 5.479473760702386e-06,
+      "loss": 0.2669,
+      "step": 7718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.258315300024818,
+      "learning_rate": 5.478459967919149e-06,
+      "loss": 0.3362,
+      "step": 7719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6593417003266453,
+      "learning_rate": 5.477446155283758e-06,
+      "loss": 0.303,
+      "step": 7720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5581072388702726,
+      "learning_rate": 5.476432322838279e-06,
+      "loss": 0.3513,
+      "step": 7721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.589422087143522,
+      "learning_rate": 5.4754184706247745e-06,
+      "loss": 0.2976,
+      "step": 7722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0525801997020117,
+      "learning_rate": 5.474404598685315e-06,
+      "loss": 0.2846,
+      "step": 7723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.565055781579796,
+      "learning_rate": 5.473390707061965e-06,
+      "loss": 0.2857,
+      "step": 7724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6844960046250694,
+      "learning_rate": 5.4723767957967955e-06,
+      "loss": 0.2706,
+      "step": 7725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.364943055973655,
+      "learning_rate": 5.4713628649318716e-06,
+      "loss": 0.3081,
+      "step": 7726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6268223581545005,
+      "learning_rate": 5.470348914509267e-06,
+      "loss": 0.2675,
+      "step": 7727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5371517334755775,
+      "learning_rate": 5.46933494457105e-06,
+      "loss": 0.2983,
+      "step": 7728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4963837345356237,
+      "learning_rate": 5.468320955159293e-06,
+      "loss": 0.281,
+      "step": 7729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9170074890030921,
+      "learning_rate": 5.467306946316066e-06,
+      "loss": 0.2804,
+      "step": 7730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.316003764567229,
+      "learning_rate": 5.466292918083444e-06,
+      "loss": 0.2852,
+      "step": 7731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.30807583938851,
+      "learning_rate": 5.4652788705035024e-06,
+      "loss": 0.308,
+      "step": 7732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7788902829699256,
+      "learning_rate": 5.464264803618312e-06,
+      "loss": 0.304,
+      "step": 7733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0855516531344285,
+      "learning_rate": 5.463250717469951e-06,
+      "loss": 0.3121,
+      "step": 7734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.734483862790188,
+      "learning_rate": 5.462236612100496e-06,
+      "loss": 0.2936,
+      "step": 7735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.071891275691287,
+      "learning_rate": 5.461222487552022e-06,
+      "loss": 0.3151,
+      "step": 7736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.735982423941206,
+      "learning_rate": 5.460208343866607e-06,
+      "loss": 0.2907,
+      "step": 7737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4393485166060214,
+      "learning_rate": 5.4591941810863314e-06,
+      "loss": 0.2968,
+      "step": 7738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8097763811921834,
+      "learning_rate": 5.458179999253274e-06,
+      "loss": 0.4989,
+      "step": 7739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.011088761778192,
+      "learning_rate": 5.457165798409514e-06,
+      "loss": 0.2902,
+      "step": 7740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.871560217624289,
+      "learning_rate": 5.456151578597133e-06,
+      "loss": 0.3,
+      "step": 7741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.724933861169226,
+      "learning_rate": 5.455137339858212e-06,
+      "loss": 0.3002,
+      "step": 7742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.233794883202092,
+      "learning_rate": 5.454123082234837e-06,
+      "loss": 0.2932,
+      "step": 7743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.911994075382329,
+      "learning_rate": 5.4531088057690864e-06,
+      "loss": 0.2949,
+      "step": 7744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8799183901741925,
+      "learning_rate": 5.4520945105030466e-06,
+      "loss": 0.2649,
+      "step": 7745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.617560859902063,
+      "learning_rate": 5.451080196478803e-06,
+      "loss": 0.2842,
+      "step": 7746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1354696114542797,
+      "learning_rate": 5.450065863738442e-06,
+      "loss": 0.283,
+      "step": 7747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2134631563214096,
+      "learning_rate": 5.449051512324046e-06,
+      "loss": 0.3138,
+      "step": 7748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6089507927171013,
+      "learning_rate": 5.448037142277708e-06,
+      "loss": 0.49,
+      "step": 7749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.031014695423584,
+      "learning_rate": 5.447022753641511e-06,
+      "loss": 0.2648,
+      "step": 7750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.389528252136984,
+      "learning_rate": 5.446008346457549e-06,
+      "loss": 0.2884,
+      "step": 7751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8959402309699893,
+      "learning_rate": 5.444993920767905e-06,
+      "loss": 0.2959,
+      "step": 7752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9432775176127008,
+      "learning_rate": 5.443979476614674e-06,
+      "loss": 0.2958,
+      "step": 7753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6567397525583398,
+      "learning_rate": 5.442965014039947e-06,
+      "loss": 0.2832,
+      "step": 7754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.751743086406248,
+      "learning_rate": 5.441950533085814e-06,
+      "loss": 0.2994,
+      "step": 7755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3200325140976727,
+      "learning_rate": 5.440936033794368e-06,
+      "loss": 0.2856,
+      "step": 7756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5140588085294735,
+      "learning_rate": 5.439921516207704e-06,
+      "loss": 0.2817,
+      "step": 7757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.233163376378304,
+      "learning_rate": 5.438906980367914e-06,
+      "loss": 0.2879,
+      "step": 7758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.199435167670893,
+      "learning_rate": 5.437892426317095e-06,
+      "loss": 0.2917,
+      "step": 7759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0216974147693554,
+      "learning_rate": 5.436877854097338e-06,
+      "loss": 0.2944,
+      "step": 7760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7542178720395472,
+      "learning_rate": 5.435863263750747e-06,
+      "loss": 0.2779,
+      "step": 7761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0811060036110973,
+      "learning_rate": 5.434848655319414e-06,
+      "loss": 0.2844,
+      "step": 7762
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.383498166954667,
+      "learning_rate": 5.433834028845436e-06,
+      "loss": 0.2921,
+      "step": 7763
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3910709502821,
+      "learning_rate": 5.432819384370914e-06,
+      "loss": 0.2864,
+      "step": 7764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.151322561434658,
+      "learning_rate": 5.431804721937949e-06,
+      "loss": 0.2692,
+      "step": 7765
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.33066431819288,
+      "learning_rate": 5.4307900415886374e-06,
+      "loss": 0.2765,
+      "step": 7766
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1432211433391815,
+      "learning_rate": 5.429775343365082e-06,
+      "loss": 0.3102,
+      "step": 7767
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9742557435916317,
+      "learning_rate": 5.428760627309384e-06,
+      "loss": 0.2803,
+      "step": 7768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2674436633968107,
+      "learning_rate": 5.427745893463647e-06,
+      "loss": 0.2865,
+      "step": 7769
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6088395490169526,
+      "learning_rate": 5.426731141869973e-06,
+      "loss": 0.4981,
+      "step": 7770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.399780861060085,
+      "learning_rate": 5.425716372570466e-06,
+      "loss": 0.295,
+      "step": 7771
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.097867324104314,
+      "learning_rate": 5.4247015856072295e-06,
+      "loss": 0.2791,
+      "step": 7772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.747160477436948,
+      "learning_rate": 5.4236867810223715e-06,
+      "loss": 0.2953,
+      "step": 7773
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.007433151237102,
+      "learning_rate": 5.4226719588579935e-06,
+      "loss": 0.3012,
+      "step": 7774
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0237241169132663,
+      "learning_rate": 5.421657119156208e-06,
+      "loss": 0.285,
+      "step": 7775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1575521191540536,
+      "learning_rate": 5.420642261959118e-06,
+      "loss": 0.2872,
+      "step": 7776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5160786933243062,
+      "learning_rate": 5.419627387308836e-06,
+      "loss": 0.2766,
+      "step": 7777
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6131710196763749,
+      "learning_rate": 5.418612495247465e-06,
+      "loss": 0.2878,
+      "step": 7778
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3376476265177493,
+      "learning_rate": 5.4175975858171204e-06,
+      "loss": 0.2933,
+      "step": 7779
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4680007350926543,
+      "learning_rate": 5.416582659059909e-06,
+      "loss": 0.2956,
+      "step": 7780
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6099706267873062,
+      "learning_rate": 5.4155677150179446e-06,
+      "loss": 0.2824,
+      "step": 7781
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.052321759581494,
+      "learning_rate": 5.414552753733334e-06,
+      "loss": 0.2837,
+      "step": 7782
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6167219981940524,
+      "learning_rate": 5.413537775248198e-06,
+      "loss": 0.4771,
+      "step": 7783
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.557764588811983,
+      "learning_rate": 5.412522779604642e-06,
+      "loss": 0.2922,
+      "step": 7784
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.8335568018077275,
+      "learning_rate": 5.411507766844784e-06,
+      "loss": 0.3058,
+      "step": 7785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.289772327945619,
+      "learning_rate": 5.410492737010737e-06,
+      "loss": 0.2712,
+      "step": 7786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7982921294117131,
+      "learning_rate": 5.40947769014462e-06,
+      "loss": 0.2914,
+      "step": 7787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.441009830512988,
+      "learning_rate": 5.408462626288544e-06,
+      "loss": 0.2992,
+      "step": 7788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.525357207988294,
+      "learning_rate": 5.4074475454846275e-06,
+      "loss": 0.3066,
+      "step": 7789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.45928730511721,
+      "learning_rate": 5.4064324477749895e-06,
+      "loss": 0.291,
+      "step": 7790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0957578722529746,
+      "learning_rate": 5.405417333201749e-06,
+      "loss": 0.2946,
+      "step": 7791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1496594428860574,
+      "learning_rate": 5.404402201807022e-06,
+      "loss": 0.2933,
+      "step": 7792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.366107406800677,
+      "learning_rate": 5.403387053632928e-06,
+      "loss": 0.2781,
+      "step": 7793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.017692727243486,
+      "learning_rate": 5.4023718887215906e-06,
+      "loss": 0.2955,
+      "step": 7794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5367372740132124,
+      "learning_rate": 5.401356707115128e-06,
+      "loss": 0.2757,
+      "step": 7795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4266617915544213,
+      "learning_rate": 5.400341508855663e-06,
+      "loss": 0.2735,
+      "step": 7796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.657927667472044,
+      "learning_rate": 5.3993262939853175e-06,
+      "loss": 0.2901,
+      "step": 7797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.038135513034856,
+      "learning_rate": 5.3983110625462144e-06,
+      "loss": 0.3043,
+      "step": 7798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4296953850693885,
+      "learning_rate": 5.397295814580479e-06,
+      "loss": 0.3101,
+      "step": 7799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.029014493023995,
+      "learning_rate": 5.396280550130234e-06,
+      "loss": 0.2989,
+      "step": 7800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6424893600592279,
+      "learning_rate": 5.395265269237604e-06,
+      "loss": 0.2816,
+      "step": 7801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.602559189504729,
+      "learning_rate": 5.394249971944717e-06,
+      "loss": 0.3001,
+      "step": 7802
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.380024316459885,
+      "learning_rate": 5.393234658293699e-06,
+      "loss": 0.2773,
+      "step": 7803
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1857487631942005,
+      "learning_rate": 5.392219328326674e-06,
+      "loss": 0.317,
+      "step": 7804
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.544890840486099,
+      "learning_rate": 5.391203982085775e-06,
+      "loss": 0.2928,
+      "step": 7805
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6648914466207075,
+      "learning_rate": 5.390188619613127e-06,
+      "loss": 0.4924,
+      "step": 7806
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8318319675761625,
+      "learning_rate": 5.389173240950861e-06,
+      "loss": 0.2878,
+      "step": 7807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2661003354214695,
+      "learning_rate": 5.388157846141105e-06,
+      "loss": 0.3093,
+      "step": 7808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3939666974106566,
+      "learning_rate": 5.3871424352259904e-06,
+      "loss": 0.2809,
+      "step": 7809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.132518406526393,
+      "learning_rate": 5.386127008247649e-06,
+      "loss": 0.3026,
+      "step": 7810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.422275574293052,
+      "learning_rate": 5.385111565248212e-06,
+      "loss": 0.2769,
+      "step": 7811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.377147047080068,
+      "learning_rate": 5.384096106269811e-06,
+      "loss": 0.2952,
+      "step": 7812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5309259557800277,
+      "learning_rate": 5.383080631354582e-06,
+      "loss": 0.2883,
+      "step": 7813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.260658727800435,
+      "learning_rate": 5.3820651405446564e-06,
+      "loss": 0.2987,
+      "step": 7814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.189576227553726,
+      "learning_rate": 5.38104963388217e-06,
+      "loss": 0.3113,
+      "step": 7815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.321911564146936,
+      "learning_rate": 5.380034111409257e-06,
+      "loss": 0.2805,
+      "step": 7816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.417295041113762,
+      "learning_rate": 5.379018573168056e-06,
+      "loss": 0.3089,
+      "step": 7817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.347253533304037,
+      "learning_rate": 5.378003019200699e-06,
+      "loss": 0.2785,
+      "step": 7818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.114979822151295,
+      "learning_rate": 5.376987449549325e-06,
+      "loss": 0.3094,
+      "step": 7819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.763573269754107,
+      "learning_rate": 5.375971864256071e-06,
+      "loss": 0.2907,
+      "step": 7820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.257328131567649,
+      "learning_rate": 5.3749562633630795e-06,
+      "loss": 0.2863,
+      "step": 7821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.837231319199659,
+      "learning_rate": 5.373940646912485e-06,
+      "loss": 0.2845,
+      "step": 7822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0511031390317953,
+      "learning_rate": 5.372925014946428e-06,
+      "loss": 0.2895,
+      "step": 7823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6441202753609079,
+      "learning_rate": 5.371909367507051e-06,
+      "loss": 0.2964,
+      "step": 7824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.657212904994267,
+      "learning_rate": 5.370893704636495e-06,
+      "loss": 0.3129,
+      "step": 7825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.289250358737912,
+      "learning_rate": 5.369878026376899e-06,
+      "loss": 0.2884,
+      "step": 7826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.438647161835852,
+      "learning_rate": 5.368862332770406e-06,
+      "loss": 0.2834,
+      "step": 7827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.208852229409675,
+      "learning_rate": 5.36784662385916e-06,
+      "loss": 0.2716,
+      "step": 7828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9287582100892882,
+      "learning_rate": 5.366830899685306e-06,
+      "loss": 0.277,
+      "step": 7829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.903851899455363,
+      "learning_rate": 5.365815160290983e-06,
+      "loss": 0.2959,
+      "step": 7830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1247015783700065,
+      "learning_rate": 5.364799405718342e-06,
+      "loss": 0.3171,
+      "step": 7831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6676871413798984,
+      "learning_rate": 5.3637836360095255e-06,
+      "loss": 0.2803,
+      "step": 7832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7662499131575617,
+      "learning_rate": 5.3627678512066795e-06,
+      "loss": 0.2876,
+      "step": 7833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6006633156553667,
+      "learning_rate": 5.36175205135195e-06,
+      "loss": 0.4691,
+      "step": 7834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4657365438642187,
+      "learning_rate": 5.360736236487486e-06,
+      "loss": 0.2788,
+      "step": 7835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.961876106210694,
+      "learning_rate": 5.359720406655435e-06,
+      "loss": 0.2735,
+      "step": 7836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2152937648310767,
+      "learning_rate": 5.358704561897946e-06,
+      "loss": 0.2892,
+      "step": 7837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.053247394787333,
+      "learning_rate": 5.357688702257165e-06,
+      "loss": 0.2795,
+      "step": 7838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 26.465164130014912,
+      "learning_rate": 5.3566728277752474e-06,
+      "loss": 0.2893,
+      "step": 7839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5914615193943744,
+      "learning_rate": 5.355656938494339e-06,
+      "loss": 0.2733,
+      "step": 7840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.148678242380159,
+      "learning_rate": 5.354641034456595e-06,
+      "loss": 0.3147,
+      "step": 7841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5725284290527721,
+      "learning_rate": 5.353625115704161e-06,
+      "loss": 0.2893,
+      "step": 7842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.870482497199345,
+      "learning_rate": 5.352609182279195e-06,
+      "loss": 0.297,
+      "step": 7843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8492267958161406,
+      "learning_rate": 5.351593234223847e-06,
+      "loss": 0.2903,
+      "step": 7844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.461672599556936,
+      "learning_rate": 5.35057727158027e-06,
+      "loss": 0.2575,
+      "step": 7845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2814202977521445,
+      "learning_rate": 5.349561294390622e-06,
+      "loss": 0.2828,
+      "step": 7846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.742395186833296,
+      "learning_rate": 5.348545302697054e-06,
+      "loss": 0.3104,
+      "step": 7847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5456809314140165,
+      "learning_rate": 5.347529296541721e-06,
+      "loss": 0.29,
+      "step": 7848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 26.66109870909653,
+      "learning_rate": 5.346513275966782e-06,
+      "loss": 0.278,
+      "step": 7849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1719069188333964,
+      "learning_rate": 5.34549724101439e-06,
+      "loss": 0.2842,
+      "step": 7850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2886971048340783,
+      "learning_rate": 5.344481191726706e-06,
+      "loss": 0.2795,
+      "step": 7851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3802020504268517,
+      "learning_rate": 5.343465128145884e-06,
+      "loss": 0.2882,
+      "step": 7852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4542347798396273,
+      "learning_rate": 5.342449050314084e-06,
+      "loss": 0.2714,
+      "step": 7853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7576394213758326,
+      "learning_rate": 5.3414329582734635e-06,
+      "loss": 0.3048,
+      "step": 7854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9112309111662238,
+      "learning_rate": 5.340416852066185e-06,
+      "loss": 0.2774,
+      "step": 7855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.112117802862802,
+      "learning_rate": 5.339400731734404e-06,
+      "loss": 0.281,
+      "step": 7856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6640068522887566,
+      "learning_rate": 5.338384597320287e-06,
+      "loss": 0.281,
+      "step": 7857
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9509680169432204,
+      "learning_rate": 5.33736844886599e-06,
+      "loss": 0.2896,
+      "step": 7858
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3020465188496206,
+      "learning_rate": 5.336352286413678e-06,
+      "loss": 0.2813,
+      "step": 7859
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7139299680622218,
+      "learning_rate": 5.335336110005511e-06,
+      "loss": 0.2889,
+      "step": 7860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7794580784858343,
+      "learning_rate": 5.3343199196836545e-06,
+      "loss": 0.2877,
+      "step": 7861
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.419535508943437,
+      "learning_rate": 5.33330371549027e-06,
+      "loss": 0.2707,
+      "step": 7862
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.62458308913267,
+      "learning_rate": 5.332287497467523e-06,
+      "loss": 0.3135,
+      "step": 7863
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1344687018315707,
+      "learning_rate": 5.331271265657576e-06,
+      "loss": 0.3049,
+      "step": 7864
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.423922270242874,
+      "learning_rate": 5.330255020102598e-06,
+      "loss": 0.2822,
+      "step": 7865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.138042924156716,
+      "learning_rate": 5.329238760844751e-06,
+      "loss": 0.3126,
+      "step": 7866
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.442546742227438,
+      "learning_rate": 5.328222487926204e-06,
+      "loss": 0.3147,
+      "step": 7867
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.370689501851116,
+      "learning_rate": 5.327206201389121e-06,
+      "loss": 0.2864,
+      "step": 7868
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.397388091452828,
+      "learning_rate": 5.326189901275673e-06,
+      "loss": 0.2871,
+      "step": 7869
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2757360750300504,
+      "learning_rate": 5.325173587628028e-06,
+      "loss": 0.2865,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8233082627858472,
+      "learning_rate": 5.324157260488351e-06,
+      "loss": 0.3177,
+      "step": 7871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0797174636973628,
+      "learning_rate": 5.3231409198988136e-06,
+      "loss": 0.2977,
+      "step": 7872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4364150530437172,
+      "learning_rate": 5.322124565901587e-06,
+      "loss": 0.2831,
+      "step": 7873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5394900767597413,
+      "learning_rate": 5.321108198538839e-06,
+      "loss": 0.3183,
+      "step": 7874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1928337204093706,
+      "learning_rate": 5.3200918178527415e-06,
+      "loss": 0.2794,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.42904582272032,
+      "learning_rate": 5.319075423885466e-06,
+      "loss": 0.2946,
+      "step": 7876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3753258728152593,
+      "learning_rate": 5.318059016679184e-06,
+      "loss": 0.2875,
+      "step": 7877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5241915561094757,
+      "learning_rate": 5.31704259627607e-06,
+      "loss": 0.2903,
+      "step": 7878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3199319689602134,
+      "learning_rate": 5.316026162718294e-06,
+      "loss": 0.3044,
+      "step": 7879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9476814347952272,
+      "learning_rate": 5.315009716048031e-06,
+      "loss": 0.3045,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7439968772929455,
+      "learning_rate": 5.313993256307457e-06,
+      "loss": 0.2993,
+      "step": 7881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9012851842038336,
+      "learning_rate": 5.312976783538743e-06,
+      "loss": 0.2832,
+      "step": 7882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7364297307752168,
+      "learning_rate": 5.3119602977840675e-06,
+      "loss": 0.274,
+      "step": 7883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1144750913095374,
+      "learning_rate": 5.310943799085605e-06,
+      "loss": 0.285,
+      "step": 7884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7648450352078844,
+      "learning_rate": 5.309927287485533e-06,
+      "loss": 0.2915,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7324473464984051,
+      "learning_rate": 5.308910763026025e-06,
+      "loss": 0.277,
+      "step": 7886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7847826428085893,
+      "learning_rate": 5.3078942257492635e-06,
+      "loss": 0.29,
+      "step": 7887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.407069782983817,
+      "learning_rate": 5.306877675697422e-06,
+      "loss": 0.299,
+      "step": 7888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0505374188782954,
+      "learning_rate": 5.305861112912682e-06,
+      "loss": 0.3191,
+      "step": 7889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4128359405053703,
+      "learning_rate": 5.3048445374372195e-06,
+      "loss": 0.2896,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.525059847611393,
+      "learning_rate": 5.303827949313216e-06,
+      "loss": 0.2844,
+      "step": 7891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0767857423777576,
+      "learning_rate": 5.302811348582851e-06,
+      "loss": 0.2865,
+      "step": 7892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8358489094171377,
+      "learning_rate": 5.301794735288307e-06,
+      "loss": 0.3444,
+      "step": 7893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.868254505609128,
+      "learning_rate": 5.300778109471761e-06,
+      "loss": 0.2877,
+      "step": 7894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5981025833410399,
+      "learning_rate": 5.2997614711753995e-06,
+      "loss": 0.4733,
+      "step": 7895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6909587767237433,
+      "learning_rate": 5.298744820441401e-06,
+      "loss": 0.2701,
+      "step": 7896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.89792792416731,
+      "learning_rate": 5.297728157311949e-06,
+      "loss": 0.2771,
+      "step": 7897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3995807621527219,
+      "learning_rate": 5.296711481829227e-06,
+      "loss": 0.2811,
+      "step": 7898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.174499932223166,
+      "learning_rate": 5.295694794035419e-06,
+      "loss": 0.2876,
+      "step": 7899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0959527284495594,
+      "learning_rate": 5.2946780939727084e-06,
+      "loss": 0.2969,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.323792377120759,
+      "learning_rate": 5.29366138168328e-06,
+      "loss": 0.3005,
+      "step": 7901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8694081748452838,
+      "learning_rate": 5.292644657209319e-06,
+      "loss": 0.2853,
+      "step": 7902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0469038900242316,
+      "learning_rate": 5.291627920593014e-06,
+      "loss": 0.2886,
+      "step": 7903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5926515851845977,
+      "learning_rate": 5.290611171876545e-06,
+      "loss": 0.4816,
+      "step": 7904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5227893147930167,
+      "learning_rate": 5.289594411102103e-06,
+      "loss": 0.2952,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4201675984421223,
+      "learning_rate": 5.288577638311876e-06,
+      "loss": 0.2912,
+      "step": 7906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7328693828558057,
+      "learning_rate": 5.287560853548051e-06,
+      "loss": 0.2797,
+      "step": 7907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5812456041002827,
+      "learning_rate": 5.286544056852814e-06,
+      "loss": 0.489,
+      "step": 7908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4004950807853747,
+      "learning_rate": 5.285527248268354e-06,
+      "loss": 0.2968,
+      "step": 7909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3279271592857764,
+      "learning_rate": 5.2845104278368616e-06,
+      "loss": 0.2952,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.069382801279133,
+      "learning_rate": 5.283493595600529e-06,
+      "loss": 0.2866,
+      "step": 7911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8005879057685519,
+      "learning_rate": 5.28247675160154e-06,
+      "loss": 0.2826,
+      "step": 7912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8240246921078522,
+      "learning_rate": 5.281459895882091e-06,
+      "loss": 0.2896,
+      "step": 7913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5857586805162879,
+      "learning_rate": 5.28044302848437e-06,
+      "loss": 0.4779,
+      "step": 7914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.883077599717525,
+      "learning_rate": 5.279426149450571e-06,
+      "loss": 0.2655,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.706124916748602,
+      "learning_rate": 5.278409258822883e-06,
+      "loss": 0.2961,
+      "step": 7916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3641752380519125,
+      "learning_rate": 5.277392356643501e-06,
+      "loss": 0.2874,
+      "step": 7917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7134409761385343,
+      "learning_rate": 5.276375442954618e-06,
+      "loss": 0.2846,
+      "step": 7918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7934162210103406,
+      "learning_rate": 5.275358517798428e-06,
+      "loss": 0.304,
+      "step": 7919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.784469004248253,
+      "learning_rate": 5.27434158121712e-06,
+      "loss": 0.2854,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.640741575724123,
+      "learning_rate": 5.273324633252897e-06,
+      "loss": 0.2837,
+      "step": 7921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6976472680018913,
+      "learning_rate": 5.272307673947947e-06,
+      "loss": 0.2747,
+      "step": 7922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7838076336567272,
+      "learning_rate": 5.271290703344469e-06,
+      "loss": 0.2779,
+      "step": 7923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.283616882704457,
+      "learning_rate": 5.270273721484657e-06,
+      "loss": 0.2924,
+      "step": 7924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6596932629621144,
+      "learning_rate": 5.269256728410709e-06,
+      "loss": 0.2836,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3390694006704675,
+      "learning_rate": 5.268239724164819e-06,
+      "loss": 0.2877,
+      "step": 7926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4626165226004626,
+      "learning_rate": 5.267222708789189e-06,
+      "loss": 0.2914,
+      "step": 7927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5017678493186915,
+      "learning_rate": 5.266205682326013e-06,
+      "loss": 0.3004,
+      "step": 7928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4501573454757062,
+      "learning_rate": 5.265188644817492e-06,
+      "loss": 0.2717,
+      "step": 7929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9411583587568177,
+      "learning_rate": 5.264171596305821e-06,
+      "loss": 0.3036,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5920735045894463,
+      "learning_rate": 5.263154536833202e-06,
+      "loss": 0.3052,
+      "step": 7931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8451966896150458,
+      "learning_rate": 5.262137466441834e-06,
+      "loss": 0.3133,
+      "step": 7932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4059217498548913,
+      "learning_rate": 5.261120385173917e-06,
+      "loss": 0.2871,
+      "step": 7933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.705879722655072,
+      "learning_rate": 5.260103293071651e-06,
+      "loss": 0.2989,
+      "step": 7934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1456088508382196,
+      "learning_rate": 5.259086190177237e-06,
+      "loss": 0.3024,
+      "step": 7935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.79279948609646,
+      "learning_rate": 5.258069076532877e-06,
+      "loss": 0.2995,
+      "step": 7936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.050047551143473,
+      "learning_rate": 5.257051952180774e-06,
+      "loss": 0.3132,
+      "step": 7937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.05594381130633,
+      "learning_rate": 5.256034817163127e-06,
+      "loss": 0.3071,
+      "step": 7938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0684131487006554,
+      "learning_rate": 5.255017671522142e-06,
+      "loss": 0.2834,
+      "step": 7939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.241061238094606,
+      "learning_rate": 5.254000515300019e-06,
+      "loss": 0.286,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7180458792869078,
+      "learning_rate": 5.252983348538967e-06,
+      "loss": 0.2931,
+      "step": 7941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.548697770337576,
+      "learning_rate": 5.2519661712811845e-06,
+      "loss": 0.2959,
+      "step": 7942
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3352332012162123,
+      "learning_rate": 5.250948983568876e-06,
+      "loss": 0.3082,
+      "step": 7943
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7361041982240581,
+      "learning_rate": 5.249931785444251e-06,
+      "loss": 0.293,
+      "step": 7944
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7055614049768366,
+      "learning_rate": 5.248914576949512e-06,
+      "loss": 0.2985,
+      "step": 7945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4417005900917659,
+      "learning_rate": 5.2478973581268645e-06,
+      "loss": 0.29,
+      "step": 7946
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9337301910545015,
+      "learning_rate": 5.246880129018515e-06,
+      "loss": 0.2833,
+      "step": 7947
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5940934482588602,
+      "learning_rate": 5.24586288966667e-06,
+      "loss": 0.2961,
+      "step": 7948
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8112097807465304,
+      "learning_rate": 5.24484564011354e-06,
+      "loss": 0.3054,
+      "step": 7949
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.205560532580205,
+      "learning_rate": 5.2438283804013265e-06,
+      "loss": 0.2837,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.236503306564229,
+      "learning_rate": 5.242811110572243e-06,
+      "loss": 0.2845,
+      "step": 7951
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.203666350475942,
+      "learning_rate": 5.241793830668492e-06,
+      "loss": 0.2985,
+      "step": 7952
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0611771440486573,
+      "learning_rate": 5.240776540732288e-06,
+      "loss": 0.2676,
+      "step": 7953
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.653082607696945,
+      "learning_rate": 5.239759240805835e-06,
+      "loss": 0.2984,
+      "step": 7954
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.7539020266152905,
+      "learning_rate": 5.238741930931348e-06,
+      "loss": 0.2772,
+      "step": 7955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.283071163067402,
+      "learning_rate": 5.237724611151034e-06,
+      "loss": 0.2782,
+      "step": 7956
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6218217562479014,
+      "learning_rate": 5.2367072815071015e-06,
+      "loss": 0.5029,
+      "step": 7957
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.276229364350079,
+      "learning_rate": 5.235689942041765e-06,
+      "loss": 0.3007,
+      "step": 7958
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.387998062722005,
+      "learning_rate": 5.234672592797236e-06,
+      "loss": 0.2815,
+      "step": 7959
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6021124702158773,
+      "learning_rate": 5.233655233815721e-06,
+      "loss": 0.2809,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.611070341895981,
+      "learning_rate": 5.232637865139436e-06,
+      "loss": 0.2809,
+      "step": 7961
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.496121167524374,
+      "learning_rate": 5.231620486810594e-06,
+      "loss": 0.3149,
+      "step": 7962
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3004514797850226,
+      "learning_rate": 5.230603098871406e-06,
+      "loss": 0.2733,
+      "step": 7963
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.787446885363309,
+      "learning_rate": 5.229585701364086e-06,
+      "loss": 0.2961,
+      "step": 7964
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5105309186295437,
+      "learning_rate": 5.228568294330847e-06,
+      "loss": 0.2878,
+      "step": 7965
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9272804431534816,
+      "learning_rate": 5.227550877813903e-06,
+      "loss": 0.299,
+      "step": 7966
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.332172028792701,
+      "learning_rate": 5.226533451855471e-06,
+      "loss": 0.29,
+      "step": 7967
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8385924181373738,
+      "learning_rate": 5.225516016497761e-06,
+      "loss": 0.2868,
+      "step": 7968
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.238441197023188,
+      "learning_rate": 5.224498571782992e-06,
+      "loss": 0.2826,
+      "step": 7969
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7610732894069905,
+      "learning_rate": 5.223481117753379e-06,
+      "loss": 0.2874,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6148411709942464,
+      "learning_rate": 5.222463654451138e-06,
+      "loss": 0.5345,
+      "step": 7971
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.850150281025383,
+      "learning_rate": 5.221446181918484e-06,
+      "loss": 0.3108,
+      "step": 7972
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8929554689762693,
+      "learning_rate": 5.220428700197635e-06,
+      "loss": 0.2887,
+      "step": 7973
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2302322929908756,
+      "learning_rate": 5.219411209330807e-06,
+      "loss": 0.3011,
+      "step": 7974
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6597011141365123,
+      "learning_rate": 5.218393709360219e-06,
+      "loss": 0.2634,
+      "step": 7975
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8544295318833768,
+      "learning_rate": 5.217376200328087e-06,
+      "loss": 0.2901,
+      "step": 7976
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6148198615159677,
+      "learning_rate": 5.216358682276631e-06,
+      "loss": 0.2785,
+      "step": 7977
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8760216823100602,
+      "learning_rate": 5.215341155248069e-06,
+      "loss": 0.2778,
+      "step": 7978
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.19102729214334,
+      "learning_rate": 5.214323619284619e-06,
+      "loss": 0.2741,
+      "step": 7979
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5999705552751469,
+      "learning_rate": 5.213306074428503e-06,
+      "loss": 0.2764,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5029067945288253,
+      "learning_rate": 5.212288520721939e-06,
+      "loss": 0.2786,
+      "step": 7981
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.002508771973122,
+      "learning_rate": 5.2112709582071464e-06,
+      "loss": 0.2876,
+      "step": 7982
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.874911072768043,
+      "learning_rate": 5.210253386926346e-06,
+      "loss": 0.2807,
+      "step": 7983
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.293807677620967,
+      "learning_rate": 5.20923580692176e-06,
+      "loss": 0.2857,
+      "step": 7984
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.896818899562079,
+      "learning_rate": 5.208218218235609e-06,
+      "loss": 0.2899,
+      "step": 7985
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.450485937256754,
+      "learning_rate": 5.207200620910114e-06,
+      "loss": 0.2894,
+      "step": 7986
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0716587200662073,
+      "learning_rate": 5.206183014987497e-06,
+      "loss": 0.2711,
+      "step": 7987
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8071708646158895,
+      "learning_rate": 5.205165400509982e-06,
+      "loss": 0.2742,
+      "step": 7988
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2390266213904055,
+      "learning_rate": 5.2041477775197875e-06,
+      "loss": 0.2896,
+      "step": 7989
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.839432018744269,
+      "learning_rate": 5.20313014605914e-06,
+      "loss": 0.2774,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.098592260273801,
+      "learning_rate": 5.202112506170263e-06,
+      "loss": 0.2836,
+      "step": 7991
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.037684427485203,
+      "learning_rate": 5.201094857895377e-06,
+      "loss": 0.271,
+      "step": 7992
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0779139106441136,
+      "learning_rate": 5.200077201276711e-06,
+      "loss": 0.2819,
+      "step": 7993
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0179452472155095,
+      "learning_rate": 5.1990595363564845e-06,
+      "loss": 0.3158,
+      "step": 7994
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2662962933022985,
+      "learning_rate": 5.198041863176925e-06,
+      "loss": 0.293,
+      "step": 7995
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.694266813304795,
+      "learning_rate": 5.197024181780256e-06,
+      "loss": 0.2718,
+      "step": 7996
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5220341333856093,
+      "learning_rate": 5.196006492208705e-06,
+      "loss": 0.2784,
+      "step": 7997
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6307678542495885,
+      "learning_rate": 5.194988794504495e-06,
+      "loss": 0.3027,
+      "step": 7998
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6771723970344932,
+      "learning_rate": 5.193971088709855e-06,
+      "loss": 0.2854,
+      "step": 7999
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6996155697591504,
+      "learning_rate": 5.192953374867009e-06,
+      "loss": 0.2816,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.342947985914777,
+      "learning_rate": 5.1919356530181865e-06,
+      "loss": 0.277,
+      "step": 8001
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1335009395378743,
+      "learning_rate": 5.190917923205611e-06,
+      "loss": 0.2752,
+      "step": 8002
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8525786953817451,
+      "learning_rate": 5.189900185471511e-06,
+      "loss": 0.2832,
+      "step": 8003
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.18225132389599,
+      "learning_rate": 5.188882439858117e-06,
+      "loss": 0.2876,
+      "step": 8004
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.368988241224818,
+      "learning_rate": 5.187864686407656e-06,
+      "loss": 0.2769,
+      "step": 8005
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2003221974676173,
+      "learning_rate": 5.186846925162353e-06,
+      "loss": 0.2889,
+      "step": 8006
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4983689874048514,
+      "learning_rate": 5.1858291561644394e-06,
+      "loss": 0.2831,
+      "step": 8007
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8058272796197763,
+      "learning_rate": 5.184811379456145e-06,
+      "loss": 0.2903,
+      "step": 8008
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7720071199535712,
+      "learning_rate": 5.183793595079697e-06,
+      "loss": 0.2825,
+      "step": 8009
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.181826200879759,
+      "learning_rate": 5.1827758030773275e-06,
+      "loss": 0.2599,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.567614006042909,
+      "learning_rate": 5.181758003491265e-06,
+      "loss": 0.2941,
+      "step": 8011
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1988998909507154,
+      "learning_rate": 5.1807401963637404e-06,
+      "loss": 0.29,
+      "step": 8012
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.083542254299491,
+      "learning_rate": 5.179722381736983e-06,
+      "loss": 0.2762,
+      "step": 8013
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2464794092514846,
+      "learning_rate": 5.178704559653227e-06,
+      "loss": 0.271,
+      "step": 8014
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.59916161382316,
+      "learning_rate": 5.1776867301547e-06,
+      "loss": 0.5044,
+      "step": 8015
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2064660580278344,
+      "learning_rate": 5.176668893283634e-06,
+      "loss": 0.2926,
+      "step": 8016
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6907536057633934,
+      "learning_rate": 5.175651049082262e-06,
+      "loss": 0.2787,
+      "step": 8017
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.519586990692234,
+      "learning_rate": 5.174633197592818e-06,
+      "loss": 0.283,
+      "step": 8018
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.131038688080356,
+      "learning_rate": 5.1736153388575305e-06,
+      "loss": 0.2877,
+      "step": 8019
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3906065783995365,
+      "learning_rate": 5.172597472918635e-06,
+      "loss": 0.2841,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3181186580179853,
+      "learning_rate": 5.1715795998183625e-06,
+      "loss": 0.2885,
+      "step": 8021
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0147444254565037,
+      "learning_rate": 5.1705617195989495e-06,
+      "loss": 0.2849,
+      "step": 8022
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.284732829682858,
+      "learning_rate": 5.169543832302627e-06,
+      "loss": 0.2959,
+      "step": 8023
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7715806483230008,
+      "learning_rate": 5.168525937971629e-06,
+      "loss": 0.2925,
+      "step": 8024
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5507911583763665,
+      "learning_rate": 5.167508036648191e-06,
+      "loss": 0.3008,
+      "step": 8025
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8122030484431046,
+      "learning_rate": 5.166490128374548e-06,
+      "loss": 0.2793,
+      "step": 8026
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2404819603162123,
+      "learning_rate": 5.165472213192934e-06,
+      "loss": 0.3042,
+      "step": 8027
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9126359663228494,
+      "learning_rate": 5.164454291145582e-06,
+      "loss": 0.2706,
+      "step": 8028
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6862084248984586,
+      "learning_rate": 5.163436362274731e-06,
+      "loss": 0.2823,
+      "step": 8029
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8249442111586103,
+      "learning_rate": 5.162418426622615e-06,
+      "loss": 0.2798,
+      "step": 8030
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.659652593548811,
+      "learning_rate": 5.1614004842314694e-06,
+      "loss": 0.2699,
+      "step": 8031
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9987159271790569,
+      "learning_rate": 5.160382535143531e-06,
+      "loss": 0.2833,
+      "step": 8032
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9910294536543716,
+      "learning_rate": 5.159364579401036e-06,
+      "loss": 0.2881,
+      "step": 8033
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.151206820163265,
+      "learning_rate": 5.158346617046221e-06,
+      "loss": 0.2704,
+      "step": 8034
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.932043974434462,
+      "learning_rate": 5.157328648121325e-06,
+      "loss": 0.2865,
+      "step": 8035
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8742209859969425,
+      "learning_rate": 5.156310672668584e-06,
+      "loss": 0.2841,
+      "step": 8036
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6486940944055828,
+      "learning_rate": 5.155292690730235e-06,
+      "loss": 0.2925,
+      "step": 8037
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6656818332786414,
+      "learning_rate": 5.154274702348517e-06,
+      "loss": 0.2841,
+      "step": 8038
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.736002438024291,
+      "learning_rate": 5.153256707565666e-06,
+      "loss": 0.2774,
+      "step": 8039
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7506932901812695,
+      "learning_rate": 5.152238706423925e-06,
+      "loss": 0.2733,
+      "step": 8040
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6446735246618587,
+      "learning_rate": 5.151220698965526e-06,
+      "loss": 0.2824,
+      "step": 8041
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3495018044799867,
+      "learning_rate": 5.1502026852327136e-06,
+      "loss": 0.293,
+      "step": 8042
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9900894311739992,
+      "learning_rate": 5.149184665267725e-06,
+      "loss": 0.2782,
+      "step": 8043
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4473681503604716,
+      "learning_rate": 5.148166639112799e-06,
+      "loss": 0.2822,
+      "step": 8044
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.715485009361348,
+      "learning_rate": 5.1471486068101774e-06,
+      "loss": 0.3001,
+      "step": 8045
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3454430352257016,
+      "learning_rate": 5.146130568402097e-06,
+      "loss": 0.2769,
+      "step": 8046
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3760925397182366,
+      "learning_rate": 5.1451125239308e-06,
+      "loss": 0.2995,
+      "step": 8047
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4208275815335933,
+      "learning_rate": 5.144094473438528e-06,
+      "loss": 0.2735,
+      "step": 8048
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4885921654964185,
+      "learning_rate": 5.14307641696752e-06,
+      "loss": 0.2719,
+      "step": 8049
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.845783800197305,
+      "learning_rate": 5.142058354560016e-06,
+      "loss": 0.2892,
+      "step": 8050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.552123748741543,
+      "learning_rate": 5.141040286258259e-06,
+      "loss": 0.2753,
+      "step": 8051
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1764173994410405,
+      "learning_rate": 5.140022212104492e-06,
+      "loss": 0.3153,
+      "step": 8052
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3296462212815934,
+      "learning_rate": 5.139004132140953e-06,
+      "loss": 0.2999,
+      "step": 8053
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8875624634374253,
+      "learning_rate": 5.137986046409884e-06,
+      "loss": 0.2735,
+      "step": 8054
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5061306577739693,
+      "learning_rate": 5.136967954953531e-06,
+      "loss": 0.309,
+      "step": 8055
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5540241913902926,
+      "learning_rate": 5.135949857814134e-06,
+      "loss": 0.2859,
+      "step": 8056
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2405392597573455,
+      "learning_rate": 5.134931755033936e-06,
+      "loss": 0.2963,
+      "step": 8057
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.134265261415188,
+      "learning_rate": 5.13391364665518e-06,
+      "loss": 0.3058,
+      "step": 8058
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6670677602740802,
+      "learning_rate": 5.132895532720108e-06,
+      "loss": 0.2937,
+      "step": 8059
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2613904123723603,
+      "learning_rate": 5.131877413270965e-06,
+      "loss": 0.2802,
+      "step": 8060
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5338154753482607,
+      "learning_rate": 5.130859288349993e-06,
+      "loss": 0.2936,
+      "step": 8061
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9411781150630898,
+      "learning_rate": 5.129841157999438e-06,
+      "loss": 0.2738,
+      "step": 8062
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2707669150827714,
+      "learning_rate": 5.128823022261542e-06,
+      "loss": 0.2865,
+      "step": 8063
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.183724499773126,
+      "learning_rate": 5.127804881178551e-06,
+      "loss": 0.3193,
+      "step": 8064
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8548627104452602,
+      "learning_rate": 5.126786734792706e-06,
+      "loss": 0.2981,
+      "step": 8065
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4332607932074577,
+      "learning_rate": 5.1257685831462565e-06,
+      "loss": 0.3022,
+      "step": 8066
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5992390097914027,
+      "learning_rate": 5.124750426281444e-06,
+      "loss": 0.3095,
+      "step": 8067
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.319122411374217,
+      "learning_rate": 5.123732264240517e-06,
+      "loss": 0.3037,
+      "step": 8068
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5635183548913294,
+      "learning_rate": 5.122714097065714e-06,
+      "loss": 0.2961,
+      "step": 8069
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.354103677632639,
+      "learning_rate": 5.1216959247992896e-06,
+      "loss": 0.3098,
+      "step": 8070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6294266845144957,
+      "learning_rate": 5.120677747483482e-06,
+      "loss": 0.4819,
+      "step": 8071
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5833706118593205,
+      "learning_rate": 5.119659565160542e-06,
+      "loss": 0.3056,
+      "step": 8072
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2022867462856706,
+      "learning_rate": 5.1186413778727125e-06,
+      "loss": 0.2961,
+      "step": 8073
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2339430255592672,
+      "learning_rate": 5.117623185662245e-06,
+      "loss": 0.2904,
+      "step": 8074
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4727351428623927,
+      "learning_rate": 5.11660498857138e-06,
+      "loss": 0.2811,
+      "step": 8075
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4549063477747954,
+      "learning_rate": 5.115586786642367e-06,
+      "loss": 0.3009,
+      "step": 8076
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.154997404078864,
+      "learning_rate": 5.1145685799174525e-06,
+      "loss": 0.2995,
+      "step": 8077
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.298505816209827,
+      "learning_rate": 5.113550368438885e-06,
+      "loss": 0.2827,
+      "step": 8078
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4602174794329403,
+      "learning_rate": 5.112532152248911e-06,
+      "loss": 0.2928,
+      "step": 8079
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9825143170171933,
+      "learning_rate": 5.111513931389777e-06,
+      "loss": 0.299,
+      "step": 8080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.405679316456453,
+      "learning_rate": 5.110495705903734e-06,
+      "loss": 0.3207,
+      "step": 8081
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0238107680183663,
+      "learning_rate": 5.109477475833027e-06,
+      "loss": 0.2906,
+      "step": 8082
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.418260652052426,
+      "learning_rate": 5.108459241219905e-06,
+      "loss": 0.2823,
+      "step": 8083
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5686712902216726,
+      "learning_rate": 5.107441002106616e-06,
+      "loss": 0.5054,
+      "step": 8084
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1012274202739682,
+      "learning_rate": 5.106422758535408e-06,
+      "loss": 0.2993,
+      "step": 8085
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3269839688973062,
+      "learning_rate": 5.105404510548534e-06,
+      "loss": 0.2973,
+      "step": 8086
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3121186570343597,
+      "learning_rate": 5.1043862581882375e-06,
+      "loss": 0.2844,
+      "step": 8087
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5471815690013278,
+      "learning_rate": 5.103368001496769e-06,
+      "loss": 0.2817,
+      "step": 8088
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5163684845845093,
+      "learning_rate": 5.102349740516379e-06,
+      "loss": 0.2978,
+      "step": 8089
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8501363901100054,
+      "learning_rate": 5.101331475289318e-06,
+      "loss": 0.2905,
+      "step": 8090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9429442626330335,
+      "learning_rate": 5.100313205857832e-06,
+      "loss": 0.3,
+      "step": 8091
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.242598751930439,
+      "learning_rate": 5.099294932264174e-06,
+      "loss": 0.2949,
+      "step": 8092
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.958946009346706,
+      "learning_rate": 5.098276654550593e-06,
+      "loss": 0.2762,
+      "step": 8093
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0006594531298725,
+      "learning_rate": 5.097258372759339e-06,
+      "loss": 0.2834,
+      "step": 8094
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4309018910323494,
+      "learning_rate": 5.096240086932661e-06,
+      "loss": 0.2853,
+      "step": 8095
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5859847155762092,
+      "learning_rate": 5.095221797112814e-06,
+      "loss": 0.2821,
+      "step": 8096
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3161250674204685,
+      "learning_rate": 5.094203503342042e-06,
+      "loss": 0.2821,
+      "step": 8097
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6229600351794886,
+      "learning_rate": 5.093185205662602e-06,
+      "loss": 0.5152,
+      "step": 8098
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.575849273510752,
+      "learning_rate": 5.09216690411674e-06,
+      "loss": 0.2875,
+      "step": 8099
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3971618086227922,
+      "learning_rate": 5.091148598746711e-06,
+      "loss": 0.2902,
+      "step": 8100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0548110743047925,
+      "learning_rate": 5.090130289594764e-06,
+      "loss": 0.2877,
+      "step": 8101
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.410056373113431,
+      "learning_rate": 5.089111976703151e-06,
+      "loss": 0.2888,
+      "step": 8102
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2334959627674285,
+      "learning_rate": 5.088093660114125e-06,
+      "loss": 0.2724,
+      "step": 8103
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1024339018243983,
+      "learning_rate": 5.087075339869937e-06,
+      "loss": 0.2774,
+      "step": 8104
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7350354360047069,
+      "learning_rate": 5.086057016012836e-06,
+      "loss": 0.2965,
+      "step": 8105
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.461945190557998,
+      "learning_rate": 5.085038688585079e-06,
+      "loss": 0.2894,
+      "step": 8106
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1600860846477663,
+      "learning_rate": 5.0840203576289135e-06,
+      "loss": 0.2837,
+      "step": 8107
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1997379931564873,
+      "learning_rate": 5.083002023186596e-06,
+      "loss": 0.2927,
+      "step": 8108
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3115319555158724,
+      "learning_rate": 5.081983685300377e-06,
+      "loss": 0.2974,
+      "step": 8109
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0692406318852896,
+      "learning_rate": 5.080965344012509e-06,
+      "loss": 0.2764,
+      "step": 8110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.957828967322293,
+      "learning_rate": 5.079946999365244e-06,
+      "loss": 0.2705,
+      "step": 8111
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.403710494563031,
+      "learning_rate": 5.078928651400838e-06,
+      "loss": 0.2998,
+      "step": 8112
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3844751932063026,
+      "learning_rate": 5.07791030016154e-06,
+      "loss": 0.2831,
+      "step": 8113
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8789455986311219,
+      "learning_rate": 5.076891945689606e-06,
+      "loss": 0.2876,
+      "step": 8114
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2053585238370856,
+      "learning_rate": 5.075873588027288e-06,
+      "loss": 0.3017,
+      "step": 8115
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.981578176439994,
+      "learning_rate": 5.074855227216842e-06,
+      "loss": 0.2856,
+      "step": 8116
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2393565715017612,
+      "learning_rate": 5.073836863300517e-06,
+      "loss": 0.2858,
+      "step": 8117
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.124483252278634,
+      "learning_rate": 5.072818496320572e-06,
+      "loss": 0.3109,
+      "step": 8118
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7475757024501903,
+      "learning_rate": 5.071800126319256e-06,
+      "loss": 0.3005,
+      "step": 8119
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9141583768087556,
+      "learning_rate": 5.070781753338828e-06,
+      "loss": 0.2769,
+      "step": 8120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.451323318961555,
+      "learning_rate": 5.069763377421536e-06,
+      "loss": 0.3067,
+      "step": 8121
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5071921951160125,
+      "learning_rate": 5.06874499860964e-06,
+      "loss": 0.3003,
+      "step": 8122
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0897711700626984,
+      "learning_rate": 5.067726616945391e-06,
+      "loss": 0.2949,
+      "step": 8123
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4642319003419675,
+      "learning_rate": 5.066708232471045e-06,
+      "loss": 0.2986,
+      "step": 8124
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7422941442622264,
+      "learning_rate": 5.065689845228854e-06,
+      "loss": 0.2899,
+      "step": 8125
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8900938353400445,
+      "learning_rate": 5.064671455261078e-06,
+      "loss": 0.3086,
+      "step": 8126
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1285408519171547,
+      "learning_rate": 5.063653062609966e-06,
+      "loss": 0.298,
+      "step": 8127
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8505723361191253,
+      "learning_rate": 5.062634667317776e-06,
+      "loss": 0.3228,
+      "step": 8128
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.772859833035717,
+      "learning_rate": 5.061616269426761e-06,
+      "loss": 0.2894,
+      "step": 8129
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9855041274032852,
+      "learning_rate": 5.06059786897918e-06,
+      "loss": 0.2919,
+      "step": 8130
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6736157068345445,
+      "learning_rate": 5.059579466017285e-06,
+      "loss": 0.2881,
+      "step": 8131
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8385217449012525,
+      "learning_rate": 5.058561060583331e-06,
+      "loss": 0.2595,
+      "step": 8132
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.528387329673985,
+      "learning_rate": 5.057542652719574e-06,
+      "loss": 0.2834,
+      "step": 8133
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7429707142645707,
+      "learning_rate": 5.056524242468274e-06,
+      "loss": 0.2893,
+      "step": 8134
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7041324964610656,
+      "learning_rate": 5.055505829871678e-06,
+      "loss": 0.2857,
+      "step": 8135
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4335164106800034,
+      "learning_rate": 5.054487414972049e-06,
+      "loss": 0.2865,
+      "step": 8136
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9924978004561862,
+      "learning_rate": 5.053468997811641e-06,
+      "loss": 0.2778,
+      "step": 8137
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 40.865144961517984,
+      "learning_rate": 5.052450578432708e-06,
+      "loss": 0.2702,
+      "step": 8138
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3016727652926572,
+      "learning_rate": 5.051432156877508e-06,
+      "loss": 0.2788,
+      "step": 8139
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9413116312589866,
+      "learning_rate": 5.050413733188296e-06,
+      "loss": 0.298,
+      "step": 8140
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7847582140692475,
+      "learning_rate": 5.049395307407329e-06,
+      "loss": 0.2858,
+      "step": 8141
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.529547832735798,
+      "learning_rate": 5.048376879576864e-06,
+      "loss": 0.2785,
+      "step": 8142
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6373970021235005,
+      "learning_rate": 5.047358449739154e-06,
+      "loss": 0.3086,
+      "step": 8143
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.66063556871953,
+      "learning_rate": 5.046340017936459e-06,
+      "loss": 0.2812,
+      "step": 8144
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.173410028038673,
+      "learning_rate": 5.045321584211035e-06,
+      "loss": 0.2834,
+      "step": 8145
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.189909801659308,
+      "learning_rate": 5.044303148605137e-06,
+      "loss": 0.3044,
+      "step": 8146
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.449596123719175,
+      "learning_rate": 5.043284711161022e-06,
+      "loss": 0.3042,
+      "step": 8147
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1180777194657785,
+      "learning_rate": 5.042266271920949e-06,
+      "loss": 0.2785,
+      "step": 8148
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8784394804409024,
+      "learning_rate": 5.041247830927173e-06,
+      "loss": 0.3005,
+      "step": 8149
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6236132036555047,
+      "learning_rate": 5.040229388221952e-06,
+      "loss": 0.2856,
+      "step": 8150
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4877568330618107,
+      "learning_rate": 5.03921094384754e-06,
+      "loss": 0.2725,
+      "step": 8151
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.779632116067725,
+      "learning_rate": 5.038192497846198e-06,
+      "loss": 0.2945,
+      "step": 8152
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8263603784506386,
+      "learning_rate": 5.037174050260181e-06,
+      "loss": 0.2847,
+      "step": 8153
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1698830201554764,
+      "learning_rate": 5.036155601131747e-06,
+      "loss": 0.2729,
+      "step": 8154
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4927890770584638,
+      "learning_rate": 5.035137150503151e-06,
+      "loss": 0.3065,
+      "step": 8155
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5545318227740046,
+      "learning_rate": 5.034118698416654e-06,
+      "loss": 0.2747,
+      "step": 8156
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4572955541828734,
+      "learning_rate": 5.033100244914512e-06,
+      "loss": 0.2763,
+      "step": 8157
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.746315447390973,
+      "learning_rate": 5.03208179003898e-06,
+      "loss": 0.3129,
+      "step": 8158
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4664765942120763,
+      "learning_rate": 5.03106333383232e-06,
+      "loss": 0.271,
+      "step": 8159
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.394632767386923,
+      "learning_rate": 5.030044876336786e-06,
+      "loss": 0.3045,
+      "step": 8160
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9123375926286352,
+      "learning_rate": 5.029026417594637e-06,
+      "loss": 0.3159,
+      "step": 8161
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5269617215825892,
+      "learning_rate": 5.02800795764813e-06,
+      "loss": 0.2955,
+      "step": 8162
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.546546453695523,
+      "learning_rate": 5.026989496539523e-06,
+      "loss": 0.3047,
+      "step": 8163
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9162045761280904,
+      "learning_rate": 5.025971034311075e-06,
+      "loss": 0.2804,
+      "step": 8164
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3073722605086497,
+      "learning_rate": 5.024952571005041e-06,
+      "loss": 0.3066,
+      "step": 8165
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9830524460973746,
+      "learning_rate": 5.02393410666368e-06,
+      "loss": 0.2835,
+      "step": 8166
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.961502319580719,
+      "learning_rate": 5.022915641329252e-06,
+      "loss": 0.2923,
+      "step": 8167
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.189041131058403,
+      "learning_rate": 5.021897175044014e-06,
+      "loss": 0.2874,
+      "step": 8168
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6629475421529214,
+      "learning_rate": 5.020878707850222e-06,
+      "loss": 0.2778,
+      "step": 8169
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6436906042157808,
+      "learning_rate": 5.0198602397901355e-06,
+      "loss": 0.2887,
+      "step": 8170
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.706239287634321,
+      "learning_rate": 5.018841770906011e-06,
+      "loss": 0.2839,
+      "step": 8171
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.700147076277438,
+      "learning_rate": 5.017823301240111e-06,
+      "loss": 0.306,
+      "step": 8172
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.644524884182939,
+      "learning_rate": 5.016804830834687e-06,
+      "loss": 0.2715,
+      "step": 8173
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.574413984967847,
+      "learning_rate": 5.015786359732003e-06,
+      "loss": 0.2824,
+      "step": 8174
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6463092278210696,
+      "learning_rate": 5.014767887974316e-06,
+      "loss": 0.2892,
+      "step": 8175
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8302860169498425,
+      "learning_rate": 5.013749415603881e-06,
+      "loss": 0.2733,
+      "step": 8176
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6922566209092376,
+      "learning_rate": 5.0127309426629575e-06,
+      "loss": 0.2665,
+      "step": 8177
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.767475220858458,
+      "learning_rate": 5.011712469193808e-06,
+      "loss": 0.285,
+      "step": 8178
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.821881597810828,
+      "learning_rate": 5.010693995238684e-06,
+      "loss": 0.2857,
+      "step": 8179
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4949420014722998,
+      "learning_rate": 5.009675520839851e-06,
+      "loss": 0.2976,
+      "step": 8180
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9029231329131013,
+      "learning_rate": 5.008657046039559e-06,
+      "loss": 0.27,
+      "step": 8181
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8667607852679868,
+      "learning_rate": 5.007638570880073e-06,
+      "loss": 0.2902,
+      "step": 8182
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.572467918809941,
+      "learning_rate": 5.0066200954036495e-06,
+      "loss": 0.4681,
+      "step": 8183
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2587774934395273,
+      "learning_rate": 5.005601619652546e-06,
+      "loss": 0.303,
+      "step": 8184
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9537089598222606,
+      "learning_rate": 5.00458314366902e-06,
+      "loss": 0.2837,
+      "step": 8185
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4045046250409439,
+      "learning_rate": 5.003564667495334e-06,
+      "loss": 0.2768,
+      "step": 8186
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.526752482442207,
+      "learning_rate": 5.002546191173742e-06,
+      "loss": 0.2804,
+      "step": 8187
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8022644753475676,
+      "learning_rate": 5.0015277147465035e-06,
+      "loss": 0.2986,
+      "step": 8188
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6767843667420423,
+      "learning_rate": 5.000509238255877e-06,
+      "loss": 0.2828,
+      "step": 8189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3425617857572867,
+      "learning_rate": 4.999490761744123e-06,
+      "loss": 0.2862,
+      "step": 8190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.9245674360968374,
+      "learning_rate": 4.9984722852534964e-06,
+      "loss": 0.2871,
+      "step": 8191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.766118190616532,
+      "learning_rate": 4.99745380882626e-06,
+      "loss": 0.2969,
+      "step": 8192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9747232663183154,
+      "learning_rate": 4.996435332504668e-06,
+      "loss": 0.2745,
+      "step": 8193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8748079083011437,
+      "learning_rate": 4.995416856330981e-06,
+      "loss": 0.2835,
+      "step": 8194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8798777101177233,
+      "learning_rate": 4.994398380347456e-06,
+      "loss": 0.2848,
+      "step": 8195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.639808692388689,
+      "learning_rate": 4.993379904596353e-06,
+      "loss": 0.2872,
+      "step": 8196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1435094676835655,
+      "learning_rate": 4.992361429119927e-06,
+      "loss": 0.2858,
+      "step": 8197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9851074493971386,
+      "learning_rate": 4.991342953960442e-06,
+      "loss": 0.2826,
+      "step": 8198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.499481015971225,
+      "learning_rate": 4.990324479160151e-06,
+      "loss": 0.3049,
+      "step": 8199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4346083663205933,
+      "learning_rate": 4.989306004761317e-06,
+      "loss": 0.2792,
+      "step": 8200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.699142284733824,
+      "learning_rate": 4.988287530806194e-06,
+      "loss": 0.3179,
+      "step": 8201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.651240072834334,
+      "learning_rate": 4.987269057337043e-06,
+      "loss": 0.2788,
+      "step": 8202
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6379106598045439,
+      "learning_rate": 4.98625058439612e-06,
+      "loss": 0.2961,
+      "step": 8203
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0033425123299136,
+      "learning_rate": 4.985232112025688e-06,
+      "loss": 0.3004,
+      "step": 8204
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4606393446522026,
+      "learning_rate": 4.984213640267996e-06,
+      "loss": 0.2692,
+      "step": 8205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1594238111119237,
+      "learning_rate": 4.983195169165313e-06,
+      "loss": 0.3018,
+      "step": 8206
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.073785619809118,
+      "learning_rate": 4.9821766987598905e-06,
+      "loss": 0.2803,
+      "step": 8207
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6358225243717622,
+      "learning_rate": 4.98115822909399e-06,
+      "loss": 0.487,
+      "step": 8208
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.245613572813672,
+      "learning_rate": 4.980139760209867e-06,
+      "loss": 0.2846,
+      "step": 8209
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.402414324187724,
+      "learning_rate": 4.979121292149781e-06,
+      "loss": 0.2835,
+      "step": 8210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7324346028894333,
+      "learning_rate": 4.978102824955988e-06,
+      "loss": 0.2897,
+      "step": 8211
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.938265565068531,
+      "learning_rate": 4.977084358670749e-06,
+      "loss": 0.3019,
+      "step": 8212
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.07939685069403,
+      "learning_rate": 4.97606589333632e-06,
+      "loss": 0.2738,
+      "step": 8213
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.361960190281248,
+      "learning_rate": 4.975047428994961e-06,
+      "loss": 0.2816,
+      "step": 8214
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.18567816403474,
+      "learning_rate": 4.9740289656889276e-06,
+      "loss": 0.2914,
+      "step": 8215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.224564304578116,
+      "learning_rate": 4.9730105034604795e-06,
+      "loss": 0.3092,
+      "step": 8216
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.58551899278767,
+      "learning_rate": 4.971992042351872e-06,
+      "loss": 0.2799,
+      "step": 8217
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.323951453494458,
+      "learning_rate": 4.970973582405366e-06,
+      "loss": 0.3097,
+      "step": 8218
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.253398014201736,
+      "learning_rate": 4.969955123663216e-06,
+      "loss": 0.2934,
+      "step": 8219
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9966836673695352,
+      "learning_rate": 4.968936666167681e-06,
+      "loss": 0.2935,
+      "step": 8220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.897449863210299,
+      "learning_rate": 4.96791820996102e-06,
+      "loss": 0.2923,
+      "step": 8221
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8632552958564557,
+      "learning_rate": 4.96689975508549e-06,
+      "loss": 0.2731,
+      "step": 8222
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7206196721758091,
+      "learning_rate": 4.9658813015833465e-06,
+      "loss": 0.2853,
+      "step": 8223
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3362008394752314,
+      "learning_rate": 4.964862849496851e-06,
+      "loss": 0.3139,
+      "step": 8224
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8345989807494922,
+      "learning_rate": 4.9638443988682555e-06,
+      "loss": 0.3108,
+      "step": 8225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7005985005836821,
+      "learning_rate": 4.962825949739822e-06,
+      "loss": 0.2738,
+      "step": 8226
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.040241574651639,
+      "learning_rate": 4.961807502153802e-06,
+      "loss": 0.2914,
+      "step": 8227
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.763435577672746,
+      "learning_rate": 4.960789056152461e-06,
+      "loss": 0.2603,
+      "step": 8228
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.959686260414063,
+      "learning_rate": 4.9597706117780495e-06,
+      "loss": 0.3001,
+      "step": 8229
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.130706889609274,
+      "learning_rate": 4.958752169072828e-06,
+      "loss": 0.2816,
+      "step": 8230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4536967288211615,
+      "learning_rate": 4.957733728079051e-06,
+      "loss": 0.2915,
+      "step": 8231
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5668313154641167,
+      "learning_rate": 4.956715288838979e-06,
+      "loss": 0.3039,
+      "step": 8232
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.61844755821173,
+      "learning_rate": 4.955696851394865e-06,
+      "loss": 0.2838,
+      "step": 8233
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.768573779849712,
+      "learning_rate": 4.954678415788968e-06,
+      "loss": 0.2896,
+      "step": 8234
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6050610772215912,
+      "learning_rate": 4.9536599820635414e-06,
+      "loss": 0.4678,
+      "step": 8235
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3711887432190775,
+      "learning_rate": 4.952641550260846e-06,
+      "loss": 0.3181,
+      "step": 8236
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2925269815694542,
+      "learning_rate": 4.951623120423138e-06,
+      "loss": 0.2657,
+      "step": 8237
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6303030514441423,
+      "learning_rate": 4.9506046925926725e-06,
+      "loss": 0.2648,
+      "step": 8238
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3464748567621223,
+      "learning_rate": 4.949586266811705e-06,
+      "loss": 0.2856,
+      "step": 8239
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6626060977836936,
+      "learning_rate": 4.948567843122494e-06,
+      "loss": 0.277,
+      "step": 8240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3347243208653583,
+      "learning_rate": 4.947549421567294e-06,
+      "loss": 0.2907,
+      "step": 8241
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8140112395828367,
+      "learning_rate": 4.94653100218836e-06,
+      "loss": 0.3006,
+      "step": 8242
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6213584045470085,
+      "learning_rate": 4.945512585027951e-06,
+      "loss": 0.4728,
+      "step": 8243
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5861131829846289,
+      "learning_rate": 4.944494170128323e-06,
+      "loss": 0.2835,
+      "step": 8244
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6417152034342402,
+      "learning_rate": 4.9434757575317286e-06,
+      "loss": 0.289,
+      "step": 8245
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6850436567976557,
+      "learning_rate": 4.9424573472804264e-06,
+      "loss": 0.2732,
+      "step": 8246
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.001742424332056,
+      "learning_rate": 4.9414389394166705e-06,
+      "loss": 0.2648,
+      "step": 8247
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.693485053493527,
+      "learning_rate": 4.940420533982718e-06,
+      "loss": 0.2872,
+      "step": 8248
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6446310687576586,
+      "learning_rate": 4.9394021310208225e-06,
+      "loss": 0.3017,
+      "step": 8249
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.886786817674654,
+      "learning_rate": 4.938383730573239e-06,
+      "loss": 0.2839,
+      "step": 8250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9207533099181677,
+      "learning_rate": 4.937365332682225e-06,
+      "loss": 0.2939,
+      "step": 8251
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0088184978962906,
+      "learning_rate": 4.936346937390036e-06,
+      "loss": 0.3113,
+      "step": 8252
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0325608995373363,
+      "learning_rate": 4.935328544738925e-06,
+      "loss": 0.2965,
+      "step": 8253
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5485844470082665,
+      "learning_rate": 4.934310154771147e-06,
+      "loss": 0.482,
+      "step": 8254
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1495215469116555,
+      "learning_rate": 4.933291767528957e-06,
+      "loss": 0.2817,
+      "step": 8255
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5681454838604005,
+      "learning_rate": 4.9322733830546115e-06,
+      "loss": 0.2775,
+      "step": 8256
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6840039722539915,
+      "learning_rate": 4.931255001390361e-06,
+      "loss": 0.2898,
+      "step": 8257
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5878128307571517,
+      "learning_rate": 4.930236622578464e-06,
+      "loss": 0.5024,
+      "step": 8258
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5355625212747435,
+      "learning_rate": 4.929218246661174e-06,
+      "loss": 0.3032,
+      "step": 8259
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6850041002117744,
+      "learning_rate": 4.928199873680745e-06,
+      "loss": 0.2735,
+      "step": 8260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4972453730764776,
+      "learning_rate": 4.927181503679429e-06,
+      "loss": 0.2757,
+      "step": 8261
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5194446073500507,
+      "learning_rate": 4.9261631366994845e-06,
+      "loss": 0.2926,
+      "step": 8262
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8590668669251484,
+      "learning_rate": 4.925144772783161e-06,
+      "loss": 0.2843,
+      "step": 8263
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.651182949328135,
+      "learning_rate": 4.924126411972714e-06,
+      "loss": 0.2989,
+      "step": 8264
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.482248813457576,
+      "learning_rate": 4.923108054310395e-06,
+      "loss": 0.268,
+      "step": 8265
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3343332975702737,
+      "learning_rate": 4.922089699838462e-06,
+      "loss": 0.2852,
+      "step": 8266
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.48254594985263,
+      "learning_rate": 4.921071348599164e-06,
+      "loss": 0.2729,
+      "step": 8267
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.603544230043783,
+      "learning_rate": 4.9200530006347575e-06,
+      "loss": 0.2889,
+      "step": 8268
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.537772216536732,
+      "learning_rate": 4.919034655987493e-06,
+      "loss": 0.306,
+      "step": 8269
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.232419358392029,
+      "learning_rate": 4.918016314699625e-06,
+      "loss": 0.2935,
+      "step": 8270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.022585956446659,
+      "learning_rate": 4.916997976813406e-06,
+      "loss": 0.2878,
+      "step": 8271
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.722205604554446,
+      "learning_rate": 4.9159796423710865e-06,
+      "loss": 0.2765,
+      "step": 8272
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.387381502409368,
+      "learning_rate": 4.914961311414922e-06,
+      "loss": 0.2707,
+      "step": 8273
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4932612601342625,
+      "learning_rate": 4.913942983987165e-06,
+      "loss": 0.2898,
+      "step": 8274
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4789325587072022,
+      "learning_rate": 4.912924660130065e-06,
+      "loss": 0.2836,
+      "step": 8275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.592684636643195,
+      "learning_rate": 4.911906339885877e-06,
+      "loss": 0.3092,
+      "step": 8276
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7816285396426617,
+      "learning_rate": 4.91088802329685e-06,
+      "loss": 0.2907,
+      "step": 8277
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.490854989742586,
+      "learning_rate": 4.909869710405238e-06,
+      "loss": 0.2886,
+      "step": 8278
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5912026344454357,
+      "learning_rate": 4.90885140125329e-06,
+      "loss": 0.2676,
+      "step": 8279
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.210525238436266,
+      "learning_rate": 4.907833095883261e-06,
+      "loss": 0.3003,
+      "step": 8280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1620126673627773,
+      "learning_rate": 4.9068147943374e-06,
+      "loss": 0.278,
+      "step": 8281
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1962018503609535,
+      "learning_rate": 4.905796496657959e-06,
+      "loss": 0.3072,
+      "step": 8282
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5142125020454695,
+      "learning_rate": 4.904778202887189e-06,
+      "loss": 0.3122,
+      "step": 8283
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3758775597468256,
+      "learning_rate": 4.90375991306734e-06,
+      "loss": 0.2883,
+      "step": 8284
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.068861653821894,
+      "learning_rate": 4.902741627240663e-06,
+      "loss": 0.2706,
+      "step": 8285
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5981640795469874,
+      "learning_rate": 4.90172334544941e-06,
+      "loss": 0.2863,
+      "step": 8286
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.46047673929236,
+      "learning_rate": 4.900705067735827e-06,
+      "loss": 0.2893,
+      "step": 8287
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.188271235936475,
+      "learning_rate": 4.899686794142169e-06,
+      "loss": 0.2984,
+      "step": 8288
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5869922792937985,
+      "learning_rate": 4.8986685247106835e-06,
+      "loss": 0.2747,
+      "step": 8289
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.901737525071232,
+      "learning_rate": 4.897650259483623e-06,
+      "loss": 0.284,
+      "step": 8290
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7613463201423094,
+      "learning_rate": 4.8966319985032325e-06,
+      "loss": 0.2656,
+      "step": 8291
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7949183307816083,
+      "learning_rate": 4.895613741811766e-06,
+      "loss": 0.2663,
+      "step": 8292
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8372796177519661,
+      "learning_rate": 4.894595489451468e-06,
+      "loss": 0.2813,
+      "step": 8293
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9594151641872095,
+      "learning_rate": 4.893577241464592e-06,
+      "loss": 0.2773,
+      "step": 8294
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5707005784419934,
+      "learning_rate": 4.892558997893386e-06,
+      "loss": 0.2879,
+      "step": 8295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.994405121263485,
+      "learning_rate": 4.8915407587800965e-06,
+      "loss": 0.3043,
+      "step": 8296
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.889226416275823,
+      "learning_rate": 4.890522524166974e-06,
+      "loss": 0.2962,
+      "step": 8297
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.236552999968069,
+      "learning_rate": 4.889504294096268e-06,
+      "loss": 0.2787,
+      "step": 8298
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.773305178868573,
+      "learning_rate": 4.8884860686102234e-06,
+      "loss": 0.2971,
+      "step": 8299
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.168430941044765,
+      "learning_rate": 4.887467847751091e-06,
+      "loss": 0.2717,
+      "step": 8300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2006097595084317,
+      "learning_rate": 4.8864496315611175e-06,
+      "loss": 0.2749,
+      "step": 8301
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9690783700100913,
+      "learning_rate": 4.8854314200825475e-06,
+      "loss": 0.2873,
+      "step": 8302
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.004249797363292,
+      "learning_rate": 4.884413213357635e-06,
+      "loss": 0.2803,
+      "step": 8303
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8958393299053462,
+      "learning_rate": 4.883395011428622e-06,
+      "loss": 0.2962,
+      "step": 8304
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.800553602780511,
+      "learning_rate": 4.882376814337757e-06,
+      "loss": 0.2843,
+      "step": 8305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4315211587650654,
+      "learning_rate": 4.881358622127288e-06,
+      "loss": 0.2698,
+      "step": 8306
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3705326963569648,
+      "learning_rate": 4.880340434839459e-06,
+      "loss": 0.2917,
+      "step": 8307
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2516337919657996,
+      "learning_rate": 4.8793222525165205e-06,
+      "loss": 0.302,
+      "step": 8308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4027960878908383,
+      "learning_rate": 4.878304075200712e-06,
+      "loss": 0.2815,
+      "step": 8309
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1470103947421344,
+      "learning_rate": 4.8772859029342864e-06,
+      "loss": 0.2772,
+      "step": 8310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6994330713125576,
+      "learning_rate": 4.876267735759486e-06,
+      "loss": 0.279,
+      "step": 8311
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2505510396157145,
+      "learning_rate": 4.875249573718557e-06,
+      "loss": 0.2926,
+      "step": 8312
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5588934345512109,
+      "learning_rate": 4.874231416853744e-06,
+      "loss": 0.3097,
+      "step": 8313
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8267615338004741,
+      "learning_rate": 4.8732132652072956e-06,
+      "loss": 0.2931,
+      "step": 8314
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.78098828533289,
+      "learning_rate": 4.872195118821452e-06,
+      "loss": 0.2902,
+      "step": 8315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.375384396299737,
+      "learning_rate": 4.871176977738461e-06,
+      "loss": 0.3106,
+      "step": 8316
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7625365282938825,
+      "learning_rate": 4.870158842000563e-06,
+      "loss": 0.2898,
+      "step": 8317
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.473743047081472,
+      "learning_rate": 4.869140711650008e-06,
+      "loss": 0.2962,
+      "step": 8318
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.630733487511131,
+      "learning_rate": 4.868122586729036e-06,
+      "loss": 0.2954,
+      "step": 8319
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0925891721373913,
+      "learning_rate": 4.867104467279894e-06,
+      "loss": 0.29,
+      "step": 8320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0939923448910562,
+      "learning_rate": 4.866086353344822e-06,
+      "loss": 0.2712,
+      "step": 8321
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.41877420552967,
+      "learning_rate": 4.865068244966066e-06,
+      "loss": 0.2929,
+      "step": 8322
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.270981754482077,
+      "learning_rate": 4.864050142185868e-06,
+      "loss": 0.296,
+      "step": 8323
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9316351673686134,
+      "learning_rate": 4.86303204504647e-06,
+      "loss": 0.2827,
+      "step": 8324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5418993099495082,
+      "learning_rate": 4.862013953590117e-06,
+      "loss": 0.2727,
+      "step": 8325
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5715061253466334,
+      "learning_rate": 4.860995867859049e-06,
+      "loss": 0.2824,
+      "step": 8326
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6413300874583487,
+      "learning_rate": 4.85997778789551e-06,
+      "loss": 0.3233,
+      "step": 8327
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7151059470009289,
+      "learning_rate": 4.858959713741742e-06,
+      "loss": 0.2717,
+      "step": 8328
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5936853821012376,
+      "learning_rate": 4.857941645439985e-06,
+      "loss": 0.2895,
+      "step": 8329
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4025304259215643,
+      "learning_rate": 4.856923583032483e-06,
+      "loss": 0.2677,
+      "step": 8330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5654484206878563,
+      "learning_rate": 4.855905526561474e-06,
+      "loss": 0.2879,
+      "step": 8331
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.695336616262636,
+      "learning_rate": 4.8548874760692e-06,
+      "loss": 0.3043,
+      "step": 8332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7832227001250294,
+      "learning_rate": 4.8538694315979045e-06,
+      "loss": 0.2927,
+      "step": 8333
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1140181232744344,
+      "learning_rate": 4.852851393189825e-06,
+      "loss": 0.2861,
+      "step": 8334
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4286677077307632,
+      "learning_rate": 4.8518333608872015e-06,
+      "loss": 0.2594,
+      "step": 8335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9602561494411748,
+      "learning_rate": 4.850815334732277e-06,
+      "loss": 0.2874,
+      "step": 8336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7235780246174375,
+      "learning_rate": 4.849797314767288e-06,
+      "loss": 0.2986,
+      "step": 8337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9333024538245205,
+      "learning_rate": 4.848779301034476e-06,
+      "loss": 0.2795,
+      "step": 8338
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.95784289340205,
+      "learning_rate": 4.847761293576078e-06,
+      "loss": 0.2863,
+      "step": 8339
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.346495865872955,
+      "learning_rate": 4.846743292434334e-06,
+      "loss": 0.2605,
+      "step": 8340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8440387401706564,
+      "learning_rate": 4.845725297651485e-06,
+      "loss": 0.2689,
+      "step": 8341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.971613200240072,
+      "learning_rate": 4.844707309269767e-06,
+      "loss": 0.2854,
+      "step": 8342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0397420833369697,
+      "learning_rate": 4.8436893273314176e-06,
+      "loss": 0.2885,
+      "step": 8343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.888534178756261,
+      "learning_rate": 4.842671351878677e-06,
+      "loss": 0.2862,
+      "step": 8344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8523021751627184,
+      "learning_rate": 4.84165338295378e-06,
+      "loss": 0.289,
+      "step": 8345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.75321363063784,
+      "learning_rate": 4.840635420598967e-06,
+      "loss": 0.2845,
+      "step": 8346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9647520906752194,
+      "learning_rate": 4.839617464856471e-06,
+      "loss": 0.2811,
+      "step": 8347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.413440971219273,
+      "learning_rate": 4.838599515768532e-06,
+      "loss": 0.2945,
+      "step": 8348
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0070939153612004,
+      "learning_rate": 4.837581573377387e-06,
+      "loss": 0.2803,
+      "step": 8349
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0024772625960554,
+      "learning_rate": 4.836563637725271e-06,
+      "loss": 0.2965,
+      "step": 8350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.783291008563796,
+      "learning_rate": 4.835545708854419e-06,
+      "loss": 0.2852,
+      "step": 8351
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.226463281537102,
+      "learning_rate": 4.834527786807069e-06,
+      "loss": 0.2825,
+      "step": 8352
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3102782421763481,
+      "learning_rate": 4.8335098716254545e-06,
+      "loss": 0.2644,
+      "step": 8353
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7464204611587495,
+      "learning_rate": 4.832491963351809e-06,
+      "loss": 0.2997,
+      "step": 8354
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.88987909463794,
+      "learning_rate": 4.831474062028372e-06,
+      "loss": 0.2811,
+      "step": 8355
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.314611497551799,
+      "learning_rate": 4.830456167697375e-06,
+      "loss": 0.2821,
+      "step": 8356
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.34470913155694,
+      "learning_rate": 4.829438280401052e-06,
+      "loss": 0.3104,
+      "step": 8357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.271400913559649,
+      "learning_rate": 4.828420400181639e-06,
+      "loss": 0.2796,
+      "step": 8358
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0068432578888458,
+      "learning_rate": 4.827402527081368e-06,
+      "loss": 0.2887,
+      "step": 8359
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.600485994703669,
+      "learning_rate": 4.826384661142472e-06,
+      "loss": 0.2832,
+      "step": 8360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3312780804242537,
+      "learning_rate": 4.825366802407184e-06,
+      "loss": 0.2862,
+      "step": 8361
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.674285070392237,
+      "learning_rate": 4.824348950917739e-06,
+      "loss": 0.279,
+      "step": 8362
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.066978445862203,
+      "learning_rate": 4.823331106716368e-06,
+      "loss": 0.2793,
+      "step": 8363
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.894374156222564,
+      "learning_rate": 4.822313269845303e-06,
+      "loss": 0.3039,
+      "step": 8364
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.510245593542162,
+      "learning_rate": 4.821295440346775e-06,
+      "loss": 0.2798,
+      "step": 8365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0874445027030997,
+      "learning_rate": 4.820277618263018e-06,
+      "loss": 0.2804,
+      "step": 8366
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8498454825062158,
+      "learning_rate": 4.819259803636261e-06,
+      "loss": 0.2761,
+      "step": 8367
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6144161860062747,
+      "learning_rate": 4.8182419965087375e-06,
+      "loss": 0.295,
+      "step": 8368
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6204533056451464,
+      "learning_rate": 4.817224196922673e-06,
+      "loss": 0.2924,
+      "step": 8369
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.140072947789576,
+      "learning_rate": 4.8162064049203035e-06,
+      "loss": 0.2773,
+      "step": 8370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3893163555588957,
+      "learning_rate": 4.815188620543857e-06,
+      "loss": 0.2981,
+      "step": 8371
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5803892602864232,
+      "learning_rate": 4.814170843835561e-06,
+      "loss": 0.5214,
+      "step": 8372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.850281135022143,
+      "learning_rate": 4.813153074837648e-06,
+      "loss": 0.2747,
+      "step": 8373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7567157375102114,
+      "learning_rate": 4.812135313592348e-06,
+      "loss": 0.28,
+      "step": 8374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8665012917513455,
+      "learning_rate": 4.8111175601418845e-06,
+      "loss": 0.2767,
+      "step": 8375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6674636307333104,
+      "learning_rate": 4.810099814528489e-06,
+      "loss": 0.2818,
+      "step": 8376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.574159705529876,
+      "learning_rate": 4.809082076794391e-06,
+      "loss": 0.2801,
+      "step": 8377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6994142917636991,
+      "learning_rate": 4.808064346981815e-06,
+      "loss": 0.5043,
+      "step": 8378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.043551212387841,
+      "learning_rate": 4.8070466251329926e-06,
+      "loss": 0.2809,
+      "step": 8379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3474862659250526,
+      "learning_rate": 4.806028911290147e-06,
+      "loss": 0.3091,
+      "step": 8380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5123096367847348,
+      "learning_rate": 4.805011205495505e-06,
+      "loss": 0.2909,
+      "step": 8381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5323443497815594,
+      "learning_rate": 4.803993507791298e-06,
+      "loss": 0.29,
+      "step": 8382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.514766547547126,
+      "learning_rate": 4.8029758182197455e-06,
+      "loss": 0.2773,
+      "step": 8383
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.778383922553858,
+      "learning_rate": 4.801958136823076e-06,
+      "loss": 0.2793,
+      "step": 8384
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5753223131355168,
+      "learning_rate": 4.800940463643517e-06,
+      "loss": 0.4815,
+      "step": 8385
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5322787209475415,
+      "learning_rate": 4.799922798723291e-06,
+      "loss": 0.2826,
+      "step": 8386
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.590813516641714,
+      "learning_rate": 4.798905142104624e-06,
+      "loss": 0.2833,
+      "step": 8387
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.406993004851999,
+      "learning_rate": 4.79788749382974e-06,
+      "loss": 0.2867,
+      "step": 8388
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7434548670177867,
+      "learning_rate": 4.796869853940862e-06,
+      "loss": 0.2695,
+      "step": 8389
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8224793567413622,
+      "learning_rate": 4.795852222480215e-06,
+      "loss": 0.2792,
+      "step": 8390
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9544250845793802,
+      "learning_rate": 4.7948345994900205e-06,
+      "loss": 0.2935,
+      "step": 8391
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.295874486943766,
+      "learning_rate": 4.793816985012503e-06,
+      "loss": 0.2921,
+      "step": 8392
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.128313276949677,
+      "learning_rate": 4.792799379089888e-06,
+      "loss": 0.302,
+      "step": 8393
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.058216889694372,
+      "learning_rate": 4.791781781764392e-06,
+      "loss": 0.2642,
+      "step": 8394
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1364291315479154,
+      "learning_rate": 4.79076419307824e-06,
+      "loss": 0.283,
+      "step": 8395
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8379552323039976,
+      "learning_rate": 4.789746613073655e-06,
+      "loss": 0.2699,
+      "step": 8396
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.510025828512651,
+      "learning_rate": 4.788729041792855e-06,
+      "loss": 0.2553,
+      "step": 8397
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.339536887154089,
+      "learning_rate": 4.787711479278063e-06,
+      "loss": 0.2747,
+      "step": 8398
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.253674938618319,
+      "learning_rate": 4.7866939255714975e-06,
+      "loss": 0.2681,
+      "step": 8399
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5930586442139374,
+      "learning_rate": 4.78567638071538e-06,
+      "loss": 0.2904,
+      "step": 8400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.60352921632056,
+      "learning_rate": 4.784658844751933e-06,
+      "loss": 0.2972,
+      "step": 8401
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.056599153816232,
+      "learning_rate": 4.7836413177233705e-06,
+      "loss": 0.3112,
+      "step": 8402
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.321717028384552,
+      "learning_rate": 4.782623799671914e-06,
+      "loss": 0.2693,
+      "step": 8403
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7516074821540624,
+      "learning_rate": 4.781606290639783e-06,
+      "loss": 0.2831,
+      "step": 8404
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5426926052473047,
+      "learning_rate": 4.780588790669195e-06,
+      "loss": 0.2751,
+      "step": 8405
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4934233565380441,
+      "learning_rate": 4.779571299802366e-06,
+      "loss": 0.2705,
+      "step": 8406
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5869935741782923,
+      "learning_rate": 4.778553818081517e-06,
+      "loss": 0.2717,
+      "step": 8407
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6472550756050794,
+      "learning_rate": 4.777536345548863e-06,
+      "loss": 0.2861,
+      "step": 8408
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0903636854619405,
+      "learning_rate": 4.7765188822466226e-06,
+      "loss": 0.2822,
+      "step": 8409
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5875475778358343,
+      "learning_rate": 4.775501428217009e-06,
+      "loss": 0.3064,
+      "step": 8410
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6261372605388646,
+      "learning_rate": 4.77448398350224e-06,
+      "loss": 0.5134,
+      "step": 8411
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6717839644134513,
+      "learning_rate": 4.773466548144532e-06,
+      "loss": 0.2782,
+      "step": 8412
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5573152097917939,
+      "learning_rate": 4.7724491221860986e-06,
+      "loss": 0.5053,
+      "step": 8413
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.065201075386083,
+      "learning_rate": 4.771431705669154e-06,
+      "loss": 0.3012,
+      "step": 8414
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.008813283520479,
+      "learning_rate": 4.770414298635916e-06,
+      "loss": 0.2862,
+      "step": 8415
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.200652662346773,
+      "learning_rate": 4.769396901128595e-06,
+      "loss": 0.2675,
+      "step": 8416
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.614085080775041,
+      "learning_rate": 4.768379513189408e-06,
+      "loss": 0.2782,
+      "step": 8417
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5337764137942766,
+      "learning_rate": 4.7673621348605655e-06,
+      "loss": 0.2881,
+      "step": 8418
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9528811152936985,
+      "learning_rate": 4.76634476618428e-06,
+      "loss": 0.2912,
+      "step": 8419
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8610145314625597,
+      "learning_rate": 4.765327407202768e-06,
+      "loss": 0.2822,
+      "step": 8420
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7819661999567937,
+      "learning_rate": 4.7643100579582355e-06,
+      "loss": 0.2901,
+      "step": 8421
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4974709682042993,
+      "learning_rate": 4.7632927184928985e-06,
+      "loss": 0.2886,
+      "step": 8422
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.346654853336016,
+      "learning_rate": 4.762275388848968e-06,
+      "loss": 0.2819,
+      "step": 8423
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5313475112359463,
+      "learning_rate": 4.761258069068654e-06,
+      "loss": 0.3097,
+      "step": 8424
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5794129127802287,
+      "learning_rate": 4.760240759194166e-06,
+      "loss": 0.4798,
+      "step": 8425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2478992436172196,
+      "learning_rate": 4.759223459267715e-06,
+      "loss": 0.296,
+      "step": 8426
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3140798998622945,
+      "learning_rate": 4.758206169331509e-06,
+      "loss": 0.2999,
+      "step": 8427
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.209286981225102,
+      "learning_rate": 4.757188889427761e-06,
+      "loss": 0.3071,
+      "step": 8428
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.151286698593463,
+      "learning_rate": 4.756171619598674e-06,
+      "loss": 0.3238,
+      "step": 8429
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.89099243932232,
+      "learning_rate": 4.7551543598864614e-06,
+      "loss": 0.3135,
+      "step": 8430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.817848693334988,
+      "learning_rate": 4.75413711033333e-06,
+      "loss": 0.2839,
+      "step": 8431
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5286534124216655,
+      "learning_rate": 4.753119870981486e-06,
+      "loss": 0.2844,
+      "step": 8432
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4574432505536095,
+      "learning_rate": 4.752102641873136e-06,
+      "loss": 0.2969,
+      "step": 8433
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6359299468810733,
+      "learning_rate": 4.75108542305049e-06,
+      "loss": 0.2988,
+      "step": 8434
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6438591581902067,
+      "learning_rate": 4.75006821455575e-06,
+      "loss": 0.3049,
+      "step": 8435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.105956843821535,
+      "learning_rate": 4.749051016431123e-06,
+      "loss": 0.2815,
+      "step": 8436
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.813638413684079,
+      "learning_rate": 4.748033828718818e-06,
+      "loss": 0.2864,
+      "step": 8437
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9256826268681375,
+      "learning_rate": 4.747016651461035e-06,
+      "loss": 0.2766,
+      "step": 8438
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9729885408359846,
+      "learning_rate": 4.7459994846999815e-06,
+      "loss": 0.2787,
+      "step": 8439
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5754963023637483,
+      "learning_rate": 4.744982328477859e-06,
+      "loss": 0.2613,
+      "step": 8440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5413783623211113,
+      "learning_rate": 4.7439651828368736e-06,
+      "loss": 0.2698,
+      "step": 8441
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.826139696164723,
+      "learning_rate": 4.7429480478192285e-06,
+      "loss": 0.2829,
+      "step": 8442
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.77022466616211,
+      "learning_rate": 4.741930923467123e-06,
+      "loss": 0.2852,
+      "step": 8443
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6323025777956142,
+      "learning_rate": 4.740913809822763e-06,
+      "loss": 0.2783,
+      "step": 8444
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6126840257945725,
+      "learning_rate": 4.7398967069283505e-06,
+      "loss": 0.2754,
+      "step": 8445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.675572458730645,
+      "learning_rate": 4.7388796148260846e-06,
+      "loss": 0.3032,
+      "step": 8446
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.918201435085034,
+      "learning_rate": 4.737862533558168e-06,
+      "loss": 0.2745,
+      "step": 8447
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7615596289358781,
+      "learning_rate": 4.7368454631668e-06,
+      "loss": 0.2914,
+      "step": 8448
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.924231687650136,
+      "learning_rate": 4.73582840369418e-06,
+      "loss": 0.2871,
+      "step": 8449
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.0118428570949543,
+      "learning_rate": 4.734811355182511e-06,
+      "loss": 0.2607,
+      "step": 8450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.965448863629376,
+      "learning_rate": 4.7337943176739874e-06,
+      "loss": 0.2744,
+      "step": 8451
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.291794914480757,
+      "learning_rate": 4.732777291210812e-06,
+      "loss": 0.2702,
+      "step": 8452
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8498930986453157,
+      "learning_rate": 4.7317602758351814e-06,
+      "loss": 0.2725,
+      "step": 8453
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5519068068456685,
+      "learning_rate": 4.730743271589293e-06,
+      "loss": 0.2793,
+      "step": 8454
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7848523060138315,
+      "learning_rate": 4.7297262785153455e-06,
+      "loss": 0.3003,
+      "step": 8455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7848219075019645,
+      "learning_rate": 4.728709296655533e-06,
+      "loss": 0.3335,
+      "step": 8456
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9582090084489585,
+      "learning_rate": 4.727692326052054e-06,
+      "loss": 0.281,
+      "step": 8457
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.322363457982876,
+      "learning_rate": 4.726675366747104e-06,
+      "loss": 0.2695,
+      "step": 8458
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.359884895054364,
+      "learning_rate": 4.72565841878288e-06,
+      "loss": 0.2899,
+      "step": 8459
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.60113332580965,
+      "learning_rate": 4.724641482201574e-06,
+      "loss": 0.2953,
+      "step": 8460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6231979197784872,
+      "learning_rate": 4.723624557045383e-06,
+      "loss": 0.268,
+      "step": 8461
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.517094361127118,
+      "learning_rate": 4.7226076433564994e-06,
+      "loss": 0.287,
+      "step": 8462
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7999813408429162,
+      "learning_rate": 4.721590741177119e-06,
+      "loss": 0.2705,
+      "step": 8463
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5857471206262909,
+      "learning_rate": 4.720573850549431e-06,
+      "loss": 0.4761,
+      "step": 8464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8948979812232025,
+      "learning_rate": 4.719556971515631e-06,
+      "loss": 0.2996,
+      "step": 8465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.329984744611168,
+      "learning_rate": 4.718540104117909e-06,
+      "loss": 0.2931,
+      "step": 8466
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8623617760742492,
+      "learning_rate": 4.717523248398461e-06,
+      "loss": 0.2758,
+      "step": 8467
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5228386275642694,
+      "learning_rate": 4.716506404399473e-06,
+      "loss": 0.279,
+      "step": 8468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.033646923485184,
+      "learning_rate": 4.715489572163139e-06,
+      "loss": 0.27,
+      "step": 8469
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.077267248613091,
+      "learning_rate": 4.714472751731647e-06,
+      "loss": 0.3069,
+      "step": 8470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5367434960848136,
+      "learning_rate": 4.713455943147188e-06,
+      "loss": 0.266,
+      "step": 8471
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.2880178154407433,
+      "learning_rate": 4.712439146451952e-06,
+      "loss": 0.2747,
+      "step": 8472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8439083862790828,
+      "learning_rate": 4.711422361688124e-06,
+      "loss": 0.2847,
+      "step": 8473
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.403644545133131,
+      "learning_rate": 4.710405588897896e-06,
+      "loss": 0.2765,
+      "step": 8474
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.467542722163241,
+      "learning_rate": 4.709388828123457e-06,
+      "loss": 0.2785,
+      "step": 8475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4478503120744513,
+      "learning_rate": 4.708372079406989e-06,
+      "loss": 0.28,
+      "step": 8476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.252758370568109,
+      "learning_rate": 4.707355342790682e-06,
+      "loss": 0.2976,
+      "step": 8477
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5136777425886425,
+      "learning_rate": 4.706338618316721e-06,
+      "loss": 0.292,
+      "step": 8478
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.106395076529786,
+      "learning_rate": 4.705321906027293e-06,
+      "loss": 0.2699,
+      "step": 8479
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.79012790847718,
+      "learning_rate": 4.7043052059645835e-06,
+      "loss": 0.2797,
+      "step": 8480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7093815457865003,
+      "learning_rate": 4.703288518170774e-06,
+      "loss": 0.2861,
+      "step": 8481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.464103243951695,
+      "learning_rate": 4.702271842688052e-06,
+      "loss": 0.3002,
+      "step": 8482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.665304705321683,
+      "learning_rate": 4.701255179558601e-06,
+      "loss": 0.2728,
+      "step": 8483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6848328109589452,
+      "learning_rate": 4.700238528824602e-06,
+      "loss": 0.273,
+      "step": 8484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8086438140132717,
+      "learning_rate": 4.69922189052824e-06,
+      "loss": 0.2679,
+      "step": 8485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.892734079629832,
+      "learning_rate": 4.698205264711695e-06,
+      "loss": 0.3,
+      "step": 8486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5813826797534165,
+      "learning_rate": 4.6971886514171495e-06,
+      "loss": 0.2855,
+      "step": 8487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.783067685410438,
+      "learning_rate": 4.696172050686784e-06,
+      "loss": 0.2845,
+      "step": 8488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.1740685978314063,
+      "learning_rate": 4.695155462562781e-06,
+      "loss": 0.2731,
+      "step": 8489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9210350324320693,
+      "learning_rate": 4.69413888708732e-06,
+      "loss": 0.291,
+      "step": 8490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6339544719116774,
+      "learning_rate": 4.693122324302579e-06,
+      "loss": 0.2918,
+      "step": 8491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.660077704029485,
+      "learning_rate": 4.692105774250738e-06,
+      "loss": 0.2829,
+      "step": 8492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.51817554738402,
+      "learning_rate": 4.6910892369739754e-06,
+      "loss": 0.2641,
+      "step": 8493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.570369080557774,
+      "learning_rate": 4.690072712514469e-06,
+      "loss": 0.2775,
+      "step": 8494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.190856266048138,
+      "learning_rate": 4.689056200914396e-06,
+      "loss": 0.2969,
+      "step": 8495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5318523971378313,
+      "learning_rate": 4.6880397022159325e-06,
+      "loss": 0.2933,
+      "step": 8496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.398498445276512,
+      "learning_rate": 4.687023216461257e-06,
+      "loss": 0.2805,
+      "step": 8497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4772995611057593,
+      "learning_rate": 4.686006743692544e-06,
+      "loss": 0.2819,
+      "step": 8498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7652774759334076,
+      "learning_rate": 4.68499028395197e-06,
+      "loss": 0.2809,
+      "step": 8499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.805621268942833,
+      "learning_rate": 4.683973837281708e-06,
+      "loss": 0.2728,
+      "step": 8500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9497376117748084,
+      "learning_rate": 4.682957403723933e-06,
+      "loss": 0.2872,
+      "step": 8501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7000536329925047,
+      "learning_rate": 4.681940983320818e-06,
+      "loss": 0.2906,
+      "step": 8502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.1535056506735675,
+      "learning_rate": 4.680924576114535e-06,
+      "loss": 0.282,
+      "step": 8503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.450309175314244,
+      "learning_rate": 4.679908182147259e-06,
+      "loss": 0.2887,
+      "step": 8504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8818322464896178,
+      "learning_rate": 4.6788918014611624e-06,
+      "loss": 0.2849,
+      "step": 8505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.470227527852286,
+      "learning_rate": 4.677875434098414e-06,
+      "loss": 0.318,
+      "step": 8506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.937588863534305,
+      "learning_rate": 4.676859080101188e-06,
+      "loss": 0.2813,
+      "step": 8507
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3713852251171992,
+      "learning_rate": 4.675842739511651e-06,
+      "loss": 0.2837,
+      "step": 8508
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4273832632733572,
+      "learning_rate": 4.674826412371975e-06,
+      "loss": 0.2943,
+      "step": 8509
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.864045484347699,
+      "learning_rate": 4.673810098724329e-06,
+      "loss": 0.2909,
+      "step": 8510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.155734353607268,
+      "learning_rate": 4.6727937986108794e-06,
+      "loss": 0.2982,
+      "step": 8511
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6988717509513624,
+      "learning_rate": 4.671777512073797e-06,
+      "loss": 0.2829,
+      "step": 8512
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9494067114925544,
+      "learning_rate": 4.67076123915525e-06,
+      "loss": 0.2889,
+      "step": 8513
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6166359579235476,
+      "learning_rate": 4.669744979897404e-06,
+      "loss": 0.2766,
+      "step": 8514
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.32629855727391,
+      "learning_rate": 4.668728734342425e-06,
+      "loss": 0.2683,
+      "step": 8515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9156123254111672,
+      "learning_rate": 4.667712502532479e-06,
+      "loss": 0.2899,
+      "step": 8516
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.110190444078173,
+      "learning_rate": 4.666696284509731e-06,
+      "loss": 0.279,
+      "step": 8517
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9956441228668982,
+      "learning_rate": 4.6656800803163454e-06,
+      "loss": 0.2739,
+      "step": 8518
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.25414985881702,
+      "learning_rate": 4.664663889994489e-06,
+      "loss": 0.2828,
+      "step": 8519
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.939016030513565,
+      "learning_rate": 4.6636477135863224e-06,
+      "loss": 0.2696,
+      "step": 8520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9488688939407426,
+      "learning_rate": 4.662631551134011e-06,
+      "loss": 0.303,
+      "step": 8521
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.78436442800251,
+      "learning_rate": 4.6616154026797144e-06,
+      "loss": 0.3002,
+      "step": 8522
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.224794063484492,
+      "learning_rate": 4.660599268265597e-06,
+      "loss": 0.278,
+      "step": 8523
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.548502201923166,
+      "learning_rate": 4.659583147933817e-06,
+      "loss": 0.2826,
+      "step": 8524
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 41.696948042233466,
+      "learning_rate": 4.6585670417265365e-06,
+      "loss": 0.2909,
+      "step": 8525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2166487367959773,
+      "learning_rate": 4.657550949685917e-06,
+      "loss": 0.2808,
+      "step": 8526
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3152271681091539,
+      "learning_rate": 4.656534871854118e-06,
+      "loss": 0.2922,
+      "step": 8527
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.357387346133596,
+      "learning_rate": 4.655518808273296e-06,
+      "loss": 0.2826,
+      "step": 8528
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.937390647985258,
+      "learning_rate": 4.654502758985611e-06,
+      "loss": 0.2791,
+      "step": 8529
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.403687022326223,
+      "learning_rate": 4.65348672403322e-06,
+      "loss": 0.3071,
+      "step": 8530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.448216136206693,
+      "learning_rate": 4.652470703458281e-06,
+      "loss": 0.2832,
+      "step": 8531
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3510141476195585,
+      "learning_rate": 4.651454697302949e-06,
+      "loss": 0.2922,
+      "step": 8532
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2209394625057297,
+      "learning_rate": 4.650438705609379e-06,
+      "loss": 0.2792,
+      "step": 8533
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6012992133253265,
+      "learning_rate": 4.6494227284197296e-06,
+      "loss": 0.4771,
+      "step": 8534
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5696914695097803,
+      "learning_rate": 4.648406765776154e-06,
+      "loss": 0.2923,
+      "step": 8535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9863923207164877,
+      "learning_rate": 4.647390817720806e-06,
+      "loss": 0.2801,
+      "step": 8536
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9554050504720806,
+      "learning_rate": 4.6463748842958416e-06,
+      "loss": 0.2992,
+      "step": 8537
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.619655729795758,
+      "learning_rate": 4.645358965543409e-06,
+      "loss": 0.2754,
+      "step": 8538
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2776661114982213,
+      "learning_rate": 4.644343061505663e-06,
+      "loss": 0.2914,
+      "step": 8539
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.154509107004463,
+      "learning_rate": 4.643327172224753e-06,
+      "loss": 0.271,
+      "step": 8540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8772663486449142,
+      "learning_rate": 4.6423112977428346e-06,
+      "loss": 0.3015,
+      "step": 8541
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.500139646785539,
+      "learning_rate": 4.641295438102056e-06,
+      "loss": 0.2648,
+      "step": 8542
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1114096276150285,
+      "learning_rate": 4.6402795933445665e-06,
+      "loss": 0.2748,
+      "step": 8543
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1839224960513643,
+      "learning_rate": 4.639263763512516e-06,
+      "loss": 0.2672,
+      "step": 8544
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8680031295894155,
+      "learning_rate": 4.638247948648052e-06,
+      "loss": 0.2775,
+      "step": 8545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.729256112303656,
+      "learning_rate": 4.637232148793323e-06,
+      "loss": 0.2943,
+      "step": 8546
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6163734714470552,
+      "learning_rate": 4.636216363990478e-06,
+      "loss": 0.2851,
+      "step": 8547
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.069947633542292,
+      "learning_rate": 4.635200594281658e-06,
+      "loss": 0.2773,
+      "step": 8548
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2162562662701955,
+      "learning_rate": 4.634184839709018e-06,
+      "loss": 0.277,
+      "step": 8549
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.268725482968977,
+      "learning_rate": 4.6331691003146966e-06,
+      "loss": 0.2577,
+      "step": 8550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2527727220316223,
+      "learning_rate": 4.632153376140841e-06,
+      "loss": 0.269,
+      "step": 8551
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.20018551401685,
+      "learning_rate": 4.631137667229596e-06,
+      "loss": 0.3038,
+      "step": 8552
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7370076469343152,
+      "learning_rate": 4.630121973623104e-06,
+      "loss": 0.3108,
+      "step": 8553
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4379113774010683,
+      "learning_rate": 4.629106295363508e-06,
+      "loss": 0.2833,
+      "step": 8554
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6625491943937982,
+      "learning_rate": 4.628090632492949e-06,
+      "loss": 0.28,
+      "step": 8555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0031481828937756,
+      "learning_rate": 4.627074985053572e-06,
+      "loss": 0.2916,
+      "step": 8556
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3079447897823093,
+      "learning_rate": 4.626059353087516e-06,
+      "loss": 0.2753,
+      "step": 8557
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3593400223302623,
+      "learning_rate": 4.625043736636922e-06,
+      "loss": 0.2905,
+      "step": 8558
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9397299322249433,
+      "learning_rate": 4.62402813574393e-06,
+      "loss": 0.2906,
+      "step": 8559
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4514322496734615,
+      "learning_rate": 4.623012550450677e-06,
+      "loss": 0.278,
+      "step": 8560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9483003155099303,
+      "learning_rate": 4.621996980799305e-06,
+      "loss": 0.2826,
+      "step": 8561
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9815518749981667,
+      "learning_rate": 4.620981426831948e-06,
+      "loss": 0.287,
+      "step": 8562
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2575559472488433,
+      "learning_rate": 4.6199658885907425e-06,
+      "loss": 0.293,
+      "step": 8563
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.102273372198595,
+      "learning_rate": 4.61895036611783e-06,
+      "loss": 0.2909,
+      "step": 8564
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1884664624238077,
+      "learning_rate": 4.617934859455344e-06,
+      "loss": 0.2628,
+      "step": 8565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9237841388999217,
+      "learning_rate": 4.6169193686454185e-06,
+      "loss": 0.2933,
+      "step": 8566
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.712097895034254,
+      "learning_rate": 4.61590389373019e-06,
+      "loss": 0.2899,
+      "step": 8567
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9398340986304743,
+      "learning_rate": 4.61488843475179e-06,
+      "loss": 0.2946,
+      "step": 8568
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.362201687229322,
+      "learning_rate": 4.613872991752354e-06,
+      "loss": 0.2809,
+      "step": 8569
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9868417860586314,
+      "learning_rate": 4.6128575647740095e-06,
+      "loss": 0.2718,
+      "step": 8570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8410643381187213,
+      "learning_rate": 4.611842153858896e-06,
+      "loss": 0.2704,
+      "step": 8571
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.913388566325249,
+      "learning_rate": 4.61082675904914e-06,
+      "loss": 0.283,
+      "step": 8572
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.162397071141212,
+      "learning_rate": 4.609811380386874e-06,
+      "loss": 0.2895,
+      "step": 8573
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.719229953542822,
+      "learning_rate": 4.608796017914226e-06,
+      "loss": 0.2624,
+      "step": 8574
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.228916571775909,
+      "learning_rate": 4.607780671673327e-06,
+      "loss": 0.2829,
+      "step": 8575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7571849859213218,
+      "learning_rate": 4.6067653417063035e-06,
+      "loss": 0.3053,
+      "step": 8576
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.437977724296435,
+      "learning_rate": 4.6057500280552854e-06,
+      "loss": 0.2789,
+      "step": 8577
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9557853163063963,
+      "learning_rate": 4.604734730762396e-06,
+      "loss": 0.2942,
+      "step": 8578
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.121689168314926,
+      "learning_rate": 4.6037194498697675e-06,
+      "loss": 0.2935,
+      "step": 8579
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.756789307169671,
+      "learning_rate": 4.602704185419523e-06,
+      "loss": 0.2775,
+      "step": 8580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.070867974225797,
+      "learning_rate": 4.601688937453787e-06,
+      "loss": 0.2829,
+      "step": 8581
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5464122002278814,
+      "learning_rate": 4.600673706014684e-06,
+      "loss": 0.4769,
+      "step": 8582
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4424936918505127,
+      "learning_rate": 4.59965849114434e-06,
+      "loss": 0.3037,
+      "step": 8583
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.458360579085679,
+      "learning_rate": 4.598643292884874e-06,
+      "loss": 0.2863,
+      "step": 8584
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6193533008497647,
+      "learning_rate": 4.59762811127841e-06,
+      "loss": 0.2804,
+      "step": 8585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4594848985223432,
+      "learning_rate": 4.596612946367071e-06,
+      "loss": 0.2713,
+      "step": 8586
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6832068829104752,
+      "learning_rate": 4.59559779819298e-06,
+      "loss": 0.2624,
+      "step": 8587
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8235682189158133,
+      "learning_rate": 4.594582666798253e-06,
+      "loss": 0.2949,
+      "step": 8588
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7972655430515903,
+      "learning_rate": 4.593567552225011e-06,
+      "loss": 0.2872,
+      "step": 8589
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1422914693986663,
+      "learning_rate": 4.592552454515373e-06,
+      "loss": 0.2717,
+      "step": 8590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.070577279731365,
+      "learning_rate": 4.5915373737114585e-06,
+      "loss": 0.2867,
+      "step": 8591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1193049675765288,
+      "learning_rate": 4.590522309855383e-06,
+      "loss": 0.2759,
+      "step": 8592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2639417689430745,
+      "learning_rate": 4.589507262989263e-06,
+      "loss": 0.2853,
+      "step": 8593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.916129817849262,
+      "learning_rate": 4.5884922331552165e-06,
+      "loss": 0.28,
+      "step": 8594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9396402468979705,
+      "learning_rate": 4.587477220395359e-06,
+      "loss": 0.2951,
+      "step": 8595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5468369879825565,
+      "learning_rate": 4.586462224751805e-06,
+      "loss": 0.2859,
+      "step": 8596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.114272870505431,
+      "learning_rate": 4.585447246266667e-06,
+      "loss": 0.2912,
+      "step": 8597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.880106452110881,
+      "learning_rate": 4.584432284982058e-06,
+      "loss": 0.2962,
+      "step": 8598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9331625466461235,
+      "learning_rate": 4.583417340940094e-06,
+      "loss": 0.2841,
+      "step": 8599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.71786524563575,
+      "learning_rate": 4.58240241418288e-06,
+      "loss": 0.2853,
+      "step": 8600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1148639066727943,
+      "learning_rate": 4.581387504752536e-06,
+      "loss": 0.2777,
+      "step": 8601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0081180780315173,
+      "learning_rate": 4.580372612691166e-06,
+      "loss": 0.2582,
+      "step": 8602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.70801635144531,
+      "learning_rate": 4.579357738040882e-06,
+      "loss": 0.2769,
+      "step": 8603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9732370667667531,
+      "learning_rate": 4.5783428808437934e-06,
+      "loss": 0.2713,
+      "step": 8604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.900251136523736,
+      "learning_rate": 4.577328041142008e-06,
+      "loss": 0.2843,
+      "step": 8605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.325128911545758,
+      "learning_rate": 4.576313218977632e-06,
+      "loss": 0.2766,
+      "step": 8606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.347966589831407,
+      "learning_rate": 4.575298414392771e-06,
+      "loss": 0.2701,
+      "step": 8607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.127965411152177,
+      "learning_rate": 4.574283627429535e-06,
+      "loss": 0.2671,
+      "step": 8608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9182318503140041,
+      "learning_rate": 4.573268858130028e-06,
+      "loss": 0.2719,
+      "step": 8609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.347908355016946,
+      "learning_rate": 4.5722541065363535e-06,
+      "loss": 0.2667,
+      "step": 8610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8857029220855384,
+      "learning_rate": 4.571239372690617e-06,
+      "loss": 0.2738,
+      "step": 8611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7631583278078047,
+      "learning_rate": 4.570224656634919e-06,
+      "loss": 0.2803,
+      "step": 8612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9975087766146427,
+      "learning_rate": 4.569209958411365e-06,
+      "loss": 0.277,
+      "step": 8613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6197380296160668,
+      "learning_rate": 4.568195278062053e-06,
+      "loss": 0.4757,
+      "step": 8614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8655645199727724,
+      "learning_rate": 4.567180615629086e-06,
+      "loss": 0.3062,
+      "step": 8615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.266578233175302,
+      "learning_rate": 4.566165971154564e-06,
+      "loss": 0.2906,
+      "step": 8616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.562224386288989,
+      "learning_rate": 4.565151344680588e-06,
+      "loss": 0.2779,
+      "step": 8617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4590352152499,
+      "learning_rate": 4.564136736249254e-06,
+      "loss": 0.2596,
+      "step": 8618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6523264225530647,
+      "learning_rate": 4.5631221459026624e-06,
+      "loss": 0.2881,
+      "step": 8619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.93516897016039,
+      "learning_rate": 4.562107573682908e-06,
+      "loss": 0.314,
+      "step": 8620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2328858142846673,
+      "learning_rate": 4.561093019632089e-06,
+      "loss": 0.2709,
+      "step": 8621
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.551435138906967,
+      "learning_rate": 4.560078483792297e-06,
+      "loss": 0.4583,
+      "step": 8622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.183199839452602,
+      "learning_rate": 4.5590639662056325e-06,
+      "loss": 0.2828,
+      "step": 8623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9929203831223479,
+      "learning_rate": 4.5580494669141865e-06,
+      "loss": 0.2738,
+      "step": 8624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0960387672284257,
+      "learning_rate": 4.557034985960054e-06,
+      "loss": 0.3031,
+      "step": 8625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5282488495606374,
+      "learning_rate": 4.556020523385326e-06,
+      "loss": 0.2897,
+      "step": 8626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.578240727402863,
+      "learning_rate": 4.5550060792320965e-06,
+      "loss": 0.2672,
+      "step": 8627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5666210547579098,
+      "learning_rate": 4.5539916535424535e-06,
+      "loss": 0.2606,
+      "step": 8628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.456100286133945,
+      "learning_rate": 4.552977246358491e-06,
+      "loss": 0.2853,
+      "step": 8629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6108797312051854,
+      "learning_rate": 4.551962857722293e-06,
+      "loss": 0.4985,
+      "step": 8630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6290358899769215,
+      "learning_rate": 4.5509484876759545e-06,
+      "loss": 0.2777,
+      "step": 8631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7857611995949587,
+      "learning_rate": 4.54993413626156e-06,
+      "loss": 0.2962,
+      "step": 8632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5804072172844503,
+      "learning_rate": 4.548919803521199e-06,
+      "loss": 0.2651,
+      "step": 8633
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.814613053214455,
+      "learning_rate": 4.547905489496955e-06,
+      "loss": 0.2728,
+      "step": 8634
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3240438288757073,
+      "learning_rate": 4.546891194230917e-06,
+      "loss": 0.265,
+      "step": 8635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.843873010889801,
+      "learning_rate": 4.545876917765166e-06,
+      "loss": 0.2854,
+      "step": 8636
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5127610501912625,
+      "learning_rate": 4.544862660141788e-06,
+      "loss": 0.3024,
+      "step": 8637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8378733164015084,
+      "learning_rate": 4.543848421402868e-06,
+      "loss": 0.2815,
+      "step": 8638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3257112494158623,
+      "learning_rate": 4.542834201590488e-06,
+      "loss": 0.3038,
+      "step": 8639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0907842299810633,
+      "learning_rate": 4.541820000746727e-06,
+      "loss": 0.2982,
+      "step": 8640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.189431225467172,
+      "learning_rate": 4.54080581891367e-06,
+      "loss": 0.2771,
+      "step": 8641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.114439143240404,
+      "learning_rate": 4.5397916561333945e-06,
+      "loss": 0.2768,
+      "step": 8642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3184204342694437,
+      "learning_rate": 4.538777512447981e-06,
+      "loss": 0.2953,
+      "step": 8643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8734650242845112,
+      "learning_rate": 4.537763387899506e-06,
+      "loss": 0.2931,
+      "step": 8644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7485818078402564,
+      "learning_rate": 4.5367492825300495e-06,
+      "loss": 0.2725,
+      "step": 8645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3865100544345346,
+      "learning_rate": 4.535735196381688e-06,
+      "loss": 0.2777,
+      "step": 8646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.404794384400355,
+      "learning_rate": 4.5347211294965e-06,
+      "loss": 0.2811,
+      "step": 8647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6905924900884721,
+      "learning_rate": 4.533707081916557e-06,
+      "loss": 0.3042,
+      "step": 8648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.273842475985826,
+      "learning_rate": 4.532693053683937e-06,
+      "loss": 0.2846,
+      "step": 8649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.868565034561557,
+      "learning_rate": 4.53167904484071e-06,
+      "loss": 0.2742,
+      "step": 8650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.248031544343279,
+      "learning_rate": 4.530665055428953e-06,
+      "loss": 0.2874,
+      "step": 8651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.511332066634212,
+      "learning_rate": 4.529651085490735e-06,
+      "loss": 0.2816,
+      "step": 8652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.139553668217882,
+      "learning_rate": 4.528637135068128e-06,
+      "loss": 0.2804,
+      "step": 8653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8447021782049204,
+      "learning_rate": 4.527623204203205e-06,
+      "loss": 0.2768,
+      "step": 8654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.548620663837192,
+      "learning_rate": 4.526609292938036e-06,
+      "loss": 0.32,
+      "step": 8655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9007032332417215,
+      "learning_rate": 4.525595401314686e-06,
+      "loss": 0.2812,
+      "step": 8656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7562137069187282,
+      "learning_rate": 4.524581529375227e-06,
+      "loss": 0.2723,
+      "step": 8657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.464324448956869,
+      "learning_rate": 4.523567677161724e-06,
+      "loss": 0.2802,
+      "step": 8658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9539615556465253,
+      "learning_rate": 4.522553844716245e-06,
+      "loss": 0.2672,
+      "step": 8659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6783096749471333,
+      "learning_rate": 4.521540032080852e-06,
+      "loss": 0.2769,
+      "step": 8660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.525224386916247,
+      "learning_rate": 4.520526239297615e-06,
+      "loss": 0.2747,
+      "step": 8661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4201583037172347,
+      "learning_rate": 4.519512466408596e-06,
+      "loss": 0.2856,
+      "step": 8662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2608409999456804,
+      "learning_rate": 4.5184987134558595e-06,
+      "loss": 0.2971,
+      "step": 8663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8139226028308588,
+      "learning_rate": 4.517484980481465e-06,
+      "loss": 0.2911,
+      "step": 8664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8104233666573968,
+      "learning_rate": 4.516471267527477e-06,
+      "loss": 0.2688,
+      "step": 8665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.159298667844816,
+      "learning_rate": 4.515457574635954e-06,
+      "loss": 0.2754,
+      "step": 8666
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8085537988998825,
+      "learning_rate": 4.514443901848955e-06,
+      "loss": 0.2794,
+      "step": 8667
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.823481615533778,
+      "learning_rate": 4.513430249208543e-06,
+      "loss": 0.2566,
+      "step": 8668
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.183569794940634,
+      "learning_rate": 4.512416616756775e-06,
+      "loss": 0.2781,
+      "step": 8669
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4572901475172415,
+      "learning_rate": 4.511403004535707e-06,
+      "loss": 0.2845,
+      "step": 8670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1257860807033913,
+      "learning_rate": 4.510389412587396e-06,
+      "loss": 0.2743,
+      "step": 8671
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.586562221477153,
+      "learning_rate": 4.509375840953898e-06,
+      "loss": 0.2644,
+      "step": 8672
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8402364882793383,
+      "learning_rate": 4.508362289677269e-06,
+      "loss": 0.283,
+      "step": 8673
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.456371909412483,
+      "learning_rate": 4.507348758799559e-06,
+      "loss": 0.2768,
+      "step": 8674
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8813355872225273,
+      "learning_rate": 4.506335248362824e-06,
+      "loss": 0.2696,
+      "step": 8675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8061554021897575,
+      "learning_rate": 4.505321758409119e-06,
+      "loss": 0.2823,
+      "step": 8676
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3721196467245311,
+      "learning_rate": 4.504308288980492e-06,
+      "loss": 0.2671,
+      "step": 8677
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2092868208521197,
+      "learning_rate": 4.503294840118994e-06,
+      "loss": 0.2712,
+      "step": 8678
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8938703630067009,
+      "learning_rate": 4.502281411866677e-06,
+      "loss": 0.2878,
+      "step": 8679
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.299576991819874,
+      "learning_rate": 4.501268004265586e-06,
+      "loss": 0.2706,
+      "step": 8680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.936327453854064,
+      "learning_rate": 4.500254617357773e-06,
+      "loss": 0.2829,
+      "step": 8681
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0996755835709644,
+      "learning_rate": 4.4992412511852816e-06,
+      "loss": 0.3129,
+      "step": 8682
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.186719977759636,
+      "learning_rate": 4.498227905790162e-06,
+      "loss": 0.2874,
+      "step": 8683
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1423388689125566,
+      "learning_rate": 4.497214581214456e-06,
+      "loss": 0.2692,
+      "step": 8684
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7107853309303507,
+      "learning_rate": 4.4962012775002125e-06,
+      "loss": 0.2653,
+      "step": 8685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6882321677791206,
+      "learning_rate": 4.495187994689471e-06,
+      "loss": 0.2743,
+      "step": 8686
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.5269752474096725,
+      "learning_rate": 4.494174732824279e-06,
+      "loss": 0.2619,
+      "step": 8687
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.587270876376558,
+      "learning_rate": 4.493161491946673e-06,
+      "loss": 0.2748,
+      "step": 8688
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4339995786329136,
+      "learning_rate": 4.492148272098696e-06,
+      "loss": 0.2663,
+      "step": 8689
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0538978301953326,
+      "learning_rate": 4.4911350733223915e-06,
+      "loss": 0.2745,
+      "step": 8690
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0367939560783195,
+      "learning_rate": 4.490121895659796e-06,
+      "loss": 0.2801,
+      "step": 8691
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.317636489632323,
+      "learning_rate": 4.489108739152948e-06,
+      "loss": 0.2782,
+      "step": 8692
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.0480680365327215,
+      "learning_rate": 4.488095603843888e-06,
+      "loss": 0.2802,
+      "step": 8693
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.730880327492681,
+      "learning_rate": 4.487082489774649e-06,
+      "loss": 0.3069,
+      "step": 8694
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.811687343272325,
+      "learning_rate": 4.486069396987269e-06,
+      "loss": 0.2839,
+      "step": 8695
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6116323535684252,
+      "learning_rate": 4.485056325523783e-06,
+      "loss": 0.2679,
+      "step": 8696
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.489179353127172,
+      "learning_rate": 4.484043275426223e-06,
+      "loss": 0.286,
+      "step": 8697
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.955007266093121,
+      "learning_rate": 4.483030246736625e-06,
+      "loss": 0.2889,
+      "step": 8698
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.05642462748354,
+      "learning_rate": 4.48201723949702e-06,
+      "loss": 0.2564,
+      "step": 8699
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.624203864039519,
+      "learning_rate": 4.4810042537494395e-06,
+      "loss": 0.2807,
+      "step": 8700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.792966610432871,
+      "learning_rate": 4.479991289535916e-06,
+      "loss": 0.2711,
+      "step": 8701
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1525156680549467,
+      "learning_rate": 4.4789783468984765e-06,
+      "loss": 0.2988,
+      "step": 8702
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 19.965924416228063,
+      "learning_rate": 4.477965425879152e-06,
+      "loss": 0.2902,
+      "step": 8703
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.090974749486585,
+      "learning_rate": 4.476952526519966e-06,
+      "loss": 0.263,
+      "step": 8704
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9306939845629139,
+      "learning_rate": 4.47593964886295e-06,
+      "loss": 0.2849,
+      "step": 8705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.230335887825449,
+      "learning_rate": 4.47492679295013e-06,
+      "loss": 0.2798,
+      "step": 8706
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8523561843967347,
+      "learning_rate": 4.47391395882353e-06,
+      "loss": 0.2971,
+      "step": 8707
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.903418956748041,
+      "learning_rate": 4.472901146525174e-06,
+      "loss": 0.2805,
+      "step": 8708
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2876029265816054,
+      "learning_rate": 4.471888356097086e-06,
+      "loss": 0.2949,
+      "step": 8709
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.753283396738876,
+      "learning_rate": 4.470875587581287e-06,
+      "loss": 0.2638,
+      "step": 8710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.397146123534283,
+      "learning_rate": 4.4698628410198015e-06,
+      "loss": 0.2911,
+      "step": 8711
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6834451016165375,
+      "learning_rate": 4.468850116454645e-06,
+      "loss": 0.2852,
+      "step": 8712
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5940769876826701,
+      "learning_rate": 4.467837413927843e-06,
+      "loss": 0.2776,
+      "step": 8713
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.068817197256334,
+      "learning_rate": 4.466824733481411e-06,
+      "loss": 0.284,
+      "step": 8714
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.733355617713016,
+      "learning_rate": 4.46581207515737e-06,
+      "loss": 0.2746,
+      "step": 8715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4566253155911313,
+      "learning_rate": 4.464799438997733e-06,
+      "loss": 0.3043,
+      "step": 8716
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6146027774224536,
+      "learning_rate": 4.463786825044519e-06,
+      "loss": 0.5033,
+      "step": 8717
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.60328592098261,
+      "learning_rate": 4.462774233339741e-06,
+      "loss": 0.2617,
+      "step": 8718
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.744840156791795,
+      "learning_rate": 4.461761663925413e-06,
+      "loss": 0.2869,
+      "step": 8719
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.196057371140877,
+      "learning_rate": 4.460749116843553e-06,
+      "loss": 0.2602,
+      "step": 8720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9060712691560484,
+      "learning_rate": 4.459736592136167e-06,
+      "loss": 0.2886,
+      "step": 8721
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8881644284223496,
+      "learning_rate": 4.45872408984527e-06,
+      "loss": 0.275,
+      "step": 8722
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7432649745383522,
+      "learning_rate": 4.457711610012873e-06,
+      "loss": 0.2992,
+      "step": 8723
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.916697148134185,
+      "learning_rate": 4.456699152680984e-06,
+      "loss": 0.2965,
+      "step": 8724
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5853925670611175,
+      "learning_rate": 4.455686717891613e-06,
+      "loss": 0.4762,
+      "step": 8725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7172402368895858,
+      "learning_rate": 4.454674305686766e-06,
+      "loss": 0.2704,
+      "step": 8726
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.559030240890836,
+      "learning_rate": 4.45366191610845e-06,
+      "loss": 0.2662,
+      "step": 8727
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1223966740884235,
+      "learning_rate": 4.452649549198672e-06,
+      "loss": 0.2764,
+      "step": 8728
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6184665383907286,
+      "learning_rate": 4.451637204999436e-06,
+      "loss": 0.277,
+      "step": 8729
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.869302526381741,
+      "learning_rate": 4.450624883552747e-06,
+      "loss": 0.3031,
+      "step": 8730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.422092786800748,
+      "learning_rate": 4.449612584900608e-06,
+      "loss": 0.2922,
+      "step": 8731
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5763176131167882,
+      "learning_rate": 4.4486003090850195e-06,
+      "loss": 0.2778,
+      "step": 8732
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9535099900192123,
+      "learning_rate": 4.447588056147985e-06,
+      "loss": 0.2831,
+      "step": 8733
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1141862189935123,
+      "learning_rate": 4.4465758261315e-06,
+      "loss": 0.2727,
+      "step": 8734
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1301601391388307,
+      "learning_rate": 4.445563619077569e-06,
+      "loss": 0.2897,
+      "step": 8735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.302629049170874,
+      "learning_rate": 4.444551435028189e-06,
+      "loss": 0.2682,
+      "step": 8736
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.269731032538597,
+      "learning_rate": 4.443539274025356e-06,
+      "loss": 0.2707,
+      "step": 8737
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3597455198382655,
+      "learning_rate": 4.442527136111067e-06,
+      "loss": 0.2982,
+      "step": 8738
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2546564141912344,
+      "learning_rate": 4.441515021327317e-06,
+      "loss": 0.2788,
+      "step": 8739
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8816917240269984,
+      "learning_rate": 4.4405029297161e-06,
+      "loss": 0.297,
+      "step": 8740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.307853514275158,
+      "learning_rate": 4.439490861319413e-06,
+      "loss": 0.2669,
+      "step": 8741
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4396761296655378,
+      "learning_rate": 4.4384788161792425e-06,
+      "loss": 0.2729,
+      "step": 8742
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.150460184328523,
+      "learning_rate": 4.437466794337584e-06,
+      "loss": 0.2988,
+      "step": 8743
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8233458776763678,
+      "learning_rate": 4.4364547958364275e-06,
+      "loss": 0.2809,
+      "step": 8744
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.339442174771447,
+      "learning_rate": 4.4354428207177645e-06,
+      "loss": 0.3095,
+      "step": 8745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.346649406151598,
+      "learning_rate": 4.434430869023579e-06,
+      "loss": 0.2827,
+      "step": 8746
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5648129658430627,
+      "learning_rate": 4.433418940795863e-06,
+      "loss": 0.2657,
+      "step": 8747
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.066333816262808,
+      "learning_rate": 4.432407036076601e-06,
+      "loss": 0.2871,
+      "step": 8748
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1392619710472855,
+      "learning_rate": 4.4313951549077775e-06,
+      "loss": 0.2633,
+      "step": 8749
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.482440548432782,
+      "learning_rate": 4.430383297331381e-06,
+      "loss": 0.2965,
+      "step": 8750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.566228056856757,
+      "learning_rate": 4.429371463389393e-06,
+      "loss": 0.2913,
+      "step": 8751
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.598308007124853,
+      "learning_rate": 4.428359653123796e-06,
+      "loss": 0.2866,
+      "step": 8752
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1237908125520835,
+      "learning_rate": 4.427347866576573e-06,
+      "loss": 0.2841,
+      "step": 8753
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.917361363393779,
+      "learning_rate": 4.4263361037897046e-06,
+      "loss": 0.2552,
+      "step": 8754
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9872444843215287,
+      "learning_rate": 4.4253243648051705e-06,
+      "loss": 0.2882,
+      "step": 8755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.092550239934306,
+      "learning_rate": 4.424312649664948e-06,
+      "loss": 0.2674,
+      "step": 8756
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.619481056358797,
+      "learning_rate": 4.423300958411014e-06,
+      "loss": 0.2811,
+      "step": 8757
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3155184754507414,
+      "learning_rate": 4.422289291085351e-06,
+      "loss": 0.2899,
+      "step": 8758
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6327497334036932,
+      "learning_rate": 4.42127764772993e-06,
+      "loss": 0.4672,
+      "step": 8759
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.089554601225153,
+      "learning_rate": 4.420266028386728e-06,
+      "loss": 0.2847,
+      "step": 8760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.346104769242135,
+      "learning_rate": 4.419254433097719e-06,
+      "loss": 0.2635,
+      "step": 8761
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.890118108497835,
+      "learning_rate": 4.418242861904873e-06,
+      "loss": 0.2842,
+      "step": 8762
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.0512350898967515,
+      "learning_rate": 4.417231314850166e-06,
+      "loss": 0.279,
+      "step": 8763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2090497839069054,
+      "learning_rate": 4.416219791975565e-06,
+      "loss": 0.2671,
+      "step": 8764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9640017675332144,
+      "learning_rate": 4.415208293323042e-06,
+      "loss": 0.3007,
+      "step": 8765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5219195963378986,
+      "learning_rate": 4.414196818934567e-06,
+      "loss": 0.2573,
+      "step": 8766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.841370004026359,
+      "learning_rate": 4.413185368852106e-06,
+      "loss": 0.2699,
+      "step": 8767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8130023330384395,
+      "learning_rate": 4.412173943117626e-06,
+      "loss": 0.2773,
+      "step": 8768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7112533878540717,
+      "learning_rate": 4.4111625417730935e-06,
+      "loss": 0.2785,
+      "step": 8769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.498945737985453,
+      "learning_rate": 4.410151164860472e-06,
+      "loss": 0.2918,
+      "step": 8770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.3917418633594,
+      "learning_rate": 4.409139812421726e-06,
+      "loss": 0.2571,
+      "step": 8771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9056012879771713,
+      "learning_rate": 4.408128484498821e-06,
+      "loss": 0.3047,
+      "step": 8772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.110824954788999,
+      "learning_rate": 4.407117181133715e-06,
+      "loss": 0.2934,
+      "step": 8773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.58552081856733,
+      "learning_rate": 4.406105902368371e-06,
+      "loss": 0.2676,
+      "step": 8774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6280054243402897,
+      "learning_rate": 4.405094648244747e-06,
+      "loss": 0.2647,
+      "step": 8775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9021394426162868,
+      "learning_rate": 4.404083418804803e-06,
+      "loss": 0.2641,
+      "step": 8776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3452257948415998,
+      "learning_rate": 4.403072214090498e-06,
+      "loss": 0.2935,
+      "step": 8777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.024300497890209,
+      "learning_rate": 4.402061034143785e-06,
+      "loss": 0.2634,
+      "step": 8778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0979430973519455,
+      "learning_rate": 4.401049879006621e-06,
+      "loss": 0.2896,
+      "step": 8779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0786174457969815,
+      "learning_rate": 4.400038748720964e-06,
+      "loss": 0.2682,
+      "step": 8780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.778758056238694,
+      "learning_rate": 4.399027643328764e-06,
+      "loss": 0.2667,
+      "step": 8781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.407984165891581,
+      "learning_rate": 4.3980165628719745e-06,
+      "loss": 0.3092,
+      "step": 8782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9337897479889632,
+      "learning_rate": 4.397005507392547e-06,
+      "loss": 0.2704,
+      "step": 8783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0966101736019462,
+      "learning_rate": 4.3959944769324314e-06,
+      "loss": 0.2716,
+      "step": 8784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5932500690387283,
+      "learning_rate": 4.394983471533579e-06,
+      "loss": 0.2712,
+      "step": 8785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7693085950654157,
+      "learning_rate": 4.393972491237935e-06,
+      "loss": 0.2749,
+      "step": 8786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6338923319208116,
+      "learning_rate": 4.39296153608745e-06,
+      "loss": 0.2894,
+      "step": 8787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.283903438910121,
+      "learning_rate": 4.39195060612407e-06,
+      "loss": 0.2868,
+      "step": 8788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.030213942018656,
+      "learning_rate": 4.390939701389738e-06,
+      "loss": 0.2699,
+      "step": 8789
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5366393441986386,
+      "learning_rate": 4.3899288219264e-06,
+      "loss": 0.2844,
+      "step": 8790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4356533143359775,
+      "learning_rate": 4.388917967776e-06,
+      "loss": 0.2797,
+      "step": 8791
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.782377717338023,
+      "learning_rate": 4.387907138980477e-06,
+      "loss": 0.287,
+      "step": 8792
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.123170864027156,
+      "learning_rate": 4.386896335581776e-06,
+      "loss": 0.2809,
+      "step": 8793
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6215816364133904,
+      "learning_rate": 4.385885557621833e-06,
+      "loss": 0.2708,
+      "step": 8794
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.036613786061605,
+      "learning_rate": 4.38487480514259e-06,
+      "loss": 0.2908,
+      "step": 8795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3803151073275663,
+      "learning_rate": 4.383864078185984e-06,
+      "loss": 0.2701,
+      "step": 8796
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.813758549658407,
+      "learning_rate": 4.382853376793951e-06,
+      "loss": 0.2751,
+      "step": 8797
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.309428422198222,
+      "learning_rate": 4.381842701008428e-06,
+      "loss": 0.2689,
+      "step": 8798
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4856878017942488,
+      "learning_rate": 4.38083205087135e-06,
+      "loss": 0.2734,
+      "step": 8799
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.587352160696565,
+      "learning_rate": 4.379821426424649e-06,
+      "loss": 0.2836,
+      "step": 8800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6788180503058727,
+      "learning_rate": 4.378810827710258e-06,
+      "loss": 0.2842,
+      "step": 8801
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7877170572081649,
+      "learning_rate": 4.3778002547701115e-06,
+      "loss": 0.2987,
+      "step": 8802
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8127992907502553,
+      "learning_rate": 4.376789707646136e-06,
+      "loss": 0.2597,
+      "step": 8803
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6606962822774591,
+      "learning_rate": 4.375779186380264e-06,
+      "loss": 0.2674,
+      "step": 8804
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6529512964717805,
+      "learning_rate": 4.37476869101442e-06,
+      "loss": 0.2755,
+      "step": 8805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1804116028790723,
+      "learning_rate": 4.373758221590534e-06,
+      "loss": 0.2697,
+      "step": 8806
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.659638487284497,
+      "learning_rate": 4.372747778150534e-06,
+      "loss": 0.4995,
+      "step": 8807
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5290014777933707,
+      "learning_rate": 4.37173736073634e-06,
+      "loss": 0.303,
+      "step": 8808
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.305633194643116,
+      "learning_rate": 4.3707269693898785e-06,
+      "loss": 0.2826,
+      "step": 8809
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0251981831796755,
+      "learning_rate": 4.3697166041530745e-06,
+      "loss": 0.3124,
+      "step": 8810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.292296163971953,
+      "learning_rate": 4.368706265067848e-06,
+      "loss": 0.296,
+      "step": 8811
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.16034555572653,
+      "learning_rate": 4.3676959521761204e-06,
+      "loss": 0.3045,
+      "step": 8812
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2779825250555747,
+      "learning_rate": 4.36668566551981e-06,
+      "loss": 0.2639,
+      "step": 8813
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.272932253024524,
+      "learning_rate": 4.365675405140835e-06,
+      "loss": 0.2909,
+      "step": 8814
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0181629381687625,
+      "learning_rate": 4.364665171081117e-06,
+      "loss": 0.319,
+      "step": 8815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6001049146500168,
+      "learning_rate": 4.363654963382566e-06,
+      "loss": 0.2653,
+      "step": 8816
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.862646015552912,
+      "learning_rate": 4.362644782087103e-06,
+      "loss": 0.2967,
+      "step": 8817
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7613298739018228,
+      "learning_rate": 4.361634627236641e-06,
+      "loss": 0.2885,
+      "step": 8818
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3554023763187777,
+      "learning_rate": 4.36062449887309e-06,
+      "loss": 0.2969,
+      "step": 8819
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3492156090735654,
+      "learning_rate": 4.3596143970383665e-06,
+      "loss": 0.2937,
+      "step": 8820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0430566626109665,
+      "learning_rate": 4.358604321774378e-06,
+      "loss": 0.2853,
+      "step": 8821
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.331127963768257,
+      "learning_rate": 4.357594273123034e-06,
+      "loss": 0.2727,
+      "step": 8822
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7288435665347155,
+      "learning_rate": 4.356584251126247e-06,
+      "loss": 0.2853,
+      "step": 8823
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4801098710765923,
+      "learning_rate": 4.35557425582592e-06,
+      "loss": 0.2593,
+      "step": 8824
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.534546151544299,
+      "learning_rate": 4.354564287263963e-06,
+      "loss": 0.289,
+      "step": 8825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.500854742667218,
+      "learning_rate": 4.353554345482281e-06,
+      "loss": 0.3003,
+      "step": 8826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0811475203848486,
+      "learning_rate": 4.352544430522776e-06,
+      "loss": 0.2637,
+      "step": 8827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.736678915037675,
+      "learning_rate": 4.3515345424273534e-06,
+      "loss": 0.2727,
+      "step": 8828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.779011418387543,
+      "learning_rate": 4.350524681237914e-06,
+      "loss": 0.2915,
+      "step": 8829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3772966415982157,
+      "learning_rate": 4.349514846996359e-06,
+      "loss": 0.3121,
+      "step": 8830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9069514685270286,
+      "learning_rate": 4.348505039744588e-06,
+      "loss": 0.264,
+      "step": 8831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8079596990685507,
+      "learning_rate": 4.347495259524502e-06,
+      "loss": 0.2663,
+      "step": 8832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0552600442669373,
+      "learning_rate": 4.346485506377995e-06,
+      "loss": 0.29,
+      "step": 8833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.443785518207296,
+      "learning_rate": 4.345475780346966e-06,
+      "loss": 0.2826,
+      "step": 8834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4957452462973535,
+      "learning_rate": 4.344466081473308e-06,
+      "loss": 0.2745,
+      "step": 8835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.79804967424767,
+      "learning_rate": 4.343456409798919e-06,
+      "loss": 0.2953,
+      "step": 8836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.847971924603,
+      "learning_rate": 4.342446765365688e-06,
+      "loss": 0.2739,
+      "step": 8837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6603864587335897,
+      "learning_rate": 4.341437148215509e-06,
+      "loss": 0.2813,
+      "step": 8838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9345727328991653,
+      "learning_rate": 4.3404275583902715e-06,
+      "loss": 0.2838,
+      "step": 8839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8273485211933544,
+      "learning_rate": 4.339417995931868e-06,
+      "loss": 0.2805,
+      "step": 8840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9221728253766586,
+      "learning_rate": 4.3384084608821836e-06,
+      "loss": 0.2709,
+      "step": 8841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3001131084931226,
+      "learning_rate": 4.337398953283108e-06,
+      "loss": 0.2908,
+      "step": 8842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.327751492623303,
+      "learning_rate": 4.3363894731765275e-06,
+      "loss": 0.2832,
+      "step": 8843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.286516408866927,
+      "learning_rate": 4.335380020604325e-06,
+      "loss": 0.2774,
+      "step": 8844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4334218433812853,
+      "learning_rate": 4.334370595608388e-06,
+      "loss": 0.2909,
+      "step": 8845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.558465768562449,
+      "learning_rate": 4.333361198230594e-06,
+      "loss": 0.2876,
+      "step": 8846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5918837871139515,
+      "learning_rate": 4.33235182851283e-06,
+      "loss": 0.2767,
+      "step": 8847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7756730051623848,
+      "learning_rate": 4.331342486496974e-06,
+      "loss": 0.2787,
+      "step": 8848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.72648946975082,
+      "learning_rate": 4.330333172224906e-06,
+      "loss": 0.2697,
+      "step": 8849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9080727141380536,
+      "learning_rate": 4.3293238857385055e-06,
+      "loss": 0.2665,
+      "step": 8850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5495870138713592,
+      "learning_rate": 4.328314627079647e-06,
+      "loss": 0.2699,
+      "step": 8851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.228494027164655,
+      "learning_rate": 4.327305396290208e-06,
+      "loss": 0.2666,
+      "step": 8852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.40738067116132,
+      "learning_rate": 4.326296193412061e-06,
+      "loss": 0.2687,
+      "step": 8853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4099924326658404,
+      "learning_rate": 4.325287018487085e-06,
+      "loss": 0.2666,
+      "step": 8854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.716089079601977,
+      "learning_rate": 4.324277871557146e-06,
+      "loss": 0.2682,
+      "step": 8855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.406387412827769,
+      "learning_rate": 4.323268752664121e-06,
+      "loss": 0.2712,
+      "step": 8856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.047449486050684,
+      "learning_rate": 4.322259661849876e-06,
+      "loss": 0.263,
+      "step": 8857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.474439262709888,
+      "learning_rate": 4.321250599156282e-06,
+      "loss": 0.2917,
+      "step": 8858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1979414166924056,
+      "learning_rate": 4.320241564625206e-06,
+      "loss": 0.2574,
+      "step": 8859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9528934305107515,
+      "learning_rate": 4.319232558298514e-06,
+      "loss": 0.2685,
+      "step": 8860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.164371315133676,
+      "learning_rate": 4.318223580218072e-06,
+      "loss": 0.298,
+      "step": 8861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1918782637115277,
+      "learning_rate": 4.3172146304257466e-06,
+      "loss": 0.2712,
+      "step": 8862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.402434207344038,
+      "learning_rate": 4.316205708963398e-06,
+      "loss": 0.2665,
+      "step": 8863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.443865349213514,
+      "learning_rate": 4.315196815872889e-06,
+      "loss": 0.2689,
+      "step": 8864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0935529911608937,
+      "learning_rate": 4.314187951196081e-06,
+      "loss": 0.2977,
+      "step": 8865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4900474685702436,
+      "learning_rate": 4.313179114974832e-06,
+      "loss": 0.2671,
+      "step": 8866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5291984547588007,
+      "learning_rate": 4.312170307251003e-06,
+      "loss": 0.2703,
+      "step": 8867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4134436391843623,
+      "learning_rate": 4.311161528066447e-06,
+      "loss": 0.2716,
+      "step": 8868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.422658943864349,
+      "learning_rate": 4.310152777463025e-06,
+      "loss": 0.2814,
+      "step": 8869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5609013701217902,
+      "learning_rate": 4.309144055482589e-06,
+      "loss": 0.273,
+      "step": 8870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8882123327606752,
+      "learning_rate": 4.308135362166993e-06,
+      "loss": 0.2802,
+      "step": 8871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5452605078974324,
+      "learning_rate": 4.307126697558091e-06,
+      "loss": 0.2589,
+      "step": 8872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0462801740054424,
+      "learning_rate": 4.306118061697732e-06,
+      "loss": 0.2643,
+      "step": 8873
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.617369301677123,
+      "learning_rate": 4.305109454627766e-06,
+      "loss": 0.2598,
+      "step": 8874
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2837312884593506,
+      "learning_rate": 4.304100876390045e-06,
+      "loss": 0.2671,
+      "step": 8875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.102618332814836,
+      "learning_rate": 4.303092327026412e-06,
+      "loss": 0.2781,
+      "step": 8876
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.642870964127251,
+      "learning_rate": 4.302083806578718e-06,
+      "loss": 0.2976,
+      "step": 8877
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2466194788050067,
+      "learning_rate": 4.301075315088808e-06,
+      "loss": 0.2836,
+      "step": 8878
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6094421454054106,
+      "learning_rate": 4.300066852598522e-06,
+      "loss": 0.2735,
+      "step": 8879
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7641210259148083,
+      "learning_rate": 4.299058419149707e-06,
+      "loss": 0.304,
+      "step": 8880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.497019961234568,
+      "learning_rate": 4.298050014784203e-06,
+      "loss": 0.2897,
+      "step": 8881
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2130726264740694,
+      "learning_rate": 4.29704163954385e-06,
+      "loss": 0.2581,
+      "step": 8882
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7827414494481095,
+      "learning_rate": 4.296033293470487e-06,
+      "loss": 0.2575,
+      "step": 8883
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5952871810474307,
+      "learning_rate": 4.295024976605955e-06,
+      "loss": 0.2678,
+      "step": 8884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.048186536648519,
+      "learning_rate": 4.2940166889920885e-06,
+      "loss": 0.305,
+      "step": 8885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.468594996183448,
+      "learning_rate": 4.293008430670724e-06,
+      "loss": 0.2753,
+      "step": 8886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.04924145427867,
+      "learning_rate": 4.2920002016836944e-06,
+      "loss": 0.2693,
+      "step": 8887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8709365030415537,
+      "learning_rate": 4.290992002072836e-06,
+      "loss": 0.2706,
+      "step": 8888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.874629366645945,
+      "learning_rate": 4.2899838318799765e-06,
+      "loss": 0.2766,
+      "step": 8889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.879841030053625,
+      "learning_rate": 4.288975691146949e-06,
+      "loss": 0.271,
+      "step": 8890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1328135006619013,
+      "learning_rate": 4.287967579915583e-06,
+      "loss": 0.2801,
+      "step": 8891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8558413617193286,
+      "learning_rate": 4.286959498227708e-06,
+      "loss": 0.2657,
+      "step": 8892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7925751896906728,
+      "learning_rate": 4.285951446125149e-06,
+      "loss": 0.26,
+      "step": 8893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.128898122164901,
+      "learning_rate": 4.284943423649734e-06,
+      "loss": 0.3001,
+      "step": 8894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1364360496462225,
+      "learning_rate": 4.283935430843285e-06,
+      "loss": 0.2593,
+      "step": 8895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7862988839976337,
+      "learning_rate": 4.2829274677476284e-06,
+      "loss": 0.261,
+      "step": 8896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.728330977776241,
+      "learning_rate": 4.281919534404584e-06,
+      "loss": 0.2665,
+      "step": 8897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9728779970840908,
+      "learning_rate": 4.280911630855972e-06,
+      "loss": 0.2592,
+      "step": 8898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7561037680395684,
+      "learning_rate": 4.279903757143615e-06,
+      "loss": 0.2888,
+      "step": 8899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9675289154909317,
+      "learning_rate": 4.278895913309331e-06,
+      "loss": 0.2886,
+      "step": 8900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.267841875167877,
+      "learning_rate": 4.277888099394935e-06,
+      "loss": 0.2786,
+      "step": 8901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.322292873054597,
+      "learning_rate": 4.276880315442246e-06,
+      "loss": 0.2762,
+      "step": 8902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0458923668148548,
+      "learning_rate": 4.275872561493076e-06,
+      "loss": 0.2754,
+      "step": 8903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.526948251766585,
+      "learning_rate": 4.274864837589241e-06,
+      "loss": 0.2759,
+      "step": 8904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3858840883045325,
+      "learning_rate": 4.27385714377255e-06,
+      "loss": 0.2814,
+      "step": 8905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.315365768078168,
+      "learning_rate": 4.272849480084815e-06,
+      "loss": 0.2815,
+      "step": 8906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.38894179386223,
+      "learning_rate": 4.271841846567848e-06,
+      "loss": 0.2845,
+      "step": 8907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2669536190395756,
+      "learning_rate": 4.270834243263457e-06,
+      "loss": 0.2651,
+      "step": 8908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.1972565635394075,
+      "learning_rate": 4.269826670213448e-06,
+      "loss": 0.2761,
+      "step": 8909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9157004083104388,
+      "learning_rate": 4.268819127459628e-06,
+      "loss": 0.2791,
+      "step": 8910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1022674094391616,
+      "learning_rate": 4.2678116150437996e-06,
+      "loss": 0.2913,
+      "step": 8911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9327625254882868,
+      "learning_rate": 4.266804133007768e-06,
+      "loss": 0.2667,
+      "step": 8912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9456884074858392,
+      "learning_rate": 4.265796681393334e-06,
+      "loss": 0.2734,
+      "step": 8913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.111866457250431,
+      "learning_rate": 4.264789260242302e-06,
+      "loss": 0.2823,
+      "step": 8914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7224210336776686,
+      "learning_rate": 4.2637818695964695e-06,
+      "loss": 0.2656,
+      "step": 8915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3786303189828897,
+      "learning_rate": 4.262774509497635e-06,
+      "loss": 0.2583,
+      "step": 8916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6003560851926935,
+      "learning_rate": 4.261767179987595e-06,
+      "loss": 0.4711,
+      "step": 8917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0008897705973374,
+      "learning_rate": 4.260759881108146e-06,
+      "loss": 0.2676,
+      "step": 8918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5084236603405603,
+      "learning_rate": 4.259752612901082e-06,
+      "loss": 0.2782,
+      "step": 8919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2484175022206485,
+      "learning_rate": 4.258745375408197e-06,
+      "loss": 0.2773,
+      "step": 8920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9786812173023716,
+      "learning_rate": 4.257738168671282e-06,
+      "loss": 0.2699,
+      "step": 8921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5888677555861132,
+      "learning_rate": 4.256730992732131e-06,
+      "loss": 0.4822,
+      "step": 8922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9832092959576713,
+      "learning_rate": 4.25572384763253e-06,
+      "loss": 0.267,
+      "step": 8923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6721588709757083,
+      "learning_rate": 4.25471673341427e-06,
+      "loss": 0.277,
+      "step": 8924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.28239510560659,
+      "learning_rate": 4.253709650119136e-06,
+      "loss": 0.286,
+      "step": 8925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.500595624838231,
+      "learning_rate": 4.252702597788914e-06,
+      "loss": 0.2811,
+      "step": 8926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.236622128288961,
+      "learning_rate": 4.251695576465387e-06,
+      "loss": 0.2617,
+      "step": 8927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2044057139659348,
+      "learning_rate": 4.25068858619034e-06,
+      "loss": 0.2667,
+      "step": 8928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.69057199713099,
+      "learning_rate": 4.249681627005555e-06,
+      "loss": 0.2886,
+      "step": 8929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8128634196629156,
+      "learning_rate": 4.248674698952813e-06,
+      "loss": 0.2816,
+      "step": 8930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.9092676289630655,
+      "learning_rate": 4.247667802073892e-06,
+      "loss": 0.271,
+      "step": 8931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7101813743100585,
+      "learning_rate": 4.2466609364105705e-06,
+      "loss": 0.2688,
+      "step": 8932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0294173890705043,
+      "learning_rate": 4.2456541020046235e-06,
+      "loss": 0.2712,
+      "step": 8933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3269548514176295,
+      "learning_rate": 4.24464729889783e-06,
+      "loss": 0.2714,
+      "step": 8934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.065650619612148,
+      "learning_rate": 4.243640527131956e-06,
+      "loss": 0.2695,
+      "step": 8935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.756013136161544,
+      "learning_rate": 4.242633786748786e-06,
+      "loss": 0.2737,
+      "step": 8936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9746356836729337,
+      "learning_rate": 4.241627077790083e-06,
+      "loss": 0.2925,
+      "step": 8937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.188654632913122,
+      "learning_rate": 4.24062040029762e-06,
+      "loss": 0.288,
+      "step": 8938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3826716843561906,
+      "learning_rate": 4.239613754313166e-06,
+      "loss": 0.2909,
+      "step": 8939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8464110193219216,
+      "learning_rate": 4.238607139878488e-06,
+      "loss": 0.2727,
+      "step": 8940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8897168708999552,
+      "learning_rate": 4.237600557035351e-06,
+      "loss": 0.2829,
+      "step": 8941
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.811244011148771,
+      "learning_rate": 4.236594005825522e-06,
+      "loss": 0.2942,
+      "step": 8942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3708395973271745,
+      "learning_rate": 4.235587486290761e-06,
+      "loss": 0.2816,
+      "step": 8943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.848707589575137,
+      "learning_rate": 4.234580998472836e-06,
+      "loss": 0.2735,
+      "step": 8944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0818197647066383,
+      "learning_rate": 4.2335745424135035e-06,
+      "loss": 0.2895,
+      "step": 8945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8105190386405015,
+      "learning_rate": 4.232568118154526e-06,
+      "loss": 0.2802,
+      "step": 8946
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6744721580363158,
+      "learning_rate": 4.231561725737659e-06,
+      "loss": 0.303,
+      "step": 8947
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7062637016033686,
+      "learning_rate": 4.230555365204662e-06,
+      "loss": 0.2674,
+      "step": 8948
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.014579609195482,
+      "learning_rate": 4.229549036597289e-06,
+      "loss": 0.2693,
+      "step": 8949
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8115931725341627,
+      "learning_rate": 4.228542739957293e-06,
+      "loss": 0.2751,
+      "step": 8950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1619926271356884,
+      "learning_rate": 4.227536475326431e-06,
+      "loss": 0.3196,
+      "step": 8951
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.148882830322034,
+      "learning_rate": 4.226530242746454e-06,
+      "loss": 0.2791,
+      "step": 8952
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0251760402930987,
+      "learning_rate": 4.2255240422591084e-06,
+      "loss": 0.2858,
+      "step": 8953
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6610577453796638,
+      "learning_rate": 4.224517873906149e-06,
+      "loss": 0.2654,
+      "step": 8954
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2941427565915737,
+      "learning_rate": 4.223511737729317e-06,
+      "loss": 0.2887,
+      "step": 8955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6778714793567895,
+      "learning_rate": 4.222505633770365e-06,
+      "loss": 0.3064,
+      "step": 8956
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4397769365284927,
+      "learning_rate": 4.221499562071034e-06,
+      "loss": 0.2769,
+      "step": 8957
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5081654100224524,
+      "learning_rate": 4.220493522673067e-06,
+      "loss": 0.2824,
+      "step": 8958
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6314409939815784,
+      "learning_rate": 4.219487515618211e-06,
+      "loss": 0.28,
+      "step": 8959
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.368558570660478,
+      "learning_rate": 4.218481540948204e-06,
+      "loss": 0.2683,
+      "step": 8960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9065182706097263,
+      "learning_rate": 4.217475598704785e-06,
+      "loss": 0.2981,
+      "step": 8961
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1749750023851813,
+      "learning_rate": 4.216469688929695e-06,
+      "loss": 0.2622,
+      "step": 8962
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.770772405111811,
+      "learning_rate": 4.215463811664668e-06,
+      "loss": 0.3086,
+      "step": 8963
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.412456722464413,
+      "learning_rate": 4.214457966951442e-06,
+      "loss": 0.2791,
+      "step": 8964
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9763561165106647,
+      "learning_rate": 4.213452154831747e-06,
+      "loss": 0.256,
+      "step": 8965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.39744250713294,
+      "learning_rate": 4.212446375347322e-06,
+      "loss": 0.2825,
+      "step": 8966
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4996357824472555,
+      "learning_rate": 4.211440628539896e-06,
+      "loss": 0.2927,
+      "step": 8967
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.857622541435515,
+      "learning_rate": 4.210434914451199e-06,
+      "loss": 0.2789,
+      "step": 8968
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0531193896832303,
+      "learning_rate": 4.209429233122958e-06,
+      "loss": 0.2887,
+      "step": 8969
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7214215778903053,
+      "learning_rate": 4.208423584596904e-06,
+      "loss": 0.2895,
+      "step": 8970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1945404384958525,
+      "learning_rate": 4.2074179689147605e-06,
+      "loss": 0.2825,
+      "step": 8971
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.990575182429611,
+      "learning_rate": 4.206412386118255e-06,
+      "loss": 0.2554,
+      "step": 8972
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0597585020530214,
+      "learning_rate": 4.205406836249106e-06,
+      "loss": 0.2827,
+      "step": 8973
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3306231040741285,
+      "learning_rate": 4.20440131934904e-06,
+      "loss": 0.2652,
+      "step": 8974
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1147726140867236,
+      "learning_rate": 4.203395835459778e-06,
+      "loss": 0.2653,
+      "step": 8975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7720739081897614,
+      "learning_rate": 4.202390384623037e-06,
+      "loss": 0.2642,
+      "step": 8976
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.163967357881253,
+      "learning_rate": 4.201384966880535e-06,
+      "loss": 0.2689,
+      "step": 8977
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0608159511849995,
+      "learning_rate": 4.200379582273991e-06,
+      "loss": 0.2697,
+      "step": 8978
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8158680199937984,
+      "learning_rate": 4.199374230845116e-06,
+      "loss": 0.2654,
+      "step": 8979
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9732864716133736,
+      "learning_rate": 4.198368912635626e-06,
+      "loss": 0.2733,
+      "step": 8980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.561783066553702,
+      "learning_rate": 4.197363627687234e-06,
+      "loss": 0.2807,
+      "step": 8981
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.314744656264878,
+      "learning_rate": 4.196358376041653e-06,
+      "loss": 0.273,
+      "step": 8982
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.240795509974399,
+      "learning_rate": 4.195353157740589e-06,
+      "loss": 0.2938,
+      "step": 8983
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.195059031410783,
+      "learning_rate": 4.194347972825752e-06,
+      "loss": 0.3006,
+      "step": 8984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.17132524491701,
+      "learning_rate": 4.193342821338848e-06,
+      "loss": 0.2773,
+      "step": 8985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.7391202719302985,
+      "learning_rate": 4.192337703321584e-06,
+      "loss": 0.2733,
+      "step": 8986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0077722985303383,
+      "learning_rate": 4.191332618815663e-06,
+      "loss": 0.2546,
+      "step": 8987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1119538811705834,
+      "learning_rate": 4.190327567862786e-06,
+      "loss": 0.2733,
+      "step": 8988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.54282550770554,
+      "learning_rate": 4.189322550504658e-06,
+      "loss": 0.2759,
+      "step": 8989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1324939689176095,
+      "learning_rate": 4.1883175667829775e-06,
+      "loss": 0.2736,
+      "step": 8990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9892797161565217,
+      "learning_rate": 4.1873126167394425e-06,
+      "loss": 0.2995,
+      "step": 8991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.969412900225151,
+      "learning_rate": 4.186307700415752e-06,
+      "loss": 0.2571,
+      "step": 8992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8208932198518593,
+      "learning_rate": 4.185302817853599e-06,
+      "loss": 0.2683,
+      "step": 8993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.344225976821004,
+      "learning_rate": 4.18429796909468e-06,
+      "loss": 0.2904,
+      "step": 8994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1287523203770173,
+      "learning_rate": 4.183293154180684e-06,
+      "loss": 0.2871,
+      "step": 8995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5447129794893724,
+      "learning_rate": 4.1822883731533085e-06,
+      "loss": 0.2702,
+      "step": 8996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6363423048591406,
+      "learning_rate": 4.18128362605424e-06,
+      "loss": 0.2783,
+      "step": 8997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5618377162236095,
+      "learning_rate": 4.180278912925169e-06,
+      "loss": 0.29,
+      "step": 8998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.874233588117033,
+      "learning_rate": 4.1792742338077806e-06,
+      "loss": 0.2697,
+      "step": 8999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1873105727580326,
+      "learning_rate": 4.178269588743764e-06,
+      "loss": 0.2781,
+      "step": 9000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5649838974752368,
+      "learning_rate": 4.1772649777748e-06,
+      "loss": 0.2714,
+      "step": 9001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.177294078163326,
+      "learning_rate": 4.1762604009425745e-06,
+      "loss": 0.258,
+      "step": 9002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1328497363526147,
+      "learning_rate": 4.175255858288765e-06,
+      "loss": 0.2743,
+      "step": 9003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.170079281837565,
+      "learning_rate": 4.174251349855058e-06,
+      "loss": 0.2749,
+      "step": 9004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.715033219634509,
+      "learning_rate": 4.173246875683128e-06,
+      "loss": 0.2712,
+      "step": 9005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0388177094530606,
+      "learning_rate": 4.1722424358146555e-06,
+      "loss": 0.2861,
+      "step": 9006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4158034859742052,
+      "learning_rate": 4.1712380302913125e-06,
+      "loss": 0.2651,
+      "step": 9007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.455353466151032,
+      "learning_rate": 4.170233659154777e-06,
+      "loss": 0.2708,
+      "step": 9008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0803466282254988,
+      "learning_rate": 4.169229322446719e-06,
+      "loss": 0.2818,
+      "step": 9009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.300872631067211,
+      "learning_rate": 4.168225020208813e-06,
+      "loss": 0.2863,
+      "step": 9010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9927010428964396,
+      "learning_rate": 4.167220752482728e-06,
+      "loss": 0.2686,
+      "step": 9011
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8583405483567006,
+      "learning_rate": 4.166216519310134e-06,
+      "loss": 0.2753,
+      "step": 9012
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1441962938758565,
+      "learning_rate": 4.165212320732696e-06,
+      "loss": 0.2592,
+      "step": 9013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7073128293033326,
+      "learning_rate": 4.1642081567920845e-06,
+      "loss": 0.2812,
+      "step": 9014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6457581470100427,
+      "learning_rate": 4.163204027529959e-06,
+      "loss": 0.4653,
+      "step": 9015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6065210477113765,
+      "learning_rate": 4.162199932987986e-06,
+      "loss": 0.2681,
+      "step": 9016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4496132064690843,
+      "learning_rate": 4.161195873207824e-06,
+      "loss": 0.2724,
+      "step": 9017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.954544716150455,
+      "learning_rate": 4.1601918482311355e-06,
+      "loss": 0.2716,
+      "step": 9018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8725483274683463,
+      "learning_rate": 4.159187858099579e-06,
+      "loss": 0.276,
+      "step": 9019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2842182803767546,
+      "learning_rate": 4.158183902854813e-06,
+      "loss": 0.2446,
+      "step": 9020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8445995790033862,
+      "learning_rate": 4.15717998253849e-06,
+      "loss": 0.2841,
+      "step": 9021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8235877124634015,
+      "learning_rate": 4.156176097192269e-06,
+      "loss": 0.2764,
+      "step": 9022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.319956144085122,
+      "learning_rate": 4.1551722468577995e-06,
+      "loss": 0.2723,
+      "step": 9023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9463835209956808,
+      "learning_rate": 4.154168431576734e-06,
+      "loss": 0.2898,
+      "step": 9024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.209764728812881,
+      "learning_rate": 4.153164651390721e-06,
+      "loss": 0.2845,
+      "step": 9025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.926766489150122,
+      "learning_rate": 4.152160906341413e-06,
+      "loss": 0.2861,
+      "step": 9026
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.843391699472927,
+      "learning_rate": 4.151157196470454e-06,
+      "loss": 0.2769,
+      "step": 9027
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6687440898243957,
+      "learning_rate": 4.150153521819491e-06,
+      "loss": 0.2732,
+      "step": 9028
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4487680916028647,
+      "learning_rate": 4.149149882430168e-06,
+      "loss": 0.3057,
+      "step": 9029
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5382904606927403,
+      "learning_rate": 4.148146278344128e-06,
+      "loss": 0.2713,
+      "step": 9030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5606882320486197,
+      "learning_rate": 4.147142709603011e-06,
+      "loss": 0.27,
+      "step": 9031
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.189897193920214,
+      "learning_rate": 4.1461391762484574e-06,
+      "loss": 0.2991,
+      "step": 9032
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.827595602428548,
+      "learning_rate": 4.145135678322106e-06,
+      "loss": 0.2787,
+      "step": 9033
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.112619212290249,
+      "learning_rate": 4.144132215865595e-06,
+      "loss": 0.2929,
+      "step": 9034
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.160940061450189,
+      "learning_rate": 4.143128788920558e-06,
+      "loss": 0.2894,
+      "step": 9035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.106577037100918,
+      "learning_rate": 4.14212539752863e-06,
+      "loss": 0.2768,
+      "step": 9036
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7121411615911037,
+      "learning_rate": 4.141122041731443e-06,
+      "loss": 0.3053,
+      "step": 9037
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.595180974080103,
+      "learning_rate": 4.140118721570628e-06,
+      "loss": 0.2968,
+      "step": 9038
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7088594743653682,
+      "learning_rate": 4.139115437087814e-06,
+      "loss": 0.2978,
+      "step": 9039
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6442278626820217,
+      "learning_rate": 4.138112188324629e-06,
+      "loss": 0.2948,
+      "step": 9040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3567313026743775,
+      "learning_rate": 4.1371089753227e-06,
+      "loss": 0.2952,
+      "step": 9041
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5945756403391027,
+      "learning_rate": 4.136105798123654e-06,
+      "loss": 0.2747,
+      "step": 9042
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.150254403840772,
+      "learning_rate": 4.135102656769112e-06,
+      "loss": 0.2576,
+      "step": 9043
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.429389980405473,
+      "learning_rate": 4.134099551300698e-06,
+      "loss": 0.2616,
+      "step": 9044
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0269162189147427,
+      "learning_rate": 4.1330964817600305e-06,
+      "loss": 0.285,
+      "step": 9045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.05733529819904,
+      "learning_rate": 4.1320934481887305e-06,
+      "loss": 0.3027,
+      "step": 9046
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.051632157570671,
+      "learning_rate": 4.131090450628413e-06,
+      "loss": 0.3013,
+      "step": 9047
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.121106128712773,
+      "learning_rate": 4.1300874891206974e-06,
+      "loss": 0.2817,
+      "step": 9048
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.394807891657507,
+      "learning_rate": 4.129084563707197e-06,
+      "loss": 0.293,
+      "step": 9049
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.530018270277474,
+      "learning_rate": 4.128081674429526e-06,
+      "loss": 0.2618,
+      "step": 9050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.044056314133268,
+      "learning_rate": 4.127078821329294e-06,
+      "loss": 0.2705,
+      "step": 9051
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.244917727649454,
+      "learning_rate": 4.126076004448113e-06,
+      "loss": 0.2619,
+      "step": 9052
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.312164056316612,
+      "learning_rate": 4.125073223827591e-06,
+      "loss": 0.2734,
+      "step": 9053
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7450869136011407,
+      "learning_rate": 4.124070479509334e-06,
+      "loss": 0.2841,
+      "step": 9054
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.820645042683543,
+      "learning_rate": 4.1230677715349475e-06,
+      "loss": 0.2722,
+      "step": 9055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8573115699327634,
+      "learning_rate": 4.122065099946038e-06,
+      "loss": 0.2865,
+      "step": 9056
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7987228881782893,
+      "learning_rate": 4.121062464784206e-06,
+      "loss": 0.2854,
+      "step": 9057
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2720463921823826,
+      "learning_rate": 4.120059866091056e-06,
+      "loss": 0.2716,
+      "step": 9058
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.864948724792566,
+      "learning_rate": 4.119057303908183e-06,
+      "loss": 0.2764,
+      "step": 9059
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9189110251778523,
+      "learning_rate": 4.118054778277189e-06,
+      "loss": 0.2754,
+      "step": 9060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.1701046556276715,
+      "learning_rate": 4.117052289239667e-06,
+      "loss": 0.2539,
+      "step": 9061
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3569895924012774,
+      "learning_rate": 4.1160498368372124e-06,
+      "loss": 0.279,
+      "step": 9062
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6604809832646694,
+      "learning_rate": 4.1150474211114225e-06,
+      "loss": 0.2853,
+      "step": 9063
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0861832642577056,
+      "learning_rate": 4.1140450421038865e-06,
+      "loss": 0.2757,
+      "step": 9064
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.451331767900175,
+      "learning_rate": 4.113042699856195e-06,
+      "loss": 0.2689,
+      "step": 9065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.133384723763996,
+      "learning_rate": 4.112040394409939e-06,
+      "loss": 0.2833,
+      "step": 9066
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.403540064432925,
+      "learning_rate": 4.1110381258067026e-06,
+      "loss": 0.2886,
+      "step": 9067
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.025493214522345,
+      "learning_rate": 4.110035894088074e-06,
+      "loss": 0.2818,
+      "step": 9068
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3290467463766698,
+      "learning_rate": 4.109033699295637e-06,
+      "loss": 0.2886,
+      "step": 9069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.0095072490111425,
+      "learning_rate": 4.108031541470972e-06,
+      "loss": 0.2735,
+      "step": 9070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.362202204453788,
+      "learning_rate": 4.107029420655664e-06,
+      "loss": 0.2803,
+      "step": 9071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0550589270341977,
+      "learning_rate": 4.106027336891293e-06,
+      "loss": 0.2646,
+      "step": 9072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8612014518138986,
+      "learning_rate": 4.105025290219435e-06,
+      "loss": 0.2814,
+      "step": 9073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.388373069164007,
+      "learning_rate": 4.104023280681667e-06,
+      "loss": 0.2834,
+      "step": 9074
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.504939264725376,
+      "learning_rate": 4.1030213083195645e-06,
+      "loss": 0.3156,
+      "step": 9075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2076874841158785,
+      "learning_rate": 4.102019373174702e-06,
+      "loss": 0.2874,
+      "step": 9076
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9629721043413078,
+      "learning_rate": 4.101017475288648e-06,
+      "loss": 0.2761,
+      "step": 9077
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2157956782011876,
+      "learning_rate": 4.100015614702977e-06,
+      "loss": 0.2773,
+      "step": 9078
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.229573127715457,
+      "learning_rate": 4.099013791459258e-06,
+      "loss": 0.2923,
+      "step": 9079
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.148341882612355,
+      "learning_rate": 4.098012005599056e-06,
+      "loss": 0.2728,
+      "step": 9080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6428228018447255,
+      "learning_rate": 4.097010257163938e-06,
+      "loss": 0.2649,
+      "step": 9081
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.395846389958749,
+      "learning_rate": 4.09600854619547e-06,
+      "loss": 0.2696,
+      "step": 9082
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.852686782804362,
+      "learning_rate": 4.095006872735211e-06,
+      "loss": 0.2809,
+      "step": 9083
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.909755305140324,
+      "learning_rate": 4.094005236824726e-06,
+      "loss": 0.2701,
+      "step": 9084
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5149630657118824,
+      "learning_rate": 4.093003638505571e-06,
+      "loss": 0.2874,
+      "step": 9085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9605236579980447,
+      "learning_rate": 4.092002077819307e-06,
+      "loss": 0.2682,
+      "step": 9086
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.133333854294839,
+      "learning_rate": 4.09100055480749e-06,
+      "loss": 0.2695,
+      "step": 9087
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.173200404535774,
+      "learning_rate": 4.0899990695116745e-06,
+      "loss": 0.2763,
+      "step": 9088
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6469218776810701,
+      "learning_rate": 4.088997621973413e-06,
+      "loss": 0.5043,
+      "step": 9089
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6861730144194709,
+      "learning_rate": 4.0879962122342596e-06,
+      "loss": 0.2667,
+      "step": 9090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.750872360073634,
+      "learning_rate": 4.086994840335763e-06,
+      "loss": 0.2781,
+      "step": 9091
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5900024528158422,
+      "learning_rate": 4.08599350631947e-06,
+      "loss": 0.4712,
+      "step": 9092
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9922181100792156,
+      "learning_rate": 4.084992210226932e-06,
+      "loss": 0.2779,
+      "step": 9093
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8149982471392105,
+      "learning_rate": 4.083990952099692e-06,
+      "loss": 0.2883,
+      "step": 9094
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.60490040605363,
+      "learning_rate": 4.0829897319792944e-06,
+      "loss": 0.2722,
+      "step": 9095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.097376451613655,
+      "learning_rate": 4.081988549907282e-06,
+      "loss": 0.2839,
+      "step": 9096
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5063950607141985,
+      "learning_rate": 4.080987405925195e-06,
+      "loss": 0.2883,
+      "step": 9097
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3265190953944352,
+      "learning_rate": 4.079986300074573e-06,
+      "loss": 0.2712,
+      "step": 9098
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.057986856956302,
+      "learning_rate": 4.078985232396953e-06,
+      "loss": 0.2662,
+      "step": 9099
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2865199414983657,
+      "learning_rate": 4.0779842029338714e-06,
+      "loss": 0.3091,
+      "step": 9100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.199288830026781,
+      "learning_rate": 4.076983211726863e-06,
+      "loss": 0.2693,
+      "step": 9101
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.625226987848647,
+      "learning_rate": 4.075982258817462e-06,
+      "loss": 0.2732,
+      "step": 9102
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3269084514484146,
+      "learning_rate": 4.074981344247197e-06,
+      "loss": 0.2831,
+      "step": 9103
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2655176838129396,
+      "learning_rate": 4.0739804680576e-06,
+      "loss": 0.2731,
+      "step": 9104
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8790560167473216,
+      "learning_rate": 4.0729796302901975e-06,
+      "loss": 0.2747,
+      "step": 9105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7285640356748337,
+      "learning_rate": 4.071978830986518e-06,
+      "loss": 0.2942,
+      "step": 9106
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2212964584503663,
+      "learning_rate": 4.070978070188083e-06,
+      "loss": 0.3006,
+      "step": 9107
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8213856857007094,
+      "learning_rate": 4.069977347936418e-06,
+      "loss": 0.285,
+      "step": 9108
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.794564427043924,
+      "learning_rate": 4.068976664273046e-06,
+      "loss": 0.309,
+      "step": 9109
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6854725343272413,
+      "learning_rate": 4.067976019239486e-06,
+      "loss": 0.4747,
+      "step": 9110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2039729196478537,
+      "learning_rate": 4.0669754128772554e-06,
+      "loss": 0.2753,
+      "step": 9111
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.479916467364978,
+      "learning_rate": 4.065974845227874e-06,
+      "loss": 0.2869,
+      "step": 9112
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.750741494660911,
+      "learning_rate": 4.064974316332854e-06,
+      "loss": 0.2864,
+      "step": 9113
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8052909941077098,
+      "learning_rate": 4.063973826233708e-06,
+      "loss": 0.2689,
+      "step": 9114
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.745078312464633,
+      "learning_rate": 4.062973374971954e-06,
+      "loss": 0.2915,
+      "step": 9115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8686242934885526,
+      "learning_rate": 4.061972962589098e-06,
+      "loss": 0.2714,
+      "step": 9116
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.269804397596038,
+      "learning_rate": 4.06097258912665e-06,
+      "loss": 0.281,
+      "step": 9117
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.988428007549461,
+      "learning_rate": 4.059972254626118e-06,
+      "loss": 0.2704,
+      "step": 9118
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.45296716345786,
+      "learning_rate": 4.058971959129006e-06,
+      "loss": 0.2776,
+      "step": 9119
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.297886432980261,
+      "learning_rate": 4.057971702676819e-06,
+      "loss": 0.2805,
+      "step": 9120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.579428080254274,
+      "learning_rate": 4.0569714853110585e-06,
+      "loss": 0.275,
+      "step": 9121
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.50183963770546,
+      "learning_rate": 4.055971307073225e-06,
+      "loss": 0.2918,
+      "step": 9122
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.102321279641808,
+      "learning_rate": 4.054971168004822e-06,
+      "loss": 0.2697,
+      "step": 9123
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8339897951859103,
+      "learning_rate": 4.0539710681473415e-06,
+      "loss": 0.2822,
+      "step": 9124
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.964936061859733,
+      "learning_rate": 4.052971007542283e-06,
+      "loss": 0.2614,
+      "step": 9125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9912289138458814,
+      "learning_rate": 4.051970986231139e-06,
+      "loss": 0.2736,
+      "step": 9126
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8981902565682565,
+      "learning_rate": 4.0509710042554026e-06,
+      "loss": 0.285,
+      "step": 9127
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.654667301972239,
+      "learning_rate": 4.049971061656566e-06,
+      "loss": 0.4594,
+      "step": 9128
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.034644345762902,
+      "learning_rate": 4.048971158476114e-06,
+      "loss": 0.2729,
+      "step": 9129
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.933527205490952,
+      "learning_rate": 4.04797129475554e-06,
+      "loss": 0.2709,
+      "step": 9130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2155708788882147,
+      "learning_rate": 4.04697147053633e-06,
+      "loss": 0.2634,
+      "step": 9131
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.733420587729311,
+      "learning_rate": 4.045971685859964e-06,
+      "loss": 0.2842,
+      "step": 9132
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.216164533715891,
+      "learning_rate": 4.044971940767927e-06,
+      "loss": 0.2582,
+      "step": 9133
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.635077684144465,
+      "learning_rate": 4.043972235301703e-06,
+      "loss": 0.2646,
+      "step": 9134
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5178197638888022,
+      "learning_rate": 4.042972569502767e-06,
+      "loss": 0.2963,
+      "step": 9135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9129191601519608,
+      "learning_rate": 4.0419729434126e-06,
+      "loss": 0.2809,
+      "step": 9136
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.08700108767703,
+      "learning_rate": 4.040973357072676e-06,
+      "loss": 0.286,
+      "step": 9137
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.037377534653133,
+      "learning_rate": 4.039973810524471e-06,
+      "loss": 0.2813,
+      "step": 9138
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9294021769241192,
+      "learning_rate": 4.038974303809459e-06,
+      "loss": 0.2758,
+      "step": 9139
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 38.69368353543663,
+      "learning_rate": 4.03797483696911e-06,
+      "loss": 0.2897,
+      "step": 9140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.955541783425685,
+      "learning_rate": 4.036975410044892e-06,
+      "loss": 0.2692,
+      "step": 9141
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7956800707173444,
+      "learning_rate": 4.035976023078278e-06,
+      "loss": 0.2683,
+      "step": 9142
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4126732889509968,
+      "learning_rate": 4.0349766761107275e-06,
+      "loss": 0.2791,
+      "step": 9143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8318435302789828,
+      "learning_rate": 4.033977369183709e-06,
+      "loss": 0.2669,
+      "step": 9144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4843744083775534,
+      "learning_rate": 4.032978102338688e-06,
+      "loss": 0.2585,
+      "step": 9145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.755507177658057,
+      "learning_rate": 4.031978875617122e-06,
+      "loss": 0.2679,
+      "step": 9146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.047013529731842,
+      "learning_rate": 4.030979689060471e-06,
+      "loss": 0.2694,
+      "step": 9147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.280461486830635,
+      "learning_rate": 4.029980542710196e-06,
+      "loss": 0.2728,
+      "step": 9148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7130861819098677,
+      "learning_rate": 4.028981436607749e-06,
+      "loss": 0.2845,
+      "step": 9149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5905654057925287,
+      "learning_rate": 4.0279823707945885e-06,
+      "loss": 0.259,
+      "step": 9150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.786063709105976,
+      "learning_rate": 4.0269833453121644e-06,
+      "loss": 0.2669,
+      "step": 9151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.093826153277464,
+      "learning_rate": 4.025984360201929e-06,
+      "loss": 0.2681,
+      "step": 9152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.129957657546236,
+      "learning_rate": 4.0249854155053345e-06,
+      "loss": 0.2589,
+      "step": 9153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.694219205540417,
+      "learning_rate": 4.023986511263827e-06,
+      "loss": 0.2757,
+      "step": 9154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.1687010584432,
+      "learning_rate": 4.022987647518851e-06,
+      "loss": 0.2802,
+      "step": 9155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6205551709536556,
+      "learning_rate": 4.0219888243118546e-06,
+      "loss": 0.282,
+      "step": 9156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.042505089343204,
+      "learning_rate": 4.020990041684278e-06,
+      "loss": 0.2599,
+      "step": 9157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9101776779715314,
+      "learning_rate": 4.019991299677565e-06,
+      "loss": 0.2828,
+      "step": 9158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.17251905671189,
+      "learning_rate": 4.018992598333151e-06,
+      "loss": 0.2864,
+      "step": 9159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9085476506756567,
+      "learning_rate": 4.017993937692478e-06,
+      "loss": 0.2713,
+      "step": 9160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4533500041412646,
+      "learning_rate": 4.0169953177969814e-06,
+      "loss": 0.28,
+      "step": 9161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.684952679191809,
+      "learning_rate": 4.015996738688094e-06,
+      "loss": 0.2866,
+      "step": 9162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9267543954730049,
+      "learning_rate": 4.01499820040725e-06,
+      "loss": 0.2744,
+      "step": 9163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.418685357774127,
+      "learning_rate": 4.013999702995881e-06,
+      "loss": 0.2793,
+      "step": 9164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.914806246349283,
+      "learning_rate": 4.013001246495415e-06,
+      "loss": 0.2716,
+      "step": 9165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0052424753735694,
+      "learning_rate": 4.012002830947281e-06,
+      "loss": 0.2658,
+      "step": 9166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.88961432657455,
+      "learning_rate": 4.011004456392903e-06,
+      "loss": 0.2607,
+      "step": 9167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.295553631017359,
+      "learning_rate": 4.010006122873707e-06,
+      "loss": 0.2821,
+      "step": 9168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.235881630326167,
+      "learning_rate": 4.009007830431118e-06,
+      "loss": 0.2823,
+      "step": 9169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.672443743563525,
+      "learning_rate": 4.008009579106551e-06,
+      "loss": 0.487,
+      "step": 9170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.006177691577419,
+      "learning_rate": 4.007011368941429e-06,
+      "loss": 0.27,
+      "step": 9171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.546157449135116,
+      "learning_rate": 4.0060131999771715e-06,
+      "loss": 0.3116,
+      "step": 9172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.827833423456838,
+      "learning_rate": 4.00501507225519e-06,
+      "loss": 0.2665,
+      "step": 9173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.168560296788205,
+      "learning_rate": 4.0040169858169e-06,
+      "loss": 0.2701,
+      "step": 9174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.7020082264023735,
+      "learning_rate": 4.003018940703716e-06,
+      "loss": 0.2705,
+      "step": 9175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.167975200895409,
+      "learning_rate": 4.002020936957045e-06,
+      "loss": 0.2627,
+      "step": 9176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5360734183114646,
+      "learning_rate": 4.0010229746183e-06,
+      "loss": 0.2587,
+      "step": 9177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.23888902376058,
+      "learning_rate": 4.0000250537288845e-06,
+      "loss": 0.2696,
+      "step": 9178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.088208684446045,
+      "learning_rate": 3.999027174330206e-06,
+      "loss": 0.2776,
+      "step": 9179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.856449926681893,
+      "learning_rate": 3.9980293364636694e-06,
+      "loss": 0.2619,
+      "step": 9180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.48506973500211,
+      "learning_rate": 3.9970315401706726e-06,
+      "loss": 0.2817,
+      "step": 9181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0070379975397,
+      "learning_rate": 3.99603378549262e-06,
+      "loss": 0.2701,
+      "step": 9182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.975709514802917,
+      "learning_rate": 3.99503607247091e-06,
+      "loss": 0.3074,
+      "step": 9183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.600647208977965,
+      "learning_rate": 3.994038401146937e-06,
+      "loss": 0.2801,
+      "step": 9184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6339462822329724,
+      "learning_rate": 3.993040771562098e-06,
+      "loss": 0.2858,
+      "step": 9185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7211227449044615,
+      "learning_rate": 3.9920431837577876e-06,
+      "loss": 0.2728,
+      "step": 9186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1177502246771835,
+      "learning_rate": 3.991045637775393e-06,
+      "loss": 0.2637,
+      "step": 9187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6273683184959713,
+      "learning_rate": 3.99004813365631e-06,
+      "loss": 0.5095,
+      "step": 9188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.10203118302687,
+      "learning_rate": 3.98905067144192e-06,
+      "loss": 0.2816,
+      "step": 9189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.98073069918242,
+      "learning_rate": 3.988053251173615e-06,
+      "loss": 0.267,
+      "step": 9190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.354093564094669,
+      "learning_rate": 3.987055872892779e-06,
+      "loss": 0.2778,
+      "step": 9191
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5464327724788087,
+      "learning_rate": 3.986058536640793e-06,
+      "loss": 0.256,
+      "step": 9192
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2718249165470352,
+      "learning_rate": 3.985061242459039e-06,
+      "loss": 0.284,
+      "step": 9193
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.600607028959871,
+      "learning_rate": 3.984063990388899e-06,
+      "loss": 0.2738,
+      "step": 9194
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.969615068838298,
+      "learning_rate": 3.983066780471747e-06,
+      "loss": 0.2857,
+      "step": 9195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4830170588299847,
+      "learning_rate": 3.982069612748959e-06,
+      "loss": 0.2752,
+      "step": 9196
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.40807304756567,
+      "learning_rate": 3.981072487261913e-06,
+      "loss": 0.2702,
+      "step": 9197
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.8419269627637975,
+      "learning_rate": 3.9800754040519785e-06,
+      "loss": 0.2587,
+      "step": 9198
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.357213022948169,
+      "learning_rate": 3.979078363160528e-06,
+      "loss": 0.2775,
+      "step": 9199
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.630534764748263,
+      "learning_rate": 3.9780813646289286e-06,
+      "loss": 0.2736,
+      "step": 9200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.526560502368286,
+      "learning_rate": 3.977084408498549e-06,
+      "loss": 0.271,
+      "step": 9201
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5819987109373376,
+      "learning_rate": 3.976087494810754e-06,
+      "loss": 0.4673,
+      "step": 9202
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.79083350753169,
+      "learning_rate": 3.975090623606907e-06,
+      "loss": 0.2933,
+      "step": 9203
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6885841543703464,
+      "learning_rate": 3.97409379492837e-06,
+      "loss": 0.2806,
+      "step": 9204
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.983305670568188,
+      "learning_rate": 3.973097008816505e-06,
+      "loss": 0.2739,
+      "step": 9205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6457106968133617,
+      "learning_rate": 3.972100265312669e-06,
+      "loss": 0.5008,
+      "step": 9206
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7651344050525344,
+      "learning_rate": 3.971103564458219e-06,
+      "loss": 0.2857,
+      "step": 9207
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8530805703622715,
+      "learning_rate": 3.970106906294509e-06,
+      "loss": 0.2693,
+      "step": 9208
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.628739316517356,
+      "learning_rate": 3.9691102908628925e-06,
+      "loss": 0.2895,
+      "step": 9209
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.30876149085146,
+      "learning_rate": 3.968113718204722e-06,
+      "loss": 0.2686,
+      "step": 9210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.652158089528851,
+      "learning_rate": 3.967117188361345e-06,
+      "loss": 0.2991,
+      "step": 9211
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.236986397916008,
+      "learning_rate": 3.96612070137411e-06,
+      "loss": 0.2507,
+      "step": 9212
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.993317824882817,
+      "learning_rate": 3.965124257284366e-06,
+      "loss": 0.2902,
+      "step": 9213
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9551932726352708,
+      "learning_rate": 3.964127856133453e-06,
+      "loss": 0.2962,
+      "step": 9214
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4245063545873173,
+      "learning_rate": 3.963131497962715e-06,
+      "loss": 0.2711,
+      "step": 9215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.503509197929178,
+      "learning_rate": 3.9621351828134935e-06,
+      "loss": 0.2891,
+      "step": 9216
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3175082169082923,
+      "learning_rate": 3.961138910727126e-06,
+      "loss": 0.2628,
+      "step": 9217
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.206590509360579,
+      "learning_rate": 3.960142681744952e-06,
+      "loss": 0.3006,
+      "step": 9218
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.976329787919778,
+      "learning_rate": 3.959146495908303e-06,
+      "loss": 0.2751,
+      "step": 9219
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.659175816800913,
+      "learning_rate": 3.958150353258515e-06,
+      "loss": 0.2952,
+      "step": 9220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8089002270216088,
+      "learning_rate": 3.957154253836921e-06,
+      "loss": 0.2695,
+      "step": 9221
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.261169418230604,
+      "learning_rate": 3.9561581976848475e-06,
+      "loss": 0.2722,
+      "step": 9222
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8436599195476457,
+      "learning_rate": 3.955162184843625e-06,
+      "loss": 0.2913,
+      "step": 9223
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7849197509321244,
+      "learning_rate": 3.95416621535458e-06,
+      "loss": 0.2746,
+      "step": 9224
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.4996914745383165,
+      "learning_rate": 3.953170289259036e-06,
+      "loss": 0.275,
+      "step": 9225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2381976452551235,
+      "learning_rate": 3.952174406598314e-06,
+      "loss": 0.2761,
+      "step": 9226
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.158901164094756,
+      "learning_rate": 3.951178567413739e-06,
+      "loss": 0.2662,
+      "step": 9227
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.369904440501148,
+      "learning_rate": 3.950182771746629e-06,
+      "loss": 0.2514,
+      "step": 9228
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.123030064883653,
+      "learning_rate": 3.9491870196383e-06,
+      "loss": 0.2937,
+      "step": 9229
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6122154605217522,
+      "learning_rate": 3.948191311130067e-06,
+      "loss": 0.2537,
+      "step": 9230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5841996804823477,
+      "learning_rate": 3.947195646263246e-06,
+      "loss": 0.2647,
+      "step": 9231
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.3264531672810795,
+      "learning_rate": 3.946200025079147e-06,
+      "loss": 0.2782,
+      "step": 9232
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.549684787107598,
+      "learning_rate": 3.945204447619081e-06,
+      "loss": 0.2824,
+      "step": 9233
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.459502477500608,
+      "learning_rate": 3.944208913924354e-06,
+      "loss": 0.2699,
+      "step": 9234
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6854850466208346,
+      "learning_rate": 3.943213424036277e-06,
+      "loss": 0.4983,
+      "step": 9235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7813734130762415,
+      "learning_rate": 3.942217977996151e-06,
+      "loss": 0.2781,
+      "step": 9236
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 21.838722174618145,
+      "learning_rate": 3.94122257584528e-06,
+      "loss": 0.2824,
+      "step": 9237
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6116462264193507,
+      "learning_rate": 3.940227217624965e-06,
+      "loss": 0.2923,
+      "step": 9238
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8344541824735696,
+      "learning_rate": 3.939231903376505e-06,
+      "loss": 0.2806,
+      "step": 9239
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9031682065374627,
+      "learning_rate": 3.938236633141199e-06,
+      "loss": 0.265,
+      "step": 9240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1751783097786794,
+      "learning_rate": 3.937241406960338e-06,
+      "loss": 0.2911,
+      "step": 9241
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.212821549681234,
+      "learning_rate": 3.93624622487522e-06,
+      "loss": 0.2713,
+      "step": 9242
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.7039019833786515,
+      "learning_rate": 3.935251086927137e-06,
+      "loss": 0.2872,
+      "step": 9243
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.078187430760622,
+      "learning_rate": 3.934255993157375e-06,
+      "loss": 0.2821,
+      "step": 9244
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.336290319319591,
+      "learning_rate": 3.933260943607228e-06,
+      "loss": 0.2729,
+      "step": 9245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4912092543409914,
+      "learning_rate": 3.932265938317977e-06,
+      "loss": 0.2828,
+      "step": 9246
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6318907624468713,
+      "learning_rate": 3.931270977330909e-06,
+      "loss": 0.4828,
+      "step": 9247
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7379207239142243,
+      "learning_rate": 3.930276060687307e-06,
+      "loss": 0.2641,
+      "step": 9248
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6218894435067417,
+      "learning_rate": 3.92928118842845e-06,
+      "loss": 0.2833,
+      "step": 9249
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.601040136110174,
+      "learning_rate": 3.928286360595619e-06,
+      "loss": 0.5165,
+      "step": 9250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1536563858684312,
+      "learning_rate": 3.927291577230091e-06,
+      "loss": 0.2702,
+      "step": 9251
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8247434711763884,
+      "learning_rate": 3.926296838373141e-06,
+      "loss": 0.2659,
+      "step": 9252
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1103878092146062,
+      "learning_rate": 3.925302144066042e-06,
+      "loss": 0.2779,
+      "step": 9253
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.466824691038825,
+      "learning_rate": 3.924307494350066e-06,
+      "loss": 0.2753,
+      "step": 9254
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8396376075892753,
+      "learning_rate": 3.9233128892664815e-06,
+      "loss": 0.2634,
+      "step": 9255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.258611616343802,
+      "learning_rate": 3.922318328856557e-06,
+      "loss": 0.2826,
+      "step": 9256
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.947238415015567,
+      "learning_rate": 3.921323813161562e-06,
+      "loss": 0.2762,
+      "step": 9257
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0129693418965022,
+      "learning_rate": 3.9203293422227564e-06,
+      "loss": 0.2792,
+      "step": 9258
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8433568163588185,
+      "learning_rate": 3.919334916081406e-06,
+      "loss": 0.2579,
+      "step": 9259
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.605504807618717,
+      "learning_rate": 3.918340534778767e-06,
+      "loss": 0.3022,
+      "step": 9260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9318041140657272,
+      "learning_rate": 3.917346198356103e-06,
+      "loss": 0.2713,
+      "step": 9261
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.643841981750636,
+      "learning_rate": 3.916351906854665e-06,
+      "loss": 0.2656,
+      "step": 9262
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9256949886880956,
+      "learning_rate": 3.915357660315712e-06,
+      "loss": 0.2691,
+      "step": 9263
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9563109630509625,
+      "learning_rate": 3.914363458780496e-06,
+      "loss": 0.2651,
+      "step": 9264
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4508005363252714,
+      "learning_rate": 3.913369302290271e-06,
+      "loss": 0.2775,
+      "step": 9265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5971326424201586,
+      "learning_rate": 3.912375190886281e-06,
+      "loss": 0.2648,
+      "step": 9266
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.675208419701591,
+      "learning_rate": 3.911381124609778e-06,
+      "loss": 0.2537,
+      "step": 9267
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2601122342457307,
+      "learning_rate": 3.9103871035020044e-06,
+      "loss": 0.29,
+      "step": 9268
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2851296786488087,
+      "learning_rate": 3.909393127604206e-06,
+      "loss": 0.2794,
+      "step": 9269
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.941750832709662,
+      "learning_rate": 3.908399196957625e-06,
+      "loss": 0.2805,
+      "step": 9270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.407309898368578,
+      "learning_rate": 3.907405311603497e-06,
+      "loss": 0.2876,
+      "step": 9271
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.078145251398732,
+      "learning_rate": 3.906411471583065e-06,
+      "loss": 0.2695,
+      "step": 9272
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8576362429051776,
+      "learning_rate": 3.905417676937564e-06,
+      "loss": 0.2726,
+      "step": 9273
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2375411146914583,
+      "learning_rate": 3.9044239277082275e-06,
+      "loss": 0.2649,
+      "step": 9274
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5901883542414676,
+      "learning_rate": 3.903430223936289e-06,
+      "loss": 0.2771,
+      "step": 9275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.386079394630121,
+      "learning_rate": 3.9024365656629774e-06,
+      "loss": 0.2847,
+      "step": 9276
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1223109386297265,
+      "learning_rate": 3.901442952929522e-06,
+      "loss": 0.2957,
+      "step": 9277
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6301339216888115,
+      "learning_rate": 3.900449385777148e-06,
+      "loss": 0.2757,
+      "step": 9278
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.331999589760647,
+      "learning_rate": 3.899455864247085e-06,
+      "loss": 0.2602,
+      "step": 9279
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0459651310011617,
+      "learning_rate": 3.898462388380551e-06,
+      "loss": 0.2835,
+      "step": 9280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6533536651957998,
+      "learning_rate": 3.897468958218771e-06,
+      "loss": 0.2727,
+      "step": 9281
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.462847176750664,
+      "learning_rate": 3.896475573802961e-06,
+      "loss": 0.2798,
+      "step": 9282
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5149493904701377,
+      "learning_rate": 3.895482235174341e-06,
+      "loss": 0.2849,
+      "step": 9283
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.296923975090486,
+      "learning_rate": 3.894488942374123e-06,
+      "loss": 0.257,
+      "step": 9284
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.256151768772217,
+      "learning_rate": 3.893495695443522e-06,
+      "loss": 0.2639,
+      "step": 9285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.313014087149545,
+      "learning_rate": 3.89250249442375e-06,
+      "loss": 0.278,
+      "step": 9286
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.540986611801791,
+      "learning_rate": 3.891509339356018e-06,
+      "loss": 0.2769,
+      "step": 9287
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7000881897142213,
+      "learning_rate": 3.890516230281532e-06,
+      "loss": 0.2718,
+      "step": 9288
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8600218578769911,
+      "learning_rate": 3.889523167241499e-06,
+      "loss": 0.2876,
+      "step": 9289
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6955977080804243,
+      "learning_rate": 3.888530150277121e-06,
+      "loss": 0.2753,
+      "step": 9290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8290332204823654,
+      "learning_rate": 3.887537179429603e-06,
+      "loss": 0.2724,
+      "step": 9291
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2165372570674795,
+      "learning_rate": 3.886544254740141e-06,
+      "loss": 0.2516,
+      "step": 9292
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6392012056535392,
+      "learning_rate": 3.885551376249936e-06,
+      "loss": 0.2578,
+      "step": 9293
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.667023314668314,
+      "learning_rate": 3.884558544000184e-06,
+      "loss": 0.2903,
+      "step": 9294
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.238751676907104,
+      "learning_rate": 3.883565758032081e-06,
+      "loss": 0.2625,
+      "step": 9295
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.480637919262386,
+      "learning_rate": 3.882573018386816e-06,
+      "loss": 0.2831,
+      "step": 9296
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.551646106426363,
+      "learning_rate": 3.8815803251055826e-06,
+      "loss": 0.274,
+      "step": 9297
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5964922109046245,
+      "learning_rate": 3.880587678229567e-06,
+      "loss": 0.2902,
+      "step": 9298
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.608368448008926,
+      "learning_rate": 3.879595077799958e-06,
+      "loss": 0.2754,
+      "step": 9299
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9146461555348528,
+      "learning_rate": 3.878602523857938e-06,
+      "loss": 0.27,
+      "step": 9300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.861027720405857,
+      "learning_rate": 3.8776100164446905e-06,
+      "loss": 0.2795,
+      "step": 9301
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5705217704095804,
+      "learning_rate": 3.876617555601398e-06,
+      "loss": 0.2834,
+      "step": 9302
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8548233970260224,
+      "learning_rate": 3.875625141369239e-06,
+      "loss": 0.2783,
+      "step": 9303
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3720798077685035,
+      "learning_rate": 3.874632773789389e-06,
+      "loss": 0.2749,
+      "step": 9304
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3129696441353427,
+      "learning_rate": 3.8736404529030255e-06,
+      "loss": 0.2564,
+      "step": 9305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6769039219375834,
+      "learning_rate": 3.87264817875132e-06,
+      "loss": 0.2874,
+      "step": 9306
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.600291436389585,
+      "learning_rate": 3.871655951375443e-06,
+      "loss": 0.5002,
+      "step": 9307
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9061199034288974,
+      "learning_rate": 3.870663770816563e-06,
+      "loss": 0.2943,
+      "step": 9308
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.855016061376826,
+      "learning_rate": 3.869671637115853e-06,
+      "loss": 0.2887,
+      "step": 9309
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.925621228706962,
+      "learning_rate": 3.868679550314472e-06,
+      "loss": 0.2769,
+      "step": 9310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4470475541665326,
+      "learning_rate": 3.867687510453587e-06,
+      "loss": 0.2895,
+      "step": 9311
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4803229483520712,
+      "learning_rate": 3.866695517574358e-06,
+      "loss": 0.2755,
+      "step": 9312
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0185927478834387,
+      "learning_rate": 3.865703571717946e-06,
+      "loss": 0.3111,
+      "step": 9313
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.425685119607202,
+      "learning_rate": 3.864711672925506e-06,
+      "loss": 0.2812,
+      "step": 9314
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6278036288086817,
+      "learning_rate": 3.863719821238196e-06,
+      "loss": 0.2859,
+      "step": 9315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.178543980447864,
+      "learning_rate": 3.862728016697167e-06,
+      "loss": 0.2752,
+      "step": 9316
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0612224909353554,
+      "learning_rate": 3.8617362593435745e-06,
+      "loss": 0.2922,
+      "step": 9317
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5189909760652602,
+      "learning_rate": 3.860744549218566e-06,
+      "loss": 0.2694,
+      "step": 9318
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.038894841179102,
+      "learning_rate": 3.85975288636329e-06,
+      "loss": 0.2928,
+      "step": 9319
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6361349039783613,
+      "learning_rate": 3.858761270818892e-06,
+      "loss": 0.2979,
+      "step": 9320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7192020723611874,
+      "learning_rate": 3.857769702626516e-06,
+      "loss": 0.3063,
+      "step": 9321
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.736338572412326,
+      "learning_rate": 3.8567781818273034e-06,
+      "loss": 0.285,
+      "step": 9322
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5403393409699893,
+      "learning_rate": 3.855786708462394e-06,
+      "loss": 0.2597,
+      "step": 9323
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.53307827225536,
+      "learning_rate": 3.854795282572926e-06,
+      "loss": 0.2653,
+      "step": 9324
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7944944551201483,
+      "learning_rate": 3.853803904200039e-06,
+      "loss": 0.253,
+      "step": 9325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.586030407716853,
+      "learning_rate": 3.852812573384861e-06,
+      "loss": 0.2652,
+      "step": 9326
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6646869134092412,
+      "learning_rate": 3.851821290168528e-06,
+      "loss": 0.2666,
+      "step": 9327
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4260893578703877,
+      "learning_rate": 3.85083005459217e-06,
+      "loss": 0.2783,
+      "step": 9328
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8494398011020112,
+      "learning_rate": 3.8498388666969134e-06,
+      "loss": 0.2729,
+      "step": 9329
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1302486781491954,
+      "learning_rate": 3.848847726523885e-06,
+      "loss": 0.2606,
+      "step": 9330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.996773571331155,
+      "learning_rate": 3.847856634114207e-06,
+      "loss": 0.2671,
+      "step": 9331
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.847318593175028,
+      "learning_rate": 3.846865589509006e-06,
+      "loss": 0.2897,
+      "step": 9332
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1254200614702987,
+      "learning_rate": 3.8458745927494e-06,
+      "loss": 0.275,
+      "step": 9333
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.608875024556574,
+      "learning_rate": 3.844883643876507e-06,
+      "loss": 0.2774,
+      "step": 9334
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9053440485058792,
+      "learning_rate": 3.843892742931443e-06,
+      "loss": 0.2788,
+      "step": 9335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9981418936468147,
+      "learning_rate": 3.842901889955322e-06,
+      "loss": 0.2995,
+      "step": 9336
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1693773742322167,
+      "learning_rate": 3.841911084989259e-06,
+      "loss": 0.2745,
+      "step": 9337
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9074893159153605,
+      "learning_rate": 3.840920328074358e-06,
+      "loss": 0.2681,
+      "step": 9338
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.941526961208954,
+      "learning_rate": 3.839929619251734e-06,
+      "loss": 0.2647,
+      "step": 9339
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8735502187969488,
+      "learning_rate": 3.838938958562491e-06,
+      "loss": 0.2893,
+      "step": 9340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9307053582966782,
+      "learning_rate": 3.837948346047733e-06,
+      "loss": 0.2689,
+      "step": 9341
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8708470553193515,
+      "learning_rate": 3.836957781748562e-06,
+      "loss": 0.2788,
+      "step": 9342
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1371577226420304,
+      "learning_rate": 3.835967265706078e-06,
+      "loss": 0.2703,
+      "step": 9343
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7927673673217885,
+      "learning_rate": 3.834976797961379e-06,
+      "loss": 0.2856,
+      "step": 9344
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.538581634819629,
+      "learning_rate": 3.833986378555562e-06,
+      "loss": 0.2729,
+      "step": 9345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.25168465298363,
+      "learning_rate": 3.832996007529721e-06,
+      "loss": 0.2933,
+      "step": 9346
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.675468828378678,
+      "learning_rate": 3.83200568492495e-06,
+      "loss": 0.2854,
+      "step": 9347
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.805766266755283,
+      "learning_rate": 3.8310154107823375e-06,
+      "loss": 0.2628,
+      "step": 9348
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.571832877111927,
+      "learning_rate": 3.8300251851429715e-06,
+      "loss": 0.2714,
+      "step": 9349
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.098782017288707,
+      "learning_rate": 3.829035008047939e-06,
+      "loss": 0.2877,
+      "step": 9350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2974646345997423,
+      "learning_rate": 3.8280448795383245e-06,
+      "loss": 0.267,
+      "step": 9351
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.85193585986451,
+      "learning_rate": 3.827054799655207e-06,
+      "loss": 0.2624,
+      "step": 9352
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7888315975033644,
+      "learning_rate": 3.82606476843967e-06,
+      "loss": 0.2521,
+      "step": 9353
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4060659652934158,
+      "learning_rate": 3.825074785932792e-06,
+      "loss": 0.2576,
+      "step": 9354
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 15.16639271451097,
+      "learning_rate": 3.8240848521756484e-06,
+      "loss": 0.2634,
+      "step": 9355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4092035208211344,
+      "learning_rate": 3.823094967209312e-06,
+      "loss": 0.2721,
+      "step": 9356
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.46751038974512,
+      "learning_rate": 3.822105131074857e-06,
+      "loss": 0.2893,
+      "step": 9357
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.020572096130606,
+      "learning_rate": 3.8211153438133515e-06,
+      "loss": 0.2713,
+      "step": 9358
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5594628901051883,
+      "learning_rate": 3.820125605465864e-06,
+      "loss": 0.2877,
+      "step": 9359
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5046444553647635,
+      "learning_rate": 3.81913591607346e-06,
+      "loss": 0.2523,
+      "step": 9360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8256666327324984,
+      "learning_rate": 3.8181462756772056e-06,
+      "loss": 0.2721,
+      "step": 9361
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.853344318626228,
+      "learning_rate": 3.817156684318161e-06,
+      "loss": 0.2699,
+      "step": 9362
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6700246115584205,
+      "learning_rate": 3.816167142037388e-06,
+      "loss": 0.2871,
+      "step": 9363
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9158309584807425,
+      "learning_rate": 3.815177648875941e-06,
+      "loss": 0.2523,
+      "step": 9364
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.271281314895718,
+      "learning_rate": 3.81418820487488e-06,
+      "loss": 0.2864,
+      "step": 9365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6268130753348529,
+      "learning_rate": 3.813198810075255e-06,
+      "loss": 0.4973,
+      "step": 9366
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.335047972564448,
+      "learning_rate": 3.8122094645181196e-06,
+      "loss": 0.2704,
+      "step": 9367
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.566688826713257,
+      "learning_rate": 3.811220168244521e-06,
+      "loss": 0.2794,
+      "step": 9368
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3316070491854632,
+      "learning_rate": 3.8102309212955122e-06,
+      "loss": 0.271,
+      "step": 9369
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.087907916226869,
+      "learning_rate": 3.809241723712135e-06,
+      "loss": 0.2845,
+      "step": 9370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2158212205162338,
+      "learning_rate": 3.8082525755354346e-06,
+      "loss": 0.2862,
+      "step": 9371
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.623133242692415,
+      "learning_rate": 3.80726347680645e-06,
+      "loss": 0.2795,
+      "step": 9372
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.954505107845916,
+      "learning_rate": 3.8062744275662237e-06,
+      "loss": 0.2834,
+      "step": 9373
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1406731505733547,
+      "learning_rate": 3.8052854278557904e-06,
+      "loss": 0.2801,
+      "step": 9374
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.781870275502023,
+      "learning_rate": 3.8042964777161862e-06,
+      "loss": 0.279,
+      "step": 9375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0254021780680995,
+      "learning_rate": 3.8033075771884457e-06,
+      "loss": 0.2877,
+      "step": 9376
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2269111364961764,
+      "learning_rate": 3.8023187263136008e-06,
+      "loss": 0.2835,
+      "step": 9377
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5921458140843855,
+      "learning_rate": 3.8013299251326775e-06,
+      "loss": 0.4999,
+      "step": 9378
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.241316463523188,
+      "learning_rate": 3.800341173686706e-06,
+      "loss": 0.2681,
+      "step": 9379
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1789631352634182,
+      "learning_rate": 3.79935247201671e-06,
+      "loss": 0.2764,
+      "step": 9380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.931271919058685,
+      "learning_rate": 3.7983638201637124e-06,
+      "loss": 0.2557,
+      "step": 9381
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.297751134830079,
+      "learning_rate": 3.7973752181687336e-06,
+      "loss": 0.2736,
+      "step": 9382
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4650614200000651,
+      "learning_rate": 3.7963866660727918e-06,
+      "loss": 0.2677,
+      "step": 9383
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5135719897702828,
+      "learning_rate": 3.7953981639169063e-06,
+      "loss": 0.2601,
+      "step": 9384
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4514162525073404,
+      "learning_rate": 3.7944097117420913e-06,
+      "loss": 0.2797,
+      "step": 9385
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.396225022896346,
+      "learning_rate": 3.793421309589358e-06,
+      "loss": 0.2727,
+      "step": 9386
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.50773517879765,
+      "learning_rate": 3.792432957499719e-06,
+      "loss": 0.2707,
+      "step": 9387
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8732056382137225,
+      "learning_rate": 3.7914446555141805e-06,
+      "loss": 0.2551,
+      "step": 9388
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6133073243689083,
+      "learning_rate": 3.7904564036737502e-06,
+      "loss": 0.5252,
+      "step": 9389
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7486217173572725,
+      "learning_rate": 3.7894682020194296e-06,
+      "loss": 0.2639,
+      "step": 9390
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.292660267779196,
+      "learning_rate": 3.788480050592226e-06,
+      "loss": 0.2578,
+      "step": 9391
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8630531458317638,
+      "learning_rate": 3.7874919494331363e-06,
+      "loss": 0.268,
+      "step": 9392
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.87811553970909,
+      "learning_rate": 3.78650389858316e-06,
+      "loss": 0.2967,
+      "step": 9393
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5650994855817189,
+      "learning_rate": 3.7855158980832924e-06,
+      "loss": 0.2624,
+      "step": 9394
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4322746388884324,
+      "learning_rate": 3.7845279479745277e-06,
+      "loss": 0.2668,
+      "step": 9395
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.359742918628945,
+      "learning_rate": 3.7835400482978565e-06,
+      "loss": 0.2761,
+      "step": 9396
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5529438969965241,
+      "learning_rate": 3.78255219909427e-06,
+      "loss": 0.2994,
+      "step": 9397
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.203903238239566,
+      "learning_rate": 3.781564400404752e-06,
+      "loss": 0.2731,
+      "step": 9398
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.66217503373457,
+      "learning_rate": 3.780576652270295e-06,
+      "loss": 0.2969,
+      "step": 9399
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2899445082648175,
+      "learning_rate": 3.7795889547318764e-06,
+      "loss": 0.2672,
+      "step": 9400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.069135078016973,
+      "learning_rate": 3.778601307830482e-06,
+      "loss": 0.2834,
+      "step": 9401
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6697055773900062,
+      "learning_rate": 3.777613711607087e-06,
+      "loss": 0.2886,
+      "step": 9402
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8849297878403781,
+      "learning_rate": 3.776626166102672e-06,
+      "loss": 0.2937,
+      "step": 9403
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.233006061379948,
+      "learning_rate": 3.7756386713582086e-06,
+      "loss": 0.2789,
+      "step": 9404
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.848638240209906,
+      "learning_rate": 3.7746512274146707e-06,
+      "loss": 0.2647,
+      "step": 9405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6240572167892289,
+      "learning_rate": 3.773663834313031e-06,
+      "loss": 0.2654,
+      "step": 9406
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6588593622644934,
+      "learning_rate": 3.7726764920942584e-06,
+      "loss": 0.2684,
+      "step": 9407
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4118707001780626,
+      "learning_rate": 3.7716892007993166e-06,
+      "loss": 0.2892,
+      "step": 9408
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.497149809190493,
+      "learning_rate": 3.7707019604691725e-06,
+      "loss": 0.254,
+      "step": 9409
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.524310881558819,
+      "learning_rate": 3.7697147711447867e-06,
+      "loss": 0.2731,
+      "step": 9410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8771781720957033,
+      "learning_rate": 3.7687276328671215e-06,
+      "loss": 0.2852,
+      "step": 9411
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.266244121038863,
+      "learning_rate": 3.7677405456771325e-06,
+      "loss": 0.2612,
+      "step": 9412
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.647449996299619,
+      "learning_rate": 3.766753509615776e-06,
+      "loss": 0.2778,
+      "step": 9413
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6143241562048347,
+      "learning_rate": 3.7657665247240083e-06,
+      "loss": 0.2749,
+      "step": 9414
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.357830174106946,
+      "learning_rate": 3.76477959104278e-06,
+      "loss": 0.2874,
+      "step": 9415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6185836244562273,
+      "learning_rate": 3.7637927086130398e-06,
+      "loss": 0.3081,
+      "step": 9416
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9563248475437076,
+      "learning_rate": 3.762805877475737e-06,
+      "loss": 0.2863,
+      "step": 9417
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.730649850736178,
+      "learning_rate": 3.761819097671815e-06,
+      "loss": 0.2695,
+      "step": 9418
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1894437178596085,
+      "learning_rate": 3.760832369242219e-06,
+      "loss": 0.2765,
+      "step": 9419
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5623841114549686,
+      "learning_rate": 3.7598456922278855e-06,
+      "loss": 0.2895,
+      "step": 9420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4984246170610744,
+      "learning_rate": 3.75885906666976e-06,
+      "loss": 0.2524,
+      "step": 9421
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.983242143886795,
+      "learning_rate": 3.757872492608775e-06,
+      "loss": 0.2677,
+      "step": 9422
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8807899664735235,
+      "learning_rate": 3.7568859700858685e-06,
+      "loss": 0.292,
+      "step": 9423
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7306002817121755,
+      "learning_rate": 3.7558994991419694e-06,
+      "loss": 0.2591,
+      "step": 9424
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.767129311082147,
+      "learning_rate": 3.754913079818012e-06,
+      "loss": 0.3087,
+      "step": 9425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1372657968123443,
+      "learning_rate": 3.75392671215492e-06,
+      "loss": 0.2889,
+      "step": 9426
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8084331556717348,
+      "learning_rate": 3.7529403961936217e-06,
+      "loss": 0.2693,
+      "step": 9427
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3941833346339667,
+      "learning_rate": 3.751954131975042e-06,
+      "loss": 0.2737,
+      "step": 9428
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6733878253655239,
+      "learning_rate": 3.7509679195401033e-06,
+      "loss": 0.2736,
+      "step": 9429
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.475572726616073,
+      "learning_rate": 3.749981758929724e-06,
+      "loss": 0.2622,
+      "step": 9430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5976562643264937,
+      "learning_rate": 3.7489956501848214e-06,
+      "loss": 0.4603,
+      "step": 9431
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.140521257382837,
+      "learning_rate": 3.748009593346311e-06,
+      "loss": 0.2665,
+      "step": 9432
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2425698579874407,
+      "learning_rate": 3.7470235884551077e-06,
+      "loss": 0.2951,
+      "step": 9433
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6330897447653205,
+      "learning_rate": 3.74603763555212e-06,
+      "loss": 0.2741,
+      "step": 9434
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7797622857168383,
+      "learning_rate": 3.7450517346782565e-06,
+      "loss": 0.2763,
+      "step": 9435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9628850449714346,
+      "learning_rate": 3.7440658858744274e-06,
+      "loss": 0.2518,
+      "step": 9436
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.0060470636898975,
+      "learning_rate": 3.7430800891815356e-06,
+      "loss": 0.2906,
+      "step": 9437
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.486718453980096,
+      "learning_rate": 3.7420943446404835e-06,
+      "loss": 0.2733,
+      "step": 9438
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0856812294426126,
+      "learning_rate": 3.7411086522921714e-06,
+      "loss": 0.2784,
+      "step": 9439
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.869901731710971,
+      "learning_rate": 3.7401230121774966e-06,
+      "loss": 0.2764,
+      "step": 9440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.112262487038523,
+      "learning_rate": 3.739137424337357e-06,
+      "loss": 0.2543,
+      "step": 9441
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.635283562506698,
+      "learning_rate": 3.7381518888126434e-06,
+      "loss": 0.2939,
+      "step": 9442
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.366542969840427,
+      "learning_rate": 3.73716640564425e-06,
+      "loss": 0.3017,
+      "step": 9443
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4567349436988268,
+      "learning_rate": 3.736180974873065e-06,
+      "loss": 0.2892,
+      "step": 9444
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.916334200846063,
+      "learning_rate": 3.735195596539978e-06,
+      "loss": 0.2974,
+      "step": 9445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.306518722634645,
+      "learning_rate": 3.734210270685871e-06,
+      "loss": 0.2655,
+      "step": 9446
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.316169047949291,
+      "learning_rate": 3.733224997351629e-06,
+      "loss": 0.2803,
+      "step": 9447
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.0936376291264605,
+      "learning_rate": 3.7322397765781304e-06,
+      "loss": 0.2902,
+      "step": 9448
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.853870498108596,
+      "learning_rate": 3.731254608406257e-06,
+      "loss": 0.2482,
+      "step": 9449
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.200638253883915,
+      "learning_rate": 3.730269492876881e-06,
+      "loss": 0.2712,
+      "step": 9450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4939312279906427,
+      "learning_rate": 3.7292844300308808e-06,
+      "loss": 0.2799,
+      "step": 9451
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0311724594938396,
+      "learning_rate": 3.728299419909126e-06,
+      "loss": 0.281,
+      "step": 9452
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0709075397930206,
+      "learning_rate": 3.7273144625524883e-06,
+      "loss": 0.2861,
+      "step": 9453
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3837120508819816,
+      "learning_rate": 3.7263295580018326e-06,
+      "loss": 0.2674,
+      "step": 9454
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.591818620406059,
+      "learning_rate": 3.7253447062980276e-06,
+      "loss": 0.2651,
+      "step": 9455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.938159348620676,
+      "learning_rate": 3.724359907481933e-06,
+      "loss": 0.2679,
+      "step": 9456
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.725483164376001,
+      "learning_rate": 3.7233751615944115e-06,
+      "loss": 0.2777,
+      "step": 9457
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.838041008716286,
+      "learning_rate": 3.7223904686763222e-06,
+      "loss": 0.2736,
+      "step": 9458
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.664735300992799,
+      "learning_rate": 3.7214058287685228e-06,
+      "loss": 0.2697,
+      "step": 9459
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.254194483518903,
+      "learning_rate": 3.7204212419118656e-06,
+      "loss": 0.2801,
+      "step": 9460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.681588347728504,
+      "learning_rate": 3.719436708147206e-06,
+      "loss": 0.2703,
+      "step": 9461
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4553635322149021,
+      "learning_rate": 3.71845222751539e-06,
+      "loss": 0.2541,
+      "step": 9462
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7166927046328586,
+      "learning_rate": 3.7174678000572684e-06,
+      "loss": 0.2831,
+      "step": 9463
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.008685742866287,
+      "learning_rate": 3.7164834258136847e-06,
+      "loss": 0.2794,
+      "step": 9464
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4774342221177545,
+      "learning_rate": 3.7154991048254823e-06,
+      "loss": 0.2678,
+      "step": 9465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.443197475780415,
+      "learning_rate": 3.7145148371335048e-06,
+      "loss": 0.2886,
+      "step": 9466
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.467522895981137,
+      "learning_rate": 3.713530622778591e-06,
+      "loss": 0.2793,
+      "step": 9467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8569011603306602,
+      "learning_rate": 3.712546461801576e-06,
+      "loss": 0.301,
+      "step": 9468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.245168060416518,
+      "learning_rate": 3.7115623542432956e-06,
+      "loss": 0.2734,
+      "step": 9469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.025178619022494,
+      "learning_rate": 3.7105783001445806e-06,
+      "loss": 0.2636,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2278134931293527,
+      "learning_rate": 3.7095942995462636e-06,
+      "loss": 0.2647,
+      "step": 9471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1190219662361325,
+      "learning_rate": 3.7086103524891685e-06,
+      "loss": 0.2782,
+      "step": 9472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2340393525549143,
+      "learning_rate": 3.7076264590141254e-06,
+      "loss": 0.2719,
+      "step": 9473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9674372597394898,
+      "learning_rate": 3.706642619161955e-06,
+      "loss": 0.2818,
+      "step": 9474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1223506864169996,
+      "learning_rate": 3.7056588329734824e-06,
+      "loss": 0.285,
+      "step": 9475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1125003844713035,
+      "learning_rate": 3.704675100489521e-06,
+      "loss": 0.272,
+      "step": 9476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.984208235662652,
+      "learning_rate": 3.7036914217508925e-06,
+      "loss": 0.2669,
+      "step": 9477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0854133214183492,
+      "learning_rate": 3.7027077967984082e-06,
+      "loss": 0.293,
+      "step": 9478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.723829858868617,
+      "learning_rate": 3.7017242256728823e-06,
+      "loss": 0.2886,
+      "step": 9479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.370612351249497,
+      "learning_rate": 3.7007407084151227e-06,
+      "loss": 0.2492,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.740248805408432,
+      "learning_rate": 3.6997572450659393e-06,
+      "loss": 0.2811,
+      "step": 9481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5811140238654917,
+      "learning_rate": 3.6987738356661377e-06,
+      "loss": 0.4914,
+      "step": 9482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.374049072434476,
+      "learning_rate": 3.6977904802565224e-06,
+      "loss": 0.2781,
+      "step": 9483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8660932853686598,
+      "learning_rate": 3.6968071788778915e-06,
+      "loss": 0.2557,
+      "step": 9484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9796606495350515,
+      "learning_rate": 3.6958239315710467e-06,
+      "loss": 0.2699,
+      "step": 9485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.372045014189948,
+      "learning_rate": 3.6948407383767814e-06,
+      "loss": 0.2952,
+      "step": 9486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1563314601829284,
+      "learning_rate": 3.693857599335892e-06,
+      "loss": 0.2675,
+      "step": 9487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.778049127970359,
+      "learning_rate": 3.6928745144891733e-06,
+      "loss": 0.2735,
+      "step": 9488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.650237607707643,
+      "learning_rate": 3.6918914838774112e-06,
+      "loss": 0.2921,
+      "step": 9489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5079599585202224,
+      "learning_rate": 3.6909085075413944e-06,
+      "loss": 0.2912,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.219577893499021,
+      "learning_rate": 3.6899255855219103e-06,
+      "loss": 0.2749,
+      "step": 9491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6408530900837763,
+      "learning_rate": 3.688942717859739e-06,
+      "loss": 0.2782,
+      "step": 9492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8833484663984135,
+      "learning_rate": 3.687959904595665e-06,
+      "loss": 0.2658,
+      "step": 9493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4750824970868708,
+      "learning_rate": 3.686977145770464e-06,
+      "loss": 0.2938,
+      "step": 9494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5744418776824995,
+      "learning_rate": 3.6859944414249115e-06,
+      "loss": 0.2761,
+      "step": 9495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3620626512794933,
+      "learning_rate": 3.685011791599786e-06,
+      "loss": 0.2546,
+      "step": 9496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.194208138257682,
+      "learning_rate": 3.6840291963358564e-06,
+      "loss": 0.2784,
+      "step": 9497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2014091159418414,
+      "learning_rate": 3.683046655673892e-06,
+      "loss": 0.2624,
+      "step": 9498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0876981272408086,
+      "learning_rate": 3.682064169654663e-06,
+      "loss": 0.2714,
+      "step": 9499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9262082312810136,
+      "learning_rate": 3.681081738318931e-06,
+      "loss": 0.2787,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8365381110416983,
+      "learning_rate": 3.680099361707461e-06,
+      "loss": 0.2538,
+      "step": 9501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6147746977786728,
+      "learning_rate": 3.6791170398610117e-06,
+      "loss": 0.2692,
+      "step": 9502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5471520912199355,
+      "learning_rate": 3.6781347728203433e-06,
+      "loss": 0.3003,
+      "step": 9503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.770211887379422,
+      "learning_rate": 3.6771525606262106e-06,
+      "loss": 0.2785,
+      "step": 9504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.170504245820591,
+      "learning_rate": 3.676170403319369e-06,
+      "loss": 0.2679,
+      "step": 9505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6214248961645321,
+      "learning_rate": 3.6751883009405677e-06,
+      "loss": 0.2667,
+      "step": 9506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5515567298608008,
+      "learning_rate": 3.674206253530558e-06,
+      "loss": 0.2733,
+      "step": 9507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.007672022318736,
+      "learning_rate": 3.673224261130085e-06,
+      "loss": 0.259,
+      "step": 9508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1697933187202643,
+      "learning_rate": 3.6722423237798934e-06,
+      "loss": 0.2869,
+      "step": 9509
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.79572793943694,
+      "learning_rate": 3.671260441520727e-06,
+      "loss": 0.2651,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8426392257354633,
+      "learning_rate": 3.6702786143933244e-06,
+      "loss": 0.2844,
+      "step": 9511
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6975805241094903,
+      "learning_rate": 3.6692968424384246e-06,
+      "loss": 0.2553,
+      "step": 9512
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6287716843869617,
+      "learning_rate": 3.668315125696763e-06,
+      "loss": 0.2726,
+      "step": 9513
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.325283612108341,
+      "learning_rate": 3.667333464209071e-06,
+      "loss": 0.2788,
+      "step": 9514
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.702951496301354,
+      "learning_rate": 3.6663518580160816e-06,
+      "loss": 0.2747,
+      "step": 9515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.004179882663363,
+      "learning_rate": 3.665370307158521e-06,
+      "loss": 0.2756,
+      "step": 9516
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5485402154075512,
+      "learning_rate": 3.6643888116771166e-06,
+      "loss": 0.2737,
+      "step": 9517
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4038471225670404,
+      "learning_rate": 3.663407371612595e-06,
+      "loss": 0.2727,
+      "step": 9518
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7067413961004168,
+      "learning_rate": 3.6624259870056745e-06,
+      "loss": 0.2655,
+      "step": 9519
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.277148410551273,
+      "learning_rate": 3.661444657897075e-06,
+      "loss": 0.3208,
+      "step": 9520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7869677855628887,
+      "learning_rate": 3.6604633843275155e-06,
+      "loss": 0.2756,
+      "step": 9521
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4524726043992415,
+      "learning_rate": 3.6594821663377084e-06,
+      "loss": 0.2738,
+      "step": 9522
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.060214208521723,
+      "learning_rate": 3.6585010039683684e-06,
+      "loss": 0.2724,
+      "step": 9523
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8525163279935382,
+      "learning_rate": 3.6575198972602024e-06,
+      "loss": 0.2647,
+      "step": 9524
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.025691398342604,
+      "learning_rate": 3.6565388462539216e-06,
+      "loss": 0.2776,
+      "step": 9525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6410247161687948,
+      "learning_rate": 3.655557850990231e-06,
+      "loss": 0.2577,
+      "step": 9526
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0813297162239737,
+      "learning_rate": 3.6545769115098324e-06,
+      "loss": 0.2805,
+      "step": 9527
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.382361030620752,
+      "learning_rate": 3.6535960278534273e-06,
+      "loss": 0.2588,
+      "step": 9528
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6712409287031182,
+      "learning_rate": 3.6526152000617153e-06,
+      "loss": 0.2617,
+      "step": 9529
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6444792416426703,
+      "learning_rate": 3.651634428175391e-06,
+      "loss": 0.5259,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.142673954328926,
+      "learning_rate": 3.6506537122351506e-06,
+      "loss": 0.2795,
+      "step": 9531
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4073711576279704,
+      "learning_rate": 3.6496730522816826e-06,
+      "loss": 0.2696,
+      "step": 9532
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.810236169890428,
+      "learning_rate": 3.6486924483556785e-06,
+      "loss": 0.288,
+      "step": 9533
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2463740861290313,
+      "learning_rate": 3.6477119004978256e-06,
+      "loss": 0.2631,
+      "step": 9534
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6096661233310655,
+      "learning_rate": 3.646731408748808e-06,
+      "loss": 0.485,
+      "step": 9535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5564148241602025,
+      "learning_rate": 3.6457509731493067e-06,
+      "loss": 0.2662,
+      "step": 9536
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2205695796794913,
+      "learning_rate": 3.6447705937400046e-06,
+      "loss": 0.2825,
+      "step": 9537
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3763773775770396,
+      "learning_rate": 3.6437902705615767e-06,
+      "loss": 0.272,
+      "step": 9538
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6252637185109093,
+      "learning_rate": 3.642810003654699e-06,
+      "loss": 0.285,
+      "step": 9539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.600352836256971,
+      "learning_rate": 3.6418297930600453e-06,
+      "loss": 0.2809,
+      "step": 9540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8982121176160733,
+      "learning_rate": 3.6408496388182857e-06,
+      "loss": 0.2713,
+      "step": 9541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8034384724322994,
+      "learning_rate": 3.639869540970089e-06,
+      "loss": 0.261,
+      "step": 9542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.097463555484217,
+      "learning_rate": 3.6388894995561217e-06,
+      "loss": 0.2792,
+      "step": 9543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.704476350937524,
+      "learning_rate": 3.637909514617046e-06,
+      "loss": 0.2811,
+      "step": 9544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3409169881743384,
+      "learning_rate": 3.6369295861935244e-06,
+      "loss": 0.2808,
+      "step": 9545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.630134912623273,
+      "learning_rate": 3.6359497143262147e-06,
+      "loss": 0.2964,
+      "step": 9546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.659050686141471,
+      "learning_rate": 3.6349698990557726e-06,
+      "loss": 0.2805,
+      "step": 9547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6016395129545904,
+      "learning_rate": 3.6339901404228563e-06,
+      "loss": 0.2605,
+      "step": 9548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.524636413876273,
+      "learning_rate": 3.6330104384681146e-06,
+      "loss": 0.2828,
+      "step": 9549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1901263874629353,
+      "learning_rate": 3.6320307932321975e-06,
+      "loss": 0.2774,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.596512830833323,
+      "learning_rate": 3.631051204755754e-06,
+      "loss": 0.2891,
+      "step": 9551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7342481675109205,
+      "learning_rate": 3.630071673079426e-06,
+      "loss": 0.278,
+      "step": 9552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0497940911905683,
+      "learning_rate": 3.629092198243859e-06,
+      "loss": 0.277,
+      "step": 9553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.957816654030364,
+      "learning_rate": 3.6281127802896897e-06,
+      "loss": 0.2831,
+      "step": 9554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.179435272257459,
+      "learning_rate": 3.6271334192575588e-06,
+      "loss": 0.2623,
+      "step": 9555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9349456326869183,
+      "learning_rate": 3.626154115188102e-06,
+      "loss": 0.272,
+      "step": 9556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9523069537094317,
+      "learning_rate": 3.6251748681219494e-06,
+      "loss": 0.2677,
+      "step": 9557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4384347855510473,
+      "learning_rate": 3.6241956780997345e-06,
+      "loss": 0.2821,
+      "step": 9558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8262044086859577,
+      "learning_rate": 3.623216545162085e-06,
+      "loss": 0.2821,
+      "step": 9559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9774129701399263,
+      "learning_rate": 3.6222374693496252e-06,
+      "loss": 0.2746,
+      "step": 9560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2819152854959217,
+      "learning_rate": 3.621258450702982e-06,
+      "loss": 0.3035,
+      "step": 9561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9882917236724367,
+      "learning_rate": 3.620279489262772e-06,
+      "loss": 0.2748,
+      "step": 9562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.247525828893059,
+      "learning_rate": 3.6193005850696173e-06,
+      "loss": 0.2816,
+      "step": 9563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8202638992073676,
+      "learning_rate": 3.6183217381641355e-06,
+      "loss": 0.2862,
+      "step": 9564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.940772928729769,
+      "learning_rate": 3.617342948586937e-06,
+      "loss": 0.2683,
+      "step": 9565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7387636580466237,
+      "learning_rate": 3.616364216378636e-06,
+      "loss": 0.2791,
+      "step": 9566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1988999742472717,
+      "learning_rate": 3.615385541579842e-06,
+      "loss": 0.2757,
+      "step": 9567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2291481414072614,
+      "learning_rate": 3.614406924231161e-06,
+      "loss": 0.2854,
+      "step": 9568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9529787782012953,
+      "learning_rate": 3.613428364373196e-06,
+      "loss": 0.2823,
+      "step": 9569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7818943939925946,
+      "learning_rate": 3.612449862046553e-06,
+      "loss": 0.2595,
+      "step": 9570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4601553066560093,
+      "learning_rate": 3.611471417291829e-06,
+      "loss": 0.2745,
+      "step": 9571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6440577796244955,
+      "learning_rate": 3.610493030149623e-06,
+      "loss": 0.2721,
+      "step": 9572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.678286651123387,
+      "learning_rate": 3.6095147006605293e-06,
+      "loss": 0.2681,
+      "step": 9573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7834784881495682,
+      "learning_rate": 3.6085364288651394e-06,
+      "loss": 0.2722,
+      "step": 9574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.814325177259081,
+      "learning_rate": 3.6075582148040454e-06,
+      "loss": 0.2871,
+      "step": 9575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6202310688168078,
+      "learning_rate": 3.606580058517834e-06,
+      "loss": 0.5071,
+      "step": 9576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7258981426967315,
+      "learning_rate": 3.6056019600470893e-06,
+      "loss": 0.2909,
+      "step": 9577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7696629825994823,
+      "learning_rate": 3.6046239194323983e-06,
+      "loss": 0.264,
+      "step": 9578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8065173962302052,
+      "learning_rate": 3.6036459367143388e-06,
+      "loss": 0.2726,
+      "step": 9579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2862529438157462,
+      "learning_rate": 3.602668011933489e-06,
+      "loss": 0.276,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.418036462727418,
+      "learning_rate": 3.6016901451304265e-06,
+      "loss": 0.3053,
+      "step": 9581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4270508975282663,
+      "learning_rate": 3.6007123363457232e-06,
+      "loss": 0.2808,
+      "step": 9582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.536831210253516,
+      "learning_rate": 3.5997345856199506e-06,
+      "loss": 0.2839,
+      "step": 9583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9428651752459847,
+      "learning_rate": 3.5987568929936756e-06,
+      "loss": 0.2851,
+      "step": 9584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5331140393216478,
+      "learning_rate": 3.5977792585074668e-06,
+      "loss": 0.2546,
+      "step": 9585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3586872468600992,
+      "learning_rate": 3.596801682201888e-06,
+      "loss": 0.284,
+      "step": 9586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.404316042555868,
+      "learning_rate": 3.5958241641174995e-06,
+      "loss": 0.2785,
+      "step": 9587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1202416960554027,
+      "learning_rate": 3.5948467042948597e-06,
+      "loss": 0.2809,
+      "step": 9588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6431241123399385,
+      "learning_rate": 3.5938693027745276e-06,
+      "loss": 0.2767,
+      "step": 9589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3141477584046433,
+      "learning_rate": 3.5928919595970546e-06,
+      "loss": 0.2599,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.770921993973036,
+      "learning_rate": 3.5919146748029923e-06,
+      "loss": 0.2676,
+      "step": 9591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6714197358579923,
+      "learning_rate": 3.590937448432893e-06,
+      "loss": 0.2825,
+      "step": 9592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5364436102142878,
+      "learning_rate": 3.589960280527301e-06,
+      "loss": 0.278,
+      "step": 9593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.229902802414021,
+      "learning_rate": 3.5889831711267618e-06,
+      "loss": 0.2728,
+      "step": 9594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7823816207742216,
+      "learning_rate": 3.5880061202718175e-06,
+      "loss": 0.2687,
+      "step": 9595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.947545949653769,
+      "learning_rate": 3.587029128003006e-06,
+      "loss": 0.2854,
+      "step": 9596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9485674708226834,
+      "learning_rate": 3.5860521943608672e-06,
+      "loss": 0.2777,
+      "step": 9597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4068436711392398,
+      "learning_rate": 3.5850753193859333e-06,
+      "loss": 0.2802,
+      "step": 9598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1895938060414317,
+      "learning_rate": 3.584098503118737e-06,
+      "loss": 0.2553,
+      "step": 9599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.440780159511413,
+      "learning_rate": 3.5831217455998103e-06,
+      "loss": 0.2854,
+      "step": 9600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7029271831529336,
+      "learning_rate": 3.5821450468696772e-06,
+      "loss": 0.2564,
+      "step": 9601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.233549206608332,
+      "learning_rate": 3.581168406968867e-06,
+      "loss": 0.2702,
+      "step": 9602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5303859237555177,
+      "learning_rate": 3.5801918259378975e-06,
+      "loss": 0.2618,
+      "step": 9603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.476267417080039,
+      "learning_rate": 3.5792153038172916e-06,
+      "loss": 0.2846,
+      "step": 9604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.717802952989295,
+      "learning_rate": 3.5782388406475675e-06,
+      "loss": 0.2742,
+      "step": 9605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9311225811939217,
+      "learning_rate": 3.577262436469237e-06,
+      "loss": 0.2551,
+      "step": 9606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5687961319052377,
+      "learning_rate": 3.5762860913228157e-06,
+      "loss": 0.2778,
+      "step": 9607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.662154331815429,
+      "learning_rate": 3.575309805248815e-06,
+      "loss": 0.2594,
+      "step": 9608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.770985071755837,
+      "learning_rate": 3.57433357828774e-06,
+      "loss": 0.276,
+      "step": 9609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9560992930775103,
+      "learning_rate": 3.5733574104800973e-06,
+      "loss": 0.2923,
+      "step": 9610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.889595933118841,
+      "learning_rate": 3.5723813018663895e-06,
+      "loss": 0.2866,
+      "step": 9611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8200372092854165,
+      "learning_rate": 3.5714052524871166e-06,
+      "loss": 0.2617,
+      "step": 9612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.854558588932161,
+      "learning_rate": 3.5704292623827786e-06,
+      "loss": 0.2647,
+      "step": 9613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.259319595869737,
+      "learning_rate": 3.569453331593867e-06,
+      "loss": 0.294,
+      "step": 9614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.326140397985674,
+      "learning_rate": 3.5684774601608797e-06,
+      "loss": 0.2748,
+      "step": 9615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.686550438990365,
+      "learning_rate": 3.5675016481243065e-06,
+      "loss": 0.273,
+      "step": 9616
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8149565447399794,
+      "learning_rate": 3.566525895524632e-06,
+      "loss": 0.2666,
+      "step": 9617
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.734911669419096,
+      "learning_rate": 3.5655502024023447e-06,
+      "loss": 0.2823,
+      "step": 9618
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.464908315509748,
+      "learning_rate": 3.564574568797928e-06,
+      "loss": 0.267,
+      "step": 9619
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6175812496529116,
+      "learning_rate": 3.5635989947518625e-06,
+      "loss": 0.2653,
+      "step": 9620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6728828239123588,
+      "learning_rate": 3.562623480304623e-06,
+      "loss": 0.2904,
+      "step": 9621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.739819967402756,
+      "learning_rate": 3.5616480254966925e-06,
+      "loss": 0.2685,
+      "step": 9622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.673540114471451,
+      "learning_rate": 3.5606726303685378e-06,
+      "loss": 0.2773,
+      "step": 9623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8132328684780588,
+      "learning_rate": 3.5596972949606333e-06,
+      "loss": 0.2837,
+      "step": 9624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9380952332327748,
+      "learning_rate": 3.5587220193134457e-06,
+      "loss": 0.2864,
+      "step": 9625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7651821095245928,
+      "learning_rate": 3.5577468034674414e-06,
+      "loss": 0.2706,
+      "step": 9626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.231627056763443,
+      "learning_rate": 3.5567716474630853e-06,
+      "loss": 0.2685,
+      "step": 9627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6114014750004175,
+      "learning_rate": 3.5557965513408353e-06,
+      "loss": 0.2595,
+      "step": 9628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.01174626091983,
+      "learning_rate": 3.554821515141151e-06,
+      "loss": 0.2746,
+      "step": 9629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.084811627206056,
+      "learning_rate": 3.553846538904491e-06,
+      "loss": 0.2602,
+      "step": 9630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.415620312522918,
+      "learning_rate": 3.5528716226713055e-06,
+      "loss": 0.2632,
+      "step": 9631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1303684915519185,
+      "learning_rate": 3.551896766482048e-06,
+      "loss": 0.2802,
+      "step": 9632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.593530392158075,
+      "learning_rate": 3.5509219703771647e-06,
+      "loss": 0.2547,
+      "step": 9633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.328479410567316,
+      "learning_rate": 3.5499472343971027e-06,
+      "loss": 0.2509,
+      "step": 9634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.704078756234964,
+      "learning_rate": 3.5489725585823064e-06,
+      "loss": 0.2883,
+      "step": 9635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.634038761388328,
+      "learning_rate": 3.5479979429732134e-06,
+      "loss": 0.2686,
+      "step": 9636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1474182986777586,
+      "learning_rate": 3.5470233876102672e-06,
+      "loss": 0.2815,
+      "step": 9637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6120178257820572,
+      "learning_rate": 3.5460488925339013e-06,
+      "loss": 0.2823,
+      "step": 9638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6475636322091867,
+      "learning_rate": 3.545074457784549e-06,
+      "loss": 0.2595,
+      "step": 9639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5034596261966902,
+      "learning_rate": 3.5441000834026427e-06,
+      "loss": 0.2702,
+      "step": 9640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7355213612859415,
+      "learning_rate": 3.543125769428608e-06,
+      "loss": 0.2643,
+      "step": 9641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7340420068099016,
+      "learning_rate": 3.542151515902874e-06,
+      "loss": 0.2674,
+      "step": 9642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8024617141733321,
+      "learning_rate": 3.5411773228658635e-06,
+      "loss": 0.2716,
+      "step": 9643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.428585854706941,
+      "learning_rate": 3.5402031903579946e-06,
+      "loss": 0.2707,
+      "step": 9644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9927899800573674,
+      "learning_rate": 3.5392291184196903e-06,
+      "loss": 0.2814,
+      "step": 9645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8924978471352056,
+      "learning_rate": 3.5382551070913647e-06,
+      "loss": 0.2616,
+      "step": 9646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9397480744608648,
+      "learning_rate": 3.5372811564134303e-06,
+      "loss": 0.2586,
+      "step": 9647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2187808600142427,
+      "learning_rate": 3.5363072664263e-06,
+      "loss": 0.2868,
+      "step": 9648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.965577050260435,
+      "learning_rate": 3.5353334371703797e-06,
+      "loss": 0.2712,
+      "step": 9649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.48875613701296,
+      "learning_rate": 3.5343596686860757e-06,
+      "loss": 0.274,
+      "step": 9650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.033883317329177,
+      "learning_rate": 3.533385961013793e-06,
+      "loss": 0.2754,
+      "step": 9651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7437274630426374,
+      "learning_rate": 3.5324123141939327e-06,
+      "loss": 0.2636,
+      "step": 9652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.274467485172925,
+      "learning_rate": 3.5314387282668917e-06,
+      "loss": 0.2636,
+      "step": 9653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2679854576901377,
+      "learning_rate": 3.530465203273067e-06,
+      "loss": 0.2601,
+      "step": 9654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7316515382461843,
+      "learning_rate": 3.5294917392528504e-06,
+      "loss": 0.2671,
+      "step": 9655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.906097991358715,
+      "learning_rate": 3.5285183362466346e-06,
+      "loss": 0.2932,
+      "step": 9656
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4206860044962317,
+      "learning_rate": 3.527544994294806e-06,
+      "loss": 0.2917,
+      "step": 9657
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8644289231604958,
+      "learning_rate": 3.5265717134377496e-06,
+      "loss": 0.3008,
+      "step": 9658
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.173226564855597,
+      "learning_rate": 3.5255984937158505e-06,
+      "loss": 0.2708,
+      "step": 9659
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0491998347676033,
+      "learning_rate": 3.52462533516949e-06,
+      "loss": 0.2832,
+      "step": 9660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1631596140764504,
+      "learning_rate": 3.523652237839044e-06,
+      "loss": 0.2863,
+      "step": 9661
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.586519950366805,
+      "learning_rate": 3.52267920176489e-06,
+      "loss": 0.2811,
+      "step": 9662
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7611328724501245,
+      "learning_rate": 3.5217062269873986e-06,
+      "loss": 0.2673,
+      "step": 9663
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0337344588209123,
+      "learning_rate": 3.520733313546942e-06,
+      "loss": 0.2796,
+      "step": 9664
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3318295373856515,
+      "learning_rate": 3.519760461483888e-06,
+      "loss": 0.2879,
+      "step": 9665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1105205949953616,
+      "learning_rate": 3.5187876708386004e-06,
+      "loss": 0.2802,
+      "step": 9666
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4433968414646754,
+      "learning_rate": 3.517814941651444e-06,
+      "loss": 0.2717,
+      "step": 9667
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3791071439894624,
+      "learning_rate": 3.5168422739627794e-06,
+      "loss": 0.2735,
+      "step": 9668
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4088311824071043,
+      "learning_rate": 3.515869667812962e-06,
+      "loss": 0.2816,
+      "step": 9669
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6487354639853633,
+      "learning_rate": 3.5148971232423495e-06,
+      "loss": 0.2695,
+      "step": 9670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.471460897138366,
+      "learning_rate": 3.513924640291292e-06,
+      "loss": 0.2724,
+      "step": 9671
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8466820640785455,
+      "learning_rate": 3.5129522190001407e-06,
+      "loss": 0.2717,
+      "step": 9672
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8338084218164097,
+      "learning_rate": 3.5119798594092426e-06,
+      "loss": 0.2607,
+      "step": 9673
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.141658710673713,
+      "learning_rate": 3.5110075615589445e-06,
+      "loss": 0.2931,
+      "step": 9674
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.422081082535986,
+      "learning_rate": 3.5100353254895867e-06,
+      "loss": 0.2635,
+      "step": 9675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8453361788018223,
+      "learning_rate": 3.509063151241511e-06,
+      "loss": 0.2673,
+      "step": 9676
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.836042848355021,
+      "learning_rate": 3.508091038855052e-06,
+      "loss": 0.2799,
+      "step": 9677
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6202018431544734,
+      "learning_rate": 3.507118988370547e-06,
+      "loss": 0.2726,
+      "step": 9678
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5788904459655787,
+      "learning_rate": 3.5061469998283255e-06,
+      "loss": 0.4571,
+      "step": 9679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5411428286339852,
+      "learning_rate": 3.5051750732687183e-06,
+      "loss": 0.2745,
+      "step": 9680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6994969809747569,
+      "learning_rate": 3.504203208732052e-06,
+      "loss": 0.2718,
+      "step": 9681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6459733915683215,
+      "learning_rate": 3.5032314062586525e-06,
+      "loss": 0.2885,
+      "step": 9682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7305200495893007,
+      "learning_rate": 3.5022596658888404e-06,
+      "loss": 0.2721,
+      "step": 9683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9519146830725447,
+      "learning_rate": 3.501287987662936e-06,
+      "loss": 0.2606,
+      "step": 9684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.86565487359643,
+      "learning_rate": 3.500316371621253e-06,
+      "loss": 0.2788,
+      "step": 9685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6391418581472857,
+      "learning_rate": 3.4993448178041095e-06,
+      "loss": 0.4783,
+      "step": 9686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.99616721580779,
+      "learning_rate": 3.498373326251814e-06,
+      "loss": 0.2875,
+      "step": 9687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8719292714633617,
+      "learning_rate": 3.4974018970046745e-06,
+      "loss": 0.283,
+      "step": 9688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9315024307023858,
+      "learning_rate": 3.496430530103001e-06,
+      "loss": 0.274,
+      "step": 9689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.725138300741341,
+      "learning_rate": 3.4954592255870964e-06,
+      "loss": 0.2675,
+      "step": 9690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.989168976277709,
+      "learning_rate": 3.4944879834972595e-06,
+      "loss": 0.2755,
+      "step": 9691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5705142104128255,
+      "learning_rate": 3.493516803873791e-06,
+      "loss": 0.4812,
+      "step": 9692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.558959175112957,
+      "learning_rate": 3.492545686756986e-06,
+      "loss": 0.2681,
+      "step": 9693
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7260435908193597,
+      "learning_rate": 3.4915746321871384e-06,
+      "loss": 0.2738,
+      "step": 9694
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9122895850254102,
+      "learning_rate": 3.490603640204538e-06,
+      "loss": 0.2647,
+      "step": 9695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7505762332473782,
+      "learning_rate": 3.4896327108494723e-06,
+      "loss": 0.2547,
+      "step": 9696
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4773745456442398,
+      "learning_rate": 3.488661844162229e-06,
+      "loss": 0.2824,
+      "step": 9697
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4595874996958724,
+      "learning_rate": 3.487691040183091e-06,
+      "loss": 0.2674,
+      "step": 9698
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7705851454020305,
+      "learning_rate": 3.4867202989523375e-06,
+      "loss": 0.2934,
+      "step": 9699
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9124997415924883,
+      "learning_rate": 3.4857496205102475e-06,
+      "loss": 0.281,
+      "step": 9700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.965828772178735,
+      "learning_rate": 3.484779004897094e-06,
+      "loss": 0.2715,
+      "step": 9701
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1927954447603812,
+      "learning_rate": 3.4838084521531513e-06,
+      "loss": 0.268,
+      "step": 9702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.349569070821062,
+      "learning_rate": 3.4828379623186883e-06,
+      "loss": 0.2792,
+      "step": 9703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1068968748909445,
+      "learning_rate": 3.481867535433974e-06,
+      "loss": 0.2952,
+      "step": 9704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.076498174164943,
+      "learning_rate": 3.4808971715392724e-06,
+      "loss": 0.2959,
+      "step": 9705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3310854290572403,
+      "learning_rate": 3.4799268706748457e-06,
+      "loss": 0.2836,
+      "step": 9706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.888579570534031,
+      "learning_rate": 3.4789566328809523e-06,
+      "loss": 0.263,
+      "step": 9707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.048864209009855,
+      "learning_rate": 3.4779864581978506e-06,
+      "loss": 0.2666,
+      "step": 9708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.963110769760852,
+      "learning_rate": 3.477016346665793e-06,
+      "loss": 0.2819,
+      "step": 9709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2173594046331218,
+      "learning_rate": 3.4760462983250327e-06,
+      "loss": 0.2682,
+      "step": 9710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6566223404374978,
+      "learning_rate": 3.475076313215817e-06,
+      "loss": 0.2577,
+      "step": 9711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.672514898123288,
+      "learning_rate": 3.4741063913783958e-06,
+      "loss": 0.2884,
+      "step": 9712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.78725750672085,
+      "learning_rate": 3.47313653285301e-06,
+      "loss": 0.2536,
+      "step": 9713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9106984866986485,
+      "learning_rate": 3.4721667376799022e-06,
+      "loss": 0.2626,
+      "step": 9714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8114622629790127,
+      "learning_rate": 3.4711970058993092e-06,
+      "loss": 0.2593,
+      "step": 9715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.947573439174941,
+      "learning_rate": 3.4702273375514695e-06,
+      "loss": 0.3067,
+      "step": 9716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0173391992265914,
+      "learning_rate": 3.469257732676613e-06,
+      "loss": 0.2655,
+      "step": 9717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6964881703085593,
+      "learning_rate": 3.468288191314971e-06,
+      "loss": 0.2756,
+      "step": 9718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.87439735988396,
+      "learning_rate": 3.4673187135067737e-06,
+      "loss": 0.2887,
+      "step": 9719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6457496674905365,
+      "learning_rate": 3.4663492992922467e-06,
+      "loss": 0.2737,
+      "step": 9720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.40900693369108,
+      "learning_rate": 3.4653799487116097e-06,
+      "loss": 0.3005,
+      "step": 9721
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9989324760989478,
+      "learning_rate": 3.464410661805086e-06,
+      "loss": 0.2814,
+      "step": 9722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.24778928134871,
+      "learning_rate": 3.46344143861289e-06,
+      "loss": 0.2707,
+      "step": 9723
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.873259969543035,
+      "learning_rate": 3.4624722791752395e-06,
+      "loss": 0.2812,
+      "step": 9724
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9071851397361037,
+      "learning_rate": 3.4615031835323437e-06,
+      "loss": 0.2688,
+      "step": 9725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.747612671781315,
+      "learning_rate": 3.4605341517244123e-06,
+      "loss": 0.2671,
+      "step": 9726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.549920079770489,
+      "learning_rate": 3.4595651837916542e-06,
+      "loss": 0.2821,
+      "step": 9727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.629429335854162,
+      "learning_rate": 3.4585962797742745e-06,
+      "loss": 0.2571,
+      "step": 9728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3913865108676804,
+      "learning_rate": 3.457627439712472e-06,
+      "loss": 0.2669,
+      "step": 9729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.973275614830469,
+      "learning_rate": 3.456658663646447e-06,
+      "loss": 0.2744,
+      "step": 9730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1536431811532277,
+      "learning_rate": 3.4556899516163942e-06,
+      "loss": 0.2558,
+      "step": 9731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.173750887599849,
+      "learning_rate": 3.4547213036625107e-06,
+      "loss": 0.2679,
+      "step": 9732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.444293839107353,
+      "learning_rate": 3.4537527198249804e-06,
+      "loss": 0.2827,
+      "step": 9733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1970971103190347,
+      "learning_rate": 3.452784200144e-06,
+      "loss": 0.2825,
+      "step": 9734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0197608298174905,
+      "learning_rate": 3.45181574465975e-06,
+      "loss": 0.2738,
+      "step": 9735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.586690916293952,
+      "learning_rate": 3.4508473534124155e-06,
+      "loss": 0.2553,
+      "step": 9736
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.221204728485687,
+      "learning_rate": 3.4498790264421755e-06,
+      "loss": 0.2999,
+      "step": 9737
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.562557774905292,
+      "learning_rate": 3.4489107637892084e-06,
+      "loss": 0.272,
+      "step": 9738
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.060300648749852,
+      "learning_rate": 3.4479425654936883e-06,
+      "loss": 0.2861,
+      "step": 9739
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.095010236330334,
+      "learning_rate": 3.4469744315957863e-06,
+      "loss": 0.2723,
+      "step": 9740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1483076401567973,
+      "learning_rate": 3.4460063621356753e-06,
+      "loss": 0.2575,
+      "step": 9741
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8839660452427271,
+      "learning_rate": 3.445038357153521e-06,
+      "loss": 0.2767,
+      "step": 9742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5896504501694708,
+      "learning_rate": 3.4440704166894865e-06,
+      "loss": 0.2677,
+      "step": 9743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.948613565615765,
+      "learning_rate": 3.443102540783735e-06,
+      "loss": 0.2889,
+      "step": 9744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.870423470052233,
+      "learning_rate": 3.4421347294764236e-06,
+      "loss": 0.2557,
+      "step": 9745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.399463870416707,
+      "learning_rate": 3.44116698280771e-06,
+      "loss": 0.2863,
+      "step": 9746
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.383572031556528,
+      "learning_rate": 3.440199300817746e-06,
+      "loss": 0.2531,
+      "step": 9747
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3395363548219366,
+      "learning_rate": 3.4392316835466834e-06,
+      "loss": 0.2646,
+      "step": 9748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4901782868002784,
+      "learning_rate": 3.4382641310346705e-06,
+      "loss": 0.2628,
+      "step": 9749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6094170931302001,
+      "learning_rate": 3.437296643321854e-06,
+      "loss": 0.4974,
+      "step": 9750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5547683876758882,
+      "learning_rate": 3.4363292204483745e-06,
+      "loss": 0.2522,
+      "step": 9751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.152238272698963,
+      "learning_rate": 3.435361862454374e-06,
+      "loss": 0.2768,
+      "step": 9752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.340303677330552,
+      "learning_rate": 3.4343945693799885e-06,
+      "loss": 0.313,
+      "step": 9753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.167579734046241,
+      "learning_rate": 3.4334273412653534e-06,
+      "loss": 0.2723,
+      "step": 9754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9885965037663498,
+      "learning_rate": 3.4324601781505973e-06,
+      "loss": 0.2588,
+      "step": 9755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8703308316835587,
+      "learning_rate": 3.431493080075856e-06,
+      "loss": 0.2655,
+      "step": 9756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.507613676412212,
+      "learning_rate": 3.4305260470812522e-06,
+      "loss": 0.2658,
+      "step": 9757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.202107476049587,
+      "learning_rate": 3.429559079206911e-06,
+      "loss": 0.2546,
+      "step": 9758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5256971890429445,
+      "learning_rate": 3.428592176492952e-06,
+      "loss": 0.2929,
+      "step": 9759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6158143000365329,
+      "learning_rate": 3.427625338979496e-06,
+      "loss": 0.4779,
+      "step": 9760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7972835015970774,
+      "learning_rate": 3.4266585667066567e-06,
+      "loss": 0.2903,
+      "step": 9761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0395101630351227,
+      "learning_rate": 3.4256918597145484e-06,
+      "loss": 0.2757,
+      "step": 9762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.310466766664242,
+      "learning_rate": 3.424725218043279e-06,
+      "loss": 0.274,
+      "step": 9763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3011208309822484,
+      "learning_rate": 3.4237586417329604e-06,
+      "loss": 0.2813,
+      "step": 9764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5851888721427545,
+      "learning_rate": 3.4227921308236943e-06,
+      "loss": 0.2658,
+      "step": 9765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4594726188202216,
+      "learning_rate": 3.4218256853555864e-06,
+      "loss": 0.2666,
+      "step": 9766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.282973661879234,
+      "learning_rate": 3.4208593053687323e-06,
+      "loss": 0.263,
+      "step": 9767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.229247173957106,
+      "learning_rate": 3.419892990903231e-06,
+      "loss": 0.2613,
+      "step": 9768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8265981069370543,
+      "learning_rate": 3.4189267419991756e-06,
+      "loss": 0.2859,
+      "step": 9769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6153322422561114,
+      "learning_rate": 3.417960558696657e-06,
+      "loss": 0.2768,
+      "step": 9770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0136087141017507,
+      "learning_rate": 3.4169944410357657e-06,
+      "loss": 0.285,
+      "step": 9771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0237742417858455,
+      "learning_rate": 3.4160283890565877e-06,
+      "loss": 0.2685,
+      "step": 9772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7441269897520957,
+      "learning_rate": 3.4150624027992043e-06,
+      "loss": 0.2813,
+      "step": 9773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6276477488264848,
+      "learning_rate": 3.4140964823036984e-06,
+      "loss": 0.5328,
+      "step": 9774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.180546889951866,
+      "learning_rate": 3.413130627610145e-06,
+      "loss": 0.2735,
+      "step": 9775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.752192174421867,
+      "learning_rate": 3.4121648387586216e-06,
+      "loss": 0.2816,
+      "step": 9776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0246969590769095,
+      "learning_rate": 3.411199115789198e-06,
+      "loss": 0.2931,
+      "step": 9777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.73087687941042,
+      "learning_rate": 3.4102334587419437e-06,
+      "loss": 0.2611,
+      "step": 9778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5906763656306007,
+      "learning_rate": 3.4092678676569292e-06,
+      "loss": 0.2669,
+      "step": 9779
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.366577653114579,
+      "learning_rate": 3.408302342574216e-06,
+      "loss": 0.2923,
+      "step": 9780
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.000430447468467,
+      "learning_rate": 3.407336883533866e-06,
+      "loss": 0.282,
+      "step": 9781
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4360295371800174,
+      "learning_rate": 3.406371490575938e-06,
+      "loss": 0.2631,
+      "step": 9782
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5623069510926189,
+      "learning_rate": 3.405406163740487e-06,
+      "loss": 0.4546,
+      "step": 9783
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6064021606481163,
+      "learning_rate": 3.4044409030675663e-06,
+      "loss": 0.2704,
+      "step": 9784
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2011488028748056,
+      "learning_rate": 3.403475708597225e-06,
+      "loss": 0.2574,
+      "step": 9785
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1070165081930075,
+      "learning_rate": 3.402510580369514e-06,
+      "loss": 0.2781,
+      "step": 9786
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.156579431733777,
+      "learning_rate": 3.4015455184244763e-06,
+      "loss": 0.2719,
+      "step": 9787
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.668771498571321,
+      "learning_rate": 3.400580522802155e-06,
+      "loss": 0.2711,
+      "step": 9788
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9583571969718583,
+      "learning_rate": 3.399615593542588e-06,
+      "loss": 0.2656,
+      "step": 9789
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 14.246863501708958,
+      "learning_rate": 3.398650730685813e-06,
+      "loss": 0.2775,
+      "step": 9790
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3448309348479337,
+      "learning_rate": 3.3976859342718625e-06,
+      "loss": 0.2746,
+      "step": 9791
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9250642633814026,
+      "learning_rate": 3.3967212043407705e-06,
+      "loss": 0.2709,
+      "step": 9792
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1228204801377535,
+      "learning_rate": 3.39575654093256e-06,
+      "loss": 0.2697,
+      "step": 9793
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.445133103556168,
+      "learning_rate": 3.3947919440872627e-06,
+      "loss": 0.289,
+      "step": 9794
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.874824749896127,
+      "learning_rate": 3.3938274138448982e-06,
+      "loss": 0.2524,
+      "step": 9795
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5459995597759115,
+      "learning_rate": 3.392862950245488e-06,
+      "loss": 0.2621,
+      "step": 9796
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6170387526796945,
+      "learning_rate": 3.3918985533290484e-06,
+      "loss": 0.2795,
+      "step": 9797
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.712002999552004,
+      "learning_rate": 3.390934223135594e-06,
+      "loss": 0.278,
+      "step": 9798
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.586512750861614,
+      "learning_rate": 3.389969959705136e-06,
+      "loss": 0.2934,
+      "step": 9799
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2440423120988036,
+      "learning_rate": 3.3890057630776834e-06,
+      "loss": 0.2631,
+      "step": 9800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.127058916272636,
+      "learning_rate": 3.3880416332932442e-06,
+      "loss": 0.2785,
+      "step": 9801
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.004798853258625,
+      "learning_rate": 3.3870775703918212e-06,
+      "loss": 0.2769,
+      "step": 9802
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5113578287649134,
+      "learning_rate": 3.3861135744134143e-06,
+      "loss": 0.2554,
+      "step": 9803
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3826156861765981,
+      "learning_rate": 3.3851496453980225e-06,
+      "loss": 0.2639,
+      "step": 9804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.885710051223053,
+      "learning_rate": 3.3841857833856396e-06,
+      "loss": 0.2786,
+      "step": 9805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0955458839006647,
+      "learning_rate": 3.383221988416259e-06,
+      "loss": 0.2697,
+      "step": 9806
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7450305018461862,
+      "learning_rate": 3.382258260529869e-06,
+      "loss": 0.2817,
+      "step": 9807
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.645270739428072,
+      "learning_rate": 3.381294599766456e-06,
+      "loss": 0.2641,
+      "step": 9808
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6665386253635939,
+      "learning_rate": 3.3803310061660067e-06,
+      "loss": 0.2734,
+      "step": 9809
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.422786529716522,
+      "learning_rate": 3.3793674797685016e-06,
+      "loss": 0.2616,
+      "step": 9810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.273669920485398,
+      "learning_rate": 3.3784040206139175e-06,
+      "loss": 0.2683,
+      "step": 9811
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0069537499235546,
+      "learning_rate": 3.3774406287422313e-06,
+      "loss": 0.2736,
+      "step": 9812
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5082467590142463,
+      "learning_rate": 3.376477304193415e-06,
+      "loss": 0.2646,
+      "step": 9813
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.558642561368121,
+      "learning_rate": 3.37551404700744e-06,
+      "loss": 0.2674,
+      "step": 9814
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.327870894682449,
+      "learning_rate": 3.374550857224269e-06,
+      "loss": 0.265,
+      "step": 9815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.69322909734235,
+      "learning_rate": 3.3735877348838738e-06,
+      "loss": 0.2586,
+      "step": 9816
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9039186857196588,
+      "learning_rate": 3.372624680026211e-06,
+      "loss": 0.2621,
+      "step": 9817
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.720366894920928,
+      "learning_rate": 3.3716616926912414e-06,
+      "loss": 0.2874,
+      "step": 9818
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9363342186241619,
+      "learning_rate": 3.37069877291892e-06,
+      "loss": 0.2884,
+      "step": 9819
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9685218744187376,
+      "learning_rate": 3.369735920749201e-06,
+      "loss": 0.2739,
+      "step": 9820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8789735450524714,
+      "learning_rate": 3.3687731362220334e-06,
+      "loss": 0.2632,
+      "step": 9821
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.96679302862232,
+      "learning_rate": 3.3678104193773654e-06,
+      "loss": 0.2771,
+      "step": 9822
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6051793016702407,
+      "learning_rate": 3.366847770255143e-06,
+      "loss": 0.268,
+      "step": 9823
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.357721907351094,
+      "learning_rate": 3.3658851888953078e-06,
+      "loss": 0.2678,
+      "step": 9824
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5222149364058855,
+      "learning_rate": 3.364922675337798e-06,
+      "loss": 0.2628,
+      "step": 9825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.32973395785985,
+      "learning_rate": 3.3639602296225514e-06,
+      "loss": 0.2764,
+      "step": 9826
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6085685513684953,
+      "learning_rate": 3.3629978517895e-06,
+      "loss": 0.2667,
+      "step": 9827
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9163977237106014,
+      "learning_rate": 3.3620355418785767e-06,
+      "loss": 0.2632,
+      "step": 9828
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5173020236245705,
+      "learning_rate": 3.361073299929707e-06,
+      "loss": 0.2714,
+      "step": 9829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3939246286157458,
+      "learning_rate": 3.3601111259828155e-06,
+      "loss": 0.2841,
+      "step": 9830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6786020369777017,
+      "learning_rate": 3.3591490200778276e-06,
+      "loss": 0.4974,
+      "step": 9831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.85122913637047,
+      "learning_rate": 3.358186982254662e-06,
+      "loss": 0.27,
+      "step": 9832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.2280408651590315,
+      "learning_rate": 3.357225012553234e-06,
+      "loss": 0.2727,
+      "step": 9833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.391669393244888,
+      "learning_rate": 3.356263111013459e-06,
+      "loss": 0.2653,
+      "step": 9834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.045810554509706,
+      "learning_rate": 3.3553012776752468e-06,
+      "loss": 0.2593,
+      "step": 9835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2256173942111057,
+      "learning_rate": 3.354339512578506e-06,
+      "loss": 0.2686,
+      "step": 9836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6763608588122259,
+      "learning_rate": 3.3533778157631403e-06,
+      "loss": 0.28,
+      "step": 9837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.97164642342639,
+      "learning_rate": 3.3524161872690548e-06,
+      "loss": 0.2568,
+      "step": 9838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.3938486688541465,
+      "learning_rate": 3.351454627136148e-06,
+      "loss": 0.2994,
+      "step": 9839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.146615352814978,
+      "learning_rate": 3.350493135404318e-06,
+      "loss": 0.2748,
+      "step": 9840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1725227018700743,
+      "learning_rate": 3.3495317121134564e-06,
+      "loss": 0.2771,
+      "step": 9841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.92116124176794,
+      "learning_rate": 3.3485703573034567e-06,
+      "loss": 0.2912,
+      "step": 9842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.178429621380662,
+      "learning_rate": 3.347609071014205e-06,
+      "loss": 0.2731,
+      "step": 9843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.61580037997173,
+      "learning_rate": 3.3466478532855896e-06,
+      "loss": 0.2633,
+      "step": 9844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3181438531650134,
+      "learning_rate": 3.3456867041574893e-06,
+      "loss": 0.2793,
+      "step": 9845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6170399567764744,
+      "learning_rate": 3.344725623669786e-06,
+      "loss": 0.2638,
+      "step": 9846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7100224589147104,
+      "learning_rate": 3.343764611862358e-06,
+      "loss": 0.281,
+      "step": 9847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.716361876494082,
+      "learning_rate": 3.342803668775078e-06,
+      "loss": 0.2834,
+      "step": 9848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.649937776220787,
+      "learning_rate": 3.341842794447817e-06,
+      "loss": 0.2747,
+      "step": 9849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6239701754106592,
+      "learning_rate": 3.340881988920445e-06,
+      "loss": 0.2655,
+      "step": 9850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6416178897046088,
+      "learning_rate": 3.339921252232824e-06,
+      "loss": 0.2581,
+      "step": 9851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0418941412002503,
+      "learning_rate": 3.3389605844248187e-06,
+      "loss": 0.2735,
+      "step": 9852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.894837418888004,
+      "learning_rate": 3.33799998553629e-06,
+      "loss": 0.296,
+      "step": 9853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.081975530778152,
+      "learning_rate": 3.337039455607094e-06,
+      "loss": 0.2669,
+      "step": 9854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.358622137777047,
+      "learning_rate": 3.336078994677085e-06,
+      "loss": 0.2616,
+      "step": 9855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0191263319974504,
+      "learning_rate": 3.3351186027861147e-06,
+      "loss": 0.2834,
+      "step": 9856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.739503391061953,
+      "learning_rate": 3.334158279974029e-06,
+      "loss": 0.2865,
+      "step": 9857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.83563435733747,
+      "learning_rate": 3.3331980262806774e-06,
+      "loss": 0.268,
+      "step": 9858
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.471785915960166,
+      "learning_rate": 3.3322378417458985e-06,
+      "loss": 0.2512,
+      "step": 9859
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.12319852461157,
+      "learning_rate": 3.3312777264095325e-06,
+      "loss": 0.2789,
+      "step": 9860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.603471809159579,
+      "learning_rate": 3.330317680311419e-06,
+      "loss": 0.4489,
+      "step": 9861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.621121359786358,
+      "learning_rate": 3.3293577034913916e-06,
+      "loss": 0.5016,
+      "step": 9862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.979352789753031,
+      "learning_rate": 3.3283977959892798e-06,
+      "loss": 0.2651,
+      "step": 9863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6987291839572347,
+      "learning_rate": 3.3274379578449133e-06,
+      "loss": 0.2607,
+      "step": 9864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9402783902988476,
+      "learning_rate": 3.326478189098116e-06,
+      "loss": 0.2928,
+      "step": 9865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6762661068517062,
+      "learning_rate": 3.3255184897887116e-06,
+      "loss": 0.266,
+      "step": 9866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.020978401809176,
+      "learning_rate": 3.324558859956517e-06,
+      "loss": 0.2607,
+      "step": 9867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4115015915816638,
+      "learning_rate": 3.323599299641353e-06,
+      "loss": 0.2877,
+      "step": 9868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.478799202552657,
+      "learning_rate": 3.3226398088830316e-06,
+      "loss": 0.268,
+      "step": 9869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.805467424554952,
+      "learning_rate": 3.321680387721364e-06,
+      "loss": 0.2557,
+      "step": 9870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4345762879435267,
+      "learning_rate": 3.3207210361961583e-06,
+      "loss": 0.2633,
+      "step": 9871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8915575667164246,
+      "learning_rate": 3.3197617543472203e-06,
+      "loss": 0.2677,
+      "step": 9872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.935309816684142,
+      "learning_rate": 3.3188025422143497e-06,
+      "loss": 0.2802,
+      "step": 9873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.896504435668146,
+      "learning_rate": 3.3178433998373495e-06,
+      "loss": 0.2717,
+      "step": 9874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5406812601018596,
+      "learning_rate": 3.3168843272560127e-06,
+      "loss": 0.257,
+      "step": 9875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.32919562705976,
+      "learning_rate": 3.315925324510135e-06,
+      "loss": 0.2696,
+      "step": 9876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8104804398343184,
+      "learning_rate": 3.3149663916395075e-06,
+      "loss": 0.2703,
+      "step": 9877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.282822067080632,
+      "learning_rate": 3.314007528683918e-06,
+      "loss": 0.2709,
+      "step": 9878
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.894601814571518,
+      "learning_rate": 3.3130487356831496e-06,
+      "loss": 0.2748,
+      "step": 9879
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1047171441470285,
+      "learning_rate": 3.312090012676987e-06,
+      "loss": 0.2716,
+      "step": 9880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.249803689698617,
+      "learning_rate": 3.3111313597052065e-06,
+      "loss": 0.2742,
+      "step": 9881
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.659785378661244,
+      "learning_rate": 3.3101727768075854e-06,
+      "loss": 0.2567,
+      "step": 9882
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3689414848001045,
+      "learning_rate": 3.3092142640238983e-06,
+      "loss": 0.2664,
+      "step": 9883
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8799493568687033,
+      "learning_rate": 3.3082558213939142e-06,
+      "loss": 0.2563,
+      "step": 9884
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0347806542432405,
+      "learning_rate": 3.3072974489574006e-06,
+      "loss": 0.2704,
+      "step": 9885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.54993934788399,
+      "learning_rate": 3.3063391467541227e-06,
+      "loss": 0.2857,
+      "step": 9886
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.424712384660554,
+      "learning_rate": 3.3053809148238426e-06,
+      "loss": 0.2804,
+      "step": 9887
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3711495443193664,
+      "learning_rate": 3.3044227532063177e-06,
+      "loss": 0.2611,
+      "step": 9888
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.119978494852936,
+      "learning_rate": 3.3034646619413046e-06,
+      "loss": 0.269,
+      "step": 9889
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.85465927054288,
+      "learning_rate": 3.3025066410685546e-06,
+      "loss": 0.2636,
+      "step": 9890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2333958252491697,
+      "learning_rate": 3.3015486906278203e-06,
+      "loss": 0.2562,
+      "step": 9891
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.332761445734929,
+      "learning_rate": 3.300590810658848e-06,
+      "loss": 0.2783,
+      "step": 9892
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5840899591393893,
+      "learning_rate": 3.299633001201381e-06,
+      "loss": 0.2738,
+      "step": 9893
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.311613716909245,
+      "learning_rate": 3.298675262295162e-06,
+      "loss": 0.2682,
+      "step": 9894
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4500980765916665,
+      "learning_rate": 3.2977175939799265e-06,
+      "loss": 0.2909,
+      "step": 9895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5356380795765818,
+      "learning_rate": 3.2967599962954134e-06,
+      "loss": 0.2674,
+      "step": 9896
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.255907166982026,
+      "learning_rate": 3.2958024692813515e-06,
+      "loss": 0.2817,
+      "step": 9897
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8803975011197585,
+      "learning_rate": 3.2948450129774726e-06,
+      "loss": 0.2743,
+      "step": 9898
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.180125193183308,
+      "learning_rate": 3.2938876274235034e-06,
+      "loss": 0.2553,
+      "step": 9899
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.068154661473635,
+      "learning_rate": 3.292930312659167e-06,
+      "loss": 0.2903,
+      "step": 9900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4971141147919904,
+      "learning_rate": 3.291973068724184e-06,
+      "loss": 0.2838,
+      "step": 9901
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.245842240400228,
+      "learning_rate": 3.2910158956582726e-06,
+      "loss": 0.2833,
+      "step": 9902
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7865857003738088,
+      "learning_rate": 3.290058793501147e-06,
+      "loss": 0.2667,
+      "step": 9903
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6155417737874286,
+      "learning_rate": 3.289101762292517e-06,
+      "loss": 0.4693,
+      "step": 9904
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5305638923653797,
+      "learning_rate": 3.288144802072097e-06,
+      "loss": 0.2593,
+      "step": 9905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.105242003805574,
+      "learning_rate": 3.287187912879588e-06,
+      "loss": 0.252,
+      "step": 9906
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4182474802379645,
+      "learning_rate": 3.286231094754695e-06,
+      "loss": 0.2677,
+      "step": 9907
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8304821396172453,
+      "learning_rate": 3.2852743477371185e-06,
+      "loss": 0.2703,
+      "step": 9908
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.877794321716428,
+      "learning_rate": 3.284317671866555e-06,
+      "loss": 0.2678,
+      "step": 9909
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.412188533589843,
+      "learning_rate": 3.2833610671826988e-06,
+      "loss": 0.2593,
+      "step": 9910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.953949652599423,
+      "learning_rate": 3.28240453372524e-06,
+      "loss": 0.2764,
+      "step": 9911
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.846073759409722,
+      "learning_rate": 3.281448071533867e-06,
+      "loss": 0.2697,
+      "step": 9912
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9134779397970272,
+      "learning_rate": 3.2804916806482683e-06,
+      "loss": 0.2639,
+      "step": 9913
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.747388588514957,
+      "learning_rate": 3.279535361108123e-06,
+      "loss": 0.2822,
+      "step": 9914
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.358002438438269,
+      "learning_rate": 3.278579112953111e-06,
+      "loss": 0.2751,
+      "step": 9915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.095930422185279,
+      "learning_rate": 3.2776229362229106e-06,
+      "loss": 0.2527,
+      "step": 9916
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3417385104267594,
+      "learning_rate": 3.2766668309571926e-06,
+      "loss": 0.265,
+      "step": 9917
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0250674225646965,
+      "learning_rate": 3.2757107971956302e-06,
+      "loss": 0.2903,
+      "step": 9918
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6154484535404571,
+      "learning_rate": 3.2747548349778867e-06,
+      "loss": 0.5037,
+      "step": 9919
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5693137529839287,
+      "learning_rate": 3.2737989443436314e-06,
+      "loss": 0.2629,
+      "step": 9920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7442982164101064,
+      "learning_rate": 3.2728431253325243e-06,
+      "loss": 0.2834,
+      "step": 9921
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7455489369445782,
+      "learning_rate": 3.271887377984223e-06,
+      "loss": 0.2588,
+      "step": 9922
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5076871106230614,
+      "learning_rate": 3.270931702338384e-06,
+      "loss": 0.2717,
+      "step": 9923
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.382204334404688,
+      "learning_rate": 3.269976098434661e-06,
+      "loss": 0.2707,
+      "step": 9924
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0311926029643566,
+      "learning_rate": 3.2690205663127007e-06,
+      "loss": 0.2585,
+      "step": 9925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8506969859923204,
+      "learning_rate": 3.268065106012153e-06,
+      "loss": 0.2643,
+      "step": 9926
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.80756019697676,
+      "learning_rate": 3.2671097175726587e-06,
+      "loss": 0.2786,
+      "step": 9927
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8991757724178666,
+      "learning_rate": 3.266154401033861e-06,
+      "loss": 0.2704,
+      "step": 9928
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9288666279158564,
+      "learning_rate": 3.265199156435398e-06,
+      "loss": 0.2869,
+      "step": 9929
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 32.79802203209757,
+      "learning_rate": 3.2642439838169023e-06,
+      "loss": 0.2579,
+      "step": 9930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.356750385280537,
+      "learning_rate": 3.2632888832180066e-06,
+      "loss": 0.2794,
+      "step": 9931
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4159543439827265,
+      "learning_rate": 3.2623338546783407e-06,
+      "loss": 0.2729,
+      "step": 9932
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.611233782833403,
+      "learning_rate": 3.261378898237529e-06,
+      "loss": 0.2532,
+      "step": 9933
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.83625966464755,
+      "learning_rate": 3.2604240139351937e-06,
+      "loss": 0.2778,
+      "step": 9934
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0249454942762197,
+      "learning_rate": 3.2594692018109577e-06,
+      "loss": 0.2567,
+      "step": 9935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.095720306438147,
+      "learning_rate": 3.258514461904435e-06,
+      "loss": 0.2827,
+      "step": 9936
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8356246857067402,
+      "learning_rate": 3.257559794255241e-06,
+      "loss": 0.2582,
+      "step": 9937
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 15.879396767792844,
+      "learning_rate": 3.2566051989029866e-06,
+      "loss": 0.2895,
+      "step": 9938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4436324697446716,
+      "learning_rate": 3.2556506758872784e-06,
+      "loss": 0.2581,
+      "step": 9939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.428921470844976,
+      "learning_rate": 3.2546962252477234e-06,
+      "loss": 0.2653,
+      "step": 9940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.023948149755126,
+      "learning_rate": 3.2537418470239208e-06,
+      "loss": 0.2679,
+      "step": 9941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2480300983769723,
+      "learning_rate": 3.2527875412554694e-06,
+      "loss": 0.2454,
+      "step": 9942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1378039251153815,
+      "learning_rate": 3.2518333079819684e-06,
+      "loss": 0.2695,
+      "step": 9943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3900987952765609,
+      "learning_rate": 3.2508791472430073e-06,
+      "loss": 0.2566,
+      "step": 9944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7525763418817006,
+      "learning_rate": 3.2499250590781773e-06,
+      "loss": 0.2575,
+      "step": 9945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6967761651448225,
+      "learning_rate": 3.248971043527066e-06,
+      "loss": 0.2747,
+      "step": 9946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.40864027668934,
+      "learning_rate": 3.248017100629256e-06,
+      "loss": 0.2552,
+      "step": 9947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.492652892048151,
+      "learning_rate": 3.247063230424329e-06,
+      "loss": 0.2662,
+      "step": 9948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 38.11552204506903,
+      "learning_rate": 3.2461094329518593e-06,
+      "loss": 0.2578,
+      "step": 9949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9645221821642918,
+      "learning_rate": 3.245155708251426e-06,
+      "loss": 0.2744,
+      "step": 9950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1398337374814704,
+      "learning_rate": 3.244202056362601e-06,
+      "loss": 0.2595,
+      "step": 9951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3540382346625854,
+      "learning_rate": 3.2432484773249495e-06,
+      "loss": 0.272,
+      "step": 9952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.011525360598053,
+      "learning_rate": 3.2422949711780395e-06,
+      "loss": 0.26,
+      "step": 9953
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7813564347347532,
+      "learning_rate": 3.2413415379614345e-06,
+      "loss": 0.2591,
+      "step": 9954
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.539751554887144,
+      "learning_rate": 3.2403881777146905e-06,
+      "loss": 0.3004,
+      "step": 9955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.311087492806082,
+      "learning_rate": 3.2394348904773687e-06,
+      "loss": 0.2835,
+      "step": 9956
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4307233930329737,
+      "learning_rate": 3.2384816762890182e-06,
+      "loss": 0.2661,
+      "step": 9957
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.040809683042068,
+      "learning_rate": 3.2375285351891918e-06,
+      "loss": 0.2787,
+      "step": 9958
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.0616870068222015,
+      "learning_rate": 3.2365754672174386e-06,
+      "loss": 0.2668,
+      "step": 9959
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.763535396368723,
+      "learning_rate": 3.235622472413301e-06,
+      "loss": 0.2627,
+      "step": 9960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6023289537315193,
+      "learning_rate": 3.234669550816321e-06,
+      "loss": 0.2675,
+      "step": 9961
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.374333265303514,
+      "learning_rate": 3.2337167024660366e-06,
+      "loss": 0.2709,
+      "step": 9962
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3761195103993602,
+      "learning_rate": 3.2327639274019835e-06,
+      "loss": 0.254,
+      "step": 9963
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.732504353626915,
+      "learning_rate": 3.2318112256636923e-06,
+      "loss": 0.2654,
+      "step": 9964
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6397055592623355,
+      "learning_rate": 3.230858597290697e-06,
+      "loss": 0.2584,
+      "step": 9965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.623168773235153,
+      "learning_rate": 3.229906042322519e-06,
+      "loss": 0.269,
+      "step": 9966
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.852660925017179,
+      "learning_rate": 3.2289535607986843e-06,
+      "loss": 0.2669,
+      "step": 9967
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3472769934172164,
+      "learning_rate": 3.2280011527587118e-06,
+      "loss": 0.2578,
+      "step": 9968
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6888069980344453,
+      "learning_rate": 3.227048818242119e-06,
+      "loss": 0.2539,
+      "step": 9969
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9437871595464753,
+      "learning_rate": 3.2260965572884202e-06,
+      "loss": 0.2621,
+      "step": 9970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.803572970593614,
+      "learning_rate": 3.2251443699371252e-06,
+      "loss": 0.2575,
+      "step": 9971
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.299282413598395,
+      "learning_rate": 3.2241922562277416e-06,
+      "loss": 0.3004,
+      "step": 9972
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9689419717334427,
+      "learning_rate": 3.223240216199778e-06,
+      "loss": 0.2797,
+      "step": 9973
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7677124015119414,
+      "learning_rate": 3.222288249892732e-06,
+      "loss": 0.2732,
+      "step": 9974
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.337022225180417,
+      "learning_rate": 3.221336357346105e-06,
+      "loss": 0.2664,
+      "step": 9975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3705375609607087,
+      "learning_rate": 3.220384538599392e-06,
+      "loss": 0.2635,
+      "step": 9976
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.426477112735168,
+      "learning_rate": 3.2194327936920842e-06,
+      "loss": 0.2733,
+      "step": 9977
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2674385705041455,
+      "learning_rate": 3.2184811226636746e-06,
+      "loss": 0.2802,
+      "step": 9978
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6950749532968423,
+      "learning_rate": 3.217529525553645e-06,
+      "loss": 0.2738,
+      "step": 9979
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2637541576120266,
+      "learning_rate": 3.2165780024014825e-06,
+      "loss": 0.269,
+      "step": 9980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.958006121460725,
+      "learning_rate": 3.215626553246667e-06,
+      "loss": 0.2787,
+      "step": 9981
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8480662328736113,
+      "learning_rate": 3.2146751781286743e-06,
+      "loss": 0.2801,
+      "step": 9982
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2434885960103785,
+      "learning_rate": 3.213723877086979e-06,
+      "loss": 0.2725,
+      "step": 9983
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5809920577684438,
+      "learning_rate": 3.2127726501610558e-06,
+      "loss": 0.2435,
+      "step": 9984
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.75565235219549,
+      "learning_rate": 3.2118214973903673e-06,
+      "loss": 0.2713,
+      "step": 9985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3000391861556224,
+      "learning_rate": 3.2108704188143803e-06,
+      "loss": 0.2737,
+      "step": 9986
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.742276002723779,
+      "learning_rate": 3.209919414472559e-06,
+      "loss": 0.2541,
+      "step": 9987
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.87638450087813,
+      "learning_rate": 3.20896848440436e-06,
+      "loss": 0.2627,
+      "step": 9988
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 24.26284996646524,
+      "learning_rate": 3.2080176286492415e-06,
+      "loss": 0.2722,
+      "step": 9989
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5086493109460943,
+      "learning_rate": 3.2070668472466525e-06,
+      "loss": 0.2743,
+      "step": 9990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9097411949715766,
+      "learning_rate": 3.2061161402360454e-06,
+      "loss": 0.2787,
+      "step": 9991
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9428254453100957,
+      "learning_rate": 3.2051655076568666e-06,
+      "loss": 0.2745,
+      "step": 9992
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1766278325768242,
+      "learning_rate": 3.204214949548558e-06,
+      "loss": 0.2748,
+      "step": 9993
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5148475629807059,
+      "learning_rate": 3.2032644659505595e-06,
+      "loss": 0.2597,
+      "step": 9994
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8351187003270963,
+      "learning_rate": 3.2023140569023124e-06,
+      "loss": 0.2713,
+      "step": 9995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3119515675998468,
+      "learning_rate": 3.2013637224432463e-06,
+      "loss": 0.2553,
+      "step": 9996
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.5777577907954905,
+      "learning_rate": 3.200413462612796e-06,
+      "loss": 0.2628,
+      "step": 9997
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4677048711679106,
+      "learning_rate": 3.1994632774503853e-06,
+      "loss": 0.2832,
+      "step": 9998
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5972451498659193,
+      "learning_rate": 3.198513166995442e-06,
+      "loss": 0.2736,
+      "step": 9999
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.927327376868954,
+      "learning_rate": 3.197563131287389e-06,
+      "loss": 0.2639,
+      "step": 10000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.18965047883231,
+      "learning_rate": 3.19661317036564e-06,
+      "loss": 0.266,
+      "step": 10001
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.341591965281301,
+      "learning_rate": 3.195663284269615e-06,
+      "loss": 0.2678,
+      "step": 10002
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1173636223746146,
+      "learning_rate": 3.1947134730387265e-06,
+      "loss": 0.2764,
+      "step": 10003
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9686828247975843,
+      "learning_rate": 3.1937637367123814e-06,
+      "loss": 0.2741,
+      "step": 10004
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.627160398881354,
+      "learning_rate": 3.192814075329988e-06,
+      "loss": 0.2683,
+      "step": 10005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0770602123616824,
+      "learning_rate": 3.191864488930948e-06,
+      "loss": 0.2644,
+      "step": 10006
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2461568601882127,
+      "learning_rate": 3.190914977554661e-06,
+      "loss": 0.2699,
+      "step": 10007
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6242753466890067,
+      "learning_rate": 3.1899655412405266e-06,
+      "loss": 0.2572,
+      "step": 10008
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.364714501622136,
+      "learning_rate": 3.1890161800279353e-06,
+      "loss": 0.2779,
+      "step": 10009
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3796187874313297,
+      "learning_rate": 3.188066893956279e-06,
+      "loss": 0.2623,
+      "step": 10010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.772680539942646,
+      "learning_rate": 3.1871176830649473e-06,
+      "loss": 0.2629,
+      "step": 10011
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.728648012424328,
+      "learning_rate": 3.1861685473933223e-06,
+      "loss": 0.263,
+      "step": 10012
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5028407688194734,
+      "learning_rate": 3.1852194869807873e-06,
+      "loss": 0.2677,
+      "step": 10013
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4799106491738157,
+      "learning_rate": 3.1842705018667173e-06,
+      "loss": 0.2721,
+      "step": 10014
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7620365634749426,
+      "learning_rate": 3.18332159209049e-06,
+      "loss": 0.273,
+      "step": 10015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.433121238508442,
+      "learning_rate": 3.1823727576914753e-06,
+      "loss": 0.2537,
+      "step": 10016
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7569294426756594,
+      "learning_rate": 3.181423998709045e-06,
+      "loss": 0.2812,
+      "step": 10017
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.7976168662829215,
+      "learning_rate": 3.180475315182563e-06,
+      "loss": 0.2777,
+      "step": 10018
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.626302240605091,
+      "learning_rate": 3.1795267071513925e-06,
+      "loss": 0.2528,
+      "step": 10019
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.014450872227873,
+      "learning_rate": 3.178578174654891e-06,
+      "loss": 0.2717,
+      "step": 10020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9183255383889546,
+      "learning_rate": 3.1776297177324167e-06,
+      "loss": 0.2754,
+      "step": 10021
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8944432411742664,
+      "learning_rate": 3.1766813364233234e-06,
+      "loss": 0.2729,
+      "step": 10022
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5160483787146775,
+      "learning_rate": 3.175733030766959e-06,
+      "loss": 0.2736,
+      "step": 10023
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8786096699155248,
+      "learning_rate": 3.1747848008026704e-06,
+      "loss": 0.2707,
+      "step": 10024
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.078054043201778,
+      "learning_rate": 3.1738366465698043e-06,
+      "loss": 0.2744,
+      "step": 10025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.514279063968887,
+      "learning_rate": 3.1728885681076983e-06,
+      "loss": 0.2889,
+      "step": 10026
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.123097377846155,
+      "learning_rate": 3.1719405654556924e-06,
+      "loss": 0.2667,
+      "step": 10027
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8728657758142397,
+      "learning_rate": 3.1709926386531174e-06,
+      "loss": 0.2802,
+      "step": 10028
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9623826940788787,
+      "learning_rate": 3.170044787739307e-06,
+      "loss": 0.2829,
+      "step": 10029
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9078598998142404,
+      "learning_rate": 3.1690970127535904e-06,
+      "loss": 0.2637,
+      "step": 10030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6845025177566337,
+      "learning_rate": 3.168149313735288e-06,
+      "loss": 0.2784,
+      "step": 10031
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.983978912469738,
+      "learning_rate": 3.1672016907237256e-06,
+      "loss": 0.2751,
+      "step": 10032
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2511549794876786,
+      "learning_rate": 3.166254143758222e-06,
+      "loss": 0.2652,
+      "step": 10033
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1075240554115413,
+      "learning_rate": 3.165306672878089e-06,
+      "loss": 0.2564,
+      "step": 10034
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7632322212063483,
+      "learning_rate": 3.1643592781226428e-06,
+      "loss": 0.2728,
+      "step": 10035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.46186032204869,
+      "learning_rate": 3.163411959531189e-06,
+      "loss": 0.2616,
+      "step": 10036
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.881101303832084,
+      "learning_rate": 3.1624647171430355e-06,
+      "loss": 0.2651,
+      "step": 10037
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2289831230123385,
+      "learning_rate": 3.1615175509974855e-06,
+      "loss": 0.2933,
+      "step": 10038
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8019027339394116,
+      "learning_rate": 3.160570461133836e-06,
+      "loss": 0.29,
+      "step": 10039
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.696416257606188,
+      "learning_rate": 3.1596234475913855e-06,
+      "loss": 0.2814,
+      "step": 10040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6196570926025736,
+      "learning_rate": 3.1586765104094295e-06,
+      "loss": 0.4681,
+      "step": 10041
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.579448029063411,
+      "learning_rate": 3.1577296496272535e-06,
+      "loss": 0.2805,
+      "step": 10042
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3746432894832377,
+      "learning_rate": 3.156782865284148e-06,
+      "loss": 0.2717,
+      "step": 10043
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8715667524654522,
+      "learning_rate": 3.155836157419394e-06,
+      "loss": 0.2784,
+      "step": 10044
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.518868977270992,
+      "learning_rate": 3.1548895260722743e-06,
+      "loss": 0.2655,
+      "step": 10045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2942224365251014,
+      "learning_rate": 3.1539429712820634e-06,
+      "loss": 0.2673,
+      "step": 10046
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6836469079689118,
+      "learning_rate": 3.15299649308804e-06,
+      "loss": 0.2619,
+      "step": 10047
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3316655067254635,
+      "learning_rate": 3.152050091529472e-06,
+      "loss": 0.2682,
+      "step": 10048
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.145213553660683,
+      "learning_rate": 3.151103766645629e-06,
+      "loss": 0.2715,
+      "step": 10049
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.251853760793825,
+      "learning_rate": 3.150157518475774e-06,
+      "loss": 0.2514,
+      "step": 10050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5113708883773502,
+      "learning_rate": 3.149211347059169e-06,
+      "loss": 0.282,
+      "step": 10051
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7868447197110866,
+      "learning_rate": 3.1482652524350727e-06,
+      "loss": 0.2689,
+      "step": 10052
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.510469220806779,
+      "learning_rate": 3.1473192346427396e-06,
+      "loss": 0.2811,
+      "step": 10053
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4004931673113967,
+      "learning_rate": 3.146373293721422e-06,
+      "loss": 0.2644,
+      "step": 10054
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8329117459328055,
+      "learning_rate": 3.14542742971037e-06,
+      "loss": 0.2899,
+      "step": 10055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7556121061371597,
+      "learning_rate": 3.1444816426488275e-06,
+      "loss": 0.2678,
+      "step": 10056
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8984575322681536,
+      "learning_rate": 3.143535932576038e-06,
+      "loss": 0.2619,
+      "step": 10057
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4185456732644264,
+      "learning_rate": 3.1425902995312394e-06,
+      "loss": 0.2663,
+      "step": 10058
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4057435765175152,
+      "learning_rate": 3.141644743553668e-06,
+      "loss": 0.2918,
+      "step": 10059
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9093835713711051,
+      "learning_rate": 3.1406992646825583e-06,
+      "loss": 0.2776,
+      "step": 10060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8243505802516449,
+      "learning_rate": 3.1397538629571368e-06,
+      "loss": 0.2678,
+      "step": 10061
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.95825607455039,
+      "learning_rate": 3.1388085384166324e-06,
+      "loss": 0.2818,
+      "step": 10062
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.369785162060367,
+      "learning_rate": 3.1378632911002688e-06,
+      "loss": 0.2591,
+      "step": 10063
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8408484348366791,
+      "learning_rate": 3.136918121047264e-06,
+      "loss": 0.2596,
+      "step": 10064
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.530616936246068,
+      "learning_rate": 3.135973028296836e-06,
+      "loss": 0.2759,
+      "step": 10065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8773389145388941,
+      "learning_rate": 3.135028012888197e-06,
+      "loss": 0.2538,
+      "step": 10066
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.29984652337607,
+      "learning_rate": 3.134083074860559e-06,
+      "loss": 0.2716,
+      "step": 10067
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1437430964536395,
+      "learning_rate": 3.1331382142531265e-06,
+      "loss": 0.2825,
+      "step": 10068
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2273024299155457,
+      "learning_rate": 3.132193431105108e-06,
+      "loss": 0.274,
+      "step": 10069
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4882246604550748,
+      "learning_rate": 3.1312487254557006e-06,
+      "loss": 0.2665,
+      "step": 10070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3844836479646911,
+      "learning_rate": 3.1303040973441036e-06,
+      "loss": 0.2646,
+      "step": 10071
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9536467788912197,
+      "learning_rate": 3.1293595468095094e-06,
+      "loss": 0.2821,
+      "step": 10072
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6201119669717978,
+      "learning_rate": 3.1284150738911125e-06,
+      "loss": 0.2499,
+      "step": 10073
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5418702408945038,
+      "learning_rate": 3.127470678628096e-06,
+      "loss": 0.2761,
+      "step": 10074
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3283382132612167,
+      "learning_rate": 3.1265263610596475e-06,
+      "loss": 0.2641,
+      "step": 10075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4433540304588321,
+      "learning_rate": 3.125582121224947e-06,
+      "loss": 0.263,
+      "step": 10076
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 20.278285307033244,
+      "learning_rate": 3.124637959163176e-06,
+      "loss": 0.2679,
+      "step": 10077
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.388136569112997,
+      "learning_rate": 3.123693874913506e-06,
+      "loss": 0.2773,
+      "step": 10078
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8994297306023664,
+      "learning_rate": 3.12274986851511e-06,
+      "loss": 0.2762,
+      "step": 10079
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3895787241115194,
+      "learning_rate": 3.1218059400071564e-06,
+      "loss": 0.2609,
+      "step": 10080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6774317358193516,
+      "learning_rate": 3.1208620894288105e-06,
+      "loss": 0.2852,
+      "step": 10081
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7341905061504503,
+      "learning_rate": 3.1199183168192337e-06,
+      "loss": 0.2881,
+      "step": 10082
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3286729637431853,
+      "learning_rate": 3.1189746222175843e-06,
+      "loss": 0.2579,
+      "step": 10083
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.110021476189369,
+      "learning_rate": 3.11803100566302e-06,
+      "loss": 0.2665,
+      "step": 10084
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.706249935834069,
+      "learning_rate": 3.117087467194693e-06,
+      "loss": 0.2639,
+      "step": 10085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9493757887438823,
+      "learning_rate": 3.11614400685175e-06,
+      "loss": 0.2732,
+      "step": 10086
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2076273682628904,
+      "learning_rate": 3.1152006246733395e-06,
+      "loss": 0.2708,
+      "step": 10087
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.313402505219228,
+      "learning_rate": 3.114257320698602e-06,
+      "loss": 0.2534,
+      "step": 10088
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.113284925556673,
+      "learning_rate": 3.1133140949666785e-06,
+      "loss": 0.2744,
+      "step": 10089
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1397030668441617,
+      "learning_rate": 3.1123709475167043e-06,
+      "loss": 0.2732,
+      "step": 10090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6414653821745988,
+      "learning_rate": 3.11142787838781e-06,
+      "loss": 0.264,
+      "step": 10091
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.693477360207317,
+      "learning_rate": 3.110484887619129e-06,
+      "loss": 0.2702,
+      "step": 10092
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6470543237118083,
+      "learning_rate": 3.109541975249787e-06,
+      "loss": 0.2557,
+      "step": 10093
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5648910258663804,
+      "learning_rate": 3.1085991413189053e-06,
+      "loss": 0.2962,
+      "step": 10094
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0426294366035807,
+      "learning_rate": 3.1076563858656062e-06,
+      "loss": 0.2506,
+      "step": 10095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4564743363958335,
+      "learning_rate": 3.1067137089290033e-06,
+      "loss": 0.2732,
+      "step": 10096
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.19124921830546,
+      "learning_rate": 3.105771110548212e-06,
+      "loss": 0.2663,
+      "step": 10097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.372263704027683,
+      "learning_rate": 3.10482859076234e-06,
+      "loss": 0.278,
+      "step": 10098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0814643219929176,
+      "learning_rate": 3.1038861496104987e-06,
+      "loss": 0.2662,
+      "step": 10099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.246160079019944,
+      "learning_rate": 3.102943787131788e-06,
+      "loss": 0.2625,
+      "step": 10100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.450483413831347,
+      "learning_rate": 3.1020015033653094e-06,
+      "loss": 0.2438,
+      "step": 10101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.261575768600086,
+      "learning_rate": 3.10105929835016e-06,
+      "loss": 0.2801,
+      "step": 10102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8687858229308059,
+      "learning_rate": 3.100117172125433e-06,
+      "loss": 0.2806,
+      "step": 10103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.880372771729046,
+      "learning_rate": 3.099175124730218e-06,
+      "loss": 0.2568,
+      "step": 10104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.921810094277788,
+      "learning_rate": 3.0982331562036037e-06,
+      "loss": 0.3059,
+      "step": 10105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.71574008247158,
+      "learning_rate": 3.097291266584673e-06,
+      "loss": 0.2753,
+      "step": 10106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5085525891596148,
+      "learning_rate": 3.0963494559125084e-06,
+      "loss": 0.2687,
+      "step": 10107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.174098534749825,
+      "learning_rate": 3.095407724226186e-06,
+      "loss": 0.2516,
+      "step": 10108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9535102854308344,
+      "learning_rate": 3.0944660715647813e-06,
+      "loss": 0.2641,
+      "step": 10109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.172982818492106,
+      "learning_rate": 3.093524497967362e-06,
+      "loss": 0.3047,
+      "step": 10110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7053760550966017,
+      "learning_rate": 3.0925830034729986e-06,
+      "loss": 0.2637,
+      "step": 10111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1150917499177964,
+      "learning_rate": 3.091641588120754e-06,
+      "loss": 0.2561,
+      "step": 10112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8556553949819037,
+      "learning_rate": 3.0907002519496875e-06,
+      "loss": 0.2812,
+      "step": 10113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.596939683625185,
+      "learning_rate": 3.089758994998861e-06,
+      "loss": 0.2707,
+      "step": 10114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.900356263255975,
+      "learning_rate": 3.088817817307327e-06,
+      "loss": 0.2668,
+      "step": 10115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7163415576738235,
+      "learning_rate": 3.0878767189141355e-06,
+      "loss": 0.2665,
+      "step": 10116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.344543836072783,
+      "learning_rate": 3.0869356998583366e-06,
+      "loss": 0.2718,
+      "step": 10117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.340726681295385,
+      "learning_rate": 3.0859947601789724e-06,
+      "loss": 0.2615,
+      "step": 10118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.515503426889501,
+      "learning_rate": 3.0850538999150866e-06,
+      "loss": 0.2706,
+      "step": 10119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.876976324052386,
+      "learning_rate": 3.084113119105714e-06,
+      "loss": 0.2899,
+      "step": 10120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.083568949066206,
+      "learning_rate": 3.08317241778989e-06,
+      "loss": 0.2715,
+      "step": 10121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.544694145277682,
+      "learning_rate": 3.0822317960066493e-06,
+      "loss": 0.2864,
+      "step": 10122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.934966786984859,
+      "learning_rate": 3.081291253795018e-06,
+      "loss": 0.2742,
+      "step": 10123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1233943000294593,
+      "learning_rate": 3.0803507911940193e-06,
+      "loss": 0.3055,
+      "step": 10124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6936264444753946,
+      "learning_rate": 3.0794104082426772e-06,
+      "loss": 0.2738,
+      "step": 10125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3895582488985383,
+      "learning_rate": 3.078470104980008e-06,
+      "loss": 0.2519,
+      "step": 10126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.712330595434032,
+      "learning_rate": 3.0775298814450273e-06,
+      "loss": 0.2664,
+      "step": 10127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.127102241181531,
+      "learning_rate": 3.076589737676744e-06,
+      "loss": 0.2912,
+      "step": 10128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7741587075314817,
+      "learning_rate": 3.0756496737141715e-06,
+      "loss": 0.2667,
+      "step": 10129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.889646570103967,
+      "learning_rate": 3.074709689596312e-06,
+      "loss": 0.2549,
+      "step": 10130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6729520432348077,
+      "learning_rate": 3.0737697853621672e-06,
+      "loss": 0.2533,
+      "step": 10131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5657536258905018,
+      "learning_rate": 3.072829961050735e-06,
+      "loss": 0.292,
+      "step": 10132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.539708974291937,
+      "learning_rate": 3.071890216701012e-06,
+      "loss": 0.263,
+      "step": 10133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.88300859835801,
+      "learning_rate": 3.0709505523519868e-06,
+      "loss": 0.287,
+      "step": 10134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5642082145514746,
+      "learning_rate": 3.07001096804265e-06,
+      "loss": 0.2758,
+      "step": 10135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.100345082396697,
+      "learning_rate": 3.0690714638119843e-06,
+      "loss": 0.2756,
+      "step": 10136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8571735285393314,
+      "learning_rate": 3.068132039698977e-06,
+      "loss": 0.2894,
+      "step": 10137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8375961100095908,
+      "learning_rate": 3.0671926957426e-06,
+      "loss": 0.2916,
+      "step": 10138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.91512447998109,
+      "learning_rate": 3.066253431981834e-06,
+      "loss": 0.2778,
+      "step": 10139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.965821722495347,
+      "learning_rate": 3.0653142484556454e-06,
+      "loss": 0.2604,
+      "step": 10140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.48749806441734,
+      "learning_rate": 3.0643751452030065e-06,
+      "loss": 0.2872,
+      "step": 10141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7852139975220183,
+      "learning_rate": 3.0634361222628794e-06,
+      "loss": 0.2716,
+      "step": 10142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.413355486247813,
+      "learning_rate": 3.0624971796742263e-06,
+      "loss": 0.2601,
+      "step": 10143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7839414408876089,
+      "learning_rate": 3.0615583174760083e-06,
+      "loss": 0.2815,
+      "step": 10144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9117422090997844,
+      "learning_rate": 3.0606195357071795e-06,
+      "loss": 0.2891,
+      "step": 10145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6819593658141683,
+      "learning_rate": 3.0596808344066896e-06,
+      "loss": 0.265,
+      "step": 10146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.459310506845838,
+      "learning_rate": 3.0587422136134896e-06,
+      "loss": 0.269,
+      "step": 10147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7168692223380229,
+      "learning_rate": 3.0578036733665224e-06,
+      "loss": 0.2603,
+      "step": 10148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 29.877550580355468,
+      "learning_rate": 3.0568652137047312e-06,
+      "loss": 0.2778,
+      "step": 10149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.178501763688367,
+      "learning_rate": 3.0559268346670507e-06,
+      "loss": 0.2973,
+      "step": 10150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7895490436953738,
+      "learning_rate": 3.0549885362924215e-06,
+      "loss": 0.2616,
+      "step": 10151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.7808985512461115,
+      "learning_rate": 3.0540503186197724e-06,
+      "loss": 0.2733,
+      "step": 10152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.800684782977907,
+      "learning_rate": 3.0531121816880327e-06,
+      "loss": 0.3003,
+      "step": 10153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6093911880438868,
+      "learning_rate": 3.052174125536126e-06,
+      "loss": 0.492,
+      "step": 10154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.932611203987603,
+      "learning_rate": 3.0512361502029747e-06,
+      "loss": 0.2607,
+      "step": 10155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5259995957077686,
+      "learning_rate": 3.0502982557274962e-06,
+      "loss": 0.2785,
+      "step": 10156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.030242197573419,
+      "learning_rate": 3.049360442148608e-06,
+      "loss": 0.2988,
+      "step": 10157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6263975983204364,
+      "learning_rate": 3.0484227095052154e-06,
+      "loss": 0.5001,
+      "step": 10158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7789997055425444,
+      "learning_rate": 3.047485057836235e-06,
+      "loss": 0.2448,
+      "step": 10159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.440988523108026,
+      "learning_rate": 3.046547487180566e-06,
+      "loss": 0.2703,
+      "step": 10160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2293867136504932,
+      "learning_rate": 3.0456099975771126e-06,
+      "loss": 0.2588,
+      "step": 10161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1273934230670917,
+      "learning_rate": 3.044672589064771e-06,
+      "loss": 0.2485,
+      "step": 10162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.213251687312522,
+      "learning_rate": 3.043735261682438e-06,
+      "loss": 0.2592,
+      "step": 10163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.799080709975172,
+      "learning_rate": 3.0427980154690017e-06,
+      "loss": 0.277,
+      "step": 10164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.932450694159075,
+      "learning_rate": 3.041860850463352e-06,
+      "loss": 0.2712,
+      "step": 10165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.63425435114625,
+      "learning_rate": 3.0409237667043744e-06,
+      "loss": 0.2686,
+      "step": 10166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.872411582669073,
+      "learning_rate": 3.039986764230951e-06,
+      "loss": 0.3041,
+      "step": 10167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0505403463626175,
+      "learning_rate": 3.0390498430819565e-06,
+      "loss": 0.2716,
+      "step": 10168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.092470412115965,
+      "learning_rate": 3.038113003296268e-06,
+      "loss": 0.2618,
+      "step": 10169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8945188839773484,
+      "learning_rate": 3.037176244912755e-06,
+      "loss": 0.2612,
+      "step": 10170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.428301612150675,
+      "learning_rate": 3.036239567970287e-06,
+      "loss": 0.2788,
+      "step": 10171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.3656500157664695,
+      "learning_rate": 3.035302972507726e-06,
+      "loss": 0.2892,
+      "step": 10172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3679817876270866,
+      "learning_rate": 3.034366458563933e-06,
+      "loss": 0.2709,
+      "step": 10173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.598483411415566,
+      "learning_rate": 3.0334300261777693e-06,
+      "loss": 0.2705,
+      "step": 10174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.075141900099125,
+      "learning_rate": 3.032493675388087e-06,
+      "loss": 0.287,
+      "step": 10175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.685290551967383,
+      "learning_rate": 3.031557406233736e-06,
+      "loss": 0.286,
+      "step": 10176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6572913861913072,
+      "learning_rate": 3.0306212187535653e-06,
+      "loss": 0.2689,
+      "step": 10177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.208896982634066,
+      "learning_rate": 3.029685112986417e-06,
+      "loss": 0.2475,
+      "step": 10178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6373364895627884,
+      "learning_rate": 3.028749088971135e-06,
+      "loss": 0.2988,
+      "step": 10179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.907750624306396,
+      "learning_rate": 3.027813146746551e-06,
+      "loss": 0.2695,
+      "step": 10180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7924498823603048,
+      "learning_rate": 3.026877286351505e-06,
+      "loss": 0.2633,
+      "step": 10181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.356377903405441,
+      "learning_rate": 3.0259415078248246e-06,
+      "loss": 0.2504,
+      "step": 10182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2792728631645054,
+      "learning_rate": 3.025005811205339e-06,
+      "loss": 0.2803,
+      "step": 10183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.673231353724311,
+      "learning_rate": 3.024070196531869e-06,
+      "loss": 0.2573,
+      "step": 10184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8496966766462282,
+      "learning_rate": 3.023134663843236e-06,
+      "loss": 0.2724,
+      "step": 10185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0429540934790427,
+      "learning_rate": 3.0221992131782573e-06,
+      "loss": 0.292,
+      "step": 10186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5778665551464879,
+      "learning_rate": 3.021263844575747e-06,
+      "loss": 0.2662,
+      "step": 10187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.54086520150844,
+      "learning_rate": 3.020328558074511e-06,
+      "loss": 0.2547,
+      "step": 10188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.527184549260713,
+      "learning_rate": 3.0193933537133624e-06,
+      "loss": 0.2532,
+      "step": 10189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.666722457301461,
+      "learning_rate": 3.0184582315311013e-06,
+      "loss": 0.2761,
+      "step": 10190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7200665926247563,
+      "learning_rate": 3.017523191566528e-06,
+      "loss": 0.2668,
+      "step": 10191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.727789154045117,
+      "learning_rate": 3.0165882338584383e-06,
+      "loss": 0.2527,
+      "step": 10192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2410172379505586,
+      "learning_rate": 3.0156533584456268e-06,
+      "loss": 0.2678,
+      "step": 10193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.33067398381951,
+      "learning_rate": 3.0147185653668805e-06,
+      "loss": 0.2725,
+      "step": 10194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.12044012495356,
+      "learning_rate": 3.0137838546609867e-06,
+      "loss": 0.2834,
+      "step": 10195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.15527127240461,
+      "learning_rate": 3.0128492263667287e-06,
+      "loss": 0.2804,
+      "step": 10196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.012823032618822,
+      "learning_rate": 3.011914680522888e-06,
+      "loss": 0.2732,
+      "step": 10197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 23.83203618924269,
+      "learning_rate": 3.0109802171682366e-06,
+      "loss": 0.2702,
+      "step": 10198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.096726270747896,
+      "learning_rate": 3.0100458363415506e-06,
+      "loss": 0.2744,
+      "step": 10199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.472775888169146,
+      "learning_rate": 3.0091115380815962e-06,
+      "loss": 0.277,
+      "step": 10200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.624312152490665,
+      "learning_rate": 3.0081773224271417e-06,
+      "loss": 0.2604,
+      "step": 10201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.142779529305478,
+      "learning_rate": 3.007243189416946e-06,
+      "loss": 0.2689,
+      "step": 10202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.450999900769122,
+      "learning_rate": 3.006309139089769e-06,
+      "loss": 0.2885,
+      "step": 10203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0392459597264607,
+      "learning_rate": 3.005375171484369e-06,
+      "loss": 0.2577,
+      "step": 10204
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6237970373006214,
+      "learning_rate": 3.0044412866394964e-06,
+      "loss": 0.2672,
+      "step": 10205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0768314942856674,
+      "learning_rate": 3.003507484593898e-06,
+      "loss": 0.2729,
+      "step": 10206
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.674237992765155,
+      "learning_rate": 3.002573765386322e-06,
+      "loss": 0.268,
+      "step": 10207
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.139310557894765,
+      "learning_rate": 3.0016401290555065e-06,
+      "loss": 0.261,
+      "step": 10208
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8527313217119192,
+      "learning_rate": 3.000706575640193e-06,
+      "loss": 0.2751,
+      "step": 10209
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.017620907037851,
+      "learning_rate": 2.9997731051791113e-06,
+      "loss": 0.2442,
+      "step": 10210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.694812998113918,
+      "learning_rate": 2.9988397177109994e-06,
+      "loss": 0.2822,
+      "step": 10211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.196133921080475,
+      "learning_rate": 2.997906413274582e-06,
+      "loss": 0.2685,
+      "step": 10212
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6917330466268106,
+      "learning_rate": 2.9969731919085844e-06,
+      "loss": 0.2771,
+      "step": 10213
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8836634430406083,
+      "learning_rate": 2.9960400536517246e-06,
+      "loss": 0.2705,
+      "step": 10214
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.13448276858894,
+      "learning_rate": 2.9951069985427246e-06,
+      "loss": 0.2855,
+      "step": 10215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.815158030942339,
+      "learning_rate": 2.9941740266202946e-06,
+      "loss": 0.2834,
+      "step": 10216
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6404612673691276,
+      "learning_rate": 2.9932411379231486e-06,
+      "loss": 0.4989,
+      "step": 10217
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9222654985393097,
+      "learning_rate": 2.9923083324899894e-06,
+      "loss": 0.2677,
+      "step": 10218
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2963999108599005,
+      "learning_rate": 2.9913756103595258e-06,
+      "loss": 0.2842,
+      "step": 10219
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0794310772310274,
+      "learning_rate": 2.9904429715704554e-06,
+      "loss": 0.2727,
+      "step": 10220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.543411504214379,
+      "learning_rate": 2.989510416161476e-06,
+      "loss": 0.28,
+      "step": 10221
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8187753782238385,
+      "learning_rate": 2.988577944171279e-06,
+      "loss": 0.2506,
+      "step": 10222
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.950717599692712,
+      "learning_rate": 2.9876455556385576e-06,
+      "loss": 0.2727,
+      "step": 10223
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.298750541110782,
+      "learning_rate": 2.9867132506019958e-06,
+      "loss": 0.274,
+      "step": 10224
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0069221080400816,
+      "learning_rate": 2.9857810291002755e-06,
+      "loss": 0.2578,
+      "step": 10225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8974810685327945,
+      "learning_rate": 2.984848891172079e-06,
+      "loss": 0.2718,
+      "step": 10226
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0549144684644705,
+      "learning_rate": 2.9839168368560827e-06,
+      "loss": 0.2783,
+      "step": 10227
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.234453526518632,
+      "learning_rate": 2.982984866190957e-06,
+      "loss": 0.2693,
+      "step": 10228
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7125253929904005,
+      "learning_rate": 2.9820529792153717e-06,
+      "loss": 0.278,
+      "step": 10229
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8623729075608615,
+      "learning_rate": 2.9811211759679926e-06,
+      "loss": 0.2622,
+      "step": 10230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.154812623973168,
+      "learning_rate": 2.9801894564874824e-06,
+      "loss": 0.2763,
+      "step": 10231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.864391409126263,
+      "learning_rate": 2.9792578208124976e-06,
+      "loss": 0.2888,
+      "step": 10232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5593740056610697,
+      "learning_rate": 2.9783262689816963e-06,
+      "loss": 0.2681,
+      "step": 10233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.492521212032518,
+      "learning_rate": 2.977394801033728e-06,
+      "loss": 0.2676,
+      "step": 10234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6088182633682745,
+      "learning_rate": 2.976463417007244e-06,
+      "loss": 0.4982,
+      "step": 10235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6838069910757487,
+      "learning_rate": 2.975532116940885e-06,
+      "loss": 0.2583,
+      "step": 10236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.151719183837695,
+      "learning_rate": 2.974600900873296e-06,
+      "loss": 0.2623,
+      "step": 10237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7806131137092898,
+      "learning_rate": 2.9736697688431116e-06,
+      "loss": 0.2849,
+      "step": 10238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0241893268833118,
+      "learning_rate": 2.972738720888969e-06,
+      "loss": 0.2668,
+      "step": 10239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.57971094866745,
+      "learning_rate": 2.971807757049496e-06,
+      "loss": 0.2606,
+      "step": 10240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.060102433166916,
+      "learning_rate": 2.970876877363322e-06,
+      "loss": 0.2874,
+      "step": 10241
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.50574208213328,
+      "learning_rate": 2.9699460818690714e-06,
+      "loss": 0.2685,
+      "step": 10242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.948747642390405,
+      "learning_rate": 2.9690153706053638e-06,
+      "loss": 0.2691,
+      "step": 10243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7094953346386037,
+      "learning_rate": 2.968084743610815e-06,
+      "loss": 0.274,
+      "step": 10244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.663651213945327,
+      "learning_rate": 2.9671542009240406e-06,
+      "loss": 0.2665,
+      "step": 10245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.41579868964434,
+      "learning_rate": 2.966223742583648e-06,
+      "loss": 0.282,
+      "step": 10246
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6469141833717196,
+      "learning_rate": 2.965293368628244e-06,
+      "loss": 0.2562,
+      "step": 10247
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5138190901598279,
+      "learning_rate": 2.964363079096434e-06,
+      "loss": 0.2613,
+      "step": 10248
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.275875266250925,
+      "learning_rate": 2.963432874026815e-06,
+      "loss": 0.2759,
+      "step": 10249
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.69531648687304,
+      "learning_rate": 2.962502753457982e-06,
+      "loss": 0.2606,
+      "step": 10250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7325910610110424,
+      "learning_rate": 2.9615727174285307e-06,
+      "loss": 0.3142,
+      "step": 10251
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6427437877053559,
+      "learning_rate": 2.960642765977047e-06,
+      "loss": 0.4688,
+      "step": 10252
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.863669200416984,
+      "learning_rate": 2.9597128991421187e-06,
+      "loss": 0.263,
+      "step": 10253
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.272733400023962,
+      "learning_rate": 2.9587831169623244e-06,
+      "loss": 0.2607,
+      "step": 10254
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0399852713986415,
+      "learning_rate": 2.957853419476243e-06,
+      "loss": 0.2753,
+      "step": 10255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9569226797701358,
+      "learning_rate": 2.956923806722453e-06,
+      "loss": 0.2727,
+      "step": 10256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8961836722298089,
+      "learning_rate": 2.9559942787395224e-06,
+      "loss": 0.2579,
+      "step": 10257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.6855425757466405,
+      "learning_rate": 2.9550648355660195e-06,
+      "loss": 0.2668,
+      "step": 10258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.574673243926472,
+      "learning_rate": 2.9541354772405096e-06,
+      "loss": 0.2774,
+      "step": 10259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6174823727962924,
+      "learning_rate": 2.953206203801552e-06,
+      "loss": 0.2526,
+      "step": 10260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7780847535332385,
+      "learning_rate": 2.952277015287705e-06,
+      "loss": 0.2639,
+      "step": 10261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7532879482171553,
+      "learning_rate": 2.95134791173752e-06,
+      "loss": 0.2646,
+      "step": 10262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7286493146123572,
+      "learning_rate": 2.9504188931895507e-06,
+      "loss": 0.2757,
+      "step": 10263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.061230434689712,
+      "learning_rate": 2.9494899596823405e-06,
+      "loss": 0.2696,
+      "step": 10264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.080154476416327,
+      "learning_rate": 2.948561111254436e-06,
+      "loss": 0.2778,
+      "step": 10265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5940276876912574,
+      "learning_rate": 2.9476323479443736e-06,
+      "loss": 0.272,
+      "step": 10266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.435622371099773,
+      "learning_rate": 2.9467036697906914e-06,
+      "loss": 0.2613,
+      "step": 10267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9486384831096513,
+      "learning_rate": 2.9457750768319202e-06,
+      "loss": 0.2561,
+      "step": 10268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.032735319772164,
+      "learning_rate": 2.9448465691065906e-06,
+      "loss": 0.2667,
+      "step": 10269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.890455038224558,
+      "learning_rate": 2.9439181466532253e-06,
+      "loss": 0.279,
+      "step": 10270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8318093530757913,
+      "learning_rate": 2.9429898095103494e-06,
+      "loss": 0.2814,
+      "step": 10271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9060370669230478,
+      "learning_rate": 2.942061557716479e-06,
+      "loss": 0.257,
+      "step": 10272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.847142304524054,
+      "learning_rate": 2.9411333913101316e-06,
+      "loss": 0.2783,
+      "step": 10273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.158199419371395,
+      "learning_rate": 2.940205310329816e-06,
+      "loss": 0.2849,
+      "step": 10274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6207273913013245,
+      "learning_rate": 2.9392773148140406e-06,
+      "loss": 0.2559,
+      "step": 10275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6615952738715276,
+      "learning_rate": 2.9383494048013096e-06,
+      "loss": 0.2675,
+      "step": 10276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3387500118835,
+      "learning_rate": 2.937421580330123e-06,
+      "loss": 0.2515,
+      "step": 10277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1542056823758857,
+      "learning_rate": 2.9364938414389797e-06,
+      "loss": 0.2688,
+      "step": 10278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2811552349845754,
+      "learning_rate": 2.9355661881663717e-06,
+      "loss": 0.251,
+      "step": 10279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.309169932842242,
+      "learning_rate": 2.9346386205507893e-06,
+      "loss": 0.268,
+      "step": 10280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.630879173555384,
+      "learning_rate": 2.9337111386307197e-06,
+      "loss": 0.2721,
+      "step": 10281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.522778125874977,
+      "learning_rate": 2.9327837424446442e-06,
+      "loss": 0.2688,
+      "step": 10282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7498924542565566,
+      "learning_rate": 2.9318564320310444e-06,
+      "loss": 0.2521,
+      "step": 10283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3253732961167728,
+      "learning_rate": 2.9309292074283936e-06,
+      "loss": 0.2515,
+      "step": 10284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9211988222354697,
+      "learning_rate": 2.930002068675164e-06,
+      "loss": 0.2653,
+      "step": 10285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0755383269524335,
+      "learning_rate": 2.9290750158098268e-06,
+      "loss": 0.2695,
+      "step": 10286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0960936078231605,
+      "learning_rate": 2.9281480488708445e-06,
+      "loss": 0.2752,
+      "step": 10287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2600792081025394,
+      "learning_rate": 2.9272211678966804e-06,
+      "loss": 0.2849,
+      "step": 10288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3490394661035454,
+      "learning_rate": 2.9262943729257924e-06,
+      "loss": 0.2601,
+      "step": 10289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.773377022832766,
+      "learning_rate": 2.9253676639966335e-06,
+      "loss": 0.2684,
+      "step": 10290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.578923030565268,
+      "learning_rate": 2.924441041147656e-06,
+      "loss": 0.4941,
+      "step": 10291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8919456581718603,
+      "learning_rate": 2.923514504417304e-06,
+      "loss": 0.2621,
+      "step": 10292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 17.785325821684957,
+      "learning_rate": 2.9225880538440242e-06,
+      "loss": 0.282,
+      "step": 10293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8583411826395766,
+      "learning_rate": 2.921661689466257e-06,
+      "loss": 0.2702,
+      "step": 10294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3448475890108256,
+      "learning_rate": 2.9207354113224384e-06,
+      "loss": 0.2938,
+      "step": 10295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.936621432202642,
+      "learning_rate": 2.919809219451e-06,
+      "loss": 0.2579,
+      "step": 10296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.199092871572625,
+      "learning_rate": 2.918883113890371e-06,
+      "loss": 0.2857,
+      "step": 10297
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7950546614614975,
+      "learning_rate": 2.9179570946789798e-06,
+      "loss": 0.2776,
+      "step": 10298
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6884177362490773,
+      "learning_rate": 2.9170311618552467e-06,
+      "loss": 0.2708,
+      "step": 10299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.239822309094006,
+      "learning_rate": 2.916105315457588e-06,
+      "loss": 0.2713,
+      "step": 10300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8058828654033277,
+      "learning_rate": 2.9151795555244245e-06,
+      "loss": 0.2629,
+      "step": 10301
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.50441884306904,
+      "learning_rate": 2.9142538820941613e-06,
+      "loss": 0.2506,
+      "step": 10302
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8329288039954224,
+      "learning_rate": 2.913328295205211e-06,
+      "loss": 0.2677,
+      "step": 10303
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.591947523918817,
+      "learning_rate": 2.9124027948959767e-06,
+      "loss": 0.2601,
+      "step": 10304
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9945975029575576,
+      "learning_rate": 2.9114773812048558e-06,
+      "loss": 0.2811,
+      "step": 10305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.414079663222745,
+      "learning_rate": 2.9105520541702503e-06,
+      "loss": 0.2735,
+      "step": 10306
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.046007457779282,
+      "learning_rate": 2.9096268138305495e-06,
+      "loss": 0.2627,
+      "step": 10307
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7599522828206562,
+      "learning_rate": 2.908701660224147e-06,
+      "loss": 0.2625,
+      "step": 10308
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.567266171204114,
+      "learning_rate": 2.9077765933894277e-06,
+      "loss": 0.2823,
+      "step": 10309
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4579231919226723,
+      "learning_rate": 2.906851613364771e-06,
+      "loss": 0.2655,
+      "step": 10310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.240191366744254,
+      "learning_rate": 2.905926720188561e-06,
+      "loss": 0.2702,
+      "step": 10311
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9275614105475722,
+      "learning_rate": 2.905001913899171e-06,
+      "loss": 0.2667,
+      "step": 10312
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.144183315313289,
+      "learning_rate": 2.9040771945349707e-06,
+      "loss": 0.2659,
+      "step": 10313
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1892419343701666,
+      "learning_rate": 2.90315256213433e-06,
+      "loss": 0.2584,
+      "step": 10314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6502800388151524,
+      "learning_rate": 2.9022280167356167e-06,
+      "loss": 0.2601,
+      "step": 10315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7334182686629287,
+      "learning_rate": 2.901303558377188e-06,
+      "loss": 0.2769,
+      "step": 10316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2887259233707633,
+      "learning_rate": 2.9003791870974005e-06,
+      "loss": 0.2708,
+      "step": 10317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9577276044161858,
+      "learning_rate": 2.8994549029346132e-06,
+      "loss": 0.2741,
+      "step": 10318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1935604083815123,
+      "learning_rate": 2.8985307059271718e-06,
+      "loss": 0.2652,
+      "step": 10319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4808680686436608,
+      "learning_rate": 2.897606596113424e-06,
+      "loss": 0.2761,
+      "step": 10320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8842912010224175,
+      "learning_rate": 2.8966825735317113e-06,
+      "loss": 0.255,
+      "step": 10321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4303413346360598,
+      "learning_rate": 2.895758638220374e-06,
+      "loss": 0.2603,
+      "step": 10322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8270048984190406,
+      "learning_rate": 2.894834790217751e-06,
+      "loss": 0.2594,
+      "step": 10323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8392163756216793,
+      "learning_rate": 2.8939110295621707e-06,
+      "loss": 0.2733,
+      "step": 10324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4869095340523093,
+      "learning_rate": 2.892987356291962e-06,
+      "loss": 0.2696,
+      "step": 10325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.135158149137114,
+      "learning_rate": 2.892063770445451e-06,
+      "loss": 0.2605,
+      "step": 10326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7634824739944561,
+      "learning_rate": 2.8911402720609594e-06,
+      "loss": 0.282,
+      "step": 10327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7813903204132482,
+      "learning_rate": 2.8902168611768032e-06,
+      "loss": 0.2767,
+      "step": 10328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.219030505984542,
+      "learning_rate": 2.889293537831295e-06,
+      "loss": 0.2818,
+      "step": 10329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3709590337236626,
+      "learning_rate": 2.8883703020627467e-06,
+      "loss": 0.256,
+      "step": 10330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.326527557524211,
+      "learning_rate": 2.8874471539094672e-06,
+      "loss": 0.2878,
+      "step": 10331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1707468394485088,
+      "learning_rate": 2.8865240934097584e-06,
+      "loss": 0.3013,
+      "step": 10332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6931270236050118,
+      "learning_rate": 2.885601120601916e-06,
+      "loss": 0.2657,
+      "step": 10333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.010592130910668,
+      "learning_rate": 2.884678235524241e-06,
+      "loss": 0.2706,
+      "step": 10334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6936869219379527,
+      "learning_rate": 2.8837554382150233e-06,
+      "loss": 0.2831,
+      "step": 10335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5181472268339018,
+      "learning_rate": 2.882832728712551e-06,
+      "loss": 0.2701,
+      "step": 10336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.255743161270764,
+      "learning_rate": 2.8819101070551073e-06,
+      "loss": 0.2889,
+      "step": 10337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8339844827187541,
+      "learning_rate": 2.8809875732809762e-06,
+      "loss": 0.2728,
+      "step": 10338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.079500585688239,
+      "learning_rate": 2.8800651274284356e-06,
+      "loss": 0.2915,
+      "step": 10339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8361709025280701,
+      "learning_rate": 2.8791427695357586e-06,
+      "loss": 0.2862,
+      "step": 10340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6004733164058902,
+      "learning_rate": 2.8782204996412134e-06,
+      "loss": 0.2578,
+      "step": 10341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7316254866285132,
+      "learning_rate": 2.8772983177830706e-06,
+      "loss": 0.2716,
+      "step": 10342
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9777529165571632,
+      "learning_rate": 2.8763762239995903e-06,
+      "loss": 0.2781,
+      "step": 10343
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7553244499987246,
+      "learning_rate": 2.8754542183290304e-06,
+      "loss": 0.2734,
+      "step": 10344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9787242758944577,
+      "learning_rate": 2.874532300809651e-06,
+      "loss": 0.271,
+      "step": 10345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7816249472016847,
+      "learning_rate": 2.8736104714796996e-06,
+      "loss": 0.258,
+      "step": 10346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.844539267323949,
+      "learning_rate": 2.8726887303774286e-06,
+      "loss": 0.2856,
+      "step": 10347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.110903491931399,
+      "learning_rate": 2.8717670775410805e-06,
+      "loss": 0.2573,
+      "step": 10348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1961716277508603,
+      "learning_rate": 2.8708455130088946e-06,
+      "loss": 0.2925,
+      "step": 10349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.840889344725491,
+      "learning_rate": 2.8699240368191124e-06,
+      "loss": 0.2822,
+      "step": 10350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1900770769306392,
+      "learning_rate": 2.8690026490099655e-06,
+      "loss": 0.281,
+      "step": 10351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.032055040907349,
+      "learning_rate": 2.8680813496196814e-06,
+      "loss": 0.2678,
+      "step": 10352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.83485329391566,
+      "learning_rate": 2.8671601386864913e-06,
+      "loss": 0.2698,
+      "step": 10353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4372661675444647,
+      "learning_rate": 2.8662390162486125e-06,
+      "loss": 0.2813,
+      "step": 10354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8711937602281714,
+      "learning_rate": 2.865317982344269e-06,
+      "loss": 0.274,
+      "step": 10355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.647553746292898,
+      "learning_rate": 2.864397037011675e-06,
+      "loss": 0.2646,
+      "step": 10356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0758115605281473,
+      "learning_rate": 2.8634761802890387e-06,
+      "loss": 0.2634,
+      "step": 10357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8131575878304973,
+      "learning_rate": 2.862555412214572e-06,
+      "loss": 0.2681,
+      "step": 10358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9124368285371574,
+      "learning_rate": 2.8616347328264764e-06,
+      "loss": 0.2602,
+      "step": 10359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6047603012030696,
+      "learning_rate": 2.860714142162956e-06,
+      "loss": 0.4803,
+      "step": 10360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6202814522088258,
+      "learning_rate": 2.859793640262205e-06,
+      "loss": 0.2636,
+      "step": 10361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4631381294144927,
+      "learning_rate": 2.8588732271624163e-06,
+      "loss": 0.2689,
+      "step": 10362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6366857709413358,
+      "learning_rate": 2.857952902901782e-06,
+      "loss": 0.2693,
+      "step": 10363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.974490977238434,
+      "learning_rate": 2.857032667518486e-06,
+      "loss": 0.2676,
+      "step": 10364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.598651836509557,
+      "learning_rate": 2.856112521050709e-06,
+      "loss": 0.2498,
+      "step": 10365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3267187774940883,
+      "learning_rate": 2.8551924635366344e-06,
+      "loss": 0.2692,
+      "step": 10366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.980678960588676,
+      "learning_rate": 2.854272495014431e-06,
+      "loss": 0.2727,
+      "step": 10367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.860173799441823,
+      "learning_rate": 2.8533526155222757e-06,
+      "loss": 0.2635,
+      "step": 10368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0060463510588296,
+      "learning_rate": 2.8524328250983337e-06,
+      "loss": 0.2659,
+      "step": 10369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.350605986020169,
+      "learning_rate": 2.8515131237807653e-06,
+      "loss": 0.2752,
+      "step": 10370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3201057957361955,
+      "learning_rate": 2.8505935116077353e-06,
+      "loss": 0.2608,
+      "step": 10371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5239889432924736,
+      "learning_rate": 2.8496739886173994e-06,
+      "loss": 0.2675,
+      "step": 10372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6003174799229049,
+      "learning_rate": 2.848754554847907e-06,
+      "loss": 0.4657,
+      "step": 10373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.931957096457086,
+      "learning_rate": 2.8478352103374085e-06,
+      "loss": 0.2639,
+      "step": 10374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.554794346000949,
+      "learning_rate": 2.846915955124052e-06,
+      "loss": 0.2705,
+      "step": 10375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.1017749279224756,
+      "learning_rate": 2.8459967892459767e-06,
+      "loss": 0.278,
+      "step": 10376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2290570477432845,
+      "learning_rate": 2.845077712741321e-06,
+      "loss": 0.2522,
+      "step": 10377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7791977780449229,
+      "learning_rate": 2.844158725648216e-06,
+      "loss": 0.2546,
+      "step": 10378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3351526868247428,
+      "learning_rate": 2.843239828004797e-06,
+      "loss": 0.2738,
+      "step": 10379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3032516202419586,
+      "learning_rate": 2.8423210198491886e-06,
+      "loss": 0.2727,
+      "step": 10380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9390686255543936,
+      "learning_rate": 2.8414023012195113e-06,
+      "loss": 0.2549,
+      "step": 10381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0219803962271286,
+      "learning_rate": 2.8404836721538866e-06,
+      "loss": 0.2501,
+      "step": 10382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8175216309949775,
+      "learning_rate": 2.8395651326904323e-06,
+      "loss": 0.2698,
+      "step": 10383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7904621967344234,
+      "learning_rate": 2.8386466828672575e-06,
+      "loss": 0.2797,
+      "step": 10384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.842134789346314,
+      "learning_rate": 2.8377283227224717e-06,
+      "loss": 0.2784,
+      "step": 10385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.694564305938711,
+      "learning_rate": 2.8368100522941755e-06,
+      "loss": 0.2632,
+      "step": 10386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7988356164491621,
+      "learning_rate": 2.8358918716204746e-06,
+      "loss": 0.2983,
+      "step": 10387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9138366241854534,
+      "learning_rate": 2.8349737807394646e-06,
+      "loss": 0.2747,
+      "step": 10388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9613620859420138,
+      "learning_rate": 2.8340557796892353e-06,
+      "loss": 0.2824,
+      "step": 10389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8392574585700325,
+      "learning_rate": 2.833137868507879e-06,
+      "loss": 0.264,
+      "step": 10390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4377198742833721,
+      "learning_rate": 2.832220047233483e-06,
+      "loss": 0.2531,
+      "step": 10391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3127795153622928,
+      "learning_rate": 2.831302315904128e-06,
+      "loss": 0.2677,
+      "step": 10392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.99991517431462,
+      "learning_rate": 2.83038467455789e-06,
+      "loss": 0.2699,
+      "step": 10393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.820705929289855,
+      "learning_rate": 2.8294671232328473e-06,
+      "loss": 0.2634,
+      "step": 10394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.518450643734968,
+      "learning_rate": 2.8285496619670695e-06,
+      "loss": 0.2773,
+      "step": 10395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.074314846461634,
+      "learning_rate": 2.827632290798621e-06,
+      "loss": 0.2569,
+      "step": 10396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5583800976647444,
+      "learning_rate": 2.826715009765569e-06,
+      "loss": 0.2846,
+      "step": 10397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.108231671123516,
+      "learning_rate": 2.825797818905969e-06,
+      "loss": 0.2596,
+      "step": 10398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.240758915529665,
+      "learning_rate": 2.8248807182578817e-06,
+      "loss": 0.2717,
+      "step": 10399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8199852189859267,
+      "learning_rate": 2.8239637078593574e-06,
+      "loss": 0.2542,
+      "step": 10400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6174296789920071,
+      "learning_rate": 2.823046787748441e-06,
+      "loss": 0.4429,
+      "step": 10401
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.456548481553083,
+      "learning_rate": 2.8221299579631834e-06,
+      "loss": 0.2578,
+      "step": 10402
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3983486438540016,
+      "learning_rate": 2.821213218541621e-06,
+      "loss": 0.2788,
+      "step": 10403
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8787150019399343,
+      "learning_rate": 2.8202965695217906e-06,
+      "loss": 0.2491,
+      "step": 10404
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8893076754105873,
+      "learning_rate": 2.8193800109417293e-06,
+      "loss": 0.2709,
+      "step": 10405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1820426995449576,
+      "learning_rate": 2.818463542839462e-06,
+      "loss": 0.274,
+      "step": 10406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5720702494820933,
+      "learning_rate": 2.8175471652530193e-06,
+      "loss": 0.2558,
+      "step": 10407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9245355640229331,
+      "learning_rate": 2.816630878220421e-06,
+      "loss": 0.2593,
+      "step": 10408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.193600663839504,
+      "learning_rate": 2.8157146817796843e-06,
+      "loss": 0.2668,
+      "step": 10409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6029318065473126,
+      "learning_rate": 2.8147985759688267e-06,
+      "loss": 0.2584,
+      "step": 10410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7862824366550654,
+      "learning_rate": 2.8138825608258556e-06,
+      "loss": 0.2584,
+      "step": 10411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.073517752624644,
+      "learning_rate": 2.812966636388782e-06,
+      "loss": 0.2871,
+      "step": 10412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7103306706033008,
+      "learning_rate": 2.8120508026956074e-06,
+      "loss": 0.2525,
+      "step": 10413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1626339198226368,
+      "learning_rate": 2.811135059784329e-06,
+      "loss": 0.273,
+      "step": 10414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.238689838195442,
+      "learning_rate": 2.8102194076929475e-06,
+      "loss": 0.2761,
+      "step": 10415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3273465781146676,
+      "learning_rate": 2.809303846459452e-06,
+      "loss": 0.2752,
+      "step": 10416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5742835740189338,
+      "learning_rate": 2.808388376121829e-06,
+      "loss": 0.2618,
+      "step": 10417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.530856597631842,
+      "learning_rate": 2.8074729967180664e-06,
+      "loss": 0.2651,
+      "step": 10418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.828599931339612,
+      "learning_rate": 2.8065577082861416e-06,
+      "loss": 0.2969,
+      "step": 10419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.966342558520523,
+      "learning_rate": 2.805642510864036e-06,
+      "loss": 0.2579,
+      "step": 10420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.479634855617904,
+      "learning_rate": 2.80472740448972e-06,
+      "loss": 0.2649,
+      "step": 10421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.368448757489006,
+      "learning_rate": 2.8038123892011615e-06,
+      "loss": 0.2586,
+      "step": 10422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.217167700332758,
+      "learning_rate": 2.8028974650363296e-06,
+      "loss": 0.2658,
+      "step": 10423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7521901095895094,
+      "learning_rate": 2.8019826320331843e-06,
+      "loss": 0.2817,
+      "step": 10424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2625006956315874,
+      "learning_rate": 2.8010678902296822e-06,
+      "loss": 0.2641,
+      "step": 10425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 17.121993836236154,
+      "learning_rate": 2.800153239663779e-06,
+      "loss": 0.2673,
+      "step": 10426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.14308314556635,
+      "learning_rate": 2.7992386803734267e-06,
+      "loss": 0.2674,
+      "step": 10427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.809321431847288,
+      "learning_rate": 2.7983242123965708e-06,
+      "loss": 0.2715,
+      "step": 10428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.809139352899553,
+      "learning_rate": 2.7974098357711544e-06,
+      "loss": 0.2771,
+      "step": 10429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5875698629766402,
+      "learning_rate": 2.796495550535113e-06,
+      "loss": 0.2859,
+      "step": 10430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5018999272604305,
+      "learning_rate": 2.795581356726388e-06,
+      "loss": 0.2542,
+      "step": 10431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.149900104850301,
+      "learning_rate": 2.7946672543829077e-06,
+      "loss": 0.2958,
+      "step": 10432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2645565055309302,
+      "learning_rate": 2.7937532435425985e-06,
+      "loss": 0.2633,
+      "step": 10433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1242294696964263,
+      "learning_rate": 2.792839324243386e-06,
+      "loss": 0.2756,
+      "step": 10434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6295723493836345,
+      "learning_rate": 2.791925496523191e-06,
+      "loss": 0.2699,
+      "step": 10435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.182514401554479,
+      "learning_rate": 2.7910117604199305e-06,
+      "loss": 0.2775,
+      "step": 10436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.013516551641431,
+      "learning_rate": 2.7900981159715157e-06,
+      "loss": 0.25,
+      "step": 10437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1994852426873885,
+      "learning_rate": 2.7891845632158527e-06,
+      "loss": 0.2791,
+      "step": 10438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.345011476749204,
+      "learning_rate": 2.788271102190851e-06,
+      "loss": 0.262,
+      "step": 10439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1211271212366243,
+      "learning_rate": 2.7873577329344105e-06,
+      "loss": 0.2467,
+      "step": 10440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.490889428901454,
+      "learning_rate": 2.786444455484425e-06,
+      "loss": 0.2579,
+      "step": 10441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.436044397374049,
+      "learning_rate": 2.785531269878791e-06,
+      "loss": 0.271,
+      "step": 10442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.416322267236971,
+      "learning_rate": 2.784618176155399e-06,
+      "loss": 0.2669,
+      "step": 10443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2415655374759598,
+      "learning_rate": 2.783705174352135e-06,
+      "loss": 0.2725,
+      "step": 10444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3137567313680365,
+      "learning_rate": 2.78279226450688e-06,
+      "loss": 0.2933,
+      "step": 10445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.787664114700681,
+      "learning_rate": 2.7818794466575095e-06,
+      "loss": 0.2632,
+      "step": 10446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5499048197975385,
+      "learning_rate": 2.7809667208419034e-06,
+      "loss": 0.2829,
+      "step": 10447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.293831891826393,
+      "learning_rate": 2.7800540870979287e-06,
+      "loss": 0.2886,
+      "step": 10448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2140332473576787,
+      "learning_rate": 2.7791415454634507e-06,
+      "loss": 0.2527,
+      "step": 10449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3097334048601796,
+      "learning_rate": 2.778229095976336e-06,
+      "loss": 0.2625,
+      "step": 10450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0738810231232616,
+      "learning_rate": 2.7773167386744432e-06,
+      "loss": 0.2872,
+      "step": 10451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.570713387112575,
+      "learning_rate": 2.7764044735956275e-06,
+      "loss": 0.265,
+      "step": 10452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7532889903374367,
+      "learning_rate": 2.775492300777739e-06,
+      "loss": 0.2567,
+      "step": 10453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0799293435877386,
+      "learning_rate": 2.774580220258625e-06,
+      "loss": 0.2694,
+      "step": 10454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0581896280321694,
+      "learning_rate": 2.773668232076132e-06,
+      "loss": 0.2703,
+      "step": 10455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9368825258630105,
+      "learning_rate": 2.7727563362680965e-06,
+      "loss": 0.2723,
+      "step": 10456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7194825863986827,
+      "learning_rate": 2.771844532872359e-06,
+      "loss": 0.2633,
+      "step": 10457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6965752359933366,
+      "learning_rate": 2.770932821926747e-06,
+      "loss": 0.2712,
+      "step": 10458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9523782752312764,
+      "learning_rate": 2.7700212034690933e-06,
+      "loss": 0.3004,
+      "step": 10459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6194571393296509,
+      "learning_rate": 2.769109677537222e-06,
+      "loss": 0.475,
+      "step": 10460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.743386645198735,
+      "learning_rate": 2.7681982441689513e-06,
+      "loss": 0.2706,
+      "step": 10461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4932330766273485,
+      "learning_rate": 2.7672869034020978e-06,
+      "loss": 0.2814,
+      "step": 10462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.197278076937047,
+      "learning_rate": 2.766375655274479e-06,
+      "loss": 0.2517,
+      "step": 10463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9752372692344045,
+      "learning_rate": 2.765464499823899e-06,
+      "loss": 0.2739,
+      "step": 10464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5880826352631157,
+      "learning_rate": 2.7645534370881682e-06,
+      "loss": 0.2636,
+      "step": 10465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.308944477965517,
+      "learning_rate": 2.7636424671050843e-06,
+      "loss": 0.2867,
+      "step": 10466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.810625758286272,
+      "learning_rate": 2.762731589912448e-06,
+      "loss": 0.2854,
+      "step": 10467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.024849569191932,
+      "learning_rate": 2.7618208055480523e-06,
+      "loss": 0.2541,
+      "step": 10468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2611604161150793,
+      "learning_rate": 2.7609101140496863e-06,
+      "loss": 0.5138,
+      "step": 10469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8194719512935786,
+      "learning_rate": 2.7599995154551352e-06,
+      "loss": 0.2866,
+      "step": 10470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.858857505565335,
+      "learning_rate": 2.7590890098021828e-06,
+      "loss": 0.2742,
+      "step": 10471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.101557608251054,
+      "learning_rate": 2.75817859712861e-06,
+      "loss": 0.2592,
+      "step": 10472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.183406893174561,
+      "learning_rate": 2.757268277472188e-06,
+      "loss": 0.2591,
+      "step": 10473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.952509958724208,
+      "learning_rate": 2.7563580508706877e-06,
+      "loss": 0.2596,
+      "step": 10474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.351708371100462,
+      "learning_rate": 2.755447917361879e-06,
+      "loss": 0.2709,
+      "step": 10475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.05542319249982,
+      "learning_rate": 2.754537876983523e-06,
+      "loss": 0.2849,
+      "step": 10476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.067770469205034,
+      "learning_rate": 2.753627929773377e-06,
+      "loss": 0.2716,
+      "step": 10477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.784719241808555,
+      "learning_rate": 2.7527180757691973e-06,
+      "loss": 0.2702,
+      "step": 10478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4259988908413455,
+      "learning_rate": 2.7518083150087395e-06,
+      "loss": 0.2831,
+      "step": 10479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6593522983902174,
+      "learning_rate": 2.750898647529747e-06,
+      "loss": 0.2487,
+      "step": 10480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7200176513562162,
+      "learning_rate": 2.7499890733699645e-06,
+      "loss": 0.2582,
+      "step": 10481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.286478809297931,
+      "learning_rate": 2.74907959256713e-06,
+      "loss": 0.258,
+      "step": 10482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8739935265522478,
+      "learning_rate": 2.748170205158984e-06,
+      "loss": 0.2632,
+      "step": 10483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6781717455937804,
+      "learning_rate": 2.747260911183255e-06,
+      "loss": 0.2756,
+      "step": 10484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.75517122915535,
+      "learning_rate": 2.7463517106776704e-06,
+      "loss": 0.256,
+      "step": 10485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2610652072859203,
+      "learning_rate": 2.7454426036799566e-06,
+      "loss": 0.2699,
+      "step": 10486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1226346728576484,
+      "learning_rate": 2.7445335902278347e-06,
+      "loss": 0.2606,
+      "step": 10487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2975074133000453,
+      "learning_rate": 2.7436246703590206e-06,
+      "loss": 0.2949,
+      "step": 10488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 35.896181409660656,
+      "learning_rate": 2.742715844111228e-06,
+      "loss": 0.2694,
+      "step": 10489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3142204998937435,
+      "learning_rate": 2.7418071115221613e-06,
+      "loss": 0.2709,
+      "step": 10490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.391682267713271,
+      "learning_rate": 2.740898472629531e-06,
+      "loss": 0.2739,
+      "step": 10491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6642260926130596,
+      "learning_rate": 2.7399899274710346e-06,
+      "loss": 0.2461,
+      "step": 10492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.789876276801303,
+      "learning_rate": 2.7390814760843695e-06,
+      "loss": 0.2763,
+      "step": 10493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5921129456610974,
+      "learning_rate": 2.738173118507229e-06,
+      "loss": 0.2837,
+      "step": 10494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2845068057407825,
+      "learning_rate": 2.7372648547773063e-06,
+      "loss": 0.263,
+      "step": 10495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.989157271413783,
+      "learning_rate": 2.736356684932283e-06,
+      "loss": 0.2817,
+      "step": 10496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.220640561908065,
+      "learning_rate": 2.7354486090098414e-06,
+      "loss": 0.254,
+      "step": 10497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5036794369636048,
+      "learning_rate": 2.734540627047658e-06,
+      "loss": 0.2701,
+      "step": 10498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7119482713201197,
+      "learning_rate": 2.7336327390834093e-06,
+      "loss": 0.2751,
+      "step": 10499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.766943979785282,
+      "learning_rate": 2.7327249451547642e-06,
+      "loss": 0.2656,
+      "step": 10500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3265290673315775,
+      "learning_rate": 2.7318172452993864e-06,
+      "loss": 0.2688,
+      "step": 10501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4995671438417246,
+      "learning_rate": 2.7309096395549395e-06,
+      "loss": 0.2785,
+      "step": 10502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7718723163471015,
+      "learning_rate": 2.730002127959084e-06,
+      "loss": 0.2642,
+      "step": 10503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7948091013136696,
+      "learning_rate": 2.729094710549472e-06,
+      "loss": 0.2661,
+      "step": 10504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3992510558784863,
+      "learning_rate": 2.728187387363754e-06,
+      "loss": 0.266,
+      "step": 10505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4472813449806115,
+      "learning_rate": 2.727280158439575e-06,
+      "loss": 0.2615,
+      "step": 10506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.843309802113135,
+      "learning_rate": 2.726373023814581e-06,
+      "loss": 0.2763,
+      "step": 10507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.76893973000872,
+      "learning_rate": 2.7254659835264064e-06,
+      "loss": 0.2833,
+      "step": 10508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.059903131021249,
+      "learning_rate": 2.7245590376126895e-06,
+      "loss": 0.2599,
+      "step": 10509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.791626822790956,
+      "learning_rate": 2.7236521861110586e-06,
+      "loss": 0.2861,
+      "step": 10510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.1410902440432285,
+      "learning_rate": 2.722745429059144e-06,
+      "loss": 0.2682,
+      "step": 10511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5038795765002517,
+      "learning_rate": 2.721838766494566e-06,
+      "loss": 0.2631,
+      "step": 10512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.661065644468989,
+      "learning_rate": 2.720932198454944e-06,
+      "loss": 0.2746,
+      "step": 10513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.049903290024286,
+      "learning_rate": 2.720025724977892e-06,
+      "loss": 0.2755,
+      "step": 10514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.268990141101887,
+      "learning_rate": 2.719119346101023e-06,
+      "loss": 0.2765,
+      "step": 10515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.02605309083175,
+      "learning_rate": 2.7182130618619423e-06,
+      "loss": 0.2684,
+      "step": 10516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6260057395368743,
+      "learning_rate": 2.7173068722982566e-06,
+      "loss": 0.2691,
+      "step": 10517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3232572660315918,
+      "learning_rate": 2.716400777447561e-06,
+      "loss": 0.274,
+      "step": 10518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.1385260684203,
+      "learning_rate": 2.7154947773474556e-06,
+      "loss": 0.2659,
+      "step": 10519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6609671404572608,
+      "learning_rate": 2.7145888720355297e-06,
+      "loss": 0.2648,
+      "step": 10520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8010479623478184,
+      "learning_rate": 2.71368306154937e-06,
+      "loss": 0.2567,
+      "step": 10521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8557498651632571,
+      "learning_rate": 2.7127773459265604e-06,
+      "loss": 0.2804,
+      "step": 10522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6119371441864176,
+      "learning_rate": 2.71187172520468e-06,
+      "loss": 0.2511,
+      "step": 10523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1180144127404255,
+      "learning_rate": 2.710966199421309e-06,
+      "loss": 0.2618,
+      "step": 10524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1346526102771137,
+      "learning_rate": 2.7100607686140155e-06,
+      "loss": 0.2676,
+      "step": 10525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.39223189442905,
+      "learning_rate": 2.709155432820366e-06,
+      "loss": 0.2747,
+      "step": 10526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5559264437866219,
+      "learning_rate": 2.7082501920779293e-06,
+      "loss": 0.4723,
+      "step": 10527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5950884134719407,
+      "learning_rate": 2.707345046424262e-06,
+      "loss": 0.2625,
+      "step": 10528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.627215776788749,
+      "learning_rate": 2.706439995896921e-06,
+      "loss": 0.2647,
+      "step": 10529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 27.149907227318852,
+      "learning_rate": 2.705535040533457e-06,
+      "loss": 0.275,
+      "step": 10530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0678446646933226,
+      "learning_rate": 2.7046301803714194e-06,
+      "loss": 0.2686,
+      "step": 10531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4813309748440227,
+      "learning_rate": 2.703725415448354e-06,
+      "loss": 0.2779,
+      "step": 10532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2494720936144463,
+      "learning_rate": 2.7028207458017996e-06,
+      "loss": 0.2647,
+      "step": 10533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0571871680881295,
+      "learning_rate": 2.701916171469292e-06,
+      "loss": 0.2637,
+      "step": 10534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8114706590401224,
+      "learning_rate": 2.7010116924883654e-06,
+      "loss": 0.2637,
+      "step": 10535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.136596359366606,
+      "learning_rate": 2.7001073088965467e-06,
+      "loss": 0.2782,
+      "step": 10536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7784525920721663,
+      "learning_rate": 2.699203020731362e-06,
+      "loss": 0.2686,
+      "step": 10537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7372657869244117,
+      "learning_rate": 2.6982988280303255e-06,
+      "loss": 0.2945,
+      "step": 10538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6775509141565044,
+      "learning_rate": 2.6973947308309647e-06,
+      "loss": 0.2817,
+      "step": 10539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6063032286178625,
+      "learning_rate": 2.6964907291707844e-06,
+      "loss": 0.2749,
+      "step": 10540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.037502628470929,
+      "learning_rate": 2.6955868230872963e-06,
+      "loss": 0.2749,
+      "step": 10541
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3071206627776366,
+      "learning_rate": 2.6946830126180016e-06,
+      "loss": 0.2795,
+      "step": 10542
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5233088589992856,
+      "learning_rate": 2.6937792978004056e-06,
+      "loss": 0.2677,
+      "step": 10543
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1354782356786717,
+      "learning_rate": 2.6928756786720026e-06,
+      "loss": 0.2873,
+      "step": 10544
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4270493297206324,
+      "learning_rate": 2.691972155270286e-06,
+      "loss": 0.2619,
+      "step": 10545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.020419575907655,
+      "learning_rate": 2.69106872763274e-06,
+      "loss": 0.2749,
+      "step": 10546
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7673720866108846,
+      "learning_rate": 2.6901653957968577e-06,
+      "loss": 0.2802,
+      "step": 10547
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.578574726522088,
+      "learning_rate": 2.6892621598001157e-06,
+      "loss": 0.2695,
+      "step": 10548
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9129988722106914,
+      "learning_rate": 2.6883590196799913e-06,
+      "loss": 0.2724,
+      "step": 10549
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9851458303620775,
+      "learning_rate": 2.687455975473955e-06,
+      "loss": 0.275,
+      "step": 10550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2185346204376484,
+      "learning_rate": 2.6865530272194796e-06,
+      "loss": 0.2657,
+      "step": 10551
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.017259300764955,
+      "learning_rate": 2.6856501749540287e-06,
+      "loss": 0.2625,
+      "step": 10552
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8912869264547008,
+      "learning_rate": 2.6847474187150603e-06,
+      "loss": 0.256,
+      "step": 10553
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.946934685722033,
+      "learning_rate": 2.683844758540034e-06,
+      "loss": 0.2483,
+      "step": 10554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.2920732445097354,
+      "learning_rate": 2.682942194466405e-06,
+      "loss": 0.2595,
+      "step": 10555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5986369436181105,
+      "learning_rate": 2.682039726531619e-06,
+      "loss": 0.482,
+      "step": 10556
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.0339234045457175,
+      "learning_rate": 2.6811373547731224e-06,
+      "loss": 0.2664,
+      "step": 10557
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6464528118677992,
+      "learning_rate": 2.680235079228354e-06,
+      "loss": 0.2931,
+      "step": 10558
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.344006149002412,
+      "learning_rate": 2.6793328999347546e-06,
+      "loss": 0.2614,
+      "step": 10559
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5891767476746761,
+      "learning_rate": 2.6784308169297525e-06,
+      "loss": 0.4672,
+      "step": 10560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2521914885194227,
+      "learning_rate": 2.677528830250782e-06,
+      "loss": 0.2643,
+      "step": 10561
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5716718943515395,
+      "learning_rate": 2.6766269399352628e-06,
+      "loss": 0.4468,
+      "step": 10562
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8879084440995968,
+      "learning_rate": 2.6757251460206215e-06,
+      "loss": 0.2757,
+      "step": 10563
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6270430444033495,
+      "learning_rate": 2.6748234485442713e-06,
+      "loss": 0.2806,
+      "step": 10564
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.943469820556087,
+      "learning_rate": 2.6739218475436267e-06,
+      "loss": 0.2607,
+      "step": 10565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8286178782546847,
+      "learning_rate": 2.6730203430560946e-06,
+      "loss": 0.2743,
+      "step": 10566
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9813317388532967,
+      "learning_rate": 2.6721189351190835e-06,
+      "loss": 0.2587,
+      "step": 10567
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6332584518626676,
+      "learning_rate": 2.6712176237699907e-06,
+      "loss": 0.2704,
+      "step": 10568
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8949438163141363,
+      "learning_rate": 2.6703164090462164e-06,
+      "loss": 0.2613,
+      "step": 10569
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4791171219959713,
+      "learning_rate": 2.66941529098515e-06,
+      "loss": 0.2728,
+      "step": 10570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0272497452550495,
+      "learning_rate": 2.668514269624186e-06,
+      "loss": 0.253,
+      "step": 10571
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5323856618951797,
+      "learning_rate": 2.6676133450007053e-06,
+      "loss": 0.267,
+      "step": 10572
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.572834718520556,
+      "learning_rate": 2.66671251715209e-06,
+      "loss": 0.2722,
+      "step": 10573
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.872051124777162,
+      "learning_rate": 2.6658117861157146e-06,
+      "loss": 0.2553,
+      "step": 10574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.555297755025289,
+      "learning_rate": 2.6649111519289537e-06,
+      "loss": 0.281,
+      "step": 10575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6974026491636993,
+      "learning_rate": 2.66401061462918e-06,
+      "loss": 0.2612,
+      "step": 10576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7852817265602523,
+      "learning_rate": 2.663110174253754e-06,
+      "loss": 0.295,
+      "step": 10577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4251423398342096,
+      "learning_rate": 2.6622098308400364e-06,
+      "loss": 0.286,
+      "step": 10578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9080344692995987,
+      "learning_rate": 2.6613095844253866e-06,
+      "loss": 0.2485,
+      "step": 10579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5804396059148693,
+      "learning_rate": 2.6604094350471564e-06,
+      "loss": 0.2516,
+      "step": 10580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6728663237700423,
+      "learning_rate": 2.6595093827426942e-06,
+      "loss": 0.2954,
+      "step": 10581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7142380925622875,
+      "learning_rate": 2.6586094275493435e-06,
+      "loss": 0.2792,
+      "step": 10582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6355585114588722,
+      "learning_rate": 2.6577095695044452e-06,
+      "loss": 0.254,
+      "step": 10583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.617590949599532,
+      "learning_rate": 2.65680980864534e-06,
+      "loss": 0.2618,
+      "step": 10584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.231354730375861,
+      "learning_rate": 2.655910145009358e-06,
+      "loss": 0.2841,
+      "step": 10585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0582960429211345,
+      "learning_rate": 2.6550105786338255e-06,
+      "loss": 0.2628,
+      "step": 10586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.722263420641252,
+      "learning_rate": 2.6541111095560713e-06,
+      "loss": 0.2611,
+      "step": 10587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7153358268931276,
+      "learning_rate": 2.6532117378134138e-06,
+      "loss": 0.2621,
+      "step": 10588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.917108995509193,
+      "learning_rate": 2.6523124634431698e-06,
+      "loss": 0.2591,
+      "step": 10589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4969706127629552,
+      "learning_rate": 2.6514132864826477e-06,
+      "loss": 0.2656,
+      "step": 10590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.9993401405330005,
+      "learning_rate": 2.6505142069691636e-06,
+      "loss": 0.2675,
+      "step": 10591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.058974857078737,
+      "learning_rate": 2.6496152249400187e-06,
+      "loss": 0.2601,
+      "step": 10592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.701421722075308,
+      "learning_rate": 2.648716340432512e-06,
+      "loss": 0.2703,
+      "step": 10593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6556566591839421,
+      "learning_rate": 2.647817553483939e-06,
+      "loss": 0.2735,
+      "step": 10594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7339461124670978,
+      "learning_rate": 2.646918864131596e-06,
+      "loss": 0.261,
+      "step": 10595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.31932599983908,
+      "learning_rate": 2.6460202724127693e-06,
+      "loss": 0.2598,
+      "step": 10596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7623526986928681,
+      "learning_rate": 2.645121778364742e-06,
+      "loss": 0.2889,
+      "step": 10597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1739857003218686,
+      "learning_rate": 2.644223382024791e-06,
+      "loss": 0.2743,
+      "step": 10598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.187989754663518,
+      "learning_rate": 2.6433250834301998e-06,
+      "loss": 0.2596,
+      "step": 10599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4144606729819666,
+      "learning_rate": 2.6424268826182377e-06,
+      "loss": 0.2605,
+      "step": 10600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.590019006914779,
+      "learning_rate": 2.6415287796261707e-06,
+      "loss": 0.2806,
+      "step": 10601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5424578813493213,
+      "learning_rate": 2.640630774491262e-06,
+      "loss": 0.2729,
+      "step": 10602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7529690132675753,
+      "learning_rate": 2.639732867250776e-06,
+      "loss": 0.2659,
+      "step": 10603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8640262209601803,
+      "learning_rate": 2.6388350579419646e-06,
+      "loss": 0.2669,
+      "step": 10604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.213846652562735,
+      "learning_rate": 2.637937346602079e-06,
+      "loss": 0.2823,
+      "step": 10605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6387929469524227,
+      "learning_rate": 2.6370397332683684e-06,
+      "loss": 0.2788,
+      "step": 10606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.876297500016865,
+      "learning_rate": 2.636142217978078e-06,
+      "loss": 0.2602,
+      "step": 10607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.678408696886539,
+      "learning_rate": 2.6352448007684466e-06,
+      "loss": 0.2888,
+      "step": 10608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9014697504973876,
+      "learning_rate": 2.634347481676708e-06,
+      "loss": 0.2664,
+      "step": 10609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0913355328851155,
+      "learning_rate": 2.6334502607400923e-06,
+      "loss": 0.2727,
+      "step": 10610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.120909900148851,
+      "learning_rate": 2.6325531379958314e-06,
+      "loss": 0.2548,
+      "step": 10611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.507661159249386,
+      "learning_rate": 2.631656113481145e-06,
+      "loss": 0.2621,
+      "step": 10612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.049524855023579,
+      "learning_rate": 2.6307591872332514e-06,
+      "loss": 0.2565,
+      "step": 10613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9332949503970627,
+      "learning_rate": 2.6298623592893676e-06,
+      "loss": 0.2923,
+      "step": 10614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6203202242253394,
+      "learning_rate": 2.628965629686706e-06,
+      "loss": 0.2765,
+      "step": 10615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.835225537225523,
+      "learning_rate": 2.628068998462472e-06,
+      "loss": 0.2629,
+      "step": 10616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.322081098431167,
+      "learning_rate": 2.627172465653868e-06,
+      "loss": 0.2647,
+      "step": 10617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.072891592974846,
+      "learning_rate": 2.6262760312980914e-06,
+      "loss": 0.2528,
+      "step": 10618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4635736856140795,
+      "learning_rate": 2.62537969543234e-06,
+      "loss": 0.2692,
+      "step": 10619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6192277971923672,
+      "learning_rate": 2.6244834580938016e-06,
+      "loss": 0.2586,
+      "step": 10620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.180052056443318,
+      "learning_rate": 2.623587319319665e-06,
+      "loss": 0.2631,
+      "step": 10621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1416244976145387,
+      "learning_rate": 2.6226912791471103e-06,
+      "loss": 0.2495,
+      "step": 10622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4886024493190455,
+      "learning_rate": 2.6217953376133187e-06,
+      "loss": 0.2619,
+      "step": 10623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.510835148698963,
+      "learning_rate": 2.6208994947554626e-06,
+      "loss": 0.2615,
+      "step": 10624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.400618012820244,
+      "learning_rate": 2.620003750610712e-06,
+      "loss": 0.2911,
+      "step": 10625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.696640786017668,
+      "learning_rate": 2.6191081052162315e-06,
+      "loss": 0.2587,
+      "step": 10626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2871950304100317,
+      "learning_rate": 2.6182125586091867e-06,
+      "loss": 0.3058,
+      "step": 10627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6835677362491857,
+      "learning_rate": 2.6173171108267316e-06,
+      "loss": 0.2666,
+      "step": 10628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.260840460070348,
+      "learning_rate": 2.6164217619060234e-06,
+      "loss": 0.2759,
+      "step": 10629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3818412361050458,
+      "learning_rate": 2.615526511884208e-06,
+      "loss": 0.2622,
+      "step": 10630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.523975425219532,
+      "learning_rate": 2.6146313607984355e-06,
+      "loss": 0.2578,
+      "step": 10631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8168268645072094,
+      "learning_rate": 2.6137363086858435e-06,
+      "loss": 0.2821,
+      "step": 10632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7028858622010878,
+      "learning_rate": 2.612841355583571e-06,
+      "loss": 0.2742,
+      "step": 10633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2438395892141583,
+      "learning_rate": 2.6119465015287493e-06,
+      "loss": 0.2598,
+      "step": 10634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5397188637521444,
+      "learning_rate": 2.6110517465585085e-06,
+      "loss": 0.2648,
+      "step": 10635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8103106211290358,
+      "learning_rate": 2.610157090709976e-06,
+      "loss": 0.2634,
+      "step": 10636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.546450720595291,
+      "learning_rate": 2.6092625340202703e-06,
+      "loss": 0.2728,
+      "step": 10637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8677470316918625,
+      "learning_rate": 2.6083680765265073e-06,
+      "loss": 0.266,
+      "step": 10638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6099796704289258,
+      "learning_rate": 2.607473718265802e-06,
+      "loss": 0.2597,
+      "step": 10639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8982228396210488,
+      "learning_rate": 2.6065794592752623e-06,
+      "loss": 0.2763,
+      "step": 10640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7486909185828226,
+      "learning_rate": 2.6056852995919918e-06,
+      "loss": 0.268,
+      "step": 10641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8147658964351727,
+      "learning_rate": 2.604791239253089e-06,
+      "loss": 0.2676,
+      "step": 10642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 16.94077740106074,
+      "learning_rate": 2.6038972782956516e-06,
+      "loss": 0.2558,
+      "step": 10643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0246544360765015,
+      "learning_rate": 2.6030034167567752e-06,
+      "loss": 0.2569,
+      "step": 10644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.750501290100431,
+      "learning_rate": 2.6021096546735438e-06,
+      "loss": 0.2758,
+      "step": 10645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.559194212820961,
+      "learning_rate": 2.60121599208304e-06,
+      "loss": 0.2756,
+      "step": 10646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4951093445217885,
+      "learning_rate": 2.600322429022347e-06,
+      "loss": 0.2699,
+      "step": 10647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6597456654934737,
+      "learning_rate": 2.5994289655285396e-06,
+      "loss": 0.2889,
+      "step": 10648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5161201312991648,
+      "learning_rate": 2.5985356016386883e-06,
+      "loss": 0.2552,
+      "step": 10649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0867449553152335,
+      "learning_rate": 2.597642337389858e-06,
+      "loss": 0.2597,
+      "step": 10650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2875975532603436,
+      "learning_rate": 2.596749172819114e-06,
+      "loss": 0.2604,
+      "step": 10651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5626438460521015,
+      "learning_rate": 2.595856107963518e-06,
+      "loss": 0.2703,
+      "step": 10652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6844249062005776,
+      "learning_rate": 2.5949631428601218e-06,
+      "loss": 0.2707,
+      "step": 10653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.084937417918076,
+      "learning_rate": 2.594070277545975e-06,
+      "loss": 0.2796,
+      "step": 10654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.587092909145393,
+      "learning_rate": 2.593177512058127e-06,
+      "loss": 0.2475,
+      "step": 10655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6498549567788552,
+      "learning_rate": 2.5922848464336203e-06,
+      "loss": 0.4876,
+      "step": 10656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9631345189313008,
+      "learning_rate": 2.5913922807094894e-06,
+      "loss": 0.2555,
+      "step": 10657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.348123254239563,
+      "learning_rate": 2.590499814922772e-06,
+      "loss": 0.2619,
+      "step": 10658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.464586238642134,
+      "learning_rate": 2.5896074491104963e-06,
+      "loss": 0.2535,
+      "step": 10659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.929509701228109,
+      "learning_rate": 2.58871518330969e-06,
+      "loss": 0.2537,
+      "step": 10660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5794823207730463,
+      "learning_rate": 2.5878230175573743e-06,
+      "loss": 0.49,
+      "step": 10661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.523906905390259,
+      "learning_rate": 2.586930951890564e-06,
+      "loss": 0.2549,
+      "step": 10662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0591089722327425,
+      "learning_rate": 2.5860389863462765e-06,
+      "loss": 0.2622,
+      "step": 10663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.3123396773365945,
+      "learning_rate": 2.5851471209615186e-06,
+      "loss": 0.2466,
+      "step": 10664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9443522141802896,
+      "learning_rate": 2.5842553557732953e-06,
+      "loss": 0.2563,
+      "step": 10665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.588902182894885,
+      "learning_rate": 2.5833636908186064e-06,
+      "loss": 0.269,
+      "step": 10666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.672968990815488,
+      "learning_rate": 2.582472126134454e-06,
+      "loss": 0.2626,
+      "step": 10667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8489354443635064,
+      "learning_rate": 2.581580661757826e-06,
+      "loss": 0.2627,
+      "step": 10668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4598953094511153,
+      "learning_rate": 2.5806892977257126e-06,
+      "loss": 0.2863,
+      "step": 10669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6698336809258087,
+      "learning_rate": 2.579798034075095e-06,
+      "loss": 0.2629,
+      "step": 10670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.629764415185163,
+      "learning_rate": 2.5789068708429576e-06,
+      "loss": 0.2706,
+      "step": 10671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0127877473725233,
+      "learning_rate": 2.578015808066273e-06,
+      "loss": 0.2604,
+      "step": 10672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.507531804997822,
+      "learning_rate": 2.5771248457820165e-06,
+      "loss": 0.2522,
+      "step": 10673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7626399370202608,
+      "learning_rate": 2.5762339840271513e-06,
+      "loss": 0.255,
+      "step": 10674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5260960041805964,
+      "learning_rate": 2.575343222838645e-06,
+      "loss": 0.2523,
+      "step": 10675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6774592924681375,
+      "learning_rate": 2.574452562253455e-06,
+      "loss": 0.2879,
+      "step": 10676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.077359334144614,
+      "learning_rate": 2.5735620023085367e-06,
+      "loss": 0.2625,
+      "step": 10677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2258704980151838,
+      "learning_rate": 2.572671543040839e-06,
+      "loss": 0.2721,
+      "step": 10678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.357607266475697,
+      "learning_rate": 2.571781184487312e-06,
+      "loss": 0.2816,
+      "step": 10679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4169846712490002,
+      "learning_rate": 2.570890926684895e-06,
+      "loss": 0.2798,
+      "step": 10680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.784672021914526,
+      "learning_rate": 2.57000076967053e-06,
+      "loss": 0.2625,
+      "step": 10681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.240635996638068,
+      "learning_rate": 2.569110713481147e-06,
+      "loss": 0.2633,
+      "step": 10682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.615192306625452,
+      "learning_rate": 2.56822075815368e-06,
+      "loss": 0.2609,
+      "step": 10683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6451409067071499,
+      "learning_rate": 2.567330903725054e-06,
+      "loss": 0.276,
+      "step": 10684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3289556990694675,
+      "learning_rate": 2.566441150232189e-06,
+      "loss": 0.276,
+      "step": 10685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7213448987786697,
+      "learning_rate": 2.5655514977120013e-06,
+      "loss": 0.2786,
+      "step": 10686
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4652025240297863,
+      "learning_rate": 2.5646619462014062e-06,
+      "loss": 0.2752,
+      "step": 10687
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.09302677785505,
+      "learning_rate": 2.5637724957373144e-06,
+      "loss": 0.2726,
+      "step": 10688
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7267881270737377,
+      "learning_rate": 2.5628831463566285e-06,
+      "loss": 0.2611,
+      "step": 10689
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.744178529246718,
+      "learning_rate": 2.561993898096249e-06,
+      "loss": 0.2617,
+      "step": 10690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8202378836545405,
+      "learning_rate": 2.5611047509930737e-06,
+      "loss": 0.259,
+      "step": 10691
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.862782784274091,
+      "learning_rate": 2.560215705083995e-06,
+      "loss": 0.2537,
+      "step": 10692
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8719111448227204,
+      "learning_rate": 2.5593267604058998e-06,
+      "loss": 0.2604,
+      "step": 10693
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.139683591121575,
+      "learning_rate": 2.558437916995671e-06,
+      "loss": 0.2696,
+      "step": 10694
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.59879788854556,
+      "learning_rate": 2.5575491748901892e-06,
+      "loss": 0.2744,
+      "step": 10695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9827703484918457,
+      "learning_rate": 2.5566605341263322e-06,
+      "loss": 0.2725,
+      "step": 10696
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7400888471913067,
+      "learning_rate": 2.55577199474097e-06,
+      "loss": 0.2694,
+      "step": 10697
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6260521221584945,
+      "learning_rate": 2.5548835567709672e-06,
+      "loss": 0.5226,
+      "step": 10698
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9467036804749311,
+      "learning_rate": 2.5539952202531905e-06,
+      "loss": 0.2886,
+      "step": 10699
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.751481328040724,
+      "learning_rate": 2.5531069852244968e-06,
+      "loss": 0.2718,
+      "step": 10700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5923047119665359,
+      "learning_rate": 2.5522188517217404e-06,
+      "loss": 0.4505,
+      "step": 10701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.921487192719859,
+      "learning_rate": 2.551330819781769e-06,
+      "loss": 0.261,
+      "step": 10702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5990451619774673,
+      "learning_rate": 2.5504428894414323e-06,
+      "loss": 0.4841,
+      "step": 10703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3902303604684936,
+      "learning_rate": 2.5495550607375723e-06,
+      "loss": 0.2637,
+      "step": 10704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8548515651860293,
+      "learning_rate": 2.5486673337070254e-06,
+      "loss": 0.2663,
+      "step": 10705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.238012333062435,
+      "learning_rate": 2.5477797083866227e-06,
+      "loss": 0.2698,
+      "step": 10706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7506722369020096,
+      "learning_rate": 2.5468921848131984e-06,
+      "loss": 0.2593,
+      "step": 10707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8529942145505167,
+      "learning_rate": 2.546004763023574e-06,
+      "loss": 0.2758,
+      "step": 10708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5379960918388584,
+      "learning_rate": 2.5451174430545704e-06,
+      "loss": 0.2713,
+      "step": 10709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5288099794809744,
+      "learning_rate": 2.5442302249430027e-06,
+      "loss": 0.2624,
+      "step": 10710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.197753413877993,
+      "learning_rate": 2.5433431087256853e-06,
+      "loss": 0.2876,
+      "step": 10711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7506673837411795,
+      "learning_rate": 2.542456094439427e-06,
+      "loss": 0.247,
+      "step": 10712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4603022809404356,
+      "learning_rate": 2.5415691821210314e-06,
+      "loss": 0.2709,
+      "step": 10713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8888431331421702,
+      "learning_rate": 2.540682371807295e-06,
+      "loss": 0.2782,
+      "step": 10714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0378610460366953,
+      "learning_rate": 2.5397956635350164e-06,
+      "loss": 0.257,
+      "step": 10715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4467611374502765,
+      "learning_rate": 2.5389090573409863e-06,
+      "loss": 0.2741,
+      "step": 10716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.68515593439781,
+      "learning_rate": 2.5380225532619886e-06,
+      "loss": 0.2544,
+      "step": 10717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8409282143245345,
+      "learning_rate": 2.5371361513348103e-06,
+      "loss": 0.2771,
+      "step": 10718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6940132298961257,
+      "learning_rate": 2.5362498515962253e-06,
+      "loss": 0.2586,
+      "step": 10719
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.183619629229947,
+      "learning_rate": 2.5353636540830117e-06,
+      "loss": 0.2792,
+      "step": 10720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.510278437598951,
+      "learning_rate": 2.534477558831938e-06,
+      "loss": 0.259,
+      "step": 10721
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6693127198703324,
+      "learning_rate": 2.533591565879768e-06,
+      "loss": 0.265,
+      "step": 10722
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8263344850511583,
+      "learning_rate": 2.5327056752632674e-06,
+      "loss": 0.2586,
+      "step": 10723
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.214349382760798,
+      "learning_rate": 2.5318198870191877e-06,
+      "loss": 0.2747,
+      "step": 10724
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6365207180987742,
+      "learning_rate": 2.530934201184287e-06,
+      "loss": 0.2497,
+      "step": 10725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2915169082627025,
+      "learning_rate": 2.5300486177953123e-06,
+      "loss": 0.2671,
+      "step": 10726
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.352748126876919,
+      "learning_rate": 2.5291631368890047e-06,
+      "loss": 0.2549,
+      "step": 10727
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.764105496513473,
+      "learning_rate": 2.52827775850211e-06,
+      "loss": 0.2692,
+      "step": 10728
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7484368178982617,
+      "learning_rate": 2.5273924826713615e-06,
+      "loss": 0.2528,
+      "step": 10729
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6640790963285514,
+      "learning_rate": 2.526507309433488e-06,
+      "loss": 0.2734,
+      "step": 10730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.623934747375392,
+      "learning_rate": 2.5256222388252223e-06,
+      "loss": 0.2813,
+      "step": 10731
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.731381905726239,
+      "learning_rate": 2.524737270883283e-06,
+      "loss": 0.2752,
+      "step": 10732
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9969171361101767,
+      "learning_rate": 2.5238524056443923e-06,
+      "loss": 0.2743,
+      "step": 10733
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6007300200548918,
+      "learning_rate": 2.522967643145263e-06,
+      "loss": 0.493,
+      "step": 10734
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0248368372994143,
+      "learning_rate": 2.5220829834226052e-06,
+      "loss": 0.2668,
+      "step": 10735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9164470473306445,
+      "learning_rate": 2.521198426513127e-06,
+      "loss": 0.2797,
+      "step": 10736
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4311060110477762,
+      "learning_rate": 2.520313972453529e-06,
+      "loss": 0.2667,
+      "step": 10737
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.356554789475532,
+      "learning_rate": 2.519429621280507e-06,
+      "loss": 0.2399,
+      "step": 10738
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.921065855938856,
+      "learning_rate": 2.518545373030755e-06,
+      "loss": 0.2678,
+      "step": 10739
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.897986968370867,
+      "learning_rate": 2.5176612277409663e-06,
+      "loss": 0.2874,
+      "step": 10740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5972060760578377,
+      "learning_rate": 2.516777185447822e-06,
+      "loss": 0.2643,
+      "step": 10741
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8534673337783578,
+      "learning_rate": 2.5158932461880025e-06,
+      "loss": 0.2643,
+      "step": 10742
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9849075715174838,
+      "learning_rate": 2.5150094099981824e-06,
+      "loss": 0.25,
+      "step": 10743
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9097060867123994,
+      "learning_rate": 2.5141256769150384e-06,
+      "loss": 0.2774,
+      "step": 10744
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6541412249210272,
+      "learning_rate": 2.513242046975235e-06,
+      "loss": 0.2625,
+      "step": 10745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9101433073194343,
+      "learning_rate": 2.5123585202154345e-06,
+      "loss": 0.2517,
+      "step": 10746
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.61749849088729,
+      "learning_rate": 2.511475096672298e-06,
+      "loss": 0.2729,
+      "step": 10747
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8666139577702474,
+      "learning_rate": 2.5105917763824807e-06,
+      "loss": 0.2888,
+      "step": 10748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5604030068521526,
+      "learning_rate": 2.5097085593826324e-06,
+      "loss": 0.2559,
+      "step": 10749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6913263652043975,
+      "learning_rate": 2.5088254457093976e-06,
+      "loss": 0.2734,
+      "step": 10750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.887452588989263,
+      "learning_rate": 2.507942435399422e-06,
+      "loss": 0.2699,
+      "step": 10751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9885788573119678,
+      "learning_rate": 2.5070595284893405e-06,
+      "loss": 0.2784,
+      "step": 10752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.321763075651108,
+      "learning_rate": 2.5061767250157883e-06,
+      "loss": 0.2862,
+      "step": 10753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4527588060279713,
+      "learning_rate": 2.50529402501539e-06,
+      "loss": 0.2875,
+      "step": 10754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9871044033792176,
+      "learning_rate": 2.5044114285247747e-06,
+      "loss": 0.2728,
+      "step": 10755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.427113811441325,
+      "learning_rate": 2.5035289355805634e-06,
+      "loss": 0.27,
+      "step": 10756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5494336135191533,
+      "learning_rate": 2.502646546219371e-06,
+      "loss": 0.26,
+      "step": 10757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7655865115779152,
+      "learning_rate": 2.501764260477807e-06,
+      "loss": 0.2568,
+      "step": 10758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0906854762511573,
+      "learning_rate": 2.5008820783924838e-06,
+      "loss": 0.2565,
+      "step": 10759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.466375728740451,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.2751,
+      "step": 10760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.912316393244981,
+      "learning_rate": 2.4991180253369594e-06,
+      "loss": 0.2599,
+      "step": 10761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.0501135923341485,
+      "learning_rate": 2.4982361544399507e-06,
+      "loss": 0.2634,
+      "step": 10762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.086258207425374,
+      "learning_rate": 2.497354387345568e-06,
+      "loss": 0.2649,
+      "step": 10763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8297431359179088,
+      "learning_rate": 2.496472724090399e-06,
+      "loss": 0.2745,
+      "step": 10764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8082842533895456,
+      "learning_rate": 2.495591164711023e-06,
+      "loss": 0.2486,
+      "step": 10765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9832680845163195,
+      "learning_rate": 2.4947097092440158e-06,
+      "loss": 0.2703,
+      "step": 10766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5536316284314864,
+      "learning_rate": 2.4938283577259555e-06,
+      "loss": 0.2626,
+      "step": 10767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7983514948701425,
+      "learning_rate": 2.492947110193407e-06,
+      "loss": 0.2522,
+      "step": 10768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.964056558789052,
+      "learning_rate": 2.492065966682935e-06,
+      "loss": 0.2577,
+      "step": 10769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6199639346245456,
+      "learning_rate": 2.4911849272311015e-06,
+      "loss": 0.4848,
+      "step": 10770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.102287851861108,
+      "learning_rate": 2.49030399187446e-06,
+      "loss": 0.2702,
+      "step": 10771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.610894320541659,
+      "learning_rate": 2.4894231606495663e-06,
+      "loss": 0.2717,
+      "step": 10772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1510178553218715,
+      "learning_rate": 2.488542433592964e-06,
+      "loss": 0.2612,
+      "step": 10773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5313128153843305,
+      "learning_rate": 2.487661810741196e-06,
+      "loss": 0.275,
+      "step": 10774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9076731489355447,
+      "learning_rate": 2.486781292130803e-06,
+      "loss": 0.2698,
+      "step": 10775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0592369070006264,
+      "learning_rate": 2.4859008777983183e-06,
+      "loss": 0.2503,
+      "step": 10776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5654897475717329,
+      "learning_rate": 2.4850205677802703e-06,
+      "loss": 0.2477,
+      "step": 10777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.043550501093226,
+      "learning_rate": 2.484140362113187e-06,
+      "loss": 0.2689,
+      "step": 10778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.541976162692309,
+      "learning_rate": 2.4832602608335874e-06,
+      "loss": 0.2726,
+      "step": 10779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6565962060903403,
+      "learning_rate": 2.482380263977992e-06,
+      "loss": 0.2785,
+      "step": 10780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.551375349168563,
+      "learning_rate": 2.481500371582911e-06,
+      "loss": 0.2544,
+      "step": 10781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.989536025504863,
+      "learning_rate": 2.4806205836848505e-06,
+      "loss": 0.2705,
+      "step": 10782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8500592155614894,
+      "learning_rate": 2.4797409003203198e-06,
+      "loss": 0.2571,
+      "step": 10783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7112224868963137,
+      "learning_rate": 2.4788613215258134e-06,
+      "loss": 0.2846,
+      "step": 10784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6933863777923017,
+      "learning_rate": 2.4779818473378303e-06,
+      "loss": 0.266,
+      "step": 10785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.859174589384238,
+      "learning_rate": 2.4771024777928605e-06,
+      "loss": 0.2869,
+      "step": 10786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.708089033096962,
+      "learning_rate": 2.4762232129273876e-06,
+      "loss": 0.2466,
+      "step": 10787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.518585611091397,
+      "learning_rate": 2.475344052777898e-06,
+      "loss": 0.2499,
+      "step": 10788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6049145839246542,
+      "learning_rate": 2.474464997380869e-06,
+      "loss": 0.4662,
+      "step": 10789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2675838406970885,
+      "learning_rate": 2.4735860467727705e-06,
+      "loss": 0.267,
+      "step": 10790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.825313867014563,
+      "learning_rate": 2.4727072009900764e-06,
+      "loss": 0.2668,
+      "step": 10791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6622286202561916,
+      "learning_rate": 2.4718284600692476e-06,
+      "loss": 0.2778,
+      "step": 10792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9039837610390817,
+      "learning_rate": 2.4709498240467484e-06,
+      "loss": 0.2611,
+      "step": 10793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.9378279982357665,
+      "learning_rate": 2.470071292959033e-06,
+      "loss": 0.2796,
+      "step": 10794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.913454823643306,
+      "learning_rate": 2.469192866842552e-06,
+      "loss": 0.2654,
+      "step": 10795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4094865781866206,
+      "learning_rate": 2.4683145457337553e-06,
+      "loss": 0.2606,
+      "step": 10796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.386602186222407,
+      "learning_rate": 2.467436329669085e-06,
+      "loss": 0.2776,
+      "step": 10797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.55838764175605,
+      "learning_rate": 2.4665582186849783e-06,
+      "loss": 0.2859,
+      "step": 10798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9091561101391172,
+      "learning_rate": 2.465680212817871e-06,
+      "loss": 0.2752,
+      "step": 10799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9170688023309563,
+      "learning_rate": 2.464802312104195e-06,
+      "loss": 0.2616,
+      "step": 10800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.857867731191885,
+      "learning_rate": 2.4639245165803732e-06,
+      "loss": 0.2678,
+      "step": 10801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8543701442588305,
+      "learning_rate": 2.4630468262828283e-06,
+      "loss": 0.2689,
+      "step": 10802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.586511193132362,
+      "learning_rate": 2.4621692412479753e-06,
+      "loss": 0.455,
+      "step": 10803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9110652553644543,
+      "learning_rate": 2.4612917615122293e-06,
+      "loss": 0.2768,
+      "step": 10804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.711422682568751,
+      "learning_rate": 2.460414387111998e-06,
+      "loss": 0.2612,
+      "step": 10805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.575483264792797,
+      "learning_rate": 2.4595371180836823e-06,
+      "loss": 0.272,
+      "step": 10806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5967939552732896,
+      "learning_rate": 2.458659954463684e-06,
+      "loss": 0.4742,
+      "step": 10807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.608502248002645,
+      "learning_rate": 2.4577828962884005e-06,
+      "loss": 0.2655,
+      "step": 10808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.121806103608374,
+      "learning_rate": 2.45690594359422e-06,
+      "loss": 0.2745,
+      "step": 10809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.782328293865478,
+      "learning_rate": 2.4560290964175288e-06,
+      "loss": 0.2619,
+      "step": 10810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3729383561164705,
+      "learning_rate": 2.4551523547947066e-06,
+      "loss": 0.2568,
+      "step": 10811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8005132301573186,
+      "learning_rate": 2.454275718762136e-06,
+      "loss": 0.268,
+      "step": 10812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2151613944442685,
+      "learning_rate": 2.4533991883561868e-06,
+      "loss": 0.2606,
+      "step": 10813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.80747473819906,
+      "learning_rate": 2.452522763613227e-06,
+      "loss": 0.2608,
+      "step": 10814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.666388877098076,
+      "learning_rate": 2.4516464445696223e-06,
+      "loss": 0.2583,
+      "step": 10815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.739491756727603,
+      "learning_rate": 2.4507702312617353e-06,
+      "loss": 0.2749,
+      "step": 10816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.52263683453807,
+      "learning_rate": 2.4498941237259195e-06,
+      "loss": 0.2863,
+      "step": 10817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2901257801936823,
+      "learning_rate": 2.449018121998526e-06,
+      "loss": 0.2567,
+      "step": 10818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4581399080578046,
+      "learning_rate": 2.4481422261158995e-06,
+      "loss": 0.269,
+      "step": 10819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.349425912986184,
+      "learning_rate": 2.4472664361143868e-06,
+      "loss": 0.2516,
+      "step": 10820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2633289576747733,
+      "learning_rate": 2.4463907520303216e-06,
+      "loss": 0.2705,
+      "step": 10821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1854781840997743,
+      "learning_rate": 2.4455151739000415e-06,
+      "loss": 0.2529,
+      "step": 10822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9687167937030654,
+      "learning_rate": 2.4446397017598726e-06,
+      "loss": 0.2765,
+      "step": 10823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8201049864582572,
+      "learning_rate": 2.443764335646143e-06,
+      "loss": 0.2695,
+      "step": 10824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0189675172167174,
+      "learning_rate": 2.4428890755951717e-06,
+      "loss": 0.2641,
+      "step": 10825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.149857973922792,
+      "learning_rate": 2.4420139216432746e-06,
+      "loss": 0.2684,
+      "step": 10826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.691615332807052,
+      "learning_rate": 2.441138873826761e-06,
+      "loss": 0.286,
+      "step": 10827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4865271893224243,
+      "learning_rate": 2.4402639321819433e-06,
+      "loss": 0.2649,
+      "step": 10828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.659360497673168,
+      "learning_rate": 2.4393890967451194e-06,
+      "loss": 0.2525,
+      "step": 10829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.738767329789025,
+      "learning_rate": 2.4385143675525917e-06,
+      "loss": 0.2873,
+      "step": 10830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3378421747300495,
+      "learning_rate": 2.4376397446406508e-06,
+      "loss": 0.2767,
+      "step": 10831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4810079656240105,
+      "learning_rate": 2.43676522804559e-06,
+      "loss": 0.2536,
+      "step": 10832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7610103965553026,
+      "learning_rate": 2.435890817803693e-06,
+      "loss": 0.2519,
+      "step": 10833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9028691179983896,
+      "learning_rate": 2.4350165139512384e-06,
+      "loss": 0.2553,
+      "step": 10834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.611323828889131,
+      "learning_rate": 2.4341423165245064e-06,
+      "loss": 0.2689,
+      "step": 10835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1389346807689904,
+      "learning_rate": 2.433268225559765e-06,
+      "loss": 0.2633,
+      "step": 10836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6035650523608391,
+      "learning_rate": 2.4323942410932862e-06,
+      "loss": 0.488,
+      "step": 10837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7200242146295561,
+      "learning_rate": 2.4315203631613303e-06,
+      "loss": 0.2671,
+      "step": 10838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.601566766359058,
+      "learning_rate": 2.4306465918001554e-06,
+      "loss": 0.2707,
+      "step": 10839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9848085958619472,
+      "learning_rate": 2.429772927046019e-06,
+      "loss": 0.2733,
+      "step": 10840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.69827138673236,
+      "learning_rate": 2.4288993689351692e-06,
+      "loss": 0.2514,
+      "step": 10841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.054443038449273,
+      "learning_rate": 2.428025917503849e-06,
+      "loss": 0.2792,
+      "step": 10842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.841957651692076,
+      "learning_rate": 2.427152572788304e-06,
+      "loss": 0.2766,
+      "step": 10843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8254332768160941,
+      "learning_rate": 2.4262793348247663e-06,
+      "loss": 0.2604,
+      "step": 10844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8950096525597646,
+      "learning_rate": 2.425406203649473e-06,
+      "loss": 0.25,
+      "step": 10845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.593341431234992,
+      "learning_rate": 2.4245331792986486e-06,
+      "loss": 0.2486,
+      "step": 10846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5497930112050453,
+      "learning_rate": 2.4236602618085157e-06,
+      "loss": 0.26,
+      "step": 10847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6949615459239775,
+      "learning_rate": 2.4227874512152965e-06,
+      "loss": 0.2577,
+      "step": 10848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7377251655209793,
+      "learning_rate": 2.421914747555204e-06,
+      "loss": 0.2725,
+      "step": 10849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3957716170763947,
+      "learning_rate": 2.4210421508644454e-06,
+      "loss": 0.2528,
+      "step": 10850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5906823880435084,
+      "learning_rate": 2.4201696611792285e-06,
+      "loss": 0.25,
+      "step": 10851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3668835120315774,
+      "learning_rate": 2.419297278535757e-06,
+      "loss": 0.2646,
+      "step": 10852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.0248823520360695,
+      "learning_rate": 2.418425002970225e-06,
+      "loss": 0.2751,
+      "step": 10853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9113828394692174,
+      "learning_rate": 2.417552834518825e-06,
+      "loss": 0.2627,
+      "step": 10854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6343467579385385,
+      "learning_rate": 2.4166807732177435e-06,
+      "loss": 0.27,
+      "step": 10855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.444913547484067,
+      "learning_rate": 2.4158088191031666e-06,
+      "loss": 0.2646,
+      "step": 10856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.23435673794038,
+      "learning_rate": 2.414936972211272e-06,
+      "loss": 0.2681,
+      "step": 10857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.170829744020603,
+      "learning_rate": 2.4140652325782317e-06,
+      "loss": 0.2614,
+      "step": 10858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3178830668271937,
+      "learning_rate": 2.4131936002402182e-06,
+      "loss": 0.2608,
+      "step": 10859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.276373835655226,
+      "learning_rate": 2.4123220752333994e-06,
+      "loss": 0.2635,
+      "step": 10860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4365126148774165,
+      "learning_rate": 2.4114506575939333e-06,
+      "loss": 0.2683,
+      "step": 10861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.012939648087876,
+      "learning_rate": 2.4105793473579773e-06,
+      "loss": 0.2638,
+      "step": 10862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.619364225017148,
+      "learning_rate": 2.4097081445616814e-06,
+      "loss": 0.2718,
+      "step": 10863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.465368191838411,
+      "learning_rate": 2.408837049241198e-06,
+      "loss": 0.2668,
+      "step": 10864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.533203366468756,
+      "learning_rate": 2.407966061432668e-06,
+      "loss": 0.289,
+      "step": 10865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4687487469774676,
+      "learning_rate": 2.407095181172227e-06,
+      "loss": 0.2565,
+      "step": 10866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.344562654570713,
+      "learning_rate": 2.4062244084960136e-06,
+      "loss": 0.2687,
+      "step": 10867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.930673067198498,
+      "learning_rate": 2.4053537434401576e-06,
+      "loss": 0.2762,
+      "step": 10868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.551485306675878,
+      "learning_rate": 2.404483186040784e-06,
+      "loss": 0.2492,
+      "step": 10869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0665610186258525,
+      "learning_rate": 2.403612736334013e-06,
+      "loss": 0.2472,
+      "step": 10870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.064619466844578,
+      "learning_rate": 2.402742394355959e-06,
+      "loss": 0.2585,
+      "step": 10871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7820775829933004,
+      "learning_rate": 2.4018721601427387e-06,
+      "loss": 0.2896,
+      "step": 10872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8421601210465457,
+      "learning_rate": 2.4010020337304575e-06,
+      "loss": 0.2704,
+      "step": 10873
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.444179906428219,
+      "learning_rate": 2.4001320151552153e-06,
+      "loss": 0.2871,
+      "step": 10874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0461616569532914,
+      "learning_rate": 2.3992621044531146e-06,
+      "loss": 0.2598,
+      "step": 10875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6281500200261377,
+      "learning_rate": 2.39839230166025e-06,
+      "loss": 0.4726,
+      "step": 10876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.289931686135603,
+      "learning_rate": 2.3975226068127107e-06,
+      "loss": 0.2634,
+      "step": 10877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.166019674554041,
+      "learning_rate": 2.39665301994658e-06,
+      "loss": 0.2687,
+      "step": 10878
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9610345374597213,
+      "learning_rate": 2.395783541097938e-06,
+      "loss": 0.2669,
+      "step": 10879
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7195799517151467,
+      "learning_rate": 2.3949141703028643e-06,
+      "loss": 0.2563,
+      "step": 10880
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4205161530393116,
+      "learning_rate": 2.394044907597427e-06,
+      "loss": 0.272,
+      "step": 10881
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6188588342391146,
+      "learning_rate": 2.393175753017697e-06,
+      "loss": 0.4873,
+      "step": 10882
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7269059264314834,
+      "learning_rate": 2.3923067065997336e-06,
+      "loss": 0.2583,
+      "step": 10883
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.129786075912959,
+      "learning_rate": 2.391437768379598e-06,
+      "loss": 0.2668,
+      "step": 10884
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.879601366199349,
+      "learning_rate": 2.3905689383933433e-06,
+      "loss": 0.2583,
+      "step": 10885
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.90273285050398,
+      "learning_rate": 2.3897002166770183e-06,
+      "loss": 0.2636,
+      "step": 10886
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7457911686373264,
+      "learning_rate": 2.3888316032666653e-06,
+      "loss": 0.264,
+      "step": 10887
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.245829896736995,
+      "learning_rate": 2.3879630981983277e-06,
+      "loss": 0.2754,
+      "step": 10888
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.3258908249547043,
+      "learning_rate": 2.387094701508042e-06,
+      "loss": 0.2521,
+      "step": 10889
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8457514254007812,
+      "learning_rate": 2.3862264132318385e-06,
+      "loss": 0.2533,
+      "step": 10890
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3115242101589804,
+      "learning_rate": 2.385358233405741e-06,
+      "loss": 0.2682,
+      "step": 10891
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3842661648783077,
+      "learning_rate": 2.3844901620657775e-06,
+      "loss": 0.2691,
+      "step": 10892
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3848533479403673,
+      "learning_rate": 2.3836221992479626e-06,
+      "loss": 0.2574,
+      "step": 10893
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.646166234318717,
+      "learning_rate": 2.3827543449883094e-06,
+      "loss": 0.2657,
+      "step": 10894
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3394498095130616,
+      "learning_rate": 2.3818865993228256e-06,
+      "loss": 0.2471,
+      "step": 10895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3183147321099677,
+      "learning_rate": 2.3810189622875174e-06,
+      "loss": 0.2886,
+      "step": 10896
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4799604803379114,
+      "learning_rate": 2.3801514339183857e-06,
+      "loss": 0.2763,
+      "step": 10897
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.559558408132068,
+      "learning_rate": 2.3792840142514244e-06,
+      "loss": 0.2563,
+      "step": 10898
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9476723320016283,
+      "learning_rate": 2.3784167033226222e-06,
+      "loss": 0.259,
+      "step": 10899
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2389654261430936,
+      "learning_rate": 2.37754950116797e-06,
+      "loss": 0.262,
+      "step": 10900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.656437300482076,
+      "learning_rate": 2.3766824078234468e-06,
+      "loss": 0.2593,
+      "step": 10901
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7259779210563946,
+      "learning_rate": 2.3758154233250298e-06,
+      "loss": 0.2618,
+      "step": 10902
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.099447602178023,
+      "learning_rate": 2.3749485477086883e-06,
+      "loss": 0.2494,
+      "step": 10903
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8399474411979575,
+      "learning_rate": 2.374081781010398e-06,
+      "loss": 0.2732,
+      "step": 10904
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.3599037997487375,
+      "learning_rate": 2.373215123266118e-06,
+      "loss": 0.2799,
+      "step": 10905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9549760089744432,
+      "learning_rate": 2.372348574511809e-06,
+      "loss": 0.2547,
+      "step": 10906
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8140650942537564,
+      "learning_rate": 2.371482134783423e-06,
+      "loss": 0.2832,
+      "step": 10907
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3040300334623325,
+      "learning_rate": 2.370615804116914e-06,
+      "loss": 0.2495,
+      "step": 10908
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3787249135748456,
+      "learning_rate": 2.369749582548225e-06,
+      "loss": 0.2658,
+      "step": 10909
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7002413232831315,
+      "learning_rate": 2.3688834701132964e-06,
+      "loss": 0.2777,
+      "step": 10910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6570233627211397,
+      "learning_rate": 2.3680174668480656e-06,
+      "loss": 0.2641,
+      "step": 10911
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9881638843813065,
+      "learning_rate": 2.367151572788467e-06,
+      "loss": 0.2632,
+      "step": 10912
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8864654449008889,
+      "learning_rate": 2.3662857879704265e-06,
+      "loss": 0.2591,
+      "step": 10913
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.382164067425556,
+      "learning_rate": 2.3654201124298658e-06,
+      "loss": 0.252,
+      "step": 10914
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6536734335078284,
+      "learning_rate": 2.364554546202703e-06,
+      "loss": 0.4739,
+      "step": 10915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6142496272903245,
+      "learning_rate": 2.363689089324855e-06,
+      "loss": 0.4544,
+      "step": 10916
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5179567144839843,
+      "learning_rate": 2.362823741832229e-06,
+      "loss": 0.274,
+      "step": 10917
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.103286936456892,
+      "learning_rate": 2.3619585037607283e-06,
+      "loss": 0.2823,
+      "step": 10918
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6817210396512827,
+      "learning_rate": 2.3610933751462554e-06,
+      "loss": 0.2719,
+      "step": 10919
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.667014841922441,
+      "learning_rate": 2.360228356024707e-06,
+      "loss": 0.2576,
+      "step": 10920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.08097641517747,
+      "learning_rate": 2.3593634464319732e-06,
+      "loss": 0.2874,
+      "step": 10921
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3777251939030866,
+      "learning_rate": 2.35849864640394e-06,
+      "loss": 0.2536,
+      "step": 10922
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0453044671274974,
+      "learning_rate": 2.357633955976488e-06,
+      "loss": 0.2549,
+      "step": 10923
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7899662206002802,
+      "learning_rate": 2.3567693751854985e-06,
+      "loss": 0.2622,
+      "step": 10924
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7915825281831113,
+      "learning_rate": 2.3559049040668425e-06,
+      "loss": 0.2678,
+      "step": 10925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9438606617371923,
+      "learning_rate": 2.355040542656387e-06,
+      "loss": 0.2499,
+      "step": 10926
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.531104619929896,
+      "learning_rate": 2.354176290989996e-06,
+      "loss": 0.2725,
+      "step": 10927
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2581310911745267,
+      "learning_rate": 2.353312149103533e-06,
+      "loss": 0.2563,
+      "step": 10928
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.91521569417984,
+      "learning_rate": 2.3524481170328506e-06,
+      "loss": 0.2713,
+      "step": 10929
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.413535705432135,
+      "learning_rate": 2.3515841948137976e-06,
+      "loss": 0.272,
+      "step": 10930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8006589061091516,
+      "learning_rate": 2.350720382482219e-06,
+      "loss": 0.2605,
+      "step": 10931
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5129055315177493,
+      "learning_rate": 2.349856680073959e-06,
+      "loss": 0.2665,
+      "step": 10932
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6849296253493475,
+      "learning_rate": 2.34899308762485e-06,
+      "loss": 0.5027,
+      "step": 10933
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.583730405393349,
+      "learning_rate": 2.34812960517073e-06,
+      "loss": 0.2675,
+      "step": 10934
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4774859424883449,
+      "learning_rate": 2.347266232747419e-06,
+      "loss": 0.2637,
+      "step": 10935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.05368579524338,
+      "learning_rate": 2.346402970390748e-06,
+      "loss": 0.2647,
+      "step": 10936
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0026683642084535,
+      "learning_rate": 2.345539818136529e-06,
+      "loss": 0.2465,
+      "step": 10937
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2468664922163395,
+      "learning_rate": 2.344676776020579e-06,
+      "loss": 0.25,
+      "step": 10938
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3456737338948885,
+      "learning_rate": 2.343813844078704e-06,
+      "loss": 0.2531,
+      "step": 10939
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2275206115756694,
+      "learning_rate": 2.3429510223467116e-06,
+      "loss": 0.262,
+      "step": 10940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8450804172787154,
+      "learning_rate": 2.3420883108603997e-06,
+      "loss": 0.2921,
+      "step": 10941
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.667519709873651,
+      "learning_rate": 2.3412257096555663e-06,
+      "loss": 0.2575,
+      "step": 10942
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0751679166352885,
+      "learning_rate": 2.3403632187679987e-06,
+      "loss": 0.2628,
+      "step": 10943
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.46009451010248,
+      "learning_rate": 2.3395008382334873e-06,
+      "loss": 0.2603,
+      "step": 10944
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.63868910424025,
+      "learning_rate": 2.3386385680878116e-06,
+      "loss": 0.2651,
+      "step": 10945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0428530019975164,
+      "learning_rate": 2.3377764083667494e-06,
+      "loss": 0.2817,
+      "step": 10946
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.260584311841463,
+      "learning_rate": 2.33691435910607e-06,
+      "loss": 0.2916,
+      "step": 10947
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.435284089693954,
+      "learning_rate": 2.336052420341544e-06,
+      "loss": 0.2585,
+      "step": 10948
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.1945625299153075,
+      "learning_rate": 2.335190592108937e-06,
+      "loss": 0.2737,
+      "step": 10949
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6983782016735072,
+      "learning_rate": 2.3343288744440055e-06,
+      "loss": 0.2687,
+      "step": 10950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9708341111944603,
+      "learning_rate": 2.333467267382502e-06,
+      "loss": 0.2581,
+      "step": 10951
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.119651187058837,
+      "learning_rate": 2.3326057709601795e-06,
+      "loss": 0.2765,
+      "step": 10952
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8961854399545466,
+      "learning_rate": 2.331744385212782e-06,
+      "loss": 0.2429,
+      "step": 10953
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2148688811033104,
+      "learning_rate": 2.330883110176049e-06,
+      "loss": 0.268,
+      "step": 10954
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9929853204084975,
+      "learning_rate": 2.3300219458857144e-06,
+      "loss": 0.2792,
+      "step": 10955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.621514330675251,
+      "learning_rate": 2.3291608923775117e-06,
+      "loss": 0.2754,
+      "step": 10956
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.704714025645811,
+      "learning_rate": 2.3282999496871694e-06,
+      "loss": 0.2779,
+      "step": 10957
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5268878406254864,
+      "learning_rate": 2.327439117850408e-06,
+      "loss": 0.2488,
+      "step": 10958
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6423044441394652,
+      "learning_rate": 2.3265783969029425e-06,
+      "loss": 0.2549,
+      "step": 10959
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4383915248263024,
+      "learning_rate": 2.32571778688049e-06,
+      "loss": 0.243,
+      "step": 10960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7243666016070005,
+      "learning_rate": 2.3248572878187566e-06,
+      "loss": 0.2502,
+      "step": 10961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.910469847432548,
+      "learning_rate": 2.323996899753445e-06,
+      "loss": 0.2767,
+      "step": 10962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2472062476080032,
+      "learning_rate": 2.3231366227202523e-06,
+      "loss": 0.2494,
+      "step": 10963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5338759061150538,
+      "learning_rate": 2.3222764567548795e-06,
+      "loss": 0.2517,
+      "step": 10964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.02135852969192,
+      "learning_rate": 2.3214164018930124e-06,
+      "loss": 0.2633,
+      "step": 10965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0287612662433467,
+      "learning_rate": 2.3205564581703367e-06,
+      "loss": 0.2604,
+      "step": 10966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7936296361297288,
+      "learning_rate": 2.3196966256225306e-06,
+      "loss": 0.2523,
+      "step": 10967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7106428506218263,
+      "learning_rate": 2.318836904285274e-06,
+      "loss": 0.3034,
+      "step": 10968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1569842069850975,
+      "learning_rate": 2.317977294194237e-06,
+      "loss": 0.2735,
+      "step": 10969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4835992143932462,
+      "learning_rate": 2.3171177953850837e-06,
+      "loss": 0.2524,
+      "step": 10970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3607794822239407,
+      "learning_rate": 2.3162584078934785e-06,
+      "loss": 0.2833,
+      "step": 10971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6030701736115138,
+      "learning_rate": 2.315399131755081e-06,
+      "loss": 0.2819,
+      "step": 10972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.975997946367097,
+      "learning_rate": 2.314539967005541e-06,
+      "loss": 0.2594,
+      "step": 10973
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.213239373020768,
+      "learning_rate": 2.313680913680509e-06,
+      "loss": 0.2581,
+      "step": 10974
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0669890418402885,
+      "learning_rate": 2.312821971815625e-06,
+      "loss": 0.2738,
+      "step": 10975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.436783865483608,
+      "learning_rate": 2.3119631414465317e-06,
+      "loss": 0.2694,
+      "step": 10976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.107960424948507,
+      "learning_rate": 2.3111044226088625e-06,
+      "loss": 0.2808,
+      "step": 10977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5052658011789797,
+      "learning_rate": 2.3102458153382446e-06,
+      "loss": 0.2551,
+      "step": 10978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5972256617506898,
+      "learning_rate": 2.309387319670306e-06,
+      "loss": 0.4581,
+      "step": 10979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3041711980146915,
+      "learning_rate": 2.3085289356406682e-06,
+      "loss": 0.2531,
+      "step": 10980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.740272779062172,
+      "learning_rate": 2.3076706632849448e-06,
+      "loss": 0.2681,
+      "step": 10981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.0874210368604125,
+      "learning_rate": 2.3068125026387483e-06,
+      "loss": 0.2864,
+      "step": 10982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2536115526514644,
+      "learning_rate": 2.3059544537376834e-06,
+      "loss": 0.2549,
+      "step": 10983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1361431338645454,
+      "learning_rate": 2.3050965166173543e-06,
+      "loss": 0.2565,
+      "step": 10984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2496520356866707,
+      "learning_rate": 2.3042386913133558e-06,
+      "loss": 0.2632,
+      "step": 10985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9677520799895307,
+      "learning_rate": 2.303380977861284e-06,
+      "loss": 0.26,
+      "step": 10986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5742895822778318,
+      "learning_rate": 2.3025233762967238e-06,
+      "loss": 0.2553,
+      "step": 10987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8123616300708694,
+      "learning_rate": 2.3016658866552618e-06,
+      "loss": 0.2454,
+      "step": 10988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.888777549190069,
+      "learning_rate": 2.3008085089724743e-06,
+      "loss": 0.251,
+      "step": 10989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8074781999311407,
+      "learning_rate": 2.299951243283936e-06,
+      "loss": 0.2707,
+      "step": 10990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4374729526319707,
+      "learning_rate": 2.299094089625215e-06,
+      "loss": 0.2717,
+      "step": 10991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2798070048658396,
+      "learning_rate": 2.2982370480318794e-06,
+      "loss": 0.2607,
+      "step": 10992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.9999077928147075,
+      "learning_rate": 2.297380118539486e-06,
+      "loss": 0.2686,
+      "step": 10993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.64007053011502,
+      "learning_rate": 2.2965233011835936e-06,
+      "loss": 0.2602,
+      "step": 10994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.910541496731146,
+      "learning_rate": 2.295666595999749e-06,
+      "loss": 0.2669,
+      "step": 10995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.788300761654256,
+      "learning_rate": 2.294810003023503e-06,
+      "loss": 0.275,
+      "step": 10996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.442347356213636,
+      "learning_rate": 2.293953522290395e-06,
+      "loss": 0.261,
+      "step": 10997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8529495014929396,
+      "learning_rate": 2.2930971538359612e-06,
+      "loss": 0.2622,
+      "step": 10998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1671811736080213,
+      "learning_rate": 2.292240897695734e-06,
+      "loss": 0.264,
+      "step": 10999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9863585472536043,
+      "learning_rate": 2.2913847539052405e-06,
+      "loss": 0.2704,
+      "step": 11000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4190139119943348,
+      "learning_rate": 2.2905287225000062e-06,
+      "loss": 0.2787,
+      "step": 11001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.967742712789526,
+      "learning_rate": 2.2896728035155487e-06,
+      "loss": 0.2645,
+      "step": 11002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5972908958120042,
+      "learning_rate": 2.288816996987378e-06,
+      "loss": 0.2608,
+      "step": 11003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.698590185358748,
+      "learning_rate": 2.287961302951008e-06,
+      "loss": 0.2595,
+      "step": 11004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.511746720975442,
+      "learning_rate": 2.2871057214419403e-06,
+      "loss": 0.2618,
+      "step": 11005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0104081680942985,
+      "learning_rate": 2.2862502524956748e-06,
+      "loss": 0.2788,
+      "step": 11006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.437342448768188,
+      "learning_rate": 2.285394896147705e-06,
+      "loss": 0.2695,
+      "step": 11007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.567028789019159,
+      "learning_rate": 2.284539652433522e-06,
+      "loss": 0.2776,
+      "step": 11008
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3600236114539146,
+      "learning_rate": 2.283684521388615e-06,
+      "loss": 0.2813,
+      "step": 11009
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4460102376280342,
+      "learning_rate": 2.282829503048461e-06,
+      "loss": 0.2683,
+      "step": 11010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.765288002276459,
+      "learning_rate": 2.2819745974485357e-06,
+      "loss": 0.2613,
+      "step": 11011
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0838044190318,
+      "learning_rate": 2.281119804624314e-06,
+      "loss": 0.2598,
+      "step": 11012
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.331842545145046,
+      "learning_rate": 2.2802651246112606e-06,
+      "loss": 0.2604,
+      "step": 11013
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.082322749369407,
+      "learning_rate": 2.279410557444838e-06,
+      "loss": 0.2504,
+      "step": 11014
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6204593116768944,
+      "learning_rate": 2.278556103160502e-06,
+      "loss": 0.2498,
+      "step": 11015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3905597247715427,
+      "learning_rate": 2.2777017617937073e-06,
+      "loss": 0.2799,
+      "step": 11016
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.472427955573566,
+      "learning_rate": 2.2768475333799035e-06,
+      "loss": 0.247,
+      "step": 11017
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6162241556398924,
+      "learning_rate": 2.2759934179545325e-06,
+      "loss": 0.2591,
+      "step": 11018
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9127677604285067,
+      "learning_rate": 2.275139415553031e-06,
+      "loss": 0.2582,
+      "step": 11019
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7527933414260848,
+      "learning_rate": 2.2742855262108364e-06,
+      "loss": 0.2616,
+      "step": 11020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7751755045145465,
+      "learning_rate": 2.273431749963377e-06,
+      "loss": 0.2736,
+      "step": 11021
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5290407432488182,
+      "learning_rate": 2.272578086846078e-06,
+      "loss": 0.2625,
+      "step": 11022
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5421788814637516,
+      "learning_rate": 2.271724536894354e-06,
+      "loss": 0.2705,
+      "step": 11023
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8246065823332165,
+      "learning_rate": 2.270871100143629e-06,
+      "loss": 0.2605,
+      "step": 11024
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.860304452842204,
+      "learning_rate": 2.27001777662931e-06,
+      "loss": 0.2649,
+      "step": 11025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.445004033177439,
+      "learning_rate": 2.2691645663868016e-06,
+      "loss": 0.2386,
+      "step": 11026
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2881442401364325,
+      "learning_rate": 2.2683114694515045e-06,
+      "loss": 0.2618,
+      "step": 11027
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.64561475847575,
+      "learning_rate": 2.2674584858588185e-06,
+      "loss": 0.2607,
+      "step": 11028
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3218454980779675,
+      "learning_rate": 2.266605615644133e-06,
+      "loss": 0.2588,
+      "step": 11029
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9211058016416054,
+      "learning_rate": 2.265752858842834e-06,
+      "loss": 0.2549,
+      "step": 11030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5657039121450247,
+      "learning_rate": 2.2649002154903056e-06,
+      "loss": 0.2798,
+      "step": 11031
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7234951650213524,
+      "learning_rate": 2.2640476856219267e-06,
+      "loss": 0.2708,
+      "step": 11032
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.779077471648995,
+      "learning_rate": 2.263195269273069e-06,
+      "loss": 0.2506,
+      "step": 11033
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.195402374947855,
+      "learning_rate": 2.2623429664791015e-06,
+      "loss": 0.2627,
+      "step": 11034
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4381753877699035,
+      "learning_rate": 2.261490777275384e-06,
+      "loss": 0.2743,
+      "step": 11035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5370562146420013,
+      "learning_rate": 2.260638701697281e-06,
+      "loss": 0.2655,
+      "step": 11036
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.194549114731265,
+      "learning_rate": 2.259786739780143e-06,
+      "loss": 0.2479,
+      "step": 11037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5973098267399335,
+      "learning_rate": 2.2589348915593197e-06,
+      "loss": 0.2622,
+      "step": 11038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3700132494673471,
+      "learning_rate": 2.2580831570701557e-06,
+      "loss": 0.2622,
+      "step": 11039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.890398019948207,
+      "learning_rate": 2.2572315363479935e-06,
+      "loss": 0.2653,
+      "step": 11040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7266076924332996,
+      "learning_rate": 2.2563800294281667e-06,
+      "loss": 0.2753,
+      "step": 11041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7583690345518133,
+      "learning_rate": 2.2555286363460053e-06,
+      "loss": 0.2725,
+      "step": 11042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7996428411083296,
+      "learning_rate": 2.254677357136834e-06,
+      "loss": 0.2598,
+      "step": 11043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.019957369916979,
+      "learning_rate": 2.2538261918359776e-06,
+      "loss": 0.2731,
+      "step": 11044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4689716646829647,
+      "learning_rate": 2.252975140478748e-06,
+      "loss": 0.2626,
+      "step": 11045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.377654027126183,
+      "learning_rate": 2.2521242031004613e-06,
+      "loss": 0.2428,
+      "step": 11046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0002552348555516,
+      "learning_rate": 2.2512733797364202e-06,
+      "loss": 0.2699,
+      "step": 11047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6584052503371383,
+      "learning_rate": 2.25042267042193e-06,
+      "loss": 0.2887,
+      "step": 11048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.39934382994924,
+      "learning_rate": 2.249572075192288e-06,
+      "loss": 0.2663,
+      "step": 11049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5849785628171148,
+      "learning_rate": 2.2487215940827856e-06,
+      "loss": 0.4702,
+      "step": 11050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7096365971610517,
+      "learning_rate": 2.247871227128709e-06,
+      "loss": 0.2526,
+      "step": 11051
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5156501599784997,
+      "learning_rate": 2.247020974365344e-06,
+      "loss": 0.2687,
+      "step": 11052
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.368127450149746,
+      "learning_rate": 2.24617083582797e-06,
+      "loss": 0.2711,
+      "step": 11053
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.861099343005788,
+      "learning_rate": 2.2453208115518595e-06,
+      "loss": 0.2611,
+      "step": 11054
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.69917155539132,
+      "learning_rate": 2.2444709015722792e-06,
+      "loss": 0.2694,
+      "step": 11055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.808405165629587,
+      "learning_rate": 2.2436211059244977e-06,
+      "loss": 0.255,
+      "step": 11056
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9282026490070858,
+      "learning_rate": 2.2427714246437726e-06,
+      "loss": 0.2549,
+      "step": 11057
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.073363292368241,
+      "learning_rate": 2.241921857765358e-06,
+      "loss": 0.2526,
+      "step": 11058
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1626567118325903,
+      "learning_rate": 2.241072405324503e-06,
+      "loss": 0.2602,
+      "step": 11059
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5080723008362154,
+      "learning_rate": 2.240223067356454e-06,
+      "loss": 0.269,
+      "step": 11060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.632891593904277,
+      "learning_rate": 2.239373843896454e-06,
+      "loss": 0.2847,
+      "step": 11061
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.089561431675379,
+      "learning_rate": 2.2385247349797367e-06,
+      "loss": 0.2714,
+      "step": 11062
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.538838110790415,
+      "learning_rate": 2.237675740641531e-06,
+      "loss": 0.2515,
+      "step": 11063
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.681589298357147,
+      "learning_rate": 2.2368268609170673e-06,
+      "loss": 0.2563,
+      "step": 11064
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4931947400211565,
+      "learning_rate": 2.235978095841565e-06,
+      "loss": 0.2504,
+      "step": 11065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0144539321232036,
+      "learning_rate": 2.2351294454502408e-06,
+      "loss": 0.2758,
+      "step": 11066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.223518759245911,
+      "learning_rate": 2.234280909778305e-06,
+      "loss": 0.2767,
+      "step": 11067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3950419129427172,
+      "learning_rate": 2.2334324888609673e-06,
+      "loss": 0.2549,
+      "step": 11068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.276218192016725,
+      "learning_rate": 2.2325841827334306e-06,
+      "loss": 0.2688,
+      "step": 11069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.855309539941102,
+      "learning_rate": 2.2317359914308924e-06,
+      "loss": 0.279,
+      "step": 11070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.116590015760823,
+      "learning_rate": 2.2308879149885426e-06,
+      "loss": 0.2692,
+      "step": 11071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.923673125791136,
+      "learning_rate": 2.2300399534415733e-06,
+      "loss": 0.2587,
+      "step": 11072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.903104121230704,
+      "learning_rate": 2.229192106825167e-06,
+      "loss": 0.2644,
+      "step": 11073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8260474646408806,
+      "learning_rate": 2.2283443751745015e-06,
+      "loss": 0.2615,
+      "step": 11074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9848969494847946,
+      "learning_rate": 2.2274967585247485e-06,
+      "loss": 0.2506,
+      "step": 11075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6350166096238177,
+      "learning_rate": 2.2266492569110804e-06,
+      "loss": 0.254,
+      "step": 11076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.573425443333891,
+      "learning_rate": 2.225801870368662e-06,
+      "loss": 0.2526,
+      "step": 11077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4488845512042205,
+      "learning_rate": 2.2249545989326516e-06,
+      "loss": 0.2705,
+      "step": 11078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.009221734149319,
+      "learning_rate": 2.224107442638202e-06,
+      "loss": 0.2724,
+      "step": 11079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.10424937516931,
+      "learning_rate": 2.223260401520467e-06,
+      "loss": 0.2525,
+      "step": 11080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6278741200920794,
+      "learning_rate": 2.2224134756145897e-06,
+      "loss": 0.2737,
+      "step": 11081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0663359360044953,
+      "learning_rate": 2.221566664955709e-06,
+      "loss": 0.2653,
+      "step": 11082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.455547465712494,
+      "learning_rate": 2.2207199695789653e-06,
+      "loss": 0.2879,
+      "step": 11083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.793989986136283,
+      "learning_rate": 2.219873389519484e-06,
+      "loss": 0.2757,
+      "step": 11084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8067461841811359,
+      "learning_rate": 2.2190269248123958e-06,
+      "loss": 0.2644,
+      "step": 11085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9519397446729565,
+      "learning_rate": 2.21818057549282e-06,
+      "loss": 0.286,
+      "step": 11086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6867546979026484,
+      "learning_rate": 2.2173343415958715e-06,
+      "loss": 0.2628,
+      "step": 11087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.901729711202853,
+      "learning_rate": 2.2164882231566655e-06,
+      "loss": 0.259,
+      "step": 11088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6494661718749033,
+      "learning_rate": 2.2156422202103067e-06,
+      "loss": 0.2563,
+      "step": 11089
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.048678844580366,
+      "learning_rate": 2.2147963327918964e-06,
+      "loss": 0.2569,
+      "step": 11090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4641382326084147,
+      "learning_rate": 2.2139505609365352e-06,
+      "loss": 0.2874,
+      "step": 11091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.515578197114846,
+      "learning_rate": 2.213104904679312e-06,
+      "loss": 0.2614,
+      "step": 11092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5400364181574417,
+      "learning_rate": 2.212259364055318e-06,
+      "loss": 0.2852,
+      "step": 11093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0606239270475943,
+      "learning_rate": 2.2114139390996348e-06,
+      "loss": 0.2686,
+      "step": 11094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3630679878668983,
+      "learning_rate": 2.2105686298473384e-06,
+      "loss": 0.2529,
+      "step": 11095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2767080461181015,
+      "learning_rate": 2.2097234363335058e-06,
+      "loss": 0.256,
+      "step": 11096
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5891959395349309,
+      "learning_rate": 2.2088783585932024e-06,
+      "loss": 0.2578,
+      "step": 11097
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.074386376433146,
+      "learning_rate": 2.2080333966614953e-06,
+      "loss": 0.2602,
+      "step": 11098
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8349753863808815,
+      "learning_rate": 2.2071885505734413e-06,
+      "loss": 0.2693,
+      "step": 11099
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8212975785039227,
+      "learning_rate": 2.2063438203640934e-06,
+      "loss": 0.2584,
+      "step": 11100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.166344451571758,
+      "learning_rate": 2.2054992060685044e-06,
+      "loss": 0.2836,
+      "step": 11101
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9066922810415483,
+      "learning_rate": 2.204654707721717e-06,
+      "loss": 0.2511,
+      "step": 11102
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.638776594110408,
+      "learning_rate": 2.2038103253587685e-06,
+      "loss": 0.2519,
+      "step": 11103
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9514373854405695,
+      "learning_rate": 2.2029660590146983e-06,
+      "loss": 0.2416,
+      "step": 11104
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6263658028104124,
+      "learning_rate": 2.202121908724533e-06,
+      "loss": 0.2704,
+      "step": 11105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3227303615510393,
+      "learning_rate": 2.2012778745233003e-06,
+      "loss": 0.2711,
+      "step": 11106
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7892407403209454,
+      "learning_rate": 2.200433956446019e-06,
+      "loss": 0.2743,
+      "step": 11107
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6171984573005869,
+      "learning_rate": 2.199590154527706e-06,
+      "loss": 0.2783,
+      "step": 11108
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6675790926600702,
+      "learning_rate": 2.198746468803372e-06,
+      "loss": 0.2618,
+      "step": 11109
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.107905547457524,
+      "learning_rate": 2.1979028993080233e-06,
+      "loss": 0.265,
+      "step": 11110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9079694562089151,
+      "learning_rate": 2.1970594460766575e-06,
+      "loss": 0.2471,
+      "step": 11111
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3998216764906461,
+      "learning_rate": 2.196216109144275e-06,
+      "loss": 0.2611,
+      "step": 11112
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.071212472082706,
+      "learning_rate": 2.195372888545867e-06,
+      "loss": 0.2594,
+      "step": 11113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.185912615809283,
+      "learning_rate": 2.1945297843164197e-06,
+      "loss": 0.2772,
+      "step": 11114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.572312217957228,
+      "learning_rate": 2.193686796490913e-06,
+      "loss": 0.2617,
+      "step": 11115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7676879298770152,
+      "learning_rate": 2.1928439251043278e-06,
+      "loss": 0.2563,
+      "step": 11116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.734980177494873,
+      "learning_rate": 2.1920011701916337e-06,
+      "loss": 0.2686,
+      "step": 11117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.030139485054264,
+      "learning_rate": 2.1911585317877986e-06,
+      "loss": 0.2384,
+      "step": 11118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.232825620521009,
+      "learning_rate": 2.1903160099277835e-06,
+      "loss": 0.2612,
+      "step": 11119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6215912547677702,
+      "learning_rate": 2.1894736046465482e-06,
+      "loss": 0.2608,
+      "step": 11120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8951279261507745,
+      "learning_rate": 2.1886313159790463e-06,
+      "loss": 0.2862,
+      "step": 11121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1648138596848687,
+      "learning_rate": 2.187789143960225e-06,
+      "loss": 0.2668,
+      "step": 11122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.05366253846743,
+      "learning_rate": 2.186947088625025e-06,
+      "loss": 0.2593,
+      "step": 11123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.621587619191963,
+      "learning_rate": 2.1861051500083886e-06,
+      "loss": 0.4774,
+      "step": 11124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.862334799297239,
+      "learning_rate": 2.185263328145248e-06,
+      "loss": 0.267,
+      "step": 11125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6372825072994112,
+      "learning_rate": 2.184421623070531e-06,
+      "loss": 0.4961,
+      "step": 11126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.111011552477857,
+      "learning_rate": 2.1835800348191604e-06,
+      "loss": 0.2583,
+      "step": 11127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5497573113746697,
+      "learning_rate": 2.182738563426057e-06,
+      "loss": 0.4468,
+      "step": 11128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9332307363591024,
+      "learning_rate": 2.181897208926136e-06,
+      "loss": 0.2603,
+      "step": 11129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3140671813940632,
+      "learning_rate": 2.1810559713543052e-06,
+      "loss": 0.2479,
+      "step": 11130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5273597655369189,
+      "learning_rate": 2.1802148507454675e-06,
+      "loss": 0.2669,
+      "step": 11131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6823541823304715,
+      "learning_rate": 2.1793738471345253e-06,
+      "loss": 0.2507,
+      "step": 11132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5571655208137851,
+      "learning_rate": 2.1785329605563733e-06,
+      "loss": 0.2642,
+      "step": 11133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2515328013997116,
+      "learning_rate": 2.1776921910458982e-06,
+      "loss": 0.2714,
+      "step": 11134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2716633686626753,
+      "learning_rate": 2.1768515386379886e-06,
+      "loss": 0.2675,
+      "step": 11135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7139707947096543,
+      "learning_rate": 2.1760110033675215e-06,
+      "loss": 0.2651,
+      "step": 11136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.818952731486872,
+      "learning_rate": 2.175170585269375e-06,
+      "loss": 0.2651,
+      "step": 11137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7930437580722884,
+      "learning_rate": 2.174330284378419e-06,
+      "loss": 0.2592,
+      "step": 11138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.488512840589353,
+      "learning_rate": 2.1734901007295166e-06,
+      "loss": 0.263,
+      "step": 11139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3384207012645537,
+      "learning_rate": 2.172650034357532e-06,
+      "loss": 0.3014,
+      "step": 11140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7797887466403626,
+      "learning_rate": 2.1718100852973193e-06,
+      "loss": 0.2636,
+      "step": 11141
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2366899928859194,
+      "learning_rate": 2.1709702535837286e-06,
+      "loss": 0.263,
+      "step": 11142
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8869791297124512,
+      "learning_rate": 2.1701305392516083e-06,
+      "loss": 0.2718,
+      "step": 11143
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.627419730919339,
+      "learning_rate": 2.169290942335796e-06,
+      "loss": 0.2685,
+      "step": 11144
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9933536815401744,
+      "learning_rate": 2.1684514628711327e-06,
+      "loss": 0.2591,
+      "step": 11145
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.585672142831425,
+      "learning_rate": 2.167612100892448e-06,
+      "loss": 0.4837,
+      "step": 11146
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9656785182238308,
+      "learning_rate": 2.166772856434566e-06,
+      "loss": 0.2734,
+      "step": 11147
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.689638391273336,
+      "learning_rate": 2.1659337295323117e-06,
+      "loss": 0.2667,
+      "step": 11148
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6796494381982516,
+      "learning_rate": 2.1650947202204998e-06,
+      "loss": 0.2477,
+      "step": 11149
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.486513243309083,
+      "learning_rate": 2.164255828533945e-06,
+      "loss": 0.2616,
+      "step": 11150
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1551226352039694,
+      "learning_rate": 2.1634170545074525e-06,
+      "loss": 0.2679,
+      "step": 11151
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.074976292114402,
+      "learning_rate": 2.162578398175823e-06,
+      "loss": 0.2644,
+      "step": 11152
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.784326610502005,
+      "learning_rate": 2.1617398595738575e-06,
+      "loss": 0.2662,
+      "step": 11153
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0804451457232425,
+      "learning_rate": 2.1609014387363463e-06,
+      "loss": 0.253,
+      "step": 11154
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8393783950324805,
+      "learning_rate": 2.1600631356980755e-06,
+      "loss": 0.2741,
+      "step": 11155
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8965425607078283,
+      "learning_rate": 2.1592249504938317e-06,
+      "loss": 0.2648,
+      "step": 11156
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.13709796683566,
+      "learning_rate": 2.1583868831583883e-06,
+      "loss": 0.3057,
+      "step": 11157
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.521775100929573,
+      "learning_rate": 2.1575489337265226e-06,
+      "loss": 0.2547,
+      "step": 11158
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2896605781529495,
+      "learning_rate": 2.156711102233e-06,
+      "loss": 0.2598,
+      "step": 11159
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0154422851143377,
+      "learning_rate": 2.155873388712582e-06,
+      "loss": 0.2795,
+      "step": 11160
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2864212875100796,
+      "learning_rate": 2.155035793200031e-06,
+      "loss": 0.2646,
+      "step": 11161
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6538919218949895,
+      "learning_rate": 2.1541983157300983e-06,
+      "loss": 0.2664,
+      "step": 11162
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.097220743349051,
+      "learning_rate": 2.15336095633753e-06,
+      "loss": 0.2788,
+      "step": 11163
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.620666279732925,
+      "learning_rate": 2.152523715057072e-06,
+      "loss": 0.2747,
+      "step": 11164
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1825001412007596,
+      "learning_rate": 2.151686591923465e-06,
+      "loss": 0.2903,
+      "step": 11165
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.634309417557884,
+      "learning_rate": 2.1508495869714395e-06,
+      "loss": 0.2575,
+      "step": 11166
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7076290668926548,
+      "learning_rate": 2.1500127002357264e-06,
+      "loss": 0.2708,
+      "step": 11167
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.545814349863931,
+      "learning_rate": 2.1491759317510456e-06,
+      "loss": 0.2598,
+      "step": 11168
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8192730287147911,
+      "learning_rate": 2.148339281552122e-06,
+      "loss": 0.2671,
+      "step": 11169
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5983696864805788,
+      "learning_rate": 2.147502749673666e-06,
+      "loss": 0.4713,
+      "step": 11170
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4567805993697562,
+      "learning_rate": 2.146666336150386e-06,
+      "loss": 0.2584,
+      "step": 11171
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7143646039687903,
+      "learning_rate": 2.1458300410169875e-06,
+      "loss": 0.2721,
+      "step": 11172
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5423343491809198,
+      "learning_rate": 2.1449938643081717e-06,
+      "loss": 0.27,
+      "step": 11173
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.500103176962945,
+      "learning_rate": 2.1441578060586322e-06,
+      "loss": 0.2648,
+      "step": 11174
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8847348174787146,
+      "learning_rate": 2.1433218663030576e-06,
+      "loss": 0.2666,
+      "step": 11175
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5373001390888033,
+      "learning_rate": 2.142486045076131e-06,
+      "loss": 0.2703,
+      "step": 11176
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4707709473272812,
+      "learning_rate": 2.141650342412536e-06,
+      "loss": 0.2726,
+      "step": 11177
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.477765128720051,
+      "learning_rate": 2.1408147583469446e-06,
+      "loss": 0.2521,
+      "step": 11178
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.38179731331009,
+      "learning_rate": 2.139979292914026e-06,
+      "loss": 0.2706,
+      "step": 11179
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.201913602413681,
+      "learning_rate": 2.1391439461484463e-06,
+      "loss": 0.2605,
+      "step": 11180
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 13.54492519426392,
+      "learning_rate": 2.1383087180848685e-06,
+      "loss": 0.2594,
+      "step": 11181
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6837664152401797,
+      "learning_rate": 2.137473608757944e-06,
+      "loss": 0.2779,
+      "step": 11182
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.5342625439670545,
+      "learning_rate": 2.1366386182023237e-06,
+      "loss": 0.2605,
+      "step": 11183
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.238471559858704,
+      "learning_rate": 2.1358037464526516e-06,
+      "loss": 0.2699,
+      "step": 11184
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.21358807675687,
+      "learning_rate": 2.1349689935435715e-06,
+      "loss": 0.2727,
+      "step": 11185
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2865253666288954,
+      "learning_rate": 2.1341343595097163e-06,
+      "loss": 0.3045,
+      "step": 11186
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5527068162802222,
+      "learning_rate": 2.133299844385715e-06,
+      "loss": 0.2666,
+      "step": 11187
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3260381470819143,
+      "learning_rate": 2.1324654482061953e-06,
+      "loss": 0.2555,
+      "step": 11188
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6032118521629376,
+      "learning_rate": 2.131631171005779e-06,
+      "loss": 0.4699,
+      "step": 11189
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7199801215884585,
+      "learning_rate": 2.13079701281908e-06,
+      "loss": 0.2771,
+      "step": 11190
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7997946886436644,
+      "learning_rate": 2.129962973680708e-06,
+      "loss": 0.2753,
+      "step": 11191
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3796796507655373,
+      "learning_rate": 2.1291290536252716e-06,
+      "loss": 0.2567,
+      "step": 11192
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.750675715309363,
+      "learning_rate": 2.12829525268737e-06,
+      "loss": 0.2799,
+      "step": 11193
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.40052063887953,
+      "learning_rate": 2.1274615709015972e-06,
+      "loss": 0.2622,
+      "step": 11194
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4484046836847706,
+      "learning_rate": 2.126628008302548e-06,
+      "loss": 0.2883,
+      "step": 11195
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1333827805240704,
+      "learning_rate": 2.1257945649248048e-06,
+      "loss": 0.2729,
+      "step": 11196
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5699763365801473,
+      "learning_rate": 2.1249612408029516e-06,
+      "loss": 0.246,
+      "step": 11197
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9683026273894892,
+      "learning_rate": 2.124128035971563e-06,
+      "loss": 0.2814,
+      "step": 11198
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8845964088070704,
+      "learning_rate": 2.123294950465209e-06,
+      "loss": 0.268,
+      "step": 11199
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4046556680794728,
+      "learning_rate": 2.122461984318459e-06,
+      "loss": 0.2496,
+      "step": 11200
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.362065342879254,
+      "learning_rate": 2.121629137565872e-06,
+      "loss": 0.2524,
+      "step": 11201
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8259556858118575,
+      "learning_rate": 2.1207964102420032e-06,
+      "loss": 0.2635,
+      "step": 11202
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.946762179583013,
+      "learning_rate": 2.119963802381407e-06,
+      "loss": 0.2573,
+      "step": 11203
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.479033609249847,
+      "learning_rate": 2.119131314018626e-06,
+      "loss": 0.2742,
+      "step": 11204
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5973115487074803,
+      "learning_rate": 2.1182989451882056e-06,
+      "loss": 0.4903,
+      "step": 11205
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5642157796981417,
+      "learning_rate": 2.117466695924681e-06,
+      "loss": 0.2651,
+      "step": 11206
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.257603880333518,
+      "learning_rate": 2.116634566262581e-06,
+      "loss": 0.242,
+      "step": 11207
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3102517941324026,
+      "learning_rate": 2.115802556236436e-06,
+      "loss": 0.2617,
+      "step": 11208
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6392256961189067,
+      "learning_rate": 2.1149706658807646e-06,
+      "loss": 0.2691,
+      "step": 11209
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.976684480344322,
+      "learning_rate": 2.1141388952300855e-06,
+      "loss": 0.2957,
+      "step": 11210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8422656535467308,
+      "learning_rate": 2.1133072443189106e-06,
+      "loss": 0.259,
+      "step": 11211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.367002481481189,
+      "learning_rate": 2.112475713181743e-06,
+      "loss": 0.2685,
+      "step": 11212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5502750509184016,
+      "learning_rate": 2.1116443018530887e-06,
+      "loss": 0.2928,
+      "step": 11213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.695723953025711,
+      "learning_rate": 2.110813010367443e-06,
+      "loss": 0.2697,
+      "step": 11214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.120399936015536,
+      "learning_rate": 2.109981838759295e-06,
+      "loss": 0.2632,
+      "step": 11215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6510816232567276,
+      "learning_rate": 2.109150787063134e-06,
+      "loss": 0.4772,
+      "step": 11216
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6061274158898775,
+      "learning_rate": 2.108319855313443e-06,
+      "loss": 0.2485,
+      "step": 11217
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1842445092804588,
+      "learning_rate": 2.1074890435446976e-06,
+      "loss": 0.2538,
+      "step": 11218
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5670237583121522,
+      "learning_rate": 2.1066583517913696e-06,
+      "loss": 0.2578,
+      "step": 11219
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.618919175699726,
+      "learning_rate": 2.105827780087924e-06,
+      "loss": 0.2719,
+      "step": 11220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.536684002560847,
+      "learning_rate": 2.104997328468826e-06,
+      "loss": 0.2759,
+      "step": 11221
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5864349389049004,
+      "learning_rate": 2.104166996968531e-06,
+      "loss": 0.4442,
+      "step": 11222
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 25.57374727550384,
+      "learning_rate": 2.10333678562149e-06,
+      "loss": 0.2453,
+      "step": 11223
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1942965238801406,
+      "learning_rate": 2.1025066944621498e-06,
+      "loss": 0.2526,
+      "step": 11224
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5604876554963547,
+      "learning_rate": 2.101676723524955e-06,
+      "loss": 0.4653,
+      "step": 11225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.538286455308112,
+      "learning_rate": 2.1008468728443416e-06,
+      "loss": 0.2652,
+      "step": 11226
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.305232031863097,
+      "learning_rate": 2.1000171424547406e-06,
+      "loss": 0.2627,
+      "step": 11227
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5268653904148923,
+      "learning_rate": 2.0991875323905773e-06,
+      "loss": 0.2597,
+      "step": 11228
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4579551686758565,
+      "learning_rate": 2.0983580426862777e-06,
+      "loss": 0.2527,
+      "step": 11229
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6436004300885614,
+      "learning_rate": 2.0975286733762575e-06,
+      "loss": 0.2676,
+      "step": 11230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0157942786307443,
+      "learning_rate": 2.096699424494925e-06,
+      "loss": 0.2687,
+      "step": 11231
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.089680836571682,
+      "learning_rate": 2.0958702960766907e-06,
+      "loss": 0.275,
+      "step": 11232
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4347347039415697,
+      "learning_rate": 2.0950412881559584e-06,
+      "loss": 0.2839,
+      "step": 11233
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8412385516769503,
+      "learning_rate": 2.094212400767122e-06,
+      "loss": 0.2751,
+      "step": 11234
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.072658839230748,
+      "learning_rate": 2.093383633944575e-06,
+      "loss": 0.2532,
+      "step": 11235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.642079530520852,
+      "learning_rate": 2.092554987722702e-06,
+      "loss": 0.2385,
+      "step": 11236
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8786982703045854,
+      "learning_rate": 2.091726462135888e-06,
+      "loss": 0.278,
+      "step": 11237
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8000350325657724,
+      "learning_rate": 2.0908980572185093e-06,
+      "loss": 0.2574,
+      "step": 11238
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.059791199192832,
+      "learning_rate": 2.0900697730049353e-06,
+      "loss": 0.2777,
+      "step": 11239
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4467078106599707,
+      "learning_rate": 2.089241609529535e-06,
+      "loss": 0.2698,
+      "step": 11240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5454674748671775,
+      "learning_rate": 2.0884135668266725e-06,
+      "loss": 0.2583,
+      "step": 11241
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5129349252160442,
+      "learning_rate": 2.0875856449307026e-06,
+      "loss": 0.2698,
+      "step": 11242
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.078105517035438,
+      "learning_rate": 2.0867578438759766e-06,
+      "loss": 0.2515,
+      "step": 11243
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8883553602994059,
+      "learning_rate": 2.085930163696841e-06,
+      "loss": 0.2811,
+      "step": 11244
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6460025422701143,
+      "learning_rate": 2.0851026044276405e-06,
+      "loss": 0.2616,
+      "step": 11245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6729825269022631,
+      "learning_rate": 2.0842751661027087e-06,
+      "loss": 0.2549,
+      "step": 11246
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.219001016374269,
+      "learning_rate": 2.0834478487563815e-06,
+      "loss": 0.2702,
+      "step": 11247
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6866832792573418,
+      "learning_rate": 2.082620652422981e-06,
+      "loss": 0.2761,
+      "step": 11248
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.055320779195588,
+      "learning_rate": 2.081793577136833e-06,
+      "loss": 0.2475,
+      "step": 11249
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9123174739920292,
+      "learning_rate": 2.0809666229322533e-06,
+      "loss": 0.2486,
+      "step": 11250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.997780556288267,
+      "learning_rate": 2.0801397898435534e-06,
+      "loss": 0.2686,
+      "step": 11251
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1026974105828473,
+      "learning_rate": 2.0793130779050374e-06,
+      "loss": 0.2579,
+      "step": 11252
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8078411799573328,
+      "learning_rate": 2.078486487151012e-06,
+      "loss": 0.26,
+      "step": 11253
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.882044324375287,
+      "learning_rate": 2.077660017615769e-06,
+      "loss": 0.2412,
+      "step": 11254
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2613193331472154,
+      "learning_rate": 2.076833669333605e-06,
+      "loss": 0.2832,
+      "step": 11255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.408733614715769,
+      "learning_rate": 2.0760074423388015e-06,
+      "loss": 0.2847,
+      "step": 11256
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.70585575241767,
+      "learning_rate": 2.075181336665645e-06,
+      "loss": 0.2561,
+      "step": 11257
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8369894987381366,
+      "learning_rate": 2.0743553523484107e-06,
+      "loss": 0.2864,
+      "step": 11258
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9047531515855274,
+      "learning_rate": 2.0735294894213687e-06,
+      "loss": 0.2486,
+      "step": 11259
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8312429988424974,
+      "learning_rate": 2.072703747918784e-06,
+      "loss": 0.277,
+      "step": 11260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.84859456575309,
+      "learning_rate": 2.0718781278749206e-06,
+      "loss": 0.2704,
+      "step": 11261
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4552664135300197,
+      "learning_rate": 2.0710526293240367e-06,
+      "loss": 0.2571,
+      "step": 11262
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.2689442857165725,
+      "learning_rate": 2.0702272523003815e-06,
+      "loss": 0.2815,
+      "step": 11263
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4942252443609294,
+      "learning_rate": 2.0694019968381993e-06,
+      "loss": 0.2756,
+      "step": 11264
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.261708605292174,
+      "learning_rate": 2.068576862971736e-06,
+      "loss": 0.2721,
+      "step": 11265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1416618633960653,
+      "learning_rate": 2.0677518507352246e-06,
+      "loss": 0.2548,
+      "step": 11266
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.741672270038633,
+      "learning_rate": 2.066926960162896e-06,
+      "loss": 0.248,
+      "step": 11267
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3673046110835774,
+      "learning_rate": 2.066102191288979e-06,
+      "loss": 0.2728,
+      "step": 11268
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8066562817407386,
+      "learning_rate": 2.065277544147692e-06,
+      "loss": 0.2591,
+      "step": 11269
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6442037120335296,
+      "learning_rate": 2.0644530187732538e-06,
+      "loss": 0.2759,
+      "step": 11270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7962337544630333,
+      "learning_rate": 2.0636286151998737e-06,
+      "loss": 0.2592,
+      "step": 11271
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7439235779197984,
+      "learning_rate": 2.0628043334617565e-06,
+      "loss": 0.2651,
+      "step": 11272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.802187952148651,
+      "learning_rate": 2.061980173593106e-06,
+      "loss": 0.2611,
+      "step": 11273
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.046964530052051,
+      "learning_rate": 2.061156135628117e-06,
+      "loss": 0.2458,
+      "step": 11274
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7023506125150913,
+      "learning_rate": 2.060332219600978e-06,
+      "loss": 0.2495,
+      "step": 11275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4510190663030633,
+      "learning_rate": 2.059508425545876e-06,
+      "loss": 0.2583,
+      "step": 11276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5984112980284597,
+      "learning_rate": 2.058684753496995e-06,
+      "loss": 0.4544,
+      "step": 11277
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.718038916639918,
+      "learning_rate": 2.0578612034885083e-06,
+      "loss": 0.2435,
+      "step": 11278
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1972264636950234,
+      "learning_rate": 2.057037775554585e-06,
+      "loss": 0.2464,
+      "step": 11279
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3567895580061946,
+      "learning_rate": 2.056214469729391e-06,
+      "loss": 0.2606,
+      "step": 11280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6567632067021063,
+      "learning_rate": 2.055391286047088e-06,
+      "loss": 0.2594,
+      "step": 11281
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.132715962089489,
+      "learning_rate": 2.0545682245418314e-06,
+      "loss": 0.2676,
+      "step": 11282
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8983599977686207,
+      "learning_rate": 2.053745285247769e-06,
+      "loss": 0.2495,
+      "step": 11283
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2749586217527535,
+      "learning_rate": 2.0529224681990477e-06,
+      "loss": 0.2441,
+      "step": 11284
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.194913080985443,
+      "learning_rate": 2.0520997734298096e-06,
+      "loss": 0.2555,
+      "step": 11285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6124897014250361,
+      "learning_rate": 2.0512772009741883e-06,
+      "loss": 0.2616,
+      "step": 11286
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.539949427933494,
+      "learning_rate": 2.050454750866313e-06,
+      "loss": 0.2718,
+      "step": 11287
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7385771706836386,
+      "learning_rate": 2.0496324231403074e-06,
+      "loss": 0.2653,
+      "step": 11288
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6378637675240364,
+      "learning_rate": 2.048810217830295e-06,
+      "loss": 0.2604,
+      "step": 11289
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0440401002533757,
+      "learning_rate": 2.0479881349703885e-06,
+      "loss": 0.2511,
+      "step": 11290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6730871924817556,
+      "learning_rate": 2.047166174594696e-06,
+      "loss": 0.2533,
+      "step": 11291
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8096736846885213,
+      "learning_rate": 2.0463443367373232e-06,
+      "loss": 0.2552,
+      "step": 11292
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8010067054137464,
+      "learning_rate": 2.045522621432372e-06,
+      "loss": 0.2632,
+      "step": 11293
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.317427290002973,
+      "learning_rate": 2.0447010287139357e-06,
+      "loss": 0.2924,
+      "step": 11294
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7938139433949192,
+      "learning_rate": 2.0438795586161027e-06,
+      "loss": 0.2562,
+      "step": 11295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.696871152362575,
+      "learning_rate": 2.0430582111729553e-06,
+      "loss": 0.2522,
+      "step": 11296
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.303049992480189,
+      "learning_rate": 2.0422369864185764e-06,
+      "loss": 0.2691,
+      "step": 11297
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7809763358651913,
+      "learning_rate": 2.0414158843870374e-06,
+      "loss": 0.2572,
+      "step": 11298
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5836344539530396,
+      "learning_rate": 2.0405949051124098e-06,
+      "loss": 0.2917,
+      "step": 11299
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.499452163795848,
+      "learning_rate": 2.039774048628754e-06,
+      "loss": 0.3023,
+      "step": 11300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.742859695149067,
+      "learning_rate": 2.0389533149701337e-06,
+      "loss": 0.2683,
+      "step": 11301
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7488158903285846,
+      "learning_rate": 2.038132704170599e-06,
+      "loss": 0.2764,
+      "step": 11302
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8710981474400807,
+      "learning_rate": 2.037312216264199e-06,
+      "loss": 0.2616,
+      "step": 11303
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7558263431775918,
+      "learning_rate": 2.0364918512849763e-06,
+      "loss": 0.2754,
+      "step": 11304
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3723448692660163,
+      "learning_rate": 2.035671609266972e-06,
+      "loss": 0.2581,
+      "step": 11305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.655202570536015,
+      "learning_rate": 2.0348514902442163e-06,
+      "loss": 0.2574,
+      "step": 11306
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4593716364778655,
+      "learning_rate": 2.0340314942507404e-06,
+      "loss": 0.2409,
+      "step": 11307
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4099536551064702,
+      "learning_rate": 2.0332116213205643e-06,
+      "loss": 0.2351,
+      "step": 11308
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.631861540894847,
+      "learning_rate": 2.0323918714877094e-06,
+      "loss": 0.2514,
+      "step": 11309
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8491413432317427,
+      "learning_rate": 2.031572244786187e-06,
+      "loss": 0.2756,
+      "step": 11310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.418752229745969,
+      "learning_rate": 2.0307527412500044e-06,
+      "loss": 0.2906,
+      "step": 11311
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.559290461037333,
+      "learning_rate": 2.0299333609131635e-06,
+      "loss": 0.2817,
+      "step": 11312
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.14333564023226,
+      "learning_rate": 2.0291141038096616e-06,
+      "loss": 0.2633,
+      "step": 11313
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.988662410341561,
+      "learning_rate": 2.028294969973495e-06,
+      "loss": 0.2566,
+      "step": 11314
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.276707696834116,
+      "learning_rate": 2.0274759594386485e-06,
+      "loss": 0.2485,
+      "step": 11315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6027524435209456,
+      "learning_rate": 2.0266570722391023e-06,
+      "loss": 0.4605,
+      "step": 11316
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5611061983899408,
+      "learning_rate": 2.025838308408837e-06,
+      "loss": 0.2797,
+      "step": 11317
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7898125265811011,
+      "learning_rate": 2.0250196679818234e-06,
+      "loss": 0.2612,
+      "step": 11318
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8054608946122124,
+      "learning_rate": 2.024201150992028e-06,
+      "loss": 0.274,
+      "step": 11319
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8249182693775046,
+      "learning_rate": 2.02338275747341e-06,
+      "loss": 0.2681,
+      "step": 11320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.2086545351873355,
+      "learning_rate": 2.022564487459929e-06,
+      "loss": 0.2592,
+      "step": 11321
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.263025405654816,
+      "learning_rate": 2.0217463409855377e-06,
+      "loss": 0.2577,
+      "step": 11322
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6013874667329715,
+      "learning_rate": 2.0209283180841803e-06,
+      "loss": 0.2683,
+      "step": 11323
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.716110993432087,
+      "learning_rate": 2.0201104187897975e-06,
+      "loss": 0.2551,
+      "step": 11324
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1477287728078567,
+      "learning_rate": 2.0192926431363274e-06,
+      "loss": 0.2529,
+      "step": 11325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.315932900324255,
+      "learning_rate": 2.0184749911576996e-06,
+      "loss": 0.2609,
+      "step": 11326
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5035421602010106,
+      "learning_rate": 2.017657462887841e-06,
+      "loss": 0.2578,
+      "step": 11327
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.905785262196667,
+      "learning_rate": 2.016840058360667e-06,
+      "loss": 0.2752,
+      "step": 11328
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.764265828420965,
+      "learning_rate": 2.016022777610102e-06,
+      "loss": 0.2609,
+      "step": 11329
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0082251367652777,
+      "learning_rate": 2.0152056206700516e-06,
+      "loss": 0.2576,
+      "step": 11330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9534691166959397,
+      "learning_rate": 2.0143885875744217e-06,
+      "loss": 0.2507,
+      "step": 11331
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9746059108719314,
+      "learning_rate": 2.013571678357111e-06,
+      "loss": 0.2694,
+      "step": 11332
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6165579992429837,
+      "learning_rate": 2.0127548930520173e-06,
+      "loss": 0.2777,
+      "step": 11333
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.443260674923377,
+      "learning_rate": 2.0119382316930293e-06,
+      "loss": 0.265,
+      "step": 11334
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9760462315592937,
+      "learning_rate": 2.011121694314031e-06,
+      "loss": 0.2701,
+      "step": 11335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.320175705161023,
+      "learning_rate": 2.0103052809488994e-06,
+      "loss": 0.2614,
+      "step": 11336
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.020051859425412,
+      "learning_rate": 2.009488991631515e-06,
+      "loss": 0.2622,
+      "step": 11337
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.721175284752356,
+      "learning_rate": 2.0086728263957446e-06,
+      "loss": 0.2686,
+      "step": 11338
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6424330441567097,
+      "learning_rate": 2.007856785275451e-06,
+      "loss": 0.2632,
+      "step": 11339
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.101973955010961,
+      "learning_rate": 2.0070408683044927e-06,
+      "loss": 0.2554,
+      "step": 11340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7957450911433073,
+      "learning_rate": 2.0062250755167273e-06,
+      "loss": 0.2608,
+      "step": 11341
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.434905798012273,
+      "learning_rate": 2.005409406946e-06,
+      "loss": 0.2681,
+      "step": 11342
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.055028020870884,
+      "learning_rate": 2.0045938626261545e-06,
+      "loss": 0.2695,
+      "step": 11343
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4087870325594685,
+      "learning_rate": 2.0037784425910297e-06,
+      "loss": 0.2586,
+      "step": 11344
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3772419911175398,
+      "learning_rate": 2.0029631468744608e-06,
+      "loss": 0.2676,
+      "step": 11345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.628980658008442,
+      "learning_rate": 2.0021479755102746e-06,
+      "loss": 0.2662,
+      "step": 11346
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.73750429787218,
+      "learning_rate": 2.0013329285322935e-06,
+      "loss": 0.2606,
+      "step": 11347
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.216705420367553,
+      "learning_rate": 2.000518005974334e-06,
+      "loss": 0.2732,
+      "step": 11348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5501045833967493,
+      "learning_rate": 1.999703207870211e-06,
+      "loss": 0.2518,
+      "step": 11349
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.345120807856687,
+      "learning_rate": 1.998888534253732e-06,
+      "loss": 0.2445,
+      "step": 11350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5801532472428638,
+      "learning_rate": 1.9980739851586962e-06,
+      "loss": 0.467,
+      "step": 11351
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4128805356799254,
+      "learning_rate": 1.9972595606189026e-06,
+      "loss": 0.2514,
+      "step": 11352
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8872851415395644,
+      "learning_rate": 1.996445260668145e-06,
+      "loss": 0.2667,
+      "step": 11353
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.51318997529435,
+      "learning_rate": 1.995631085340209e-06,
+      "loss": 0.2538,
+      "step": 11354
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9785116508051475,
+      "learning_rate": 1.994817034668875e-06,
+      "loss": 0.2547,
+      "step": 11355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1575406168232623,
+      "learning_rate": 1.9940031086879184e-06,
+      "loss": 0.2498,
+      "step": 11356
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4897315642909783,
+      "learning_rate": 1.9931893074311145e-06,
+      "loss": 0.272,
+      "step": 11357
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7826147316133816,
+      "learning_rate": 1.9923756309322246e-06,
+      "loss": 0.2541,
+      "step": 11358
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3455543589097223,
+      "learning_rate": 1.9915620792250133e-06,
+      "loss": 0.2675,
+      "step": 11359
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5774023758944333,
+      "learning_rate": 1.9907486523432336e-06,
+      "loss": 0.4495,
+      "step": 11360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6034605147044834,
+      "learning_rate": 1.989935350320639e-06,
+      "loss": 0.265,
+      "step": 11361
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1300232647843877,
+      "learning_rate": 1.9891221731909733e-06,
+      "loss": 0.268,
+      "step": 11362
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.669427878905994,
+      "learning_rate": 1.988309120987977e-06,
+      "loss": 0.2664,
+      "step": 11363
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.704242222687707,
+      "learning_rate": 1.987496193745382e-06,
+      "loss": 0.2603,
+      "step": 11364
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.579530652329033,
+      "learning_rate": 1.986683391496923e-06,
+      "loss": 0.2534,
+      "step": 11365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7661306057677901,
+      "learning_rate": 1.9858707142763205e-06,
+      "loss": 0.2562,
+      "step": 11366
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7420720500346207,
+      "learning_rate": 1.985058162117297e-06,
+      "loss": 0.2624,
+      "step": 11367
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5833631407263254,
+      "learning_rate": 1.9842457350535642e-06,
+      "loss": 0.5086,
+      "step": 11368
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5620309514403976,
+      "learning_rate": 1.9834334331188345e-06,
+      "loss": 0.4662,
+      "step": 11369
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.294759900570698,
+      "learning_rate": 1.9826212563468095e-06,
+      "loss": 0.2521,
+      "step": 11370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5657930904808723,
+      "learning_rate": 1.981809204771188e-06,
+      "loss": 0.2622,
+      "step": 11371
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3102104453249055,
+      "learning_rate": 1.9809972784256614e-06,
+      "loss": 0.2722,
+      "step": 11372
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.113928762991423,
+      "learning_rate": 1.9801854773439206e-06,
+      "loss": 0.2465,
+      "step": 11373
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9706746452481925,
+      "learning_rate": 1.9793738015596497e-06,
+      "loss": 0.2603,
+      "step": 11374
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2277375849547645,
+      "learning_rate": 1.978562251106525e-06,
+      "loss": 0.2622,
+      "step": 11375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.907641102352029,
+      "learning_rate": 1.977750826018217e-06,
+      "loss": 0.2648,
+      "step": 11376
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.592571779688514,
+      "learning_rate": 1.9769395263283973e-06,
+      "loss": 0.2776,
+      "step": 11377
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2381136243548942,
+      "learning_rate": 1.9761283520707268e-06,
+      "loss": 0.2761,
+      "step": 11378
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9940143447905285,
+      "learning_rate": 1.9753173032788616e-06,
+      "loss": 0.2401,
+      "step": 11379
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5298983352048547,
+      "learning_rate": 1.9745063799864506e-06,
+      "loss": 0.2899,
+      "step": 11380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8051648505776428,
+      "learning_rate": 1.973695582227147e-06,
+      "loss": 0.27,
+      "step": 11381
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4991138970649578,
+      "learning_rate": 1.972884910034589e-06,
+      "loss": 0.2431,
+      "step": 11382
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0111220715082245,
+      "learning_rate": 1.972074363442413e-06,
+      "loss": 0.2606,
+      "step": 11383
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8018546620806766,
+      "learning_rate": 1.971263942484248e-06,
+      "loss": 0.2494,
+      "step": 11384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.002055503375998,
+      "learning_rate": 1.9704536471937238e-06,
+      "loss": 0.2641,
+      "step": 11385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7357738843295063,
+      "learning_rate": 1.969643477604458e-06,
+      "loss": 0.2574,
+      "step": 11386
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.028767298644523,
+      "learning_rate": 1.9688334337500677e-06,
+      "loss": 0.2617,
+      "step": 11387
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.581514765675958,
+      "learning_rate": 1.968023515664158e-06,
+      "loss": 0.2517,
+      "step": 11388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9354875005489085,
+      "learning_rate": 1.967213723380342e-06,
+      "loss": 0.2719,
+      "step": 11389
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.143422083357361,
+      "learning_rate": 1.966404056932215e-06,
+      "loss": 0.2546,
+      "step": 11390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7764018934828976,
+      "learning_rate": 1.9655945163533723e-06,
+      "loss": 0.2585,
+      "step": 11391
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.497693339571945,
+      "learning_rate": 1.964785101677401e-06,
+      "loss": 0.2645,
+      "step": 11392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2809630448123803,
+      "learning_rate": 1.9639758129378888e-06,
+      "loss": 0.2706,
+      "step": 11393
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.075029666733048,
+      "learning_rate": 1.9631666501684126e-06,
+      "loss": 0.253,
+      "step": 11394
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.643925236239441,
+      "learning_rate": 1.9623576134025447e-06,
+      "loss": 0.2568,
+      "step": 11395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7403959416229888,
+      "learning_rate": 1.9615487026738546e-06,
+      "loss": 0.2488,
+      "step": 11396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.897176321303575,
+      "learning_rate": 1.9607399180159076e-06,
+      "loss": 0.2625,
+      "step": 11397
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4038872054017089,
+      "learning_rate": 1.9599312594622594e-06,
+      "loss": 0.2656,
+      "step": 11398
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4292549886742427,
+      "learning_rate": 1.9591227270464635e-06,
+      "loss": 0.2451,
+      "step": 11399
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6982447805255887,
+      "learning_rate": 1.958314320802065e-06,
+      "loss": 0.2666,
+      "step": 11400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4206654848390734,
+      "learning_rate": 1.9575060407626094e-06,
+      "loss": 0.2529,
+      "step": 11401
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.900697119871712,
+      "learning_rate": 1.9566978869616327e-06,
+      "loss": 0.2656,
+      "step": 11402
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6884534322469396,
+      "learning_rate": 1.9558898594326635e-06,
+      "loss": 0.2742,
+      "step": 11403
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.341516010359561,
+      "learning_rate": 1.9550819582092317e-06,
+      "loss": 0.2589,
+      "step": 11404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6031833805313624,
+      "learning_rate": 1.9542741833248597e-06,
+      "loss": 0.2622,
+      "step": 11405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1997605954118207,
+      "learning_rate": 1.953466534813062e-06,
+      "loss": 0.2777,
+      "step": 11406
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.261326186835715,
+      "learning_rate": 1.952659012707348e-06,
+      "loss": 0.2638,
+      "step": 11407
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.663279408984441,
+      "learning_rate": 1.9518516170412234e-06,
+      "loss": 0.2684,
+      "step": 11408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5113837783844868,
+      "learning_rate": 1.9510443478481905e-06,
+      "loss": 0.2465,
+      "step": 11409
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8365462034653697,
+      "learning_rate": 1.950237205161741e-06,
+      "loss": 0.2643,
+      "step": 11410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4929882173889322,
+      "learning_rate": 1.9494301890153695e-06,
+      "loss": 0.2603,
+      "step": 11411
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.408961938331628,
+      "learning_rate": 1.948623299442556e-06,
+      "loss": 0.2482,
+      "step": 11412
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.737767357578204,
+      "learning_rate": 1.947816536476783e-06,
+      "loss": 0.2648,
+      "step": 11413
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.524209273130116,
+      "learning_rate": 1.9470099001515237e-06,
+      "loss": 0.2671,
+      "step": 11414
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4741269951647045,
+      "learning_rate": 1.9462033905002457e-06,
+      "loss": 0.2698,
+      "step": 11415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2131247086558754,
+      "learning_rate": 1.945397007556412e-06,
+      "loss": 0.2547,
+      "step": 11416
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4341785772433284,
+      "learning_rate": 1.9445907513534834e-06,
+      "loss": 0.2648,
+      "step": 11417
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5722984405073626,
+      "learning_rate": 1.94378462192491e-06,
+      "loss": 0.2631,
+      "step": 11418
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5805557957652592,
+      "learning_rate": 1.942978619304143e-06,
+      "loss": 0.2633,
+      "step": 11419
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.472081491767904,
+      "learning_rate": 1.9421727435246214e-06,
+      "loss": 0.2489,
+      "step": 11420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2142139857719663,
+      "learning_rate": 1.941366994619785e-06,
+      "loss": 0.2494,
+      "step": 11421
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7860826111303763,
+      "learning_rate": 1.9405613726230653e-06,
+      "loss": 0.2556,
+      "step": 11422
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8046385687067628,
+      "learning_rate": 1.939755877567888e-06,
+      "loss": 0.2595,
+      "step": 11423
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6670271164555346,
+      "learning_rate": 1.9389505094876736e-06,
+      "loss": 0.2674,
+      "step": 11424
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8890571202441133,
+      "learning_rate": 1.938145268415839e-06,
+      "loss": 0.2969,
+      "step": 11425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.9879525381207985,
+      "learning_rate": 1.9373401543857983e-06,
+      "loss": 0.2796,
+      "step": 11426
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3161419496663187,
+      "learning_rate": 1.936535167430954e-06,
+      "loss": 0.2518,
+      "step": 11427
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.232052173403798,
+      "learning_rate": 1.935730307584705e-06,
+      "loss": 0.2599,
+      "step": 11428
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9742934527172964,
+      "learning_rate": 1.9349255748804506e-06,
+      "loss": 0.2528,
+      "step": 11429
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8988202940201568,
+      "learning_rate": 1.934120969351578e-06,
+      "loss": 0.2792,
+      "step": 11430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.403931595456558,
+      "learning_rate": 1.9333164910314715e-06,
+      "loss": 0.2518,
+      "step": 11431
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.326714842309047,
+      "learning_rate": 1.9325121399535103e-06,
+      "loss": 0.2718,
+      "step": 11432
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.867530048596062,
+      "learning_rate": 1.9317079161510675e-06,
+      "loss": 0.2716,
+      "step": 11433
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.516557314740614,
+      "learning_rate": 1.9309038196575157e-06,
+      "loss": 0.2595,
+      "step": 11434
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.586599339571851,
+      "learning_rate": 1.9300998505062157e-06,
+      "loss": 0.2398,
+      "step": 11435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2522506558220243,
+      "learning_rate": 1.9292960087305234e-06,
+      "loss": 0.2481,
+      "step": 11436
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0020347706698143,
+      "learning_rate": 1.9284922943637965e-06,
+      "loss": 0.2502,
+      "step": 11437
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.79563012579662,
+      "learning_rate": 1.9276887074393793e-06,
+      "loss": 0.2682,
+      "step": 11438
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8305073671706702,
+      "learning_rate": 1.926885247990615e-06,
+      "loss": 0.2958,
+      "step": 11439
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.72451949456854,
+      "learning_rate": 1.926081916050838e-06,
+      "loss": 0.2643,
+      "step": 11440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6114740381338866,
+      "learning_rate": 1.9252787116533826e-06,
+      "loss": 0.459,
+      "step": 11441
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5079612345430435,
+      "learning_rate": 1.924475634831576e-06,
+      "loss": 0.2539,
+      "step": 11442
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.127769946153632,
+      "learning_rate": 1.9236726856187387e-06,
+      "loss": 0.279,
+      "step": 11443
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8740271212326134,
+      "learning_rate": 1.922869864048184e-06,
+      "loss": 0.2692,
+      "step": 11444
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.823193367665255,
+      "learning_rate": 1.9220671701532256e-06,
+      "loss": 0.2842,
+      "step": 11445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3000630886978786,
+      "learning_rate": 1.921264603967168e-06,
+      "loss": 0.2528,
+      "step": 11446
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1792497744263115,
+      "learning_rate": 1.9204621655233096e-06,
+      "loss": 0.2544,
+      "step": 11447
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.960125826508275,
+      "learning_rate": 1.9196598548549445e-06,
+      "loss": 0.2498,
+      "step": 11448
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0879985838242017,
+      "learning_rate": 1.9188576719953635e-06,
+      "loss": 0.274,
+      "step": 11449
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5514771715147577,
+      "learning_rate": 1.918055616977852e-06,
+      "loss": 0.2523,
+      "step": 11450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.915211518804343,
+      "learning_rate": 1.917253689835687e-06,
+      "loss": 0.2548,
+      "step": 11451
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6524053314383431,
+      "learning_rate": 1.91645189060214e-06,
+      "loss": 0.482,
+      "step": 11452
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.990368602133595,
+      "learning_rate": 1.915650219310483e-06,
+      "loss": 0.2472,
+      "step": 11453
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7150313533698642,
+      "learning_rate": 1.914848675993977e-06,
+      "loss": 0.2724,
+      "step": 11454
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.56708186980046,
+      "learning_rate": 1.914047260685878e-06,
+      "loss": 0.2543,
+      "step": 11455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1096326092215487,
+      "learning_rate": 1.91324597341944e-06,
+      "loss": 0.2535,
+      "step": 11456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1502170936499385,
+      "learning_rate": 1.9124448142279084e-06,
+      "loss": 0.2619,
+      "step": 11457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4409823053965565,
+      "learning_rate": 1.9116437831445273e-06,
+      "loss": 0.2714,
+      "step": 11458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.231188912890816,
+      "learning_rate": 1.9108428802025308e-06,
+      "loss": 0.2617,
+      "step": 11459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5516612379231187,
+      "learning_rate": 1.910042105435149e-06,
+      "loss": 0.2598,
+      "step": 11460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5415978599533364,
+      "learning_rate": 1.90924145887561e-06,
+      "loss": 0.2527,
+      "step": 11461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4079412109689216,
+      "learning_rate": 1.90844094055713e-06,
+      "loss": 0.2778,
+      "step": 11462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5578665959114903,
+      "learning_rate": 1.9076405505129298e-06,
+      "loss": 0.248,
+      "step": 11463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7101699307622105,
+      "learning_rate": 1.9068402887762133e-06,
+      "loss": 0.269,
+      "step": 11464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8690068427042108,
+      "learning_rate": 1.9060401553801887e-06,
+      "loss": 0.267,
+      "step": 11465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.802343913344231,
+      "learning_rate": 1.9052401503580541e-06,
+      "loss": 0.2572,
+      "step": 11466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9048858238915385,
+      "learning_rate": 1.904440273743003e-06,
+      "loss": 0.2667,
+      "step": 11467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5851466547212005,
+      "learning_rate": 1.903640525568221e-06,
+      "loss": 0.27,
+      "step": 11468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6106298306366249,
+      "learning_rate": 1.902840905866895e-06,
+      "loss": 0.4944,
+      "step": 11469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7960634647164426,
+      "learning_rate": 1.9020414146721995e-06,
+      "loss": 0.2557,
+      "step": 11470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.069294754634697,
+      "learning_rate": 1.9012420520173097e-06,
+      "loss": 0.265,
+      "step": 11471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1230797406007387,
+      "learning_rate": 1.9004428179353895e-06,
+      "loss": 0.249,
+      "step": 11472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.977106912460412,
+      "learning_rate": 1.8996437124596045e-06,
+      "loss": 0.2778,
+      "step": 11473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.148882914640681,
+      "learning_rate": 1.8988447356231082e-06,
+      "loss": 0.2437,
+      "step": 11474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.612463440365609,
+      "learning_rate": 1.8980458874590525e-06,
+      "loss": 0.2563,
+      "step": 11475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.251921214254672,
+      "learning_rate": 1.8972471680005805e-06,
+      "loss": 0.2567,
+      "step": 11476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6270869313081857,
+      "learning_rate": 1.8964485772808345e-06,
+      "loss": 0.2501,
+      "step": 11477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.031496906945291,
+      "learning_rate": 1.8956501153329516e-06,
+      "loss": 0.268,
+      "step": 11478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8361410734328967,
+      "learning_rate": 1.8948517821900592e-06,
+      "loss": 0.2653,
+      "step": 11479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1441496679139687,
+      "learning_rate": 1.8940535778852804e-06,
+      "loss": 0.2717,
+      "step": 11480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7183299102711644,
+      "learning_rate": 1.8932555024517363e-06,
+      "loss": 0.2746,
+      "step": 11481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1089751911758734,
+      "learning_rate": 1.8924575559225406e-06,
+      "loss": 0.2465,
+      "step": 11482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8839783331538997,
+      "learning_rate": 1.8916597383308e-06,
+      "loss": 0.2596,
+      "step": 11483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.550563791260173,
+      "learning_rate": 1.8908620497096159e-06,
+      "loss": 0.2631,
+      "step": 11484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.710879720987616,
+      "learning_rate": 1.8900644900920884e-06,
+      "loss": 0.2597,
+      "step": 11485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 32.768201741543656,
+      "learning_rate": 1.8892670595113105e-06,
+      "loss": 0.2624,
+      "step": 11486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.045250859829823,
+      "learning_rate": 1.8884697580003674e-06,
+      "loss": 0.2452,
+      "step": 11487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4852251987712686,
+      "learning_rate": 1.8876725855923394e-06,
+      "loss": 0.2683,
+      "step": 11488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6527115258742013,
+      "learning_rate": 1.8868755423203056e-06,
+      "loss": 0.2781,
+      "step": 11489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.399537261154635,
+      "learning_rate": 1.886078628217335e-06,
+      "loss": 0.2605,
+      "step": 11490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9713359301015374,
+      "learning_rate": 1.8852818433164927e-06,
+      "loss": 0.2516,
+      "step": 11491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.296147865855016,
+      "learning_rate": 1.8844851876508375e-06,
+      "loss": 0.2785,
+      "step": 11492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3306562118900755,
+      "learning_rate": 1.883688661253426e-06,
+      "loss": 0.2675,
+      "step": 11493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3282691457943727,
+      "learning_rate": 1.882892264157309e-06,
+      "loss": 0.2817,
+      "step": 11494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.276776008424065,
+      "learning_rate": 1.8820959963955282e-06,
+      "loss": 0.2816,
+      "step": 11495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.638741689277499,
+      "learning_rate": 1.8812998580011204e-06,
+      "loss": 0.268,
+      "step": 11496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.652807009176715,
+      "learning_rate": 1.880503849007123e-06,
+      "loss": 0.27,
+      "step": 11497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.67812107606071,
+      "learning_rate": 1.8797079694465619e-06,
+      "loss": 0.4994,
+      "step": 11498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6366241977833366,
+      "learning_rate": 1.8789122193524594e-06,
+      "loss": 0.4871,
+      "step": 11499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.891088314246333,
+      "learning_rate": 1.878116598757831e-06,
+      "loss": 0.2449,
+      "step": 11500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.89586591882636,
+      "learning_rate": 1.8773211076956898e-06,
+      "loss": 0.2506,
+      "step": 11501
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8680385957877492,
+      "learning_rate": 1.8765257461990444e-06,
+      "loss": 0.2644,
+      "step": 11502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5303857899235052,
+      "learning_rate": 1.875730514300893e-06,
+      "loss": 0.2511,
+      "step": 11503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1661314046330924,
+      "learning_rate": 1.8749354120342307e-06,
+      "loss": 0.2789,
+      "step": 11504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.91931618367787,
+      "learning_rate": 1.8741404394320506e-06,
+      "loss": 0.2493,
+      "step": 11505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0348641881384446,
+      "learning_rate": 1.8733455965273355e-06,
+      "loss": 0.2845,
+      "step": 11506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.931735348327818,
+      "learning_rate": 1.8725508833530638e-06,
+      "loss": 0.2584,
+      "step": 11507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.981218916565921,
+      "learning_rate": 1.8717562999422128e-06,
+      "loss": 0.2699,
+      "step": 11508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1187270794333974,
+      "learning_rate": 1.8709618463277474e-06,
+      "loss": 0.2518,
+      "step": 11509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.451284277050738,
+      "learning_rate": 1.8701675225426341e-06,
+      "loss": 0.2623,
+      "step": 11510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.391133879892232,
+      "learning_rate": 1.8693733286198296e-06,
+      "loss": 0.2484,
+      "step": 11511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.650244144192052,
+      "learning_rate": 1.8685792645922852e-06,
+      "loss": 0.2504,
+      "step": 11512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5001549512291834,
+      "learning_rate": 1.86778533049295e-06,
+      "loss": 0.2578,
+      "step": 11513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.153463694552769,
+      "learning_rate": 1.8669915263547656e-06,
+      "loss": 0.258,
+      "step": 11514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.506997653386509,
+      "learning_rate": 1.8661978522106654e-06,
+      "loss": 0.2477,
+      "step": 11515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.008297387897438,
+      "learning_rate": 1.8654043080935846e-06,
+      "loss": 0.2602,
+      "step": 11516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7498243961326454,
+      "learning_rate": 1.864610894036445e-06,
+      "loss": 0.2524,
+      "step": 11517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.510139821959095,
+      "learning_rate": 1.86381761007217e-06,
+      "loss": 0.2503,
+      "step": 11518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8953997341756814,
+      "learning_rate": 1.863024456233673e-06,
+      "loss": 0.2512,
+      "step": 11519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9859824432416895,
+      "learning_rate": 1.8622314325538615e-06,
+      "loss": 0.2664,
+      "step": 11520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.771046837551444,
+      "learning_rate": 1.8614385390656432e-06,
+      "loss": 0.2611,
+      "step": 11521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.571088068936452,
+      "learning_rate": 1.8606457758019125e-06,
+      "loss": 0.2409,
+      "step": 11522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.928979470443616,
+      "learning_rate": 1.8598531427955662e-06,
+      "loss": 0.2667,
+      "step": 11523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.703417933287143,
+      "learning_rate": 1.859060640079491e-06,
+      "loss": 0.2641,
+      "step": 11524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.716672621669686,
+      "learning_rate": 1.8582682676865666e-06,
+      "loss": 0.2802,
+      "step": 11525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.834495224753895,
+      "learning_rate": 1.8574760256496742e-06,
+      "loss": 0.2664,
+      "step": 11526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3960608525254616,
+      "learning_rate": 1.8566839140016829e-06,
+      "loss": 0.2581,
+      "step": 11527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3241034315555993,
+      "learning_rate": 1.8558919327754577e-06,
+      "loss": 0.2741,
+      "step": 11528
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.684780560119694,
+      "learning_rate": 1.8551000820038628e-06,
+      "loss": 0.2645,
+      "step": 11529
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8758240644753343,
+      "learning_rate": 1.8543083617197494e-06,
+      "loss": 0.2545,
+      "step": 11530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4815599160007857,
+      "learning_rate": 1.853516771955971e-06,
+      "loss": 0.2745,
+      "step": 11531
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8268556016603312,
+      "learning_rate": 1.852725312745371e-06,
+      "loss": 0.2628,
+      "step": 11532
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8432119921001295,
+      "learning_rate": 1.8519339841207857e-06,
+      "loss": 0.2567,
+      "step": 11533
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9344243019241505,
+      "learning_rate": 1.8511427861150526e-06,
+      "loss": 0.2548,
+      "step": 11534
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6120661362937332,
+      "learning_rate": 1.8503517187609993e-06,
+      "loss": 0.2566,
+      "step": 11535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8039736735689818,
+      "learning_rate": 1.8495607820914451e-06,
+      "loss": 0.2505,
+      "step": 11536
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5540975630774976,
+      "learning_rate": 1.8487699761392098e-06,
+      "loss": 0.2689,
+      "step": 11537
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7908464601448393,
+      "learning_rate": 1.8479793009371073e-06,
+      "loss": 0.2442,
+      "step": 11538
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2954945231916777,
+      "learning_rate": 1.8471887565179426e-06,
+      "loss": 0.2534,
+      "step": 11539
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3186157692833462,
+      "learning_rate": 1.8463983429145143e-06,
+      "loss": 0.2682,
+      "step": 11540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8801713595759486,
+      "learning_rate": 1.8456080601596216e-06,
+      "loss": 0.264,
+      "step": 11541
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.462008547163594,
+      "learning_rate": 1.844817908286054e-06,
+      "loss": 0.2373,
+      "step": 11542
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6097105168691064,
+      "learning_rate": 1.8440278873265948e-06,
+      "loss": 0.2622,
+      "step": 11543
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9709023890444757,
+      "learning_rate": 1.843237997314023e-06,
+      "loss": 0.2596,
+      "step": 11544
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8745638645026685,
+      "learning_rate": 1.8424482382811137e-06,
+      "loss": 0.2769,
+      "step": 11545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.586185787760726,
+      "learning_rate": 1.841658610260637e-06,
+      "loss": 0.2613,
+      "step": 11546
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6403014749982072,
+      "learning_rate": 1.8408691132853551e-06,
+      "loss": 0.2676,
+      "step": 11547
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2678189150366417,
+      "learning_rate": 1.8400797473880223e-06,
+      "loss": 0.3071,
+      "step": 11548
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6357951822984167,
+      "learning_rate": 1.8392905126013955e-06,
+      "loss": 0.2721,
+      "step": 11549
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.567934800131614,
+      "learning_rate": 1.8385014089582192e-06,
+      "loss": 0.2688,
+      "step": 11550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3716559900717848,
+      "learning_rate": 1.8377124364912353e-06,
+      "loss": 0.253,
+      "step": 11551
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8073247367751648,
+      "learning_rate": 1.8369235952331777e-06,
+      "loss": 0.2723,
+      "step": 11552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.31472427613521,
+      "learning_rate": 1.8361348852167777e-06,
+      "loss": 0.2721,
+      "step": 11553
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8746213927342585,
+      "learning_rate": 1.8353463064747629e-06,
+      "loss": 0.2794,
+      "step": 11554
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.976646350966091,
+      "learning_rate": 1.834557859039851e-06,
+      "loss": 0.2614,
+      "step": 11555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8511431337909376,
+      "learning_rate": 1.8337695429447545e-06,
+      "loss": 0.2523,
+      "step": 11556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.471850900623397,
+      "learning_rate": 1.832981358222185e-06,
+      "loss": 0.2729,
+      "step": 11557
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5496506970491823,
+      "learning_rate": 1.832193304904844e-06,
+      "loss": 0.2541,
+      "step": 11558
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.615265765841692,
+      "learning_rate": 1.831405383025428e-06,
+      "loss": 0.2706,
+      "step": 11559
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5479936667300176,
+      "learning_rate": 1.8306175926166325e-06,
+      "loss": 0.2746,
+      "step": 11560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.380276231801311,
+      "learning_rate": 1.8298299337111408e-06,
+      "loss": 0.2718,
+      "step": 11561
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4341596745639706,
+      "learning_rate": 1.8290424063416373e-06,
+      "loss": 0.2644,
+      "step": 11562
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.034541876921524,
+      "learning_rate": 1.828255010540797e-06,
+      "loss": 0.297,
+      "step": 11563
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7744960371721548,
+      "learning_rate": 1.8274677463412882e-06,
+      "loss": 0.254,
+      "step": 11564
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5001021581761966,
+      "learning_rate": 1.8266806137757798e-06,
+      "loss": 0.2655,
+      "step": 11565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.615301559198779,
+      "learning_rate": 1.8258936128769284e-06,
+      "loss": 0.2687,
+      "step": 11566
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.013126014624267,
+      "learning_rate": 1.8251067436773872e-06,
+      "loss": 0.2796,
+      "step": 11567
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8850472742960813,
+      "learning_rate": 1.8243200062098088e-06,
+      "loss": 0.2565,
+      "step": 11568
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5647787071432304,
+      "learning_rate": 1.8235334005068317e-06,
+      "loss": 0.2748,
+      "step": 11569
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.069001867201661,
+      "learning_rate": 1.8227469266010977e-06,
+      "loss": 0.2538,
+      "step": 11570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.263115722281822,
+      "learning_rate": 1.8219605845252376e-06,
+      "loss": 0.2987,
+      "step": 11571
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9129042617801826,
+      "learning_rate": 1.8211743743118754e-06,
+      "loss": 0.284,
+      "step": 11572
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.319808480813705,
+      "learning_rate": 1.8203882959936363e-06,
+      "loss": 0.2652,
+      "step": 11573
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.900489197518082,
+      "learning_rate": 1.8196023496031335e-06,
+      "loss": 0.2683,
+      "step": 11574
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4806769121490957,
+      "learning_rate": 1.8188165351729792e-06,
+      "loss": 0.2531,
+      "step": 11575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3055423915071636,
+      "learning_rate": 1.8180308527357776e-06,
+      "loss": 0.2787,
+      "step": 11576
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.6946897793832365,
+      "learning_rate": 1.817245302324126e-06,
+      "loss": 0.2731,
+      "step": 11577
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6195659234728494,
+      "learning_rate": 1.8164598839706216e-06,
+      "loss": 0.2789,
+      "step": 11578
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6389624269780703,
+      "learning_rate": 1.815674597707851e-06,
+      "loss": 0.2615,
+      "step": 11579
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.228590629549094,
+      "learning_rate": 1.814889443568396e-06,
+      "loss": 0.2412,
+      "step": 11580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.825630571773635,
+      "learning_rate": 1.8141044215848368e-06,
+      "loss": 0.2548,
+      "step": 11581
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.064168411558224,
+      "learning_rate": 1.8133195317897423e-06,
+      "loss": 0.2622,
+      "step": 11582
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5773045508505517,
+      "learning_rate": 1.8125347742156823e-06,
+      "loss": 0.259,
+      "step": 11583
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5084784166258076,
+      "learning_rate": 1.8117501488952166e-06,
+      "loss": 0.2521,
+      "step": 11584
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.833971890557965,
+      "learning_rate": 1.810965655860898e-06,
+      "loss": 0.2571,
+      "step": 11585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.91254258856792,
+      "learning_rate": 1.8101812951452808e-06,
+      "loss": 0.2701,
+      "step": 11586
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6042623625113431,
+      "learning_rate": 1.809397066780907e-06,
+      "loss": 0.456,
+      "step": 11587
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.617086890004137,
+      "learning_rate": 1.8086129708003142e-06,
+      "loss": 0.2647,
+      "step": 11588
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5307919202942437,
+      "learning_rate": 1.8078290072360382e-06,
+      "loss": 0.2663,
+      "step": 11589
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5452028492532603,
+      "learning_rate": 1.8070451761206082e-06,
+      "loss": 0.2675,
+      "step": 11590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4609771617984315,
+      "learning_rate": 1.806261477486545e-06,
+      "loss": 0.263,
+      "step": 11591
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8342722329839665,
+      "learning_rate": 1.8054779113663662e-06,
+      "loss": 0.258,
+      "step": 11592
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.735699545196827,
+      "learning_rate": 1.8046944777925806e-06,
+      "loss": 0.2414,
+      "step": 11593
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.20141948652523,
+      "learning_rate": 1.803911176797699e-06,
+      "loss": 0.2677,
+      "step": 11594
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.323350389575852,
+      "learning_rate": 1.8031280084142193e-06,
+      "loss": 0.2714,
+      "step": 11595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.919456952414611,
+      "learning_rate": 1.802344972674635e-06,
+      "loss": 0.2565,
+      "step": 11596
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4279110676808817,
+      "learning_rate": 1.8015620696114377e-06,
+      "loss": 0.2703,
+      "step": 11597
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.823465183847571,
+      "learning_rate": 1.8007792992571128e-06,
+      "loss": 0.2616,
+      "step": 11598
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.46668559256316,
+      "learning_rate": 1.7999966616441372e-06,
+      "loss": 0.2607,
+      "step": 11599
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.103914243666078,
+      "learning_rate": 1.7992141568049837e-06,
+      "loss": 0.288,
+      "step": 11600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7392868339710725,
+      "learning_rate": 1.798431784772119e-06,
+      "loss": 0.2563,
+      "step": 11601
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.827683844614338,
+      "learning_rate": 1.7976495455780074e-06,
+      "loss": 0.2777,
+      "step": 11602
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4846781854381141,
+      "learning_rate": 1.796867439255104e-06,
+      "loss": 0.2756,
+      "step": 11603
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.290617499398284,
+      "learning_rate": 1.7960854658358584e-06,
+      "loss": 0.2957,
+      "step": 11604
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.596333979997199,
+      "learning_rate": 1.7953036253527178e-06,
+      "loss": 0.2638,
+      "step": 11605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0672337342872313,
+      "learning_rate": 1.7945219178381236e-06,
+      "loss": 0.2683,
+      "step": 11606
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5226318927342812,
+      "learning_rate": 1.7937403433245087e-06,
+      "loss": 0.28,
+      "step": 11607
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2869851636087017,
+      "learning_rate": 1.7929589018443016e-06,
+      "loss": 0.2498,
+      "step": 11608
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5392580115799617,
+      "learning_rate": 1.7921775934299246e-06,
+      "loss": 0.2635,
+      "step": 11609
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.542464530669765,
+      "learning_rate": 1.791396418113799e-06,
+      "loss": 0.2789,
+      "step": 11610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4273074585088805,
+      "learning_rate": 1.7906153759283346e-06,
+      "loss": 0.2502,
+      "step": 11611
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5442717287376975,
+      "learning_rate": 1.789834466905937e-06,
+      "loss": 0.2487,
+      "step": 11612
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4770303384864432,
+      "learning_rate": 1.7890536910790095e-06,
+      "loss": 0.2492,
+      "step": 11613
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4491829471908595,
+      "learning_rate": 1.7882730484799494e-06,
+      "loss": 0.2712,
+      "step": 11614
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5239929949679778,
+      "learning_rate": 1.787492539141145e-06,
+      "loss": 0.27,
+      "step": 11615
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1348628007020314,
+      "learning_rate": 1.7867121630949814e-06,
+      "loss": 0.2663,
+      "step": 11616
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5677734227878761,
+      "learning_rate": 1.7859319203738357e-06,
+      "loss": 0.2536,
+      "step": 11617
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.219686016898887,
+      "learning_rate": 1.785151811010085e-06,
+      "loss": 0.2414,
+      "step": 11618
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4438572905554685,
+      "learning_rate": 1.7843718350360944e-06,
+      "loss": 0.2588,
+      "step": 11619
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.814113199465102,
+      "learning_rate": 1.7835919924842298e-06,
+      "loss": 0.26,
+      "step": 11620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7805856542310003,
+      "learning_rate": 1.7828122833868445e-06,
+      "loss": 0.2621,
+      "step": 11621
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1671388259741913,
+      "learning_rate": 1.7820327077762938e-06,
+      "loss": 0.2487,
+      "step": 11622
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.289129757929059,
+      "learning_rate": 1.7812532656849218e-06,
+      "loss": 0.2776,
+      "step": 11623
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8509117063410268,
+      "learning_rate": 1.7804739571450675e-06,
+      "loss": 0.2709,
+      "step": 11624
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2528244140635003,
+      "learning_rate": 1.7796947821890681e-06,
+      "loss": 0.268,
+      "step": 11625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.195180930040564,
+      "learning_rate": 1.7789157408492513e-06,
+      "loss": 0.2635,
+      "step": 11626
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.200946659581297,
+      "learning_rate": 1.778136833157943e-06,
+      "loss": 0.2733,
+      "step": 11627
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9669832001505725,
+      "learning_rate": 1.7773580591474599e-06,
+      "loss": 0.2679,
+      "step": 11628
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.146853320790772,
+      "learning_rate": 1.7765794188501133e-06,
+      "loss": 0.2667,
+      "step": 11629
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.671603421163086,
+      "learning_rate": 1.7758009122982144e-06,
+      "loss": 0.2478,
+      "step": 11630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4877291221639142,
+      "learning_rate": 1.7750225395240623e-06,
+      "loss": 0.2583,
+      "step": 11631
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.791835426391931,
+      "learning_rate": 1.7742443005599507e-06,
+      "loss": 0.2683,
+      "step": 11632
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.398489748969295,
+      "learning_rate": 1.7734661954381754e-06,
+      "loss": 0.2643,
+      "step": 11633
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9239909326068136,
+      "learning_rate": 1.7726882241910166e-06,
+      "loss": 0.2454,
+      "step": 11634
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.893966136002067,
+      "learning_rate": 1.771910386850757e-06,
+      "loss": 0.2524,
+      "step": 11635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3547275542984183,
+      "learning_rate": 1.7711326834496694e-06,
+      "loss": 0.2678,
+      "step": 11636
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4015095807488276,
+      "learning_rate": 1.7703551140200203e-06,
+      "loss": 0.2713,
+      "step": 11637
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.010584242813043,
+      "learning_rate": 1.769577678594076e-06,
+      "loss": 0.2388,
+      "step": 11638
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.702886554338077,
+      "learning_rate": 1.7688003772040912e-06,
+      "loss": 0.2969,
+      "step": 11639
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2686726796212513,
+      "learning_rate": 1.7680232098823164e-06,
+      "loss": 0.2788,
+      "step": 11640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6408821662912878,
+      "learning_rate": 1.7672461766609993e-06,
+      "loss": 0.2493,
+      "step": 11641
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.642616086181689,
+      "learning_rate": 1.7664692775723825e-06,
+      "loss": 0.2551,
+      "step": 11642
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6519694481770981,
+      "learning_rate": 1.7656925126486979e-06,
+      "loss": 0.2634,
+      "step": 11643
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.416757240748026,
+      "learning_rate": 1.7649158819221761e-06,
+      "loss": 0.2641,
+      "step": 11644
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.305871361033636,
+      "learning_rate": 1.7641393854250395e-06,
+      "loss": 0.2666,
+      "step": 11645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4664810327906352,
+      "learning_rate": 1.7633630231895083e-06,
+      "loss": 0.2602,
+      "step": 11646
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7818493241273152,
+      "learning_rate": 1.7625867952477943e-06,
+      "loss": 0.2669,
+      "step": 11647
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4525700460055968,
+      "learning_rate": 1.7618107016321023e-06,
+      "loss": 0.2501,
+      "step": 11648
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7809141559111528,
+      "learning_rate": 1.7610347423746365e-06,
+      "loss": 0.2684,
+      "step": 11649
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7893666663528591,
+      "learning_rate": 1.760258917507594e-06,
+      "loss": 0.25,
+      "step": 11650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5303411523326833,
+      "learning_rate": 1.7594832270631633e-06,
+      "loss": 0.261,
+      "step": 11651
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6455878897895337,
+      "learning_rate": 1.7587076710735296e-06,
+      "loss": 0.4513,
+      "step": 11652
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8996348615728103,
+      "learning_rate": 1.75793224957087e-06,
+      "loss": 0.2832,
+      "step": 11653
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2014973717587885,
+      "learning_rate": 1.7571569625873613e-06,
+      "loss": 0.2597,
+      "step": 11654
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.001370566681674,
+      "learning_rate": 1.7563818101551704e-06,
+      "loss": 0.2902,
+      "step": 11655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6365757940514447,
+      "learning_rate": 1.7556067923064578e-06,
+      "loss": 0.2573,
+      "step": 11656
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9545635215836095,
+      "learning_rate": 1.7548319090733822e-06,
+      "loss": 0.2548,
+      "step": 11657
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3930601162195142,
+      "learning_rate": 1.7540571604880963e-06,
+      "loss": 0.2507,
+      "step": 11658
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.077357826721203,
+      "learning_rate": 1.753282546582744e-06,
+      "loss": 0.2753,
+      "step": 11659
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5105846893542216,
+      "learning_rate": 1.7525080673894663e-06,
+      "loss": 0.262,
+      "step": 11660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0441988034041034,
+      "learning_rate": 1.7517337229403946e-06,
+      "loss": 0.2709,
+      "step": 11661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.696674881499434,
+      "learning_rate": 1.7509595132676627e-06,
+      "loss": 0.2694,
+      "step": 11662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.05382432379998,
+      "learning_rate": 1.750185438403391e-06,
+      "loss": 0.2597,
+      "step": 11663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4667357356081046,
+      "learning_rate": 1.7494114983796966e-06,
+      "loss": 0.2646,
+      "step": 11664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0920409162784868,
+      "learning_rate": 1.7486376932286925e-06,
+      "loss": 0.2512,
+      "step": 11665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.345179870515097,
+      "learning_rate": 1.7478640229824872e-06,
+      "loss": 0.2571,
+      "step": 11666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7376213684105677,
+      "learning_rate": 1.7470904876731804e-06,
+      "loss": 0.2741,
+      "step": 11667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.771613632062838,
+      "learning_rate": 1.7463170873328667e-06,
+      "loss": 0.2697,
+      "step": 11668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.55801121445325,
+      "learning_rate": 1.745543821993635e-06,
+      "loss": 0.2608,
+      "step": 11669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5823996231364373,
+      "learning_rate": 1.7447706916875717e-06,
+      "loss": 0.4698,
+      "step": 11670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.647196429200539,
+      "learning_rate": 1.7439976964467532e-06,
+      "loss": 0.2504,
+      "step": 11671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6926303658328,
+      "learning_rate": 1.743224836303255e-06,
+      "loss": 0.2425,
+      "step": 11672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6919745608462098,
+      "learning_rate": 1.7424521112891418e-06,
+      "loss": 0.2689,
+      "step": 11673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9286600778162801,
+      "learning_rate": 1.741679521436478e-06,
+      "loss": 0.277,
+      "step": 11674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 34.17848592255194,
+      "learning_rate": 1.740907066777318e-06,
+      "loss": 0.2728,
+      "step": 11675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7389085754189018,
+      "learning_rate": 1.7401347473437124e-06,
+      "loss": 0.2624,
+      "step": 11676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5115580443115493,
+      "learning_rate": 1.7393625631677052e-06,
+      "loss": 0.2638,
+      "step": 11677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.299529952387337,
+      "learning_rate": 1.7385905142813387e-06,
+      "loss": 0.2683,
+      "step": 11678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.678281920994348,
+      "learning_rate": 1.7378186007166426e-06,
+      "loss": 0.2594,
+      "step": 11679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6147927500913823,
+      "learning_rate": 1.7370468225056487e-06,
+      "loss": 0.2489,
+      "step": 11680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3803832366180733,
+      "learning_rate": 1.7362751796803757e-06,
+      "loss": 0.2699,
+      "step": 11681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2956701821478753,
+      "learning_rate": 1.7355036722728447e-06,
+      "loss": 0.2695,
+      "step": 11682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0852043888052307,
+      "learning_rate": 1.7347323003150647e-06,
+      "loss": 0.2632,
+      "step": 11683
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5722202040312103,
+      "learning_rate": 1.7339610638390413e-06,
+      "loss": 0.4754,
+      "step": 11684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4999099087433647,
+      "learning_rate": 1.733189962876773e-06,
+      "loss": 0.2473,
+      "step": 11685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3405561995654947,
+      "learning_rate": 1.7324189974602552e-06,
+      "loss": 0.2453,
+      "step": 11686
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7920033792237762,
+      "learning_rate": 1.731648167621479e-06,
+      "loss": 0.2598,
+      "step": 11687
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9209915036232665,
+      "learning_rate": 1.7308774733924254e-06,
+      "loss": 0.2486,
+      "step": 11688
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2248039585299622,
+      "learning_rate": 1.7301069148050708e-06,
+      "loss": 0.2375,
+      "step": 11689
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2562677500023627,
+      "learning_rate": 1.7293364918913896e-06,
+      "loss": 0.2654,
+      "step": 11690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5772938891061674,
+      "learning_rate": 1.7285662046833469e-06,
+      "loss": 0.2592,
+      "step": 11691
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.263217946340115,
+      "learning_rate": 1.727796053212903e-06,
+      "loss": 0.257,
+      "step": 11692
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4642998078322564,
+      "learning_rate": 1.7270260375120114e-06,
+      "loss": 0.2497,
+      "step": 11693
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.922779491169293,
+      "learning_rate": 1.7262561576126225e-06,
+      "loss": 0.255,
+      "step": 11694
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7178176465998325,
+      "learning_rate": 1.7254864135466825e-06,
+      "loss": 0.2653,
+      "step": 11695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.791889022442444,
+      "learning_rate": 1.7247168053461272e-06,
+      "loss": 0.269,
+      "step": 11696
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0789070804220042,
+      "learning_rate": 1.7239473330428874e-06,
+      "loss": 0.2611,
+      "step": 11697
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.733873805643619,
+      "learning_rate": 1.7231779966688938e-06,
+      "loss": 0.2638,
+      "step": 11698
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8731252908626708,
+      "learning_rate": 1.7224087962560654e-06,
+      "loss": 0.2865,
+      "step": 11699
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5683024424539835,
+      "learning_rate": 1.7216397318363175e-06,
+      "loss": 0.4845,
+      "step": 11700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3266079086383,
+      "learning_rate": 1.7208708034415566e-06,
+      "loss": 0.2603,
+      "step": 11701
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4772091036325459,
+      "learning_rate": 1.7201020111036942e-06,
+      "loss": 0.249,
+      "step": 11702
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3318413488655816,
+      "learning_rate": 1.7193333548546248e-06,
+      "loss": 0.2519,
+      "step": 11703
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.879203382531289,
+      "learning_rate": 1.7185648347262419e-06,
+      "loss": 0.2798,
+      "step": 11704
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.337589940419514,
+      "learning_rate": 1.7177964507504309e-06,
+      "loss": 0.2612,
+      "step": 11705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8062267155195184,
+      "learning_rate": 1.7170282029590762e-06,
+      "loss": 0.2572,
+      "step": 11706
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.588670748118247,
+      "learning_rate": 1.7162600913840527e-06,
+      "loss": 0.2658,
+      "step": 11707
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0250954480159216,
+      "learning_rate": 1.715492116057229e-06,
+      "loss": 0.27,
+      "step": 11708
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.308641864443634,
+      "learning_rate": 1.7147242770104716e-06,
+      "loss": 0.2613,
+      "step": 11709
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.746783481461446,
+      "learning_rate": 1.7139565742756409e-06,
+      "loss": 0.2492,
+      "step": 11710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0127040872933692,
+      "learning_rate": 1.713189007884588e-06,
+      "loss": 0.2541,
+      "step": 11711
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.461726681942587,
+      "learning_rate": 1.7124215778691616e-06,
+      "loss": 0.25,
+      "step": 11712
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7282673392088093,
+      "learning_rate": 1.7116542842612017e-06,
+      "loss": 0.2601,
+      "step": 11713
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.184984498330675,
+      "learning_rate": 1.710887127092548e-06,
+      "loss": 0.2656,
+      "step": 11714
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3939637027447924,
+      "learning_rate": 1.7101201063950295e-06,
+      "loss": 0.2477,
+      "step": 11715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0316670554870813,
+      "learning_rate": 1.7093532222004694e-06,
+      "loss": 0.2789,
+      "step": 11716
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2368565527433213,
+      "learning_rate": 1.7085864745406894e-06,
+      "loss": 0.2769,
+      "step": 11717
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6876175509679776,
+      "learning_rate": 1.707819863447504e-06,
+      "loss": 0.2523,
+      "step": 11718
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5491355963980498,
+      "learning_rate": 1.7070533889527202e-06,
+      "loss": 0.4875,
+      "step": 11719
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0802363436224196,
+      "learning_rate": 1.7062870510881401e-06,
+      "loss": 0.277,
+      "step": 11720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3939600406632313,
+      "learning_rate": 1.7055208498855591e-06,
+      "loss": 0.2505,
+      "step": 11721
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6561300208490082,
+      "learning_rate": 1.7047547853767714e-06,
+      "loss": 0.2656,
+      "step": 11722
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.085061399844055,
+      "learning_rate": 1.7039888575935588e-06,
+      "loss": 0.2596,
+      "step": 11723
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.35310945864418,
+      "learning_rate": 1.7032230665677046e-06,
+      "loss": 0.2597,
+      "step": 11724
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4761506138469251,
+      "learning_rate": 1.7024574123309795e-06,
+      "loss": 0.2599,
+      "step": 11725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4213739859126764,
+      "learning_rate": 1.7016918949151551e-06,
+      "loss": 0.2459,
+      "step": 11726
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5246435825088704,
+      "learning_rate": 1.7009265143519927e-06,
+      "loss": 0.2595,
+      "step": 11727
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3481570591323324,
+      "learning_rate": 1.7001612706732484e-06,
+      "loss": 0.2469,
+      "step": 11728
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.235773637995172,
+      "learning_rate": 1.6993961639106726e-06,
+      "loss": 0.2686,
+      "step": 11729
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.271644764279985,
+      "learning_rate": 1.6986311940960148e-06,
+      "loss": 0.2753,
+      "step": 11730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.937126375319672,
+      "learning_rate": 1.6978663612610102e-06,
+      "loss": 0.2731,
+      "step": 11731
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8566076457703664,
+      "learning_rate": 1.6971016654373979e-06,
+      "loss": 0.25,
+      "step": 11732
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.036749952673217,
+      "learning_rate": 1.6963371066569017e-06,
+      "loss": 0.2727,
+      "step": 11733
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7444668121962237,
+      "learning_rate": 1.6955726849512495e-06,
+      "loss": 0.2608,
+      "step": 11734
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2119610463202544,
+      "learning_rate": 1.6948084003521553e-06,
+      "loss": 0.2685,
+      "step": 11735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5387795986646753,
+      "learning_rate": 1.6940442528913309e-06,
+      "loss": 0.2516,
+      "step": 11736
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.508764890414543,
+      "learning_rate": 1.6932802426004812e-06,
+      "loss": 0.2688,
+      "step": 11737
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.67967061402469,
+      "learning_rate": 1.692516369511308e-06,
+      "loss": 0.2596,
+      "step": 11738
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6086632782889763,
+      "learning_rate": 1.6917526336555068e-06,
+      "loss": 0.5037,
+      "step": 11739
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.296951764295234,
+      "learning_rate": 1.6909890350647645e-06,
+      "loss": 0.2721,
+      "step": 11740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.745862100151195,
+      "learning_rate": 1.6902255737707634e-06,
+      "loss": 0.2608,
+      "step": 11741
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.730642746716277,
+      "learning_rate": 1.6894622498051838e-06,
+      "loss": 0.2485,
+      "step": 11742
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.094308521086035,
+      "learning_rate": 1.6886990631996951e-06,
+      "loss": 0.2685,
+      "step": 11743
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.691004634565247,
+      "learning_rate": 1.6879360139859646e-06,
+      "loss": 0.2427,
+      "step": 11744
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.60926938700073,
+      "learning_rate": 1.6871731021956495e-06,
+      "loss": 0.4766,
+      "step": 11745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9327936579866638,
+      "learning_rate": 1.6864103278604071e-06,
+      "loss": 0.2583,
+      "step": 11746
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.493622110827338,
+      "learning_rate": 1.685647691011887e-06,
+      "loss": 0.2675,
+      "step": 11747
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4140127398872653,
+      "learning_rate": 1.6848851916817316e-06,
+      "loss": 0.2686,
+      "step": 11748
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9979295985962144,
+      "learning_rate": 1.6841228299015755e-06,
+      "loss": 0.2646,
+      "step": 11749
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.257469315564016,
+      "learning_rate": 1.683360605703055e-06,
+      "loss": 0.2735,
+      "step": 11750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7783167662283246,
+      "learning_rate": 1.6825985191177941e-06,
+      "loss": 0.2637,
+      "step": 11751
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.997435546723059,
+      "learning_rate": 1.681836570177413e-06,
+      "loss": 0.2706,
+      "step": 11752
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.597448862879755,
+      "learning_rate": 1.681074758913523e-06,
+      "loss": 0.5039,
+      "step": 11753
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9406822339699066,
+      "learning_rate": 1.6803130853577392e-06,
+      "loss": 0.2589,
+      "step": 11754
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.901023273563546,
+      "learning_rate": 1.679551549541662e-06,
+      "loss": 0.2591,
+      "step": 11755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.578241439993665,
+      "learning_rate": 1.678790151496889e-06,
+      "loss": 0.473,
+      "step": 11756
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.754896853850894,
+      "learning_rate": 1.6780288912550102e-06,
+      "loss": 0.2598,
+      "step": 11757
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9530054450688783,
+      "learning_rate": 1.6772677688476146e-06,
+      "loss": 0.2555,
+      "step": 11758
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4590854111521858,
+      "learning_rate": 1.6765067843062816e-06,
+      "loss": 0.2727,
+      "step": 11759
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8566288786439034,
+      "learning_rate": 1.6757459376625852e-06,
+      "loss": 0.2562,
+      "step": 11760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.599038445668937,
+      "learning_rate": 1.6749852289480906e-06,
+      "loss": 0.2493,
+      "step": 11761
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1675105011173987,
+      "learning_rate": 1.6742246581943683e-06,
+      "loss": 0.2554,
+      "step": 11762
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8702437625465167,
+      "learning_rate": 1.673464225432972e-06,
+      "loss": 0.2645,
+      "step": 11763
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.628692368302644,
+      "learning_rate": 1.672703930695454e-06,
+      "loss": 0.2567,
+      "step": 11764
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5839239474989595,
+      "learning_rate": 1.6719437740133577e-06,
+      "loss": 0.4346,
+      "step": 11765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.07993418322798,
+      "learning_rate": 1.6711837554182276e-06,
+      "loss": 0.2704,
+      "step": 11766
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.081586985271671,
+      "learning_rate": 1.6704238749415958e-06,
+      "loss": 0.2715,
+      "step": 11767
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.54605044868619,
+      "learning_rate": 1.6696641326149897e-06,
+      "loss": 0.2447,
+      "step": 11768
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.49403293486929,
+      "learning_rate": 1.6689045284699347e-06,
+      "loss": 0.2713,
+      "step": 11769
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7950145411044751,
+      "learning_rate": 1.6681450625379487e-06,
+      "loss": 0.2581,
+      "step": 11770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.933289996831663,
+      "learning_rate": 1.6673857348505422e-06,
+      "loss": 0.2577,
+      "step": 11771
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0326330429046733,
+      "learning_rate": 1.6666265454392206e-06,
+      "loss": 0.2669,
+      "step": 11772
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9024021251154875,
+      "learning_rate": 1.665867494335483e-06,
+      "loss": 0.2706,
+      "step": 11773
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.053164362489364,
+      "learning_rate": 1.6651085815708268e-06,
+      "loss": 0.2647,
+      "step": 11774
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.514542001307621,
+      "learning_rate": 1.664349807176739e-06,
+      "loss": 0.2515,
+      "step": 11775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7235526814462925,
+      "learning_rate": 1.6635911711847007e-06,
+      "loss": 0.255,
+      "step": 11776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.166238222918449,
+      "learning_rate": 1.6628326736261914e-06,
+      "loss": 0.2648,
+      "step": 11777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6559977904009444,
+      "learning_rate": 1.6620743145326829e-06,
+      "loss": 0.2807,
+      "step": 11778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0426253329229977,
+      "learning_rate": 1.6613160939356403e-06,
+      "loss": 0.255,
+      "step": 11779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0517655887107757,
+      "learning_rate": 1.6605580118665233e-06,
+      "loss": 0.2503,
+      "step": 11780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1754399538743616,
+      "learning_rate": 1.6598000683567838e-06,
+      "loss": 0.2556,
+      "step": 11781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.901653787557663,
+      "learning_rate": 1.659042263437874e-06,
+      "loss": 0.2483,
+      "step": 11782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.345049395981367,
+      "learning_rate": 1.6582845971412332e-06,
+      "loss": 0.2777,
+      "step": 11783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.06937098899403,
+      "learning_rate": 1.657527069498302e-06,
+      "loss": 0.2586,
+      "step": 11784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5680557772512367,
+      "learning_rate": 1.6567696805405075e-06,
+      "loss": 0.2572,
+      "step": 11785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3874081282609536,
+      "learning_rate": 1.656012430299279e-06,
+      "loss": 0.2771,
+      "step": 11786
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5989031663760926,
+      "learning_rate": 1.655255318806035e-06,
+      "loss": 0.4963,
+      "step": 11787
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.529049548014022,
+      "learning_rate": 1.6544983460921882e-06,
+      "loss": 0.2653,
+      "step": 11788
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.802033404813695,
+      "learning_rate": 1.6537415121891454e-06,
+      "loss": 0.2643,
+      "step": 11789
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6588807491390738,
+      "learning_rate": 1.6529848171283113e-06,
+      "loss": 0.2607,
+      "step": 11790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8993431460926697,
+      "learning_rate": 1.652228260941084e-06,
+      "loss": 0.249,
+      "step": 11791
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 14.431290294460897,
+      "learning_rate": 1.6514718436588517e-06,
+      "loss": 0.2689,
+      "step": 11792
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.036317463653162,
+      "learning_rate": 1.6507155653129992e-06,
+      "loss": 0.2694,
+      "step": 11793
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8506399393182689,
+      "learning_rate": 1.6499594259349083e-06,
+      "loss": 0.2653,
+      "step": 11794
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4949799470416039,
+      "learning_rate": 1.6492034255559514e-06,
+      "loss": 0.2744,
+      "step": 11795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.313066645298156,
+      "learning_rate": 1.6484475642074955e-06,
+      "loss": 0.264,
+      "step": 11796
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4319119523253767,
+      "learning_rate": 1.6476918419209021e-06,
+      "loss": 0.2578,
+      "step": 11797
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.21520285802145,
+      "learning_rate": 1.6469362587275283e-06,
+      "loss": 0.2688,
+      "step": 11798
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9102041030243637,
+      "learning_rate": 1.6461808146587266e-06,
+      "loss": 0.2512,
+      "step": 11799
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.13520231046298,
+      "learning_rate": 1.6454255097458399e-06,
+      "loss": 0.2572,
+      "step": 11800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6482321469633137,
+      "learning_rate": 1.6446703440202054e-06,
+      "loss": 0.4889,
+      "step": 11801
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4670376716034883,
+      "learning_rate": 1.6439153175131601e-06,
+      "loss": 0.2662,
+      "step": 11802
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.500349371351017,
+      "learning_rate": 1.6431604302560289e-06,
+      "loss": 0.2621,
+      "step": 11803
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.568830629198864,
+      "learning_rate": 1.6424056822801343e-06,
+      "loss": 0.27,
+      "step": 11804
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.689698718055636,
+      "learning_rate": 1.6416510736167894e-06,
+      "loss": 0.2718,
+      "step": 11805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.305639380884586,
+      "learning_rate": 1.6408966042973073e-06,
+      "loss": 0.2599,
+      "step": 11806
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9843904255661617,
+      "learning_rate": 1.6401422743529927e-06,
+      "loss": 0.2596,
+      "step": 11807
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5982928678592767,
+      "learning_rate": 1.6393880838151432e-06,
+      "loss": 0.2589,
+      "step": 11808
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4961203988435854,
+      "learning_rate": 1.6386340327150496e-06,
+      "loss": 0.2616,
+      "step": 11809
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2307769427273816,
+      "learning_rate": 1.6378801210840018e-06,
+      "loss": 0.2721,
+      "step": 11810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9924991045431384,
+      "learning_rate": 1.6371263489532795e-06,
+      "loss": 0.2602,
+      "step": 11811
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.355843185123422,
+      "learning_rate": 1.6363727163541587e-06,
+      "loss": 0.2713,
+      "step": 11812
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6402361533211867,
+      "learning_rate": 1.6356192233179063e-06,
+      "loss": 0.2499,
+      "step": 11813
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.775833491056217,
+      "learning_rate": 1.634865869875788e-06,
+      "loss": 0.2841,
+      "step": 11814
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4126877447225055,
+      "learning_rate": 1.6341126560590637e-06,
+      "loss": 0.2692,
+      "step": 11815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.721410768568857,
+      "learning_rate": 1.633359581898984e-06,
+      "loss": 0.2507,
+      "step": 11816
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.588412549744532,
+      "learning_rate": 1.6326066474267931e-06,
+      "loss": 0.4395,
+      "step": 11817
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3592591381615065,
+      "learning_rate": 1.6318538526737354e-06,
+      "loss": 0.2556,
+      "step": 11818
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.297946060006865,
+      "learning_rate": 1.631101197671044e-06,
+      "loss": 0.2519,
+      "step": 11819
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6660246422459932,
+      "learning_rate": 1.6303486824499459e-06,
+      "loss": 0.2548,
+      "step": 11820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7470054571205338,
+      "learning_rate": 1.6295963070416658e-06,
+      "loss": 0.2657,
+      "step": 11821
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6014955834137027,
+      "learning_rate": 1.628844071477424e-06,
+      "loss": 0.4609,
+      "step": 11822
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.769016713587465,
+      "learning_rate": 1.628091975788429e-06,
+      "loss": 0.2657,
+      "step": 11823
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8453180889738028,
+      "learning_rate": 1.627340020005887e-06,
+      "loss": 0.291,
+      "step": 11824
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4959257455667023,
+      "learning_rate": 1.6265882041609964e-06,
+      "loss": 0.2688,
+      "step": 11825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7455441204424988,
+      "learning_rate": 1.6258365282849547e-06,
+      "loss": 0.2446,
+      "step": 11826
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5213502263259033,
+      "learning_rate": 1.6250849924089485e-06,
+      "loss": 0.2907,
+      "step": 11827
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9003721308085535,
+      "learning_rate": 1.6243335965641593e-06,
+      "loss": 0.2483,
+      "step": 11828
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6456645060294284,
+      "learning_rate": 1.6235823407817647e-06,
+      "loss": 0.246,
+      "step": 11829
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3052556292761452,
+      "learning_rate": 1.622831225092938e-06,
+      "loss": 0.2571,
+      "step": 11830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7321240346034084,
+      "learning_rate": 1.6220802495288418e-06,
+      "loss": 0.2507,
+      "step": 11831
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1755243859919093,
+      "learning_rate": 1.6213294141206366e-06,
+      "loss": 0.2687,
+      "step": 11832
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.039563658824441,
+      "learning_rate": 1.6205787188994726e-06,
+      "loss": 0.2647,
+      "step": 11833
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.4653985842077635,
+      "learning_rate": 1.6198281638965023e-06,
+      "loss": 0.2693,
+      "step": 11834
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6229795598659216,
+      "learning_rate": 1.619077749142864e-06,
+      "loss": 0.2536,
+      "step": 11835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 14.362800910237805,
+      "learning_rate": 1.618327474669696e-06,
+      "loss": 0.2458,
+      "step": 11836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.165323961044263,
+      "learning_rate": 1.6175773405081263e-06,
+      "loss": 0.2558,
+      "step": 11837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0797768971008512,
+      "learning_rate": 1.6168273466892826e-06,
+      "loss": 0.2724,
+      "step": 11838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.51796264846945,
+      "learning_rate": 1.6160774932442808e-06,
+      "loss": 0.2655,
+      "step": 11839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.81021565265868,
+      "learning_rate": 1.6153277802042349e-06,
+      "loss": 0.2753,
+      "step": 11840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.32776950753152,
+      "learning_rate": 1.6145782076002493e-06,
+      "loss": 0.2506,
+      "step": 11841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8041643534287357,
+      "learning_rate": 1.6138287754634285e-06,
+      "loss": 0.2615,
+      "step": 11842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9553419239251053,
+      "learning_rate": 1.6130794838248653e-06,
+      "loss": 0.2461,
+      "step": 11843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8993415032975691,
+      "learning_rate": 1.6123303327156515e-06,
+      "loss": 0.241,
+      "step": 11844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.476463975922597,
+      "learning_rate": 1.6115813221668676e-06,
+      "loss": 0.252,
+      "step": 11845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.791918695056272,
+      "learning_rate": 1.6108324522095948e-06,
+      "loss": 0.269,
+      "step": 11846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.34204523919045,
+      "learning_rate": 1.6100837228749034e-06,
+      "loss": 0.2535,
+      "step": 11847
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9681715046276151,
+      "learning_rate": 1.6093351341938595e-06,
+      "loss": 0.2697,
+      "step": 11848
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.92936775419886,
+      "learning_rate": 1.6085866861975219e-06,
+      "loss": 0.2582,
+      "step": 11849
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7529375084635115,
+      "learning_rate": 1.6078383789169466e-06,
+      "loss": 0.2626,
+      "step": 11850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7072302835746838,
+      "learning_rate": 1.6070902123831832e-06,
+      "loss": 0.2662,
+      "step": 11851
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.57082915701216,
+      "learning_rate": 1.6063421866272738e-06,
+      "loss": 0.2618,
+      "step": 11852
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8261968009035916,
+      "learning_rate": 1.6055943016802533e-06,
+      "loss": 0.2644,
+      "step": 11853
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4569429665757117,
+      "learning_rate": 1.604846557573156e-06,
+      "loss": 0.2488,
+      "step": 11854
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.946417091387337,
+      "learning_rate": 1.6040989543370057e-06,
+      "loss": 0.2365,
+      "step": 11855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.243356464107158,
+      "learning_rate": 1.603351492002821e-06,
+      "loss": 0.249,
+      "step": 11856
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.093287370996489,
+      "learning_rate": 1.602604170601615e-06,
+      "loss": 0.2695,
+      "step": 11857
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5898220441347646,
+      "learning_rate": 1.601856990164396e-06,
+      "loss": 0.2597,
+      "step": 11858
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.709914051757555,
+      "learning_rate": 1.6011099507221677e-06,
+      "loss": 0.2553,
+      "step": 11859
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9656589445698858,
+      "learning_rate": 1.600363052305925e-06,
+      "loss": 0.2483,
+      "step": 11860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1406403928513473,
+      "learning_rate": 1.599616294946656e-06,
+      "loss": 0.2502,
+      "step": 11861
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5859280108116744,
+      "learning_rate": 1.5988696786753489e-06,
+      "loss": 0.2568,
+      "step": 11862
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.451305252896568,
+      "learning_rate": 1.5981232035229789e-06,
+      "loss": 0.2579,
+      "step": 11863
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8968391425248419,
+      "learning_rate": 1.59737686952052e-06,
+      "loss": 0.2645,
+      "step": 11864
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7087956113154734,
+      "learning_rate": 1.596630676698937e-06,
+      "loss": 0.2586,
+      "step": 11865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6320847648591372,
+      "learning_rate": 1.5958846250891924e-06,
+      "loss": 0.2636,
+      "step": 11866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3935022032319897,
+      "learning_rate": 1.5951387147222426e-06,
+      "loss": 0.2676,
+      "step": 11867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.972545154743655,
+      "learning_rate": 1.5943929456290353e-06,
+      "loss": 0.2584,
+      "step": 11868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3620123371947725,
+      "learning_rate": 1.5936473178405115e-06,
+      "loss": 0.2434,
+      "step": 11869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7552358800792407,
+      "learning_rate": 1.5929018313876133e-06,
+      "loss": 0.2584,
+      "step": 11870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.484971098012521,
+      "learning_rate": 1.5921564863012696e-06,
+      "loss": 0.2621,
+      "step": 11871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3294578172874192,
+      "learning_rate": 1.5914112826124046e-06,
+      "loss": 0.2528,
+      "step": 11872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7504983993422945,
+      "learning_rate": 1.5906662203519413e-06,
+      "loss": 0.2433,
+      "step": 11873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3484787834414336,
+      "learning_rate": 1.5899212995507907e-06,
+      "loss": 0.272,
+      "step": 11874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7115203724763024,
+      "learning_rate": 1.5891765202398636e-06,
+      "loss": 0.2535,
+      "step": 11875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4853380230261126,
+      "learning_rate": 1.5884318824500616e-06,
+      "loss": 0.2553,
+      "step": 11876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5846041938247355,
+      "learning_rate": 1.5876873862122781e-06,
+      "loss": 0.2719,
+      "step": 11877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4196277911506971,
+      "learning_rate": 1.586943031557408e-06,
+      "loss": 0.2557,
+      "step": 11878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9084763581467403,
+      "learning_rate": 1.586198818516334e-06,
+      "loss": 0.2528,
+      "step": 11879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.710566703156637,
+      "learning_rate": 1.5854547471199328e-06,
+      "loss": 0.2591,
+      "step": 11880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1133888050406626,
+      "learning_rate": 1.5847108173990804e-06,
+      "loss": 0.2901,
+      "step": 11881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8562106526254234,
+      "learning_rate": 1.583967029384641e-06,
+      "loss": 0.2629,
+      "step": 11882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6122073234864452,
+      "learning_rate": 1.5832233831074783e-06,
+      "loss": 0.4742,
+      "step": 11883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6017867089945995,
+      "learning_rate": 1.582479878598447e-06,
+      "loss": 0.4793,
+      "step": 11884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6468932831399952,
+      "learning_rate": 1.581736515888394e-06,
+      "loss": 0.2553,
+      "step": 11885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9967885114687465,
+      "learning_rate": 1.5809932950081657e-06,
+      "loss": 0.2653,
+      "step": 11886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.266315630618689,
+      "learning_rate": 1.5802502159885968e-06,
+      "loss": 0.2681,
+      "step": 11887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.986349448661057,
+      "learning_rate": 1.5795072788605226e-06,
+      "loss": 0.2631,
+      "step": 11888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7236149444049063,
+      "learning_rate": 1.5787644836547667e-06,
+      "loss": 0.2822,
+      "step": 11889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.077695773204747,
+      "learning_rate": 1.5780218304021477e-06,
+      "loss": 0.263,
+      "step": 11890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5311655300179265,
+      "learning_rate": 1.5772793191334828e-06,
+      "loss": 0.255,
+      "step": 11891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4024394966952487,
+      "learning_rate": 1.576536949879578e-06,
+      "loss": 0.2611,
+      "step": 11892
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.950961361803954,
+      "learning_rate": 1.5757947226712344e-06,
+      "loss": 0.2727,
+      "step": 11893
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 18.33122662785955,
+      "learning_rate": 1.5750526375392517e-06,
+      "loss": 0.281,
+      "step": 11894
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.494641768361291,
+      "learning_rate": 1.5743106945144166e-06,
+      "loss": 0.2587,
+      "step": 11895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.407395698512057,
+      "learning_rate": 1.5735688936275174e-06,
+      "loss": 0.2778,
+      "step": 11896
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5668321288737272,
+      "learning_rate": 1.5728272349093287e-06,
+      "loss": 0.2511,
+      "step": 11897
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.810810671072288,
+      "learning_rate": 1.5720857183906275e-06,
+      "loss": 0.2566,
+      "step": 11898
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.810282556503871,
+      "learning_rate": 1.571344344102178e-06,
+      "loss": 0.2534,
+      "step": 11899
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1878011512487316,
+      "learning_rate": 1.5706031120747417e-06,
+      "loss": 0.2653,
+      "step": 11900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.0712484006038085,
+      "learning_rate": 1.569862022339072e-06,
+      "loss": 0.2483,
+      "step": 11901
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6527144597464218,
+      "learning_rate": 1.56912107492592e-06,
+      "loss": 0.273,
+      "step": 11902
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 19.838911551780072,
+      "learning_rate": 1.56838026986603e-06,
+      "loss": 0.2706,
+      "step": 11903
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.522449482689555,
+      "learning_rate": 1.567639607190138e-06,
+      "loss": 0.238,
+      "step": 11904
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.014349755394319,
+      "learning_rate": 1.566899086928973e-06,
+      "loss": 0.247,
+      "step": 11905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2942124436971643,
+      "learning_rate": 1.5661587091132646e-06,
+      "loss": 0.2423,
+      "step": 11906
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6652853533826246,
+      "learning_rate": 1.565418473773731e-06,
+      "loss": 0.2477,
+      "step": 11907
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.688960854504309,
+      "learning_rate": 1.564678380941085e-06,
+      "loss": 0.2634,
+      "step": 11908
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8093549149604553,
+      "learning_rate": 1.5639384306460331e-06,
+      "loss": 0.254,
+      "step": 11909
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.751705599615996,
+      "learning_rate": 1.5631986229192791e-06,
+      "loss": 0.253,
+      "step": 11910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4181956495410795,
+      "learning_rate": 1.5624589577915205e-06,
+      "loss": 0.2694,
+      "step": 11911
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.585333299902547,
+      "learning_rate": 1.5617194352934451e-06,
+      "loss": 0.2748,
+      "step": 11912
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3754365359386904,
+      "learning_rate": 1.5609800554557358e-06,
+      "loss": 0.2587,
+      "step": 11913
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7493901033412564,
+      "learning_rate": 1.5602408183090744e-06,
+      "loss": 0.2659,
+      "step": 11914
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.479838596165756,
+      "learning_rate": 1.559501723884131e-06,
+      "loss": 0.2382,
+      "step": 11915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.45159273101338,
+      "learning_rate": 1.5587627722115717e-06,
+      "loss": 0.244,
+      "step": 11916
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7023586831820152,
+      "learning_rate": 1.5580239633220563e-06,
+      "loss": 0.2616,
+      "step": 11917
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8896552588110056,
+      "learning_rate": 1.55728529724624e-06,
+      "loss": 0.2639,
+      "step": 11918
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9517282312415587,
+      "learning_rate": 1.5565467740147732e-06,
+      "loss": 0.2377,
+      "step": 11919
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.248304202353867,
+      "learning_rate": 1.555808393658297e-06,
+      "loss": 0.2753,
+      "step": 11920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7899906901154115,
+      "learning_rate": 1.5550701562074465e-06,
+      "loss": 0.2597,
+      "step": 11921
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.769585045134498,
+      "learning_rate": 1.5543320616928558e-06,
+      "loss": 0.2622,
+      "step": 11922
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4658665214876905,
+      "learning_rate": 1.5535941101451478e-06,
+      "loss": 0.2566,
+      "step": 11923
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0309553236705917,
+      "learning_rate": 1.5528563015949421e-06,
+      "loss": 0.2774,
+      "step": 11924
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5888897895927225,
+      "learning_rate": 1.5521186360728497e-06,
+      "loss": 0.262,
+      "step": 11925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3547565026352504,
+      "learning_rate": 1.5513811136094786e-06,
+      "loss": 0.2545,
+      "step": 11926
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.266784297481978,
+      "learning_rate": 1.5506437342354324e-06,
+      "loss": 0.2558,
+      "step": 11927
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7020897530161574,
+      "learning_rate": 1.549906497981305e-06,
+      "loss": 0.2874,
+      "step": 11928
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.027743530206067,
+      "learning_rate": 1.549169404877683e-06,
+      "loss": 0.2871,
+      "step": 11929
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3002851894086886,
+      "learning_rate": 1.5484324549551532e-06,
+      "loss": 0.2446,
+      "step": 11930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.940854903658385,
+      "learning_rate": 1.5476956482442918e-06,
+      "loss": 0.2544,
+      "step": 11931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6021044647928193,
+      "learning_rate": 1.5469589847756677e-06,
+      "loss": 0.2539,
+      "step": 11932
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2905678702315992,
+      "learning_rate": 1.5462224645798507e-06,
+      "loss": 0.2675,
+      "step": 11933
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2728942236191934,
+      "learning_rate": 1.5454860876873968e-06,
+      "loss": 0.2612,
+      "step": 11934
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.296256772914394,
+      "learning_rate": 1.5447498541288625e-06,
+      "loss": 0.246,
+      "step": 11935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1979331063387524,
+      "learning_rate": 1.544013763934794e-06,
+      "loss": 0.2654,
+      "step": 11936
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.21921664885315,
+      "learning_rate": 1.543277817135731e-06,
+      "loss": 0.2635,
+      "step": 11937
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.553450963655312,
+      "learning_rate": 1.5425420137622133e-06,
+      "loss": 0.2692,
+      "step": 11938
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.189599600989339,
+      "learning_rate": 1.5418063538447687e-06,
+      "loss": 0.248,
+      "step": 11939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.383964627773125,
+      "learning_rate": 1.5410708374139189e-06,
+      "loss": 0.2725,
+      "step": 11940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4591173037880572,
+      "learning_rate": 1.5403354645001857e-06,
+      "loss": 0.2633,
+      "step": 11941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7464732086978818,
+      "learning_rate": 1.5396002351340772e-06,
+      "loss": 0.2695,
+      "step": 11942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1924013514986886,
+      "learning_rate": 1.5388651493461026e-06,
+      "loss": 0.2667,
+      "step": 11943
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8687920252622625,
+      "learning_rate": 1.5381302071667614e-06,
+      "loss": 0.2557,
+      "step": 11944
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.254162357203387,
+      "learning_rate": 1.5373954086265447e-06,
+      "loss": 0.2754,
+      "step": 11945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1932413521027496,
+      "learning_rate": 1.5366607537559448e-06,
+      "loss": 0.2463,
+      "step": 11946
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.168190153008269,
+      "learning_rate": 1.5359262425854398e-06,
+      "loss": 0.2597,
+      "step": 11947
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.760150342410197,
+      "learning_rate": 1.5351918751455097e-06,
+      "loss": 0.2689,
+      "step": 11948
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5582380661426294,
+      "learning_rate": 1.5344576514666231e-06,
+      "loss": 0.2531,
+      "step": 11949
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2917949160450024,
+      "learning_rate": 1.533723571579242e-06,
+      "loss": 0.2466,
+      "step": 11950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0579349956275412,
+      "learning_rate": 1.532989635513829e-06,
+      "loss": 0.2569,
+      "step": 11951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1387940862761026,
+      "learning_rate": 1.5322558433008332e-06,
+      "loss": 0.2787,
+      "step": 11952
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0218715524721236,
+      "learning_rate": 1.531522194970701e-06,
+      "loss": 0.2627,
+      "step": 11953
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.740251383808904,
+      "learning_rate": 1.5307886905538732e-06,
+      "loss": 0.2422,
+      "step": 11954
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.594114810697703,
+      "learning_rate": 1.530055330080787e-06,
+      "loss": 0.4549,
+      "step": 11955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8247060917469438,
+      "learning_rate": 1.5293221135818676e-06,
+      "loss": 0.2722,
+      "step": 11956
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.896839741503334,
+      "learning_rate": 1.5285890410875388e-06,
+      "loss": 0.2472,
+      "step": 11957
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6456790709071374,
+      "learning_rate": 1.527856112628215e-06,
+      "loss": 0.2497,
+      "step": 11958
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.44152185672861,
+      "learning_rate": 1.5271233282343095e-06,
+      "loss": 0.2591,
+      "step": 11959
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0892971632742032,
+      "learning_rate": 1.5263906879362256e-06,
+      "loss": 0.2578,
+      "step": 11960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.194558456264757,
+      "learning_rate": 1.525658191764361e-06,
+      "loss": 0.2721,
+      "step": 11961
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8897807708386882,
+      "learning_rate": 1.5249258397491086e-06,
+      "loss": 0.2713,
+      "step": 11962
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0664841821110844,
+      "learning_rate": 1.5241936319208572e-06,
+      "loss": 0.2652,
+      "step": 11963
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6448856221330757,
+      "learning_rate": 1.5234615683099857e-06,
+      "loss": 0.2601,
+      "step": 11964
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.678796882725787,
+      "learning_rate": 1.5227296489468685e-06,
+      "loss": 0.2756,
+      "step": 11965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.664808175604224,
+      "learning_rate": 1.5219978738618734e-06,
+      "loss": 0.2668,
+      "step": 11966
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.100819538826289,
+      "learning_rate": 1.5212662430853659e-06,
+      "loss": 0.2499,
+      "step": 11967
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.376328737191698,
+      "learning_rate": 1.5205347566477002e-06,
+      "loss": 0.2637,
+      "step": 11968
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6641584539080455,
+      "learning_rate": 1.5198034145792267e-06,
+      "loss": 0.4767,
+      "step": 11969
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7424025291555487,
+      "learning_rate": 1.5190722169102906e-06,
+      "loss": 0.2749,
+      "step": 11970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6825061395834442,
+      "learning_rate": 1.518341163671233e-06,
+      "loss": 0.2453,
+      "step": 11971
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.900240666346348,
+      "learning_rate": 1.5176102548923844e-06,
+      "loss": 0.2641,
+      "step": 11972
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.948203795265517,
+      "learning_rate": 1.5168794906040719e-06,
+      "loss": 0.2842,
+      "step": 11973
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9603099547812366,
+      "learning_rate": 1.516148870836614e-06,
+      "loss": 0.2641,
+      "step": 11974
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1220814185419328,
+      "learning_rate": 1.5154183956203295e-06,
+      "loss": 0.2746,
+      "step": 11975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.771850191668032,
+      "learning_rate": 1.514688064985525e-06,
+      "loss": 0.2763,
+      "step": 11976
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.574735579899599,
+      "learning_rate": 1.513957878962502e-06,
+      "loss": 0.249,
+      "step": 11977
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5782235861767826,
+      "learning_rate": 1.5132278375815578e-06,
+      "loss": 0.2788,
+      "step": 11978
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5446073858393137,
+      "learning_rate": 1.512497940872986e-06,
+      "loss": 0.2519,
+      "step": 11979
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8054213731930635,
+      "learning_rate": 1.5117681888670694e-06,
+      "loss": 0.2471,
+      "step": 11980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4743759970123511,
+      "learning_rate": 1.5110385815940842e-06,
+      "loss": 0.2548,
+      "step": 11981
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.678778934250006,
+      "learning_rate": 1.5103091190843072e-06,
+      "loss": 0.2675,
+      "step": 11982
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3988365123464415,
+      "learning_rate": 1.509579801368003e-06,
+      "loss": 0.2404,
+      "step": 11983
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.505055998955773,
+      "learning_rate": 1.5088506284754312e-06,
+      "loss": 0.2514,
+      "step": 11984
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.609320783732682,
+      "learning_rate": 1.5081216004368492e-06,
+      "loss": 0.2704,
+      "step": 11985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7426797782797734,
+      "learning_rate": 1.5073927172825032e-06,
+      "loss": 0.2583,
+      "step": 11986
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4989185183015086,
+      "learning_rate": 1.506663979042638e-06,
+      "loss": 0.2446,
+      "step": 11987
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5553763225014214,
+      "learning_rate": 1.505935385747489e-06,
+      "loss": 0.2661,
+      "step": 11988
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9648034406003163,
+      "learning_rate": 1.505206937427286e-06,
+      "loss": 0.2724,
+      "step": 11989
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6727122932410934,
+      "learning_rate": 1.5044786341122557e-06,
+      "loss": 0.2687,
+      "step": 11990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.0541108125247995,
+      "learning_rate": 1.503750475832616e-06,
+      "loss": 0.2917,
+      "step": 11991
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.864789485460652,
+      "learning_rate": 1.5030224626185774e-06,
+      "loss": 0.2374,
+      "step": 11992
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.556706876194909,
+      "learning_rate": 1.5022945945003499e-06,
+      "loss": 0.2578,
+      "step": 11993
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.16915453716689,
+      "learning_rate": 1.50156687150813e-06,
+      "loss": 0.2727,
+      "step": 11994
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9587952436999463,
+      "learning_rate": 1.5008392936721166e-06,
+      "loss": 0.2529,
+      "step": 11995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7195241674442108,
+      "learning_rate": 1.5001118610224963e-06,
+      "loss": 0.2389,
+      "step": 11996
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7931812262722704,
+      "learning_rate": 1.4993845735894496e-06,
+      "loss": 0.2544,
+      "step": 11997
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2405025422970355,
+      "learning_rate": 1.4986574314031566e-06,
+      "loss": 0.2645,
+      "step": 11998
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.135625269821901,
+      "learning_rate": 1.497930434493784e-06,
+      "loss": 0.2536,
+      "step": 11999
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.046505744040854,
+      "learning_rate": 1.4972035828914993e-06,
+      "loss": 0.2862,
+      "step": 12000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.482562277878135,
+      "learning_rate": 1.4964768766264604e-06,
+      "loss": 0.2631,
+      "step": 12001
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.534614146089112,
+      "learning_rate": 1.4957503157288166e-06,
+      "loss": 0.2569,
+      "step": 12002
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8568030182913553,
+      "learning_rate": 1.4950239002287181e-06,
+      "loss": 0.2724,
+      "step": 12003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.555922049773706,
+      "learning_rate": 1.494297630156304e-06,
+      "loss": 0.2596,
+      "step": 12004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.725616201416361,
+      "learning_rate": 1.493571505541706e-06,
+      "loss": 0.2715,
+      "step": 12005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7002847467613185,
+      "learning_rate": 1.492845526415056e-06,
+      "loss": 0.2787,
+      "step": 12006
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.678468794217633,
+      "learning_rate": 1.492119692806473e-06,
+      "loss": 0.2497,
+      "step": 12007
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8387834655868671,
+      "learning_rate": 1.4913940047460762e-06,
+      "loss": 0.2665,
+      "step": 12008
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.716906784132446,
+      "learning_rate": 1.490668462263974e-06,
+      "loss": 0.2576,
+      "step": 12009
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.013935443864974,
+      "learning_rate": 1.4899430653902691e-06,
+      "loss": 0.2543,
+      "step": 12010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.024479615880311,
+      "learning_rate": 1.4892178141550628e-06,
+      "loss": 0.2492,
+      "step": 12011
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.08530901693931,
+      "learning_rate": 1.4884927085884448e-06,
+      "loss": 0.265,
+      "step": 12012
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.111331303976927,
+      "learning_rate": 1.4877677487205005e-06,
+      "loss": 0.2605,
+      "step": 12013
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5905043143237334,
+      "learning_rate": 1.4870429345813103e-06,
+      "loss": 0.2578,
+      "step": 12014
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6271409889503476,
+      "learning_rate": 1.4863182662009501e-06,
+      "loss": 0.4487,
+      "step": 12015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5999760615757777,
+      "learning_rate": 1.4855937436094858e-06,
+      "loss": 0.2424,
+      "step": 12016
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.238525805627581,
+      "learning_rate": 1.4848693668369795e-06,
+      "loss": 0.2494,
+      "step": 12017
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4332015326435306,
+      "learning_rate": 1.4841451359134857e-06,
+      "loss": 0.2514,
+      "step": 12018
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7978355334402403,
+      "learning_rate": 1.483421050869056e-06,
+      "loss": 0.257,
+      "step": 12019
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8016530563684148,
+      "learning_rate": 1.482697111733733e-06,
+      "loss": 0.2761,
+      "step": 12020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0149743446607356,
+      "learning_rate": 1.4819733185375535e-06,
+      "loss": 0.2535,
+      "step": 12021
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.885860580385384,
+      "learning_rate": 1.481249671310549e-06,
+      "loss": 0.2835,
+      "step": 12022
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4484556718172874,
+      "learning_rate": 1.4805261700827472e-06,
+      "loss": 0.2524,
+      "step": 12023
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5016746816025455,
+      "learning_rate": 1.4798028148841665e-06,
+      "loss": 0.2574,
+      "step": 12024
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.199610801733965,
+      "learning_rate": 1.4790796057448187e-06,
+      "loss": 0.2754,
+      "step": 12025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1407007334441586,
+      "learning_rate": 1.478356542694711e-06,
+      "loss": 0.2673,
+      "step": 12026
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7195045978267984,
+      "learning_rate": 1.4776336257638468e-06,
+      "loss": 0.26,
+      "step": 12027
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8504353959269848,
+      "learning_rate": 1.4769108549822198e-06,
+      "loss": 0.2772,
+      "step": 12028
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6572654675745015,
+      "learning_rate": 1.476188230379818e-06,
+      "loss": 0.2517,
+      "step": 12029
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.972166544547642,
+      "learning_rate": 1.475465751986625e-06,
+      "loss": 0.2586,
+      "step": 12030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5642743584507386,
+      "learning_rate": 1.4747434198326206e-06,
+      "loss": 0.252,
+      "step": 12031
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.171954189885917,
+      "learning_rate": 1.474021233947772e-06,
+      "loss": 0.2424,
+      "step": 12032
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.604920851024164,
+      "learning_rate": 1.4732991943620462e-06,
+      "loss": 0.2734,
+      "step": 12033
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.744629193997247,
+      "learning_rate": 1.4725773011053995e-06,
+      "loss": 0.2484,
+      "step": 12034
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3283157992310137,
+      "learning_rate": 1.4718555542077873e-06,
+      "loss": 0.279,
+      "step": 12035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6727049308694935,
+      "learning_rate": 1.471133953699153e-06,
+      "loss": 0.2636,
+      "step": 12036
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5159553532366852,
+      "learning_rate": 1.4704124996094405e-06,
+      "loss": 0.2327,
+      "step": 12037
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6535073658917429,
+      "learning_rate": 1.4696911919685813e-06,
+      "loss": 0.465,
+      "step": 12038
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.195212118883088,
+      "learning_rate": 1.4689700308065064e-06,
+      "loss": 0.2502,
+      "step": 12039
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.714089820462613,
+      "learning_rate": 1.4682490161531371e-06,
+      "loss": 0.2472,
+      "step": 12040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.225937081733685,
+      "learning_rate": 1.467528148038389e-06,
+      "loss": 0.2574,
+      "step": 12041
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3797596864482804,
+      "learning_rate": 1.4668074264921712e-06,
+      "loss": 0.2474,
+      "step": 12042
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1504471511916403,
+      "learning_rate": 1.4660868515443899e-06,
+      "loss": 0.2754,
+      "step": 12043
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9441453582088524,
+      "learning_rate": 1.4653664232249409e-06,
+      "loss": 0.2717,
+      "step": 12044
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.650606618098762,
+      "learning_rate": 1.4646461415637188e-06,
+      "loss": 0.2508,
+      "step": 12045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.818647293254394,
+      "learning_rate": 1.4639260065906058e-06,
+      "loss": 0.2645,
+      "step": 12046
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7636787860325356,
+      "learning_rate": 1.4632060183354857e-06,
+      "loss": 0.2701,
+      "step": 12047
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5771368792371883,
+      "learning_rate": 1.4624861768282295e-06,
+      "loss": 0.2503,
+      "step": 12048
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9217027774237119,
+      "learning_rate": 1.4617664820987055e-06,
+      "loss": 0.272,
+      "step": 12049
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1391593192395453,
+      "learning_rate": 1.461046934176773e-06,
+      "loss": 0.2544,
+      "step": 12050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8396394512248049,
+      "learning_rate": 1.460327533092289e-06,
+      "loss": 0.2571,
+      "step": 12051
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4925167561067028,
+      "learning_rate": 1.4596082788751048e-06,
+      "loss": 0.2481,
+      "step": 12052
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.106952728761021,
+      "learning_rate": 1.4588891715550612e-06,
+      "loss": 0.2565,
+      "step": 12053
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.475499050212135,
+      "learning_rate": 1.458170211161994e-06,
+      "loss": 0.2689,
+      "step": 12054
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.554015332841206,
+      "learning_rate": 1.457451397725737e-06,
+      "loss": 0.2559,
+      "step": 12055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2286089315703848,
+      "learning_rate": 1.4567327312761143e-06,
+      "loss": 0.2537,
+      "step": 12056
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.140349767536782,
+      "learning_rate": 1.4560142118429427e-06,
+      "loss": 0.2764,
+      "step": 12057
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9077386950447266,
+      "learning_rate": 1.4552958394560374e-06,
+      "loss": 0.2356,
+      "step": 12058
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.411234553571343,
+      "learning_rate": 1.4545776141452023e-06,
+      "loss": 0.2842,
+      "step": 12059
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1401287558122895,
+      "learning_rate": 1.4538595359402413e-06,
+      "loss": 0.2646,
+      "step": 12060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.198233857607244,
+      "learning_rate": 1.4531416048709469e-06,
+      "loss": 0.2531,
+      "step": 12061
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2902543845869912,
+      "learning_rate": 1.4524238209671048e-06,
+      "loss": 0.2575,
+      "step": 12062
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6376816292524472,
+      "learning_rate": 1.4517061842585017e-06,
+      "loss": 0.482,
+      "step": 12063
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2892471408943904,
+      "learning_rate": 1.450988694774912e-06,
+      "loss": 0.2568,
+      "step": 12064
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6052649285732987,
+      "learning_rate": 1.4502713525461026e-06,
+      "loss": 0.2605,
+      "step": 12065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.933754656876215,
+      "learning_rate": 1.44955415760184e-06,
+      "loss": 0.2531,
+      "step": 12066
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5154627525102058,
+      "learning_rate": 1.448837109971884e-06,
+      "loss": 0.2529,
+      "step": 12067
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2899602501426988,
+      "learning_rate": 1.448120209685983e-06,
+      "loss": 0.2479,
+      "step": 12068
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 14.110370146331876,
+      "learning_rate": 1.4474034567738838e-06,
+      "loss": 0.2645,
+      "step": 12069
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6665665184508645,
+      "learning_rate": 1.4466868512653237e-06,
+      "loss": 0.2937,
+      "step": 12070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.774512981607251,
+      "learning_rate": 1.4459703931900393e-06,
+      "loss": 0.2538,
+      "step": 12071
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3921342498113027,
+      "learning_rate": 1.4452540825777555e-06,
+      "loss": 0.2513,
+      "step": 12072
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.697737022695345,
+      "learning_rate": 1.4445379194581926e-06,
+      "loss": 0.2451,
+      "step": 12073
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.600425139139422,
+      "learning_rate": 1.4438219038610662e-06,
+      "loss": 0.2547,
+      "step": 12074
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.585123484955184,
+      "learning_rate": 1.4431060358160876e-06,
+      "loss": 0.2588,
+      "step": 12075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4003655430597806,
+      "learning_rate": 1.4423903153529568e-06,
+      "loss": 0.2881,
+      "step": 12076
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3755810571745266,
+      "learning_rate": 1.441674742501371e-06,
+      "loss": 0.2571,
+      "step": 12077
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3908024382707085,
+      "learning_rate": 1.440959317291019e-06,
+      "loss": 0.2651,
+      "step": 12078
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4494320693999985,
+      "learning_rate": 1.4402440397515882e-06,
+      "loss": 0.2506,
+      "step": 12079
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9483648144851375,
+      "learning_rate": 1.4395289099127545e-06,
+      "loss": 0.2616,
+      "step": 12080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.104228117768375,
+      "learning_rate": 1.4388139278041895e-06,
+      "loss": 0.2622,
+      "step": 12081
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2158394746627743,
+      "learning_rate": 1.4380990934555593e-06,
+      "loss": 0.2713,
+      "step": 12082
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.209765771543538,
+      "learning_rate": 1.4373844068965265e-06,
+      "loss": 0.2498,
+      "step": 12083
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2188412155029993,
+      "learning_rate": 1.4366698681567425e-06,
+      "loss": 0.2683,
+      "step": 12084
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2706651879775706,
+      "learning_rate": 1.4359554772658551e-06,
+      "loss": 0.2465,
+      "step": 12085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3405066611028733,
+      "learning_rate": 1.4352412342535033e-06,
+      "loss": 0.2461,
+      "step": 12086
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7765449780426272,
+      "learning_rate": 1.434527139149326e-06,
+      "loss": 0.2667,
+      "step": 12087
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.955786494324836,
+      "learning_rate": 1.4338131919829513e-06,
+      "loss": 0.2804,
+      "step": 12088
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6086075855513613,
+      "learning_rate": 1.4330993927839993e-06,
+      "loss": 0.2731,
+      "step": 12089
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2540326874061916,
+      "learning_rate": 1.4323857415820892e-06,
+      "loss": 0.261,
+      "step": 12090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.211682084578564,
+      "learning_rate": 1.4316722384068332e-06,
+      "loss": 0.2623,
+      "step": 12091
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7568744545590538,
+      "learning_rate": 1.430958883287834e-06,
+      "loss": 0.2656,
+      "step": 12092
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8693378545721617,
+      "learning_rate": 1.43024567625469e-06,
+      "loss": 0.2701,
+      "step": 12093
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9474574766802821,
+      "learning_rate": 1.4295326173369921e-06,
+      "loss": 0.2653,
+      "step": 12094
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2457017927086524,
+      "learning_rate": 1.4288197065643288e-06,
+      "loss": 0.2696,
+      "step": 12095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5463452648486415,
+      "learning_rate": 1.428106943966278e-06,
+      "loss": 0.4737,
+      "step": 12096
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7727890902270838,
+      "learning_rate": 1.4273943295724163e-06,
+      "loss": 0.2806,
+      "step": 12097
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3453058333942383,
+      "learning_rate": 1.4266818634123075e-06,
+      "loss": 0.2552,
+      "step": 12098
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.686475443139553,
+      "learning_rate": 1.4259695455155165e-06,
+      "loss": 0.2444,
+      "step": 12099
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.097105243851308,
+      "learning_rate": 1.4252573759115974e-06,
+      "loss": 0.2453,
+      "step": 12100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.442296196789242,
+      "learning_rate": 1.4245453546300996e-06,
+      "loss": 0.2646,
+      "step": 12101
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.737063829014486,
+      "learning_rate": 1.4238334817005638e-06,
+      "loss": 0.2488,
+      "step": 12102
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.618681622374901,
+      "learning_rate": 1.4231217571525308e-06,
+      "loss": 0.2622,
+      "step": 12103
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5502227468010201,
+      "learning_rate": 1.4224101810155273e-06,
+      "loss": 0.2547,
+      "step": 12104
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0551417447986835,
+      "learning_rate": 1.4216987533190819e-06,
+      "loss": 0.2794,
+      "step": 12105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6893047780363037,
+      "learning_rate": 1.4209874740927087e-06,
+      "loss": 0.2388,
+      "step": 12106
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.8125832767489305,
+      "learning_rate": 1.4202763433659244e-06,
+      "loss": 0.2748,
+      "step": 12107
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8689003970630766,
+      "learning_rate": 1.419565361168233e-06,
+      "loss": 0.2619,
+      "step": 12108
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.820352185620478,
+      "learning_rate": 1.4188545275291343e-06,
+      "loss": 0.2524,
+      "step": 12109
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.904757860928734,
+      "learning_rate": 1.4181438424781203e-06,
+      "loss": 0.2708,
+      "step": 12110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.219140879375936,
+      "learning_rate": 1.4174333060446804e-06,
+      "loss": 0.2796,
+      "step": 12111
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 21.511571905586756,
+      "learning_rate": 1.4167229182582974e-06,
+      "loss": 0.2649,
+      "step": 12112
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5516098362282018,
+      "learning_rate": 1.4160126791484453e-06,
+      "loss": 0.4471,
+      "step": 12113
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6622975944093763,
+      "learning_rate": 1.4153025887445914e-06,
+      "loss": 0.2423,
+      "step": 12114
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.50915692108032,
+      "learning_rate": 1.4145926470762023e-06,
+      "loss": 0.2729,
+      "step": 12115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.390954474437639,
+      "learning_rate": 1.4138828541727322e-06,
+      "loss": 0.2811,
+      "step": 12116
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.882459526932715,
+      "learning_rate": 1.413173210063632e-06,
+      "loss": 0.2581,
+      "step": 12117
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.78501533400747,
+      "learning_rate": 1.4124637147783431e-06,
+      "loss": 0.2893,
+      "step": 12118
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6551772270319498,
+      "learning_rate": 1.4117543683463103e-06,
+      "loss": 0.2491,
+      "step": 12119
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4012756677228824,
+      "learning_rate": 1.4110451707969624e-06,
+      "loss": 0.2481,
+      "step": 12120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0419703125098643,
+      "learning_rate": 1.4103361221597244e-06,
+      "loss": 0.2601,
+      "step": 12121
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0914647375675712,
+      "learning_rate": 1.4096272224640155e-06,
+      "loss": 0.2561,
+      "step": 12122
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2375686612246555,
+      "learning_rate": 1.4089184717392524e-06,
+      "loss": 0.2406,
+      "step": 12123
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9218048095779081,
+      "learning_rate": 1.4082098700148395e-06,
+      "loss": 0.2569,
+      "step": 12124
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7098076528240505,
+      "learning_rate": 1.407501417320179e-06,
+      "loss": 0.2744,
+      "step": 12125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9171778187326898,
+      "learning_rate": 1.406793113684663e-06,
+      "loss": 0.2599,
+      "step": 12126
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.74314241740096,
+      "learning_rate": 1.4060849591376858e-06,
+      "loss": 0.2549,
+      "step": 12127
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.414918815760029,
+      "learning_rate": 1.405376953708627e-06,
+      "loss": 0.2585,
+      "step": 12128
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.799812980432534,
+      "learning_rate": 1.4046690974268634e-06,
+      "loss": 0.2424,
+      "step": 12129
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.629016152305734,
+      "learning_rate": 1.4039613903217636e-06,
+      "loss": 0.2597,
+      "step": 12130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.118408806726114,
+      "learning_rate": 1.403253832422694e-06,
+      "loss": 0.2555,
+      "step": 12131
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.47675570741331,
+      "learning_rate": 1.402546423759012e-06,
+      "loss": 0.2481,
+      "step": 12132
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5083457030073633,
+      "learning_rate": 1.401839164360067e-06,
+      "loss": 0.2457,
+      "step": 12133
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1849445597091224,
+      "learning_rate": 1.401132054255206e-06,
+      "loss": 0.2577,
+      "step": 12134
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4755413355219726,
+      "learning_rate": 1.4004250934737701e-06,
+      "loss": 0.26,
+      "step": 12135
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.156483687758241,
+      "learning_rate": 1.39971828204509e-06,
+      "loss": 0.2516,
+      "step": 12136
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6256868513634082,
+      "learning_rate": 1.3990116199984938e-06,
+      "loss": 0.2482,
+      "step": 12137
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5934940225949321,
+      "learning_rate": 1.3983051073632996e-06,
+      "loss": 0.2511,
+      "step": 12138
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.487029862492722,
+      "learning_rate": 1.397598744168826e-06,
+      "loss": 0.2706,
+      "step": 12139
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.862388876202757,
+      "learning_rate": 1.3968925304443786e-06,
+      "loss": 0.2494,
+      "step": 12140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6227209940447236,
+      "learning_rate": 1.3961864662192586e-06,
+      "loss": 0.2535,
+      "step": 12141
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7130382213619153,
+      "learning_rate": 1.3954805515227632e-06,
+      "loss": 0.2531,
+      "step": 12142
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8278220899417228,
+      "learning_rate": 1.3947747863841836e-06,
+      "loss": 0.2528,
+      "step": 12143
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.251253788843478,
+      "learning_rate": 1.3940691708328018e-06,
+      "loss": 0.2574,
+      "step": 12144
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.746734438790287,
+      "learning_rate": 1.3933637048978949e-06,
+      "loss": 0.2588,
+      "step": 12145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9579740225844458,
+      "learning_rate": 1.3926583886087326e-06,
+      "loss": 0.2477,
+      "step": 12146
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.138227823451893,
+      "learning_rate": 1.3919532219945819e-06,
+      "loss": 0.2992,
+      "step": 12147
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7109840867803783,
+      "learning_rate": 1.3912482050846999e-06,
+      "loss": 0.2637,
+      "step": 12148
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.216299194263243,
+      "learning_rate": 1.390543337908341e-06,
+      "loss": 0.2548,
+      "step": 12149
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7964161006614783,
+      "learning_rate": 1.3898386204947485e-06,
+      "loss": 0.2665,
+      "step": 12150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9030536138531213,
+      "learning_rate": 1.3891340528731656e-06,
+      "loss": 0.2543,
+      "step": 12151
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.386004139164735,
+      "learning_rate": 1.3884296350728245e-06,
+      "loss": 0.2494,
+      "step": 12152
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3586158587829322,
+      "learning_rate": 1.3877253671229528e-06,
+      "loss": 0.256,
+      "step": 12153
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5827677368568623,
+      "learning_rate": 1.38702124905277e-06,
+      "loss": 0.4691,
+      "step": 12154
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9524956657439987,
+      "learning_rate": 1.386317280891495e-06,
+      "loss": 0.2708,
+      "step": 12155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.794127078629825,
+      "learning_rate": 1.3856134626683325e-06,
+      "loss": 0.2545,
+      "step": 12156
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9505565417195476,
+      "learning_rate": 1.384909794412489e-06,
+      "loss": 0.2647,
+      "step": 12157
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3767627075695774,
+      "learning_rate": 1.3842062761531577e-06,
+      "loss": 0.2529,
+      "step": 12158
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.275406547439901,
+      "learning_rate": 1.3835029079195322e-06,
+      "loss": 0.2542,
+      "step": 12159
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1390473320247203,
+      "learning_rate": 1.3827996897407942e-06,
+      "loss": 0.2385,
+      "step": 12160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.502283166060426,
+      "learning_rate": 1.3820966216461224e-06,
+      "loss": 0.25,
+      "step": 12161
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8176713406240028,
+      "learning_rate": 1.3813937036646858e-06,
+      "loss": 0.2561,
+      "step": 12162
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0244744902021257,
+      "learning_rate": 1.3806909358256522e-06,
+      "loss": 0.2474,
+      "step": 12163
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7662667869668818,
+      "learning_rate": 1.3799883181581818e-06,
+      "loss": 0.2703,
+      "step": 12164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.787122337335751,
+      "learning_rate": 1.3792858506914258e-06,
+      "loss": 0.2507,
+      "step": 12165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.411214800296559,
+      "learning_rate": 1.37858353345453e-06,
+      "loss": 0.2721,
+      "step": 12166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.065825198070482,
+      "learning_rate": 1.3778813664766366e-06,
+      "loss": 0.2403,
+      "step": 12167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5864386815362713,
+      "learning_rate": 1.377179349786879e-06,
+      "loss": 0.4672,
+      "step": 12168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9330363884101904,
+      "learning_rate": 1.376477483414385e-06,
+      "loss": 0.2511,
+      "step": 12169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3594106557510997,
+      "learning_rate": 1.3757757673882753e-06,
+      "loss": 0.2415,
+      "step": 12170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.854869228502983,
+      "learning_rate": 1.375074201737666e-06,
+      "loss": 0.2635,
+      "step": 12171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5622857479687187,
+      "learning_rate": 1.3743727864916684e-06,
+      "loss": 0.2404,
+      "step": 12172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9452416758872502,
+      "learning_rate": 1.3736715216793833e-06,
+      "loss": 0.2608,
+      "step": 12173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7397170903180896,
+      "learning_rate": 1.3729704073299066e-06,
+      "loss": 0.2683,
+      "step": 12174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.022305899674779,
+      "learning_rate": 1.3722694434723316e-06,
+      "loss": 0.2855,
+      "step": 12175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9460262754557436,
+      "learning_rate": 1.3715686301357407e-06,
+      "loss": 0.2422,
+      "step": 12176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.843969987891146,
+      "learning_rate": 1.3708679673492127e-06,
+      "loss": 0.2627,
+      "step": 12177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0233249351876323,
+      "learning_rate": 1.3701674551418154e-06,
+      "loss": 0.2664,
+      "step": 12178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.773220153771355,
+      "learning_rate": 1.3694670935426202e-06,
+      "loss": 0.2419,
+      "step": 12179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8184339760275756,
+      "learning_rate": 1.3687668825806843e-06,
+      "loss": 0.241,
+      "step": 12180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2042115949749004,
+      "learning_rate": 1.3680668222850607e-06,
+      "loss": 0.2504,
+      "step": 12181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.52442515121413,
+      "learning_rate": 1.3673669126847938e-06,
+      "loss": 0.2895,
+      "step": 12182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2367914963257904,
+      "learning_rate": 1.3666671538089272e-06,
+      "loss": 0.2564,
+      "step": 12183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6387767623544527,
+      "learning_rate": 1.365967545686494e-06,
+      "loss": 0.4548,
+      "step": 12184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5423614370986392,
+      "learning_rate": 1.3652680883465224e-06,
+      "loss": 0.2467,
+      "step": 12185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.296864002626352,
+      "learning_rate": 1.364568781818031e-06,
+      "loss": 0.2464,
+      "step": 12186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.202580176727354,
+      "learning_rate": 1.3638696261300404e-06,
+      "loss": 0.2654,
+      "step": 12187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1249299824419166,
+      "learning_rate": 1.3631706213115582e-06,
+      "loss": 0.2424,
+      "step": 12188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.350336052881772,
+      "learning_rate": 1.3624717673915861e-06,
+      "loss": 0.2368,
+      "step": 12189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4282564488224514,
+      "learning_rate": 1.36177306439912e-06,
+      "loss": 0.2518,
+      "step": 12190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3358017943237215,
+      "learning_rate": 1.3610745123631536e-06,
+      "loss": 0.2449,
+      "step": 12191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.099503338369686,
+      "learning_rate": 1.360376111312669e-06,
+      "loss": 0.2403,
+      "step": 12192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.835473065478738,
+      "learning_rate": 1.359677861276642e-06,
+      "loss": 0.2733,
+      "step": 12193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5521528226802335,
+      "learning_rate": 1.3589797622840473e-06,
+      "loss": 0.4471,
+      "step": 12194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8520327327932877,
+      "learning_rate": 1.3582818143638505e-06,
+      "loss": 0.2594,
+      "step": 12195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6606673204036535,
+      "learning_rate": 1.35758401754501e-06,
+      "loss": 0.2568,
+      "step": 12196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.789776641886652,
+      "learning_rate": 1.3568863718564779e-06,
+      "loss": 0.26,
+      "step": 12197
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3393705210298186,
+      "learning_rate": 1.3561888773271997e-06,
+      "loss": 0.2544,
+      "step": 12198
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2725920425656425,
+      "learning_rate": 1.355491533986119e-06,
+      "loss": 0.2607,
+      "step": 12199
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4027854019975927,
+      "learning_rate": 1.354794341862166e-06,
+      "loss": 0.2545,
+      "step": 12200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6602622518563193,
+      "learning_rate": 1.3540973009842722e-06,
+      "loss": 0.2488,
+      "step": 12201
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.4201286451726185,
+      "learning_rate": 1.3534004113813553e-06,
+      "loss": 0.2449,
+      "step": 12202
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4548751633048314,
+      "learning_rate": 1.3527036730823344e-06,
+      "loss": 0.2549,
+      "step": 12203
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.878016111572727,
+      "learning_rate": 1.3520070861161162e-06,
+      "loss": 0.2718,
+      "step": 12204
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.554610789370885,
+      "learning_rate": 1.3513106505116036e-06,
+      "loss": 0.2627,
+      "step": 12205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8104370302578667,
+      "learning_rate": 1.350614366297691e-06,
+      "loss": 0.2473,
+      "step": 12206
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2328859510418466,
+      "learning_rate": 1.3499182335032728e-06,
+      "loss": 0.2394,
+      "step": 12207
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4119699133315216,
+      "learning_rate": 1.3492222521572286e-06,
+      "loss": 0.2797,
+      "step": 12208
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9954328470563016,
+      "learning_rate": 1.3485264222884387e-06,
+      "loss": 0.2661,
+      "step": 12209
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8542275818763985,
+      "learning_rate": 1.3478307439257721e-06,
+      "loss": 0.2546,
+      "step": 12210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0227989591725586,
+      "learning_rate": 1.347135217098096e-06,
+      "loss": 0.2864,
+      "step": 12211
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7970596585819967,
+      "learning_rate": 1.3464398418342683e-06,
+      "loss": 0.2425,
+      "step": 12212
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4076250465106765,
+      "learning_rate": 1.3457446181631412e-06,
+      "loss": 0.255,
+      "step": 12213
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0487167145072678,
+      "learning_rate": 1.3450495461135587e-06,
+      "loss": 0.2798,
+      "step": 12214
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9664566338913183,
+      "learning_rate": 1.3443546257143624e-06,
+      "loss": 0.2515,
+      "step": 12215
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9933489856074946,
+      "learning_rate": 1.3436598569943877e-06,
+      "loss": 0.272,
+      "step": 12216
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.363127307583459,
+      "learning_rate": 1.34296523998246e-06,
+      "loss": 0.2543,
+      "step": 12217
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7804329629619244,
+      "learning_rate": 1.3422707747073982e-06,
+      "loss": 0.26,
+      "step": 12218
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 18.699902909772035,
+      "learning_rate": 1.3415764611980203e-06,
+      "loss": 0.2512,
+      "step": 12219
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5363615952587026,
+      "learning_rate": 1.3408822994831334e-06,
+      "loss": 0.2604,
+      "step": 12220
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.743496592571209,
+      "learning_rate": 1.340188289591539e-06,
+      "loss": 0.2669,
+      "step": 12221
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5999052530662206,
+      "learning_rate": 1.339494431552032e-06,
+      "loss": 0.269,
+      "step": 12222
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2818387282556643,
+      "learning_rate": 1.3388007253934027e-06,
+      "loss": 0.2514,
+      "step": 12223
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7594046784324404,
+      "learning_rate": 1.3381071711444354e-06,
+      "loss": 0.2566,
+      "step": 12224
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.996730874594099,
+      "learning_rate": 1.3374137688339061e-06,
+      "loss": 0.2631,
+      "step": 12225
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6209726289735795,
+      "learning_rate": 1.3367205184905834e-06,
+      "loss": 0.238,
+      "step": 12226
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4512645326127482,
+      "learning_rate": 1.3360274201432345e-06,
+      "loss": 0.2676,
+      "step": 12227
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.004577670521368,
+      "learning_rate": 1.3353344738206158e-06,
+      "loss": 0.2577,
+      "step": 12228
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9752401514118758,
+      "learning_rate": 1.3346416795514794e-06,
+      "loss": 0.2645,
+      "step": 12229
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8238158477170305,
+      "learning_rate": 1.3339490373645686e-06,
+      "loss": 0.2532,
+      "step": 12230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5715346091088411,
+      "learning_rate": 1.3332565472886238e-06,
+      "loss": 0.4747,
+      "step": 12231
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0124216246458646,
+      "learning_rate": 1.3325642093523789e-06,
+      "loss": 0.247,
+      "step": 12232
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4418951142527257,
+      "learning_rate": 1.3318720235845594e-06,
+      "loss": 0.253,
+      "step": 12233
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4855711853991473,
+      "learning_rate": 1.3311799900138834e-06,
+      "loss": 0.2633,
+      "step": 12234
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.7269823728543825,
+      "learning_rate": 1.3304881086690674e-06,
+      "loss": 0.262,
+      "step": 12235
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.111751311674048,
+      "learning_rate": 1.3297963795788177e-06,
+      "loss": 0.2404,
+      "step": 12236
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.866616775899338,
+      "learning_rate": 1.3291048027718357e-06,
+      "loss": 0.2553,
+      "step": 12237
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0361375820927377,
+      "learning_rate": 1.3284133782768139e-06,
+      "loss": 0.2482,
+      "step": 12238
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8277084526024296,
+      "learning_rate": 1.3277221061224422e-06,
+      "loss": 0.2697,
+      "step": 12239
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.760103770365269,
+      "learning_rate": 1.3270309863374047e-06,
+      "loss": 0.277,
+      "step": 12240
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4116259303005694,
+      "learning_rate": 1.3263400189503756e-06,
+      "loss": 0.2468,
+      "step": 12241
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8392725729239117,
+      "learning_rate": 1.3256492039900226e-06,
+      "loss": 0.241,
+      "step": 12242
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.103783520812081,
+      "learning_rate": 1.324958541485012e-06,
+      "loss": 0.259,
+      "step": 12243
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6020325443230078,
+      "learning_rate": 1.3242680314639995e-06,
+      "loss": 0.486,
+      "step": 12244
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3984756000123766,
+      "learning_rate": 1.3235776739556333e-06,
+      "loss": 0.247,
+      "step": 12245
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2871965411899904,
+      "learning_rate": 1.3228874689885606e-06,
+      "loss": 0.2402,
+      "step": 12246
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.720329837754432,
+      "learning_rate": 1.3221974165914176e-06,
+      "loss": 0.2578,
+      "step": 12247
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7224414669061514,
+      "learning_rate": 1.3215075167928372e-06,
+      "loss": 0.2637,
+      "step": 12248
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.820060765595058,
+      "learning_rate": 1.3208177696214436e-06,
+      "loss": 0.2714,
+      "step": 12249
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5356019136963701,
+      "learning_rate": 1.3201281751058542e-06,
+      "loss": 0.2641,
+      "step": 12250
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9641856076478836,
+      "learning_rate": 1.319438733274685e-06,
+      "loss": 0.2455,
+      "step": 12251
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2092214727781436,
+      "learning_rate": 1.3187494441565397e-06,
+      "loss": 0.2526,
+      "step": 12252
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.519156337722903,
+      "learning_rate": 1.318060307780017e-06,
+      "loss": 0.275,
+      "step": 12253
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9369238548468752,
+      "learning_rate": 1.3173713241737123e-06,
+      "loss": 0.2549,
+      "step": 12254
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.776159112521937,
+      "learning_rate": 1.3166824933662136e-06,
+      "loss": 0.2591,
+      "step": 12255
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7435280964426205,
+      "learning_rate": 1.3159938153861012e-06,
+      "loss": 0.2569,
+      "step": 12256
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.043634005003201,
+      "learning_rate": 1.3153052902619484e-06,
+      "loss": 0.266,
+      "step": 12257
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.939685880575297,
+      "learning_rate": 1.3146169180223228e-06,
+      "loss": 0.257,
+      "step": 12258
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0139655057295855,
+      "learning_rate": 1.3139286986957884e-06,
+      "loss": 0.2564,
+      "step": 12259
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6765394533407212,
+      "learning_rate": 1.3132406323108981e-06,
+      "loss": 0.2587,
+      "step": 12260
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5876611684976215,
+      "learning_rate": 1.3125527188962034e-06,
+      "loss": 0.4666,
+      "step": 12261
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0890686446704105,
+      "learning_rate": 1.311864958480245e-06,
+      "loss": 0.2331,
+      "step": 12262
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.677254633816732,
+      "learning_rate": 1.311177351091562e-06,
+      "loss": 0.2422,
+      "step": 12263
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.256836932168832,
+      "learning_rate": 1.3104898967586827e-06,
+      "loss": 0.2553,
+      "step": 12264
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.70247783766268,
+      "learning_rate": 1.3098025955101306e-06,
+      "loss": 0.2522,
+      "step": 12265
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1079012304757425,
+      "learning_rate": 1.3091154473744221e-06,
+      "loss": 0.2481,
+      "step": 12266
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.603033998295756,
+      "learning_rate": 1.308428452380071e-06,
+      "loss": 0.4413,
+      "step": 12267
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8033667127370485,
+      "learning_rate": 1.3077416105555784e-06,
+      "loss": 0.2582,
+      "step": 12268
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.134625154010529,
+      "learning_rate": 1.3070549219294465e-06,
+      "loss": 0.25,
+      "step": 12269
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9638394449868155,
+      "learning_rate": 1.306368386530164e-06,
+      "loss": 0.2741,
+      "step": 12270
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6311224043929804,
+      "learning_rate": 1.3056820043862189e-06,
+      "loss": 0.2553,
+      "step": 12271
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.800949375937788,
+      "learning_rate": 1.3049957755260894e-06,
+      "loss": 0.2474,
+      "step": 12272
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7937288608334936,
+      "learning_rate": 1.3043096999782478e-06,
+      "loss": 0.2493,
+      "step": 12273
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.947387471232567,
+      "learning_rate": 1.3036237777711602e-06,
+      "loss": 0.2383,
+      "step": 12274
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8362005445986116,
+      "learning_rate": 1.3029380089332872e-06,
+      "loss": 0.2557,
+      "step": 12275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2221529467055463,
+      "learning_rate": 1.3022523934930848e-06,
+      "loss": 0.2772,
+      "step": 12276
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.750850222222142,
+      "learning_rate": 1.301566931478998e-06,
+      "loss": 0.2386,
+      "step": 12277
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.210715233519939,
+      "learning_rate": 1.3008816229194671e-06,
+      "loss": 0.278,
+      "step": 12278
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.228814960391766,
+      "learning_rate": 1.3001964678429296e-06,
+      "loss": 0.2579,
+      "step": 12279
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7137110584683724,
+      "learning_rate": 1.2995114662778125e-06,
+      "loss": 0.2445,
+      "step": 12280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6586462192318272,
+      "learning_rate": 1.2988266182525371e-06,
+      "loss": 0.25,
+      "step": 12281
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5299095831505582,
+      "learning_rate": 1.298141923795518e-06,
+      "loss": 0.2341,
+      "step": 12282
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6383193506290523,
+      "learning_rate": 1.2974573829351655e-06,
+      "loss": 0.2546,
+      "step": 12283
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1167957018806236,
+      "learning_rate": 1.2967729956998843e-06,
+      "loss": 0.2629,
+      "step": 12284
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.167159529146566,
+      "learning_rate": 1.296088762118069e-06,
+      "loss": 0.2414,
+      "step": 12285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7300635099688708,
+      "learning_rate": 1.2954046822181088e-06,
+      "loss": 0.2661,
+      "step": 12286
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9307077429699162,
+      "learning_rate": 1.2947207560283891e-06,
+      "loss": 0.2538,
+      "step": 12287
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.343178622654274,
+      "learning_rate": 1.2940369835772875e-06,
+      "loss": 0.2707,
+      "step": 12288
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.000069593865243,
+      "learning_rate": 1.2933533648931734e-06,
+      "loss": 0.2533,
+      "step": 12289
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9158547480528179,
+      "learning_rate": 1.2926699000044107e-06,
+      "loss": 0.2614,
+      "step": 12290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.064007732610284,
+      "learning_rate": 1.2919865889393584e-06,
+      "loss": 0.266,
+      "step": 12291
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5719989869828415,
+      "learning_rate": 1.2913034317263706e-06,
+      "loss": 0.2428,
+      "step": 12292
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5747876282402975,
+      "learning_rate": 1.2906204283937907e-06,
+      "loss": 0.4773,
+      "step": 12293
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.215520242756292,
+      "learning_rate": 1.2899375789699564e-06,
+      "loss": 0.2596,
+      "step": 12294
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8201404682568871,
+      "learning_rate": 1.289254883483203e-06,
+      "loss": 0.2761,
+      "step": 12295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6073309541184808,
+      "learning_rate": 1.2885723419618556e-06,
+      "loss": 0.2527,
+      "step": 12296
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1361266884470442,
+      "learning_rate": 1.2878899544342327e-06,
+      "loss": 0.2513,
+      "step": 12297
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8028098344253858,
+      "learning_rate": 1.2872077209286505e-06,
+      "loss": 0.2404,
+      "step": 12298
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.785329958891657,
+      "learning_rate": 1.2865256414734133e-06,
+      "loss": 0.2312,
+      "step": 12299
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.398529969598262,
+      "learning_rate": 1.285843716096824e-06,
+      "loss": 0.2621,
+      "step": 12300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7311487904016185,
+      "learning_rate": 1.2851619448271762e-06,
+      "loss": 0.2635,
+      "step": 12301
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3297448433351207,
+      "learning_rate": 1.2844803276927565e-06,
+      "loss": 0.2564,
+      "step": 12302
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5766723875318966,
+      "learning_rate": 1.283798864721849e-06,
+      "loss": 0.2797,
+      "step": 12303
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9294270427827505,
+      "learning_rate": 1.2831175559427266e-06,
+      "loss": 0.2548,
+      "step": 12304
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5845101405312683,
+      "learning_rate": 1.2824364013836583e-06,
+      "loss": 0.2635,
+      "step": 12305
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.954355978769204,
+      "learning_rate": 1.2817554010729071e-06,
+      "loss": 0.2688,
+      "step": 12306
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4917984473922121,
+      "learning_rate": 1.2810745550387283e-06,
+      "loss": 0.264,
+      "step": 12307
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4676256062330664,
+      "learning_rate": 1.2803938633093726e-06,
+      "loss": 0.2606,
+      "step": 12308
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3007111312262816,
+      "learning_rate": 1.2797133259130822e-06,
+      "loss": 0.2567,
+      "step": 12309
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7365384012426694,
+      "learning_rate": 1.2790329428780928e-06,
+      "loss": 0.2528,
+      "step": 12310
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.859494170895407,
+      "learning_rate": 1.2783527142326375e-06,
+      "loss": 0.262,
+      "step": 12311
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.078863950762435,
+      "learning_rate": 1.277672640004936e-06,
+      "loss": 0.2674,
+      "step": 12312
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7138783608599124,
+      "learning_rate": 1.2769927202232108e-06,
+      "loss": 0.2524,
+      "step": 12313
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8801935359782573,
+      "learning_rate": 1.2763129549156699e-06,
+      "loss": 0.2678,
+      "step": 12314
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.53751842020365,
+      "learning_rate": 1.2756333441105168e-06,
+      "loss": 0.2458,
+      "step": 12315
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.082930999696426,
+      "learning_rate": 1.2749538878359535e-06,
+      "loss": 0.2641,
+      "step": 12316
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.973673385915936,
+      "learning_rate": 1.2742745861201694e-06,
+      "loss": 0.2432,
+      "step": 12317
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.012117341277662,
+      "learning_rate": 1.273595438991349e-06,
+      "loss": 0.2942,
+      "step": 12318
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8174064867377344,
+      "learning_rate": 1.2729164464776743e-06,
+      "loss": 0.2483,
+      "step": 12319
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6989239722003788,
+      "learning_rate": 1.2722376086073147e-06,
+      "loss": 0.281,
+      "step": 12320
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6236687822976983,
+      "learning_rate": 1.27155892540844e-06,
+      "loss": 0.259,
+      "step": 12321
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7435312816188404,
+      "learning_rate": 1.270880396909207e-06,
+      "loss": 0.2619,
+      "step": 12322
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9706493887070782,
+      "learning_rate": 1.2702020231377688e-06,
+      "loss": 0.2336,
+      "step": 12323
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4683597384257465,
+      "learning_rate": 1.2695238041222752e-06,
+      "loss": 0.2698,
+      "step": 12324
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.937780816532654,
+      "learning_rate": 1.2688457398908648e-06,
+      "loss": 0.2518,
+      "step": 12325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1077805243028296,
+      "learning_rate": 1.2681678304716711e-06,
+      "loss": 0.2603,
+      "step": 12326
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5625326232647218,
+      "learning_rate": 1.267490075892822e-06,
+      "loss": 0.4493,
+      "step": 12327
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3462421567618548,
+      "learning_rate": 1.2668124761824408e-06,
+      "loss": 0.2449,
+      "step": 12328
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.261398561192962,
+      "learning_rate": 1.266135031368641e-06,
+      "loss": 0.2699,
+      "step": 12329
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.170633701689669,
+      "learning_rate": 1.265457741479531e-06,
+      "loss": 0.2718,
+      "step": 12330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8752811960831397,
+      "learning_rate": 1.2647806065432106e-06,
+      "loss": 0.2518,
+      "step": 12331
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.166894900555235,
+      "learning_rate": 1.2641036265877794e-06,
+      "loss": 0.2643,
+      "step": 12332
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.247543311408234,
+      "learning_rate": 1.2634268016413242e-06,
+      "loss": 0.2607,
+      "step": 12333
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9844969605356477,
+      "learning_rate": 1.2627501317319263e-06,
+      "loss": 0.2614,
+      "step": 12334
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2404791224035177,
+      "learning_rate": 1.2620736168876636e-06,
+      "loss": 0.2507,
+      "step": 12335
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.941104051602157,
+      "learning_rate": 1.2613972571366074e-06,
+      "loss": 0.2586,
+      "step": 12336
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.378846833590021,
+      "learning_rate": 1.2607210525068192e-06,
+      "loss": 0.2526,
+      "step": 12337
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9412812644753084,
+      "learning_rate": 1.260045003026355e-06,
+      "loss": 0.2595,
+      "step": 12338
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.836719651860733,
+      "learning_rate": 1.2593691087232684e-06,
+      "loss": 0.2706,
+      "step": 12339
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.541883754239091,
+      "learning_rate": 1.258693369625601e-06,
+      "loss": 0.2536,
+      "step": 12340
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9858546006136708,
+      "learning_rate": 1.2580177857613912e-06,
+      "loss": 0.2424,
+      "step": 12341
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.818343775609208,
+      "learning_rate": 1.2573423571586686e-06,
+      "loss": 0.2537,
+      "step": 12342
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8340129739737918,
+      "learning_rate": 1.2566670838454593e-06,
+      "loss": 0.2549,
+      "step": 12343
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5880251809977763,
+      "learning_rate": 1.2559919658497827e-06,
+      "loss": 0.4537,
+      "step": 12344
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0419660037749505,
+      "learning_rate": 1.2553170031996497e-06,
+      "loss": 0.2734,
+      "step": 12345
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1628444714305917,
+      "learning_rate": 1.2546421959230642e-06,
+      "loss": 0.2404,
+      "step": 12346
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.122535391561523,
+      "learning_rate": 1.253967544048028e-06,
+      "loss": 0.2551,
+      "step": 12347
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.0228905299660935,
+      "learning_rate": 1.2532930476025322e-06,
+      "loss": 0.2608,
+      "step": 12348
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1976743069358076,
+      "learning_rate": 1.2526187066145628e-06,
+      "loss": 0.2584,
+      "step": 12349
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.140037319583642,
+      "learning_rate": 1.2519445211120979e-06,
+      "loss": 0.2546,
+      "step": 12350
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8626642140723708,
+      "learning_rate": 1.2512704911231116e-06,
+      "loss": 0.2392,
+      "step": 12351
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.315816455665586,
+      "learning_rate": 1.2505966166755729e-06,
+      "loss": 0.2531,
+      "step": 12352
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7528098609029947,
+      "learning_rate": 1.2499228977974403e-06,
+      "loss": 0.2682,
+      "step": 12353
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9548536727600647,
+      "learning_rate": 1.2492493345166662e-06,
+      "loss": 0.2473,
+      "step": 12354
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5180813166607314,
+      "learning_rate": 1.2485759268612008e-06,
+      "loss": 0.2503,
+      "step": 12355
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4406103045390592,
+      "learning_rate": 1.2479026748589834e-06,
+      "loss": 0.25,
+      "step": 12356
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.0824957613167765,
+      "learning_rate": 1.2472295785379468e-06,
+      "loss": 0.2502,
+      "step": 12357
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5232142161234588,
+      "learning_rate": 1.2465566379260224e-06,
+      "loss": 0.2542,
+      "step": 12358
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9319094125273095,
+      "learning_rate": 1.245883853051128e-06,
+      "loss": 0.2372,
+      "step": 12359
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8080425389440837,
+      "learning_rate": 1.2452112239411823e-06,
+      "loss": 0.2629,
+      "step": 12360
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.304213693016038,
+      "learning_rate": 1.2445387506240925e-06,
+      "loss": 0.2732,
+      "step": 12361
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6106698475194454,
+      "learning_rate": 1.243866433127759e-06,
+      "loss": 0.4669,
+      "step": 12362
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8446023964467877,
+      "learning_rate": 1.24319427148008e-06,
+      "loss": 0.2544,
+      "step": 12363
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3624951310690876,
+      "learning_rate": 1.242522265708942e-06,
+      "loss": 0.2857,
+      "step": 12364
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.60000752409501,
+      "learning_rate": 1.2418504158422306e-06,
+      "loss": 0.2747,
+      "step": 12365
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.652813906098466,
+      "learning_rate": 1.2411787219078208e-06,
+      "loss": 0.2423,
+      "step": 12366
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5945675052839787,
+      "learning_rate": 1.2405071839335808e-06,
+      "loss": 0.262,
+      "step": 12367
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.565938368415108,
+      "learning_rate": 1.2398358019473766e-06,
+      "loss": 0.2546,
+      "step": 12368
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.755968406847944,
+      "learning_rate": 1.2391645759770642e-06,
+      "loss": 0.2445,
+      "step": 12369
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 178.19725017159877,
+      "learning_rate": 1.238493506050492e-06,
+      "loss": 0.2667,
+      "step": 12370
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.518734467800095,
+      "learning_rate": 1.2378225921955067e-06,
+      "loss": 0.2489,
+      "step": 12371
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.685700762759292,
+      "learning_rate": 1.237151834439943e-06,
+      "loss": 0.2652,
+      "step": 12372
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6633170610302996,
+      "learning_rate": 1.2364812328116348e-06,
+      "loss": 0.2425,
+      "step": 12373
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9514821613397824,
+      "learning_rate": 1.235810787338405e-06,
+      "loss": 0.2413,
+      "step": 12374
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.374962238376874,
+      "learning_rate": 1.2351404980480702e-06,
+      "loss": 0.2653,
+      "step": 12375
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5858438771644625,
+      "learning_rate": 1.2344703649684442e-06,
+      "loss": 0.4414,
+      "step": 12376
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.555637482244403,
+      "learning_rate": 1.233800388127332e-06,
+      "loss": 0.2658,
+      "step": 12377
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9046070394648476,
+      "learning_rate": 1.233130567552529e-06,
+      "loss": 0.2615,
+      "step": 12378
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2725242366830396,
+      "learning_rate": 1.2324609032718298e-06,
+      "loss": 0.2602,
+      "step": 12379
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9590298443285727,
+      "learning_rate": 1.2317913953130213e-06,
+      "loss": 0.272,
+      "step": 12380
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9062195970508713,
+      "learning_rate": 1.2311220437038801e-06,
+      "loss": 0.2589,
+      "step": 12381
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7911703758876762,
+      "learning_rate": 1.23045284847218e-06,
+      "loss": 0.2562,
+      "step": 12382
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5286150250647546,
+      "learning_rate": 1.2297838096456855e-06,
+      "loss": 0.2563,
+      "step": 12383
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4972484247860378,
+      "learning_rate": 1.2291149272521586e-06,
+      "loss": 0.2602,
+      "step": 12384
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0914695433560055,
+      "learning_rate": 1.2284462013193515e-06,
+      "loss": 0.2404,
+      "step": 12385
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8861475027800798,
+      "learning_rate": 1.227777631875009e-06,
+      "loss": 0.2666,
+      "step": 12386
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5058317852938243,
+      "learning_rate": 1.2271092189468726e-06,
+      "loss": 0.2568,
+      "step": 12387
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8664571856773016,
+      "learning_rate": 1.2264409625626778e-06,
+      "loss": 0.245,
+      "step": 12388
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.626670678213925,
+      "learning_rate": 1.2257728627501497e-06,
+      "loss": 0.2518,
+      "step": 12389
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7524761963452473,
+      "learning_rate": 1.2251049195370097e-06,
+      "loss": 0.2582,
+      "step": 12390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7858765545749833,
+      "learning_rate": 1.2244371329509697e-06,
+      "loss": 0.2594,
+      "step": 12391
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6927565945282785,
+      "learning_rate": 1.223769503019741e-06,
+      "loss": 0.2558,
+      "step": 12392
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.001672851441979,
+      "learning_rate": 1.2231020297710228e-06,
+      "loss": 0.2422,
+      "step": 12393
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.661360873683687,
+      "learning_rate": 1.2224347132325082e-06,
+      "loss": 0.2527,
+      "step": 12394
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.009360135903299,
+      "learning_rate": 1.2217675534318873e-06,
+      "loss": 0.243,
+      "step": 12395
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2687522923484273,
+      "learning_rate": 1.2211005503968431e-06,
+      "loss": 0.2624,
+      "step": 12396
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4004277469686612,
+      "learning_rate": 1.220433704155049e-06,
+      "loss": 0.2477,
+      "step": 12397
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0374917725442665,
+      "learning_rate": 1.219767014734174e-06,
+      "loss": 0.2686,
+      "step": 12398
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6373522856638223,
+      "learning_rate": 1.2191004821618785e-06,
+      "loss": 0.2533,
+      "step": 12399
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.590581844065551,
+      "learning_rate": 1.2184341064658206e-06,
+      "loss": 0.2631,
+      "step": 12400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.877908650692595,
+      "learning_rate": 1.2177678876736488e-06,
+      "loss": 0.254,
+      "step": 12401
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.574956335709526,
+      "learning_rate": 1.2171018258130042e-06,
+      "loss": 0.2649,
+      "step": 12402
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.123557344062702,
+      "learning_rate": 1.2164359209115235e-06,
+      "loss": 0.244,
+      "step": 12403
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2841673830322133,
+      "learning_rate": 1.2157701729968384e-06,
+      "loss": 0.2629,
+      "step": 12404
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.658671307641624,
+      "learning_rate": 1.2151045820965708e-06,
+      "loss": 0.2447,
+      "step": 12405
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.439038942895115,
+      "learning_rate": 1.2144391482383361e-06,
+      "loss": 0.2476,
+      "step": 12406
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.109420604086303,
+      "learning_rate": 1.2137738714497437e-06,
+      "loss": 0.2548,
+      "step": 12407
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7114240748355343,
+      "learning_rate": 1.2131087517584006e-06,
+      "loss": 0.2651,
+      "step": 12408
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9308033531199176,
+      "learning_rate": 1.2124437891918995e-06,
+      "loss": 0.2484,
+      "step": 12409
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7548605238780082,
+      "learning_rate": 1.2117789837778349e-06,
+      "loss": 0.2356,
+      "step": 12410
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.974220225322346,
+      "learning_rate": 1.2111143355437877e-06,
+      "loss": 0.234,
+      "step": 12411
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.752797791187074,
+      "learning_rate": 1.2104498445173373e-06,
+      "loss": 0.2339,
+      "step": 12412
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6263168882674981,
+      "learning_rate": 1.2097855107260542e-06,
+      "loss": 0.4699,
+      "step": 12413
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.156443661231618,
+      "learning_rate": 1.209121334197501e-06,
+      "loss": 0.247,
+      "step": 12414
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4142469559438777,
+      "learning_rate": 1.2084573149592382e-06,
+      "loss": 0.2503,
+      "step": 12415
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.647771114225083,
+      "learning_rate": 1.2077934530388163e-06,
+      "loss": 0.2396,
+      "step": 12416
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.529543479823659,
+      "learning_rate": 1.2071297484637785e-06,
+      "loss": 0.2468,
+      "step": 12417
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6622057967595985,
+      "learning_rate": 1.2064662012616651e-06,
+      "loss": 0.2584,
+      "step": 12418
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.868284889800844,
+      "learning_rate": 1.2058028114600061e-06,
+      "loss": 0.2506,
+      "step": 12419
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3575051245893266,
+      "learning_rate": 1.205139579086329e-06,
+      "loss": 0.2731,
+      "step": 12420
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8052124583456362,
+      "learning_rate": 1.204476504168151e-06,
+      "loss": 0.244,
+      "step": 12421
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7006314857029459,
+      "learning_rate": 1.203813586732983e-06,
+      "loss": 0.2552,
+      "step": 12422
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.815932145949014,
+      "learning_rate": 1.2031508268083342e-06,
+      "loss": 0.2782,
+      "step": 12423
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5145252146341597,
+      "learning_rate": 1.2024882244216996e-06,
+      "loss": 0.2438,
+      "step": 12424
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.940945168592251,
+      "learning_rate": 1.201825779600575e-06,
+      "loss": 0.2598,
+      "step": 12425
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6301880384540195,
+      "learning_rate": 1.2011634923724452e-06,
+      "loss": 0.2485,
+      "step": 12426
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.843311714911491,
+      "learning_rate": 1.2005013627647889e-06,
+      "loss": 0.2532,
+      "step": 12427
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7614695281699693,
+      "learning_rate": 1.1998393908050803e-06,
+      "loss": 0.2537,
+      "step": 12428
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1405152867619286,
+      "learning_rate": 1.1991775765207854e-06,
+      "loss": 0.2679,
+      "step": 12429
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5735871136892312,
+      "learning_rate": 1.1985159199393626e-06,
+      "loss": 0.4516,
+      "step": 12430
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7317408857345176,
+      "learning_rate": 1.1978544210882675e-06,
+      "loss": 0.2497,
+      "step": 12431
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.075251173038843,
+      "learning_rate": 1.197193079994945e-06,
+      "loss": 0.267,
+      "step": 12432
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.828177396460023,
+      "learning_rate": 1.1965318966868372e-06,
+      "loss": 0.2445,
+      "step": 12433
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.91106666692495,
+      "learning_rate": 1.1958708711913769e-06,
+      "loss": 0.2403,
+      "step": 12434
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7164893516252615,
+      "learning_rate": 1.1952100035359893e-06,
+      "loss": 0.261,
+      "step": 12435
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9739400176144315,
+      "learning_rate": 1.1945492937480984e-06,
+      "loss": 0.2515,
+      "step": 12436
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6913900386011396,
+      "learning_rate": 1.1938887418551164e-06,
+      "loss": 0.254,
+      "step": 12437
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.442235477648228,
+      "learning_rate": 1.1932283478844497e-06,
+      "loss": 0.2649,
+      "step": 12438
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.726568134779693,
+      "learning_rate": 1.1925681118635008e-06,
+      "loss": 0.2729,
+      "step": 12439
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6070856432939178,
+      "learning_rate": 1.1919080338196642e-06,
+      "loss": 0.4923,
+      "step": 12440
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9357918761025683,
+      "learning_rate": 1.1912481137803277e-06,
+      "loss": 0.2582,
+      "step": 12441
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.239781688494264,
+      "learning_rate": 1.1905883517728723e-06,
+      "loss": 0.2741,
+      "step": 12442
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.548832100515837,
+      "learning_rate": 1.1899287478246707e-06,
+      "loss": 0.2451,
+      "step": 12443
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9318255389905616,
+      "learning_rate": 1.1892693019630946e-06,
+      "loss": 0.2755,
+      "step": 12444
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.309551825458991,
+      "learning_rate": 1.1886100142155032e-06,
+      "loss": 0.2632,
+      "step": 12445
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5611418476221632,
+      "learning_rate": 1.1879508846092513e-06,
+      "loss": 0.2738,
+      "step": 12446
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5195417147947374,
+      "learning_rate": 1.1872919131716875e-06,
+      "loss": 0.2497,
+      "step": 12447
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8920029144962827,
+      "learning_rate": 1.1866330999301562e-06,
+      "loss": 0.2658,
+      "step": 12448
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4493648452887045,
+      "learning_rate": 1.1859744449119908e-06,
+      "loss": 0.2571,
+      "step": 12449
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5764234114088442,
+      "learning_rate": 1.1853159481445203e-06,
+      "loss": 0.2499,
+      "step": 12450
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9210018274176985,
+      "learning_rate": 1.184657609655065e-06,
+      "loss": 0.2657,
+      "step": 12451
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6734108429947123,
+      "learning_rate": 1.183999429470944e-06,
+      "loss": 0.2585,
+      "step": 12452
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.85496893988578,
+      "learning_rate": 1.1833414076194643e-06,
+      "loss": 0.2543,
+      "step": 12453
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.874318287772399,
+      "learning_rate": 1.1826835441279277e-06,
+      "loss": 0.2703,
+      "step": 12454
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.020251792916572,
+      "learning_rate": 1.182025839023631e-06,
+      "loss": 0.2456,
+      "step": 12455
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.054236271080544,
+      "learning_rate": 1.1813682923338654e-06,
+      "loss": 0.2484,
+      "step": 12456
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5462757259698257,
+      "learning_rate": 1.1807109040859115e-06,
+      "loss": 0.256,
+      "step": 12457
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7216708233070572,
+      "learning_rate": 1.1800536743070467e-06,
+      "loss": 0.2622,
+      "step": 12458
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7179250263832087,
+      "learning_rate": 1.1793966030245379e-06,
+      "loss": 0.2536,
+      "step": 12459
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6603924928332737,
+      "learning_rate": 1.1787396902656518e-06,
+      "loss": 0.4963,
+      "step": 12460
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5371285628577036,
+      "learning_rate": 1.1780829360576418e-06,
+      "loss": 0.2527,
+      "step": 12461
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1436323176442866,
+      "learning_rate": 1.1774263404277607e-06,
+      "loss": 0.2524,
+      "step": 12462
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.620969723792433,
+      "learning_rate": 1.1767699034032492e-06,
+      "loss": 0.2527,
+      "step": 12463
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.06020484986667,
+      "learning_rate": 1.1761136250113465e-06,
+      "loss": 0.2586,
+      "step": 12464
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.324309916458656,
+      "learning_rate": 1.1754575052792815e-06,
+      "loss": 0.2812,
+      "step": 12465
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1097163696447896,
+      "learning_rate": 1.1748015442342781e-06,
+      "loss": 0.273,
+      "step": 12466
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5635493235170816,
+      "learning_rate": 1.1741457419035507e-06,
+      "loss": 0.4458,
+      "step": 12467
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.33909884006406,
+      "learning_rate": 1.1734900983143137e-06,
+      "loss": 0.2516,
+      "step": 12468
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.627532320798,
+      "learning_rate": 1.1728346134937684e-06,
+      "loss": 0.2681,
+      "step": 12469
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4333070540487776,
+      "learning_rate": 1.1721792874691134e-06,
+      "loss": 0.2567,
+      "step": 12470
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9691859001338756,
+      "learning_rate": 1.1715241202675376e-06,
+      "loss": 0.269,
+      "step": 12471
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.214259415331407,
+      "learning_rate": 1.170869111916228e-06,
+      "loss": 0.247,
+      "step": 12472
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.141149431017653,
+      "learning_rate": 1.1702142624423597e-06,
+      "loss": 0.2471,
+      "step": 12473
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4431629562547268,
+      "learning_rate": 1.169559571873104e-06,
+      "loss": 0.2388,
+      "step": 12474
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9320162319467629,
+      "learning_rate": 1.1689050402356245e-06,
+      "loss": 0.2676,
+      "step": 12475
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9550619027480893,
+      "learning_rate": 1.1682506675570794e-06,
+      "loss": 0.2668,
+      "step": 12476
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.423839090169735,
+      "learning_rate": 1.1675964538646212e-06,
+      "loss": 0.2612,
+      "step": 12477
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.703081067910517,
+      "learning_rate": 1.1669423991853934e-06,
+      "loss": 0.2656,
+      "step": 12478
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.809074591000523,
+      "learning_rate": 1.1662885035465326e-06,
+      "loss": 0.2561,
+      "step": 12479
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.289841843814653,
+      "learning_rate": 1.1656347669751721e-06,
+      "loss": 0.2365,
+      "step": 12480
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7508959446791594,
+      "learning_rate": 1.1649811894984365e-06,
+      "loss": 0.2742,
+      "step": 12481
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.768419559671135,
+      "learning_rate": 1.1643277711434426e-06,
+      "loss": 0.2486,
+      "step": 12482
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1725298500932095,
+      "learning_rate": 1.1636745119373006e-06,
+      "loss": 0.2634,
+      "step": 12483
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0440031235292655,
+      "learning_rate": 1.1630214119071176e-06,
+      "loss": 0.2552,
+      "step": 12484
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.165666677740684,
+      "learning_rate": 1.1623684710799931e-06,
+      "loss": 0.2787,
+      "step": 12485
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.634598689959782,
+      "learning_rate": 1.1617156894830168e-06,
+      "loss": 0.2526,
+      "step": 12486
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0799976191475413,
+      "learning_rate": 1.1610630671432726e-06,
+      "loss": 0.2522,
+      "step": 12487
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.526831303344769,
+      "learning_rate": 1.160410604087842e-06,
+      "loss": 0.275,
+      "step": 12488
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7336969226906678,
+      "learning_rate": 1.1597583003437957e-06,
+      "loss": 0.2441,
+      "step": 12489
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3347651756990528,
+      "learning_rate": 1.1591061559381983e-06,
+      "loss": 0.2344,
+      "step": 12490
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5842669488335628,
+      "learning_rate": 1.1584541708981056e-06,
+      "loss": 0.2454,
+      "step": 12491
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0914782853210023,
+      "learning_rate": 1.157802345250576e-06,
+      "loss": 0.241,
+      "step": 12492
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6067020423351305,
+      "learning_rate": 1.1571506790226512e-06,
+      "loss": 0.5115,
+      "step": 12493
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.577052080039558,
+      "learning_rate": 1.1564991722413703e-06,
+      "loss": 0.2581,
+      "step": 12494
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.141169498888766,
+      "learning_rate": 1.1558478249337645e-06,
+      "loss": 0.2558,
+      "step": 12495
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8072534450279496,
+      "learning_rate": 1.155196637126862e-06,
+      "loss": 0.2558,
+      "step": 12496
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4883261753788135,
+      "learning_rate": 1.1545456088476798e-06,
+      "loss": 0.2548,
+      "step": 12497
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1176064660172833,
+      "learning_rate": 1.1538947401232292e-06,
+      "loss": 0.2452,
+      "step": 12498
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.44912751435392,
+      "learning_rate": 1.1532440309805172e-06,
+      "loss": 0.2758,
+      "step": 12499
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.273582491757086,
+      "learning_rate": 1.1525934814465445e-06,
+      "loss": 0.2393,
+      "step": 12500
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9041350921919935,
+      "learning_rate": 1.151943091548302e-06,
+      "loss": 0.2604,
+      "step": 12501
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.287163244391985,
+      "learning_rate": 1.1512928613127755e-06,
+      "loss": 0.2495,
+      "step": 12502
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4595723358799626,
+      "learning_rate": 1.1506427907669433e-06,
+      "loss": 0.256,
+      "step": 12503
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.49651984953011,
+      "learning_rate": 1.1499928799377797e-06,
+      "loss": 0.2595,
+      "step": 12504
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.125624539039244,
+      "learning_rate": 1.1493431288522506e-06,
+      "loss": 0.2559,
+      "step": 12505
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8762300636033735,
+      "learning_rate": 1.1486935375373127e-06,
+      "loss": 0.2578,
+      "step": 12506
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5016800503813323,
+      "learning_rate": 1.1480441060199205e-06,
+      "loss": 0.2544,
+      "step": 12507
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.838339903371633,
+      "learning_rate": 1.147394834327022e-06,
+      "loss": 0.2664,
+      "step": 12508
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.881716108332491,
+      "learning_rate": 1.1467457224855545e-06,
+      "loss": 0.2537,
+      "step": 12509
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5807947566489926,
+      "learning_rate": 1.1460967705224513e-06,
+      "loss": 0.2605,
+      "step": 12510
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.26284734727908,
+      "learning_rate": 1.1454479784646366e-06,
+      "loss": 0.2579,
+      "step": 12511
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.8058990245876085,
+      "learning_rate": 1.1447993463390338e-06,
+      "loss": 0.2557,
+      "step": 12512
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6603955876008762,
+      "learning_rate": 1.1441508741725532e-06,
+      "loss": 0.2607,
+      "step": 12513
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8662390847806956,
+      "learning_rate": 1.1435025619921003e-06,
+      "loss": 0.27,
+      "step": 12514
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8993136860139344,
+      "learning_rate": 1.1428544098245764e-06,
+      "loss": 0.2605,
+      "step": 12515
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6618781794788233,
+      "learning_rate": 1.1422064176968751e-06,
+      "loss": 0.2393,
+      "step": 12516
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6535574441274803,
+      "learning_rate": 1.1415585856358818e-06,
+      "loss": 0.2459,
+      "step": 12517
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1107706388054064,
+      "learning_rate": 1.140910913668476e-06,
+      "loss": 0.254,
+      "step": 12518
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0057155683324996,
+      "learning_rate": 1.1402634018215297e-06,
+      "loss": 0.2567,
+      "step": 12519
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8805554651096124,
+      "learning_rate": 1.1396160501219122e-06,
+      "loss": 0.256,
+      "step": 12520
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5254401353977964,
+      "learning_rate": 1.1389688585964797e-06,
+      "loss": 0.2589,
+      "step": 12521
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7727760794380036,
+      "learning_rate": 1.1383218272720886e-06,
+      "loss": 0.2486,
+      "step": 12522
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.226380985944135,
+      "learning_rate": 1.1376749561755829e-06,
+      "loss": 0.251,
+      "step": 12523
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9691395237304765,
+      "learning_rate": 1.137028245333805e-06,
+      "loss": 0.2513,
+      "step": 12524
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7968276456284897,
+      "learning_rate": 1.1363816947735866e-06,
+      "loss": 0.2797,
+      "step": 12525
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7319762883552918,
+      "learning_rate": 1.135735304521754e-06,
+      "loss": 0.2488,
+      "step": 12526
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3259678274218527,
+      "learning_rate": 1.1350890746051257e-06,
+      "loss": 0.2511,
+      "step": 12527
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.774371210727824,
+      "learning_rate": 1.1344430050505174e-06,
+      "loss": 0.2683,
+      "step": 12528
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4778816333882983,
+      "learning_rate": 1.1337970958847354e-06,
+      "loss": 0.2398,
+      "step": 12529
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.35855918601479,
+      "learning_rate": 1.1331513471345796e-06,
+      "loss": 0.2563,
+      "step": 12530
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.002575286983129,
+      "learning_rate": 1.1325057588268406e-06,
+      "loss": 0.2638,
+      "step": 12531
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.794344514719857,
+      "learning_rate": 1.1318603309883092e-06,
+      "loss": 0.2573,
+      "step": 12532
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0473068150855536,
+      "learning_rate": 1.131215063645763e-06,
+      "loss": 0.2479,
+      "step": 12533
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1101361089676454,
+      "learning_rate": 1.1305699568259754e-06,
+      "loss": 0.2535,
+      "step": 12534
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6986855224429824,
+      "learning_rate": 1.129925010555712e-06,
+      "loss": 0.2493,
+      "step": 12535
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.41802514819793,
+      "learning_rate": 1.129280224861734e-06,
+      "loss": 0.2528,
+      "step": 12536
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1709270847637714,
+      "learning_rate": 1.128635599770796e-06,
+      "loss": 0.2532,
+      "step": 12537
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6679563241448176,
+      "learning_rate": 1.1279911353096428e-06,
+      "loss": 0.2451,
+      "step": 12538
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.737360594545524,
+      "learning_rate": 1.127346831505014e-06,
+      "loss": 0.2513,
+      "step": 12539
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.782237902939371,
+      "learning_rate": 1.126702688383645e-06,
+      "loss": 0.2483,
+      "step": 12540
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1581764216894004,
+      "learning_rate": 1.1260587059722612e-06,
+      "loss": 0.2604,
+      "step": 12541
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9506341887736045,
+      "learning_rate": 1.1254148842975826e-06,
+      "loss": 0.2657,
+      "step": 12542
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.031156869127606,
+      "learning_rate": 1.1247712233863196e-06,
+      "loss": 0.243,
+      "step": 12543
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8372615209296899,
+      "learning_rate": 1.1241277232651849e-06,
+      "loss": 0.2634,
+      "step": 12544
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.997029139268055,
+      "learning_rate": 1.123484383960875e-06,
+      "loss": 0.2414,
+      "step": 12545
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.031356522291658,
+      "learning_rate": 1.1228412055000831e-06,
+      "loss": 0.2791,
+      "step": 12546
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.188305805998532,
+      "learning_rate": 1.1221981879094952e-06,
+      "loss": 0.2515,
+      "step": 12547
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.713710686893574,
+      "learning_rate": 1.121555331215794e-06,
+      "loss": 0.2685,
+      "step": 12548
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6204778214657698,
+      "learning_rate": 1.12091263544565e-06,
+      "loss": 0.4421,
+      "step": 12549
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6060995304773358,
+      "learning_rate": 1.1202701006257317e-06,
+      "loss": 0.248,
+      "step": 12550
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.884199085345835,
+      "learning_rate": 1.119627726782695e-06,
+      "loss": 0.2599,
+      "step": 12551
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.031984087381864,
+      "learning_rate": 1.1189855139431988e-06,
+      "loss": 0.2416,
+      "step": 12552
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.219641225185397,
+      "learning_rate": 1.1183434621338874e-06,
+      "loss": 0.267,
+      "step": 12553
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.19017248474657,
+      "learning_rate": 1.1177015713814005e-06,
+      "loss": 0.2626,
+      "step": 12554
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9210577488538396,
+      "learning_rate": 1.1170598417123695e-06,
+      "loss": 0.2715,
+      "step": 12555
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2658083660306367,
+      "learning_rate": 1.116418273153424e-06,
+      "loss": 0.239,
+      "step": 12556
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.335857299370904,
+      "learning_rate": 1.1157768657311824e-06,
+      "loss": 0.2561,
+      "step": 12557
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8964327199140771,
+      "learning_rate": 1.1151356194722563e-06,
+      "loss": 0.2807,
+      "step": 12558
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.270531759566121,
+      "learning_rate": 1.114494534403253e-06,
+      "loss": 0.2648,
+      "step": 12559
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6292387873161762,
+      "learning_rate": 1.1138536105507752e-06,
+      "loss": 0.2628,
+      "step": 12560
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.33056391057903,
+      "learning_rate": 1.1132128479414133e-06,
+      "loss": 0.2614,
+      "step": 12561
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.3726849490375415,
+      "learning_rate": 1.1125722466017547e-06,
+      "loss": 0.2387,
+      "step": 12562
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.069500013054947,
+      "learning_rate": 1.1119318065583763e-06,
+      "loss": 0.256,
+      "step": 12563
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.369525702333648,
+      "learning_rate": 1.111291527837855e-06,
+      "loss": 0.2503,
+      "step": 12564
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6172731811989232,
+      "learning_rate": 1.110651410466755e-06,
+      "loss": 0.2294,
+      "step": 12565
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9636983040958742,
+      "learning_rate": 1.1100114544716351e-06,
+      "loss": 0.2402,
+      "step": 12566
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0941481280855103,
+      "learning_rate": 1.1093716598790494e-06,
+      "loss": 0.2619,
+      "step": 12567
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8640338042927724,
+      "learning_rate": 1.1087320267155448e-06,
+      "loss": 0.2475,
+      "step": 12568
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.731232259530004,
+      "learning_rate": 1.108092555007661e-06,
+      "loss": 0.2579,
+      "step": 12569
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.035099166033582,
+      "learning_rate": 1.1074532447819291e-06,
+      "loss": 0.2555,
+      "step": 12570
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.88827441752297,
+      "learning_rate": 1.1068140960648753e-06,
+      "loss": 0.2681,
+      "step": 12571
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.257072743708595,
+      "learning_rate": 1.1061751088830208e-06,
+      "loss": 0.2593,
+      "step": 12572
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.754039672030549,
+      "learning_rate": 1.1055362832628757e-06,
+      "loss": 0.2535,
+      "step": 12573
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4491884704594167,
+      "learning_rate": 1.1048976192309496e-06,
+      "loss": 0.2379,
+      "step": 12574
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9714295966204913,
+      "learning_rate": 1.1042591168137379e-06,
+      "loss": 0.2552,
+      "step": 12575
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.122563903119853,
+      "learning_rate": 1.103620776037736e-06,
+      "loss": 0.2591,
+      "step": 12576
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7153754520571782,
+      "learning_rate": 1.1029825969294294e-06,
+      "loss": 0.2519,
+      "step": 12577
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.915400174354112,
+      "learning_rate": 1.102344579515297e-06,
+      "loss": 0.2725,
+      "step": 12578
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9739363400590335,
+      "learning_rate": 1.1017067238218093e-06,
+      "loss": 0.2466,
+      "step": 12579
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3522173238319866,
+      "learning_rate": 1.1010690298754352e-06,
+      "loss": 0.2969,
+      "step": 12580
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8681438226119749,
+      "learning_rate": 1.1004314977026304e-06,
+      "loss": 0.256,
+      "step": 12581
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1370796301878596,
+      "learning_rate": 1.0997941273298512e-06,
+      "loss": 0.2456,
+      "step": 12582
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5404171831350435,
+      "learning_rate": 1.09915691878354e-06,
+      "loss": 0.2436,
+      "step": 12583
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.030863537508889,
+      "learning_rate": 1.0985198720901375e-06,
+      "loss": 0.2489,
+      "step": 12584
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.387414855463944,
+      "learning_rate": 1.097882987276076e-06,
+      "loss": 0.2555,
+      "step": 12585
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8833909598958247,
+      "learning_rate": 1.09724626436778e-06,
+      "loss": 0.2542,
+      "step": 12586
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.31885258932891,
+      "learning_rate": 1.0966097033916674e-06,
+      "loss": 0.2561,
+      "step": 12587
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6444723961562331,
+      "learning_rate": 1.095973304374151e-06,
+      "loss": 0.456,
+      "step": 12588
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0304281248864497,
+      "learning_rate": 1.095337067341638e-06,
+      "loss": 0.2607,
+      "step": 12589
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4028968508352064,
+      "learning_rate": 1.094700992320526e-06,
+      "loss": 0.2495,
+      "step": 12590
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6004150796113565,
+      "learning_rate": 1.0940650793372048e-06,
+      "loss": 0.4702,
+      "step": 12591
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.704736975204606,
+      "learning_rate": 1.0934293284180625e-06,
+      "loss": 0.2513,
+      "step": 12592
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7708386570462173,
+      "learning_rate": 1.092793739589476e-06,
+      "loss": 0.2751,
+      "step": 12593
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.935491697614958,
+      "learning_rate": 1.0921583128778174e-06,
+      "loss": 0.256,
+      "step": 12594
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.163138681186147,
+      "learning_rate": 1.0915230483094502e-06,
+      "loss": 0.245,
+      "step": 12595
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0448580586558363,
+      "learning_rate": 1.090887945910734e-06,
+      "loss": 0.2569,
+      "step": 12596
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8075072237919652,
+      "learning_rate": 1.0902530057080218e-06,
+      "loss": 0.2554,
+      "step": 12597
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3769456760994443,
+      "learning_rate": 1.0896182277276568e-06,
+      "loss": 0.2421,
+      "step": 12598
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.635367036395204,
+      "learning_rate": 1.0889836119959757e-06,
+      "loss": 0.2763,
+      "step": 12599
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.465053926785206,
+      "learning_rate": 1.0883491585393125e-06,
+      "loss": 0.2391,
+      "step": 12600
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4718448703491538,
+      "learning_rate": 1.0877148673839905e-06,
+      "loss": 0.2541,
+      "step": 12601
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.782588444553286,
+      "learning_rate": 1.0870807385563282e-06,
+      "loss": 0.2515,
+      "step": 12602
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.755272247053251,
+      "learning_rate": 1.0864467720826343e-06,
+      "loss": 0.2696,
+      "step": 12603
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.60891908528279,
+      "learning_rate": 1.0858129679892148e-06,
+      "loss": 0.4695,
+      "step": 12604
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.766288661741712,
+      "learning_rate": 1.0851793263023696e-06,
+      "loss": 0.2431,
+      "step": 12605
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.057605425557707,
+      "learning_rate": 1.0845458470483877e-06,
+      "loss": 0.268,
+      "step": 12606
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.766783735190372,
+      "learning_rate": 1.083912530253552e-06,
+      "loss": 0.2559,
+      "step": 12607
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.691451251144815,
+      "learning_rate": 1.0832793759441418e-06,
+      "loss": 0.2425,
+      "step": 12608
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.444297540711089,
+      "learning_rate": 1.082646384146428e-06,
+      "loss": 0.2727,
+      "step": 12609
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9744189851229572,
+      "learning_rate": 1.0820135548866718e-06,
+      "loss": 0.2696,
+      "step": 12610
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9574124572550855,
+      "learning_rate": 1.0813808881911326e-06,
+      "loss": 0.2555,
+      "step": 12611
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.865698936407434,
+      "learning_rate": 1.0807483840860616e-06,
+      "loss": 0.2522,
+      "step": 12612
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.47363020299997,
+      "learning_rate": 1.080116042597702e-06,
+      "loss": 0.2427,
+      "step": 12613
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5894927783966755,
+      "learning_rate": 1.0794838637522898e-06,
+      "loss": 0.2488,
+      "step": 12614
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6714157842975013,
+      "learning_rate": 1.0788518475760545e-06,
+      "loss": 0.2562,
+      "step": 12615
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.059916192120151,
+      "learning_rate": 1.0782199940952226e-06,
+      "loss": 0.2521,
+      "step": 12616
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.582203056058463,
+      "learning_rate": 1.0775883033360085e-06,
+      "loss": 0.2539,
+      "step": 12617
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.281674729655202,
+      "learning_rate": 1.0769567753246214e-06,
+      "loss": 0.2754,
+      "step": 12618
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.247061740126943,
+      "learning_rate": 1.076325410087266e-06,
+      "loss": 0.245,
+      "step": 12619
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.107378748262202,
+      "learning_rate": 1.0756942076501397e-06,
+      "loss": 0.2466,
+      "step": 12620
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.070681722068169,
+      "learning_rate": 1.0750631680394314e-06,
+      "loss": 0.2543,
+      "step": 12621
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.894514451643714,
+      "learning_rate": 1.0744322912813231e-06,
+      "loss": 0.2681,
+      "step": 12622
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.6935646765427474,
+      "learning_rate": 1.0738015774019911e-06,
+      "loss": 0.2756,
+      "step": 12623
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4813043818811578,
+      "learning_rate": 1.0731710264276062e-06,
+      "loss": 0.2463,
+      "step": 12624
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0690950187614274,
+      "learning_rate": 1.072540638384329e-06,
+      "loss": 0.2559,
+      "step": 12625
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1003674811340676,
+      "learning_rate": 1.0719104132983176e-06,
+      "loss": 0.273,
+      "step": 12626
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.999234616830355,
+      "learning_rate": 1.0712803511957199e-06,
+      "loss": 0.2639,
+      "step": 12627
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0810651801883537,
+      "learning_rate": 1.0706504521026788e-06,
+      "loss": 0.2828,
+      "step": 12628
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.252220017339744,
+      "learning_rate": 1.0700207160453308e-06,
+      "loss": 0.2584,
+      "step": 12629
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5991030471073953,
+      "learning_rate": 1.0693911430498032e-06,
+      "loss": 0.243,
+      "step": 12630
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.270072849526293,
+      "learning_rate": 1.0687617331422173e-06,
+      "loss": 0.2957,
+      "step": 12631
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0615147930890165,
+      "learning_rate": 1.0681324863486907e-06,
+      "loss": 0.2353,
+      "step": 12632
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.942022182797392,
+      "learning_rate": 1.0675034026953302e-06,
+      "loss": 0.2611,
+      "step": 12633
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7247293771471623,
+      "learning_rate": 1.06687448220824e-06,
+      "loss": 0.2438,
+      "step": 12634
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 27.532396668252304,
+      "learning_rate": 1.0662457249135116e-06,
+      "loss": 0.2613,
+      "step": 12635
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9627645819310615,
+      "learning_rate": 1.0656171308372371e-06,
+      "loss": 0.2532,
+      "step": 12636
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.860449922467492,
+      "learning_rate": 1.0649887000054954e-06,
+      "loss": 0.2438,
+      "step": 12637
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.428910228790399,
+      "learning_rate": 1.0643604324443623e-06,
+      "loss": 0.2636,
+      "step": 12638
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5791223517173654,
+      "learning_rate": 1.0637323281799045e-06,
+      "loss": 0.2629,
+      "step": 12639
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5656433746105467,
+      "learning_rate": 1.063104387238183e-06,
+      "loss": 0.2398,
+      "step": 12640
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.084940454890411,
+      "learning_rate": 1.0624766096452555e-06,
+      "loss": 0.243,
+      "step": 12641
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.509478982840512,
+      "learning_rate": 1.0618489954271667e-06,
+      "loss": 0.244,
+      "step": 12642
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.604940557742458,
+      "learning_rate": 1.0612215446099566e-06,
+      "loss": 0.2649,
+      "step": 12643
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.152044277074773,
+      "learning_rate": 1.0605942572196626e-06,
+      "loss": 0.2692,
+      "step": 12644
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2724483867519862,
+      "learning_rate": 1.0599671332823096e-06,
+      "loss": 0.2498,
+      "step": 12645
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3261794034612644,
+      "learning_rate": 1.0593401728239183e-06,
+      "loss": 0.2654,
+      "step": 12646
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.697217798764649,
+      "learning_rate": 1.0587133758705015e-06,
+      "loss": 0.257,
+      "step": 12647
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.556527071138105,
+      "learning_rate": 1.0580867424480674e-06,
+      "loss": 0.4191,
+      "step": 12648
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7960712602951898,
+      "learning_rate": 1.0574602725826171e-06,
+      "loss": 0.2576,
+      "step": 12649
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3832907326687363,
+      "learning_rate": 1.0568339663001431e-06,
+      "loss": 0.2487,
+      "step": 12650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4440824776907264,
+      "learning_rate": 1.0562078236266304e-06,
+      "loss": 0.2546,
+      "step": 12651
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7054232843073094,
+      "learning_rate": 1.0555818445880612e-06,
+      "loss": 0.2566,
+      "step": 12652
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1606381253869293,
+      "learning_rate": 1.0549560292104071e-06,
+      "loss": 0.2563,
+      "step": 12653
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4857746976529542,
+      "learning_rate": 1.054330377519635e-06,
+      "loss": 0.2505,
+      "step": 12654
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5759929981336405,
+      "learning_rate": 1.0537048895417024e-06,
+      "loss": 0.2412,
+      "step": 12655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4120537014752426,
+      "learning_rate": 1.0530795653025634e-06,
+      "loss": 0.2384,
+      "step": 12656
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0359148928023334,
+      "learning_rate": 1.052454404828165e-06,
+      "loss": 0.267,
+      "step": 12657
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.715593916937375,
+      "learning_rate": 1.0518294081444447e-06,
+      "loss": 0.2517,
+      "step": 12658
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7637411676576837,
+      "learning_rate": 1.0512045752773336e-06,
+      "loss": 0.2485,
+      "step": 12659
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9585179326587279,
+      "learning_rate": 1.0505799062527605e-06,
+      "loss": 0.258,
+      "step": 12660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7517473444612268,
+      "learning_rate": 1.0499554010966418e-06,
+      "loss": 0.2686,
+      "step": 12661
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1278934042656314,
+      "learning_rate": 1.0493310598348894e-06,
+      "loss": 0.2586,
+      "step": 12662
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4294070183837997,
+      "learning_rate": 1.048706882493407e-06,
+      "loss": 0.2745,
+      "step": 12663
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.563678841463182,
+      "learning_rate": 1.0480828690980949e-06,
+      "loss": 0.2446,
+      "step": 12664
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.356532435630444,
+      "learning_rate": 1.0474590196748447e-06,
+      "loss": 0.2536,
+      "step": 12665
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3055337489752814,
+      "learning_rate": 1.0468353342495407e-06,
+      "loss": 0.2638,
+      "step": 12666
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.271229517955468,
+      "learning_rate": 1.0462118128480586e-06,
+      "loss": 0.2704,
+      "step": 12667
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.103510597477528,
+      "learning_rate": 1.0455884554962725e-06,
+      "loss": 0.2555,
+      "step": 12668
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7146463056873202,
+      "learning_rate": 1.0449652622200458e-06,
+      "loss": 0.2393,
+      "step": 12669
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4774842234541814,
+      "learning_rate": 1.044342233045233e-06,
+      "loss": 0.2438,
+      "step": 12670
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0275772087169353,
+      "learning_rate": 1.0437193679976887e-06,
+      "loss": 0.2474,
+      "step": 12671
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1172927523323017,
+      "learning_rate": 1.0430966671032538e-06,
+      "loss": 0.2519,
+      "step": 12672
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3934315297416164,
+      "learning_rate": 1.0424741303877678e-06,
+      "loss": 0.2507,
+      "step": 12673
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9443693351825124,
+      "learning_rate": 1.0418517578770594e-06,
+      "loss": 0.2457,
+      "step": 12674
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.69091874738381,
+      "learning_rate": 1.041229549596951e-06,
+      "loss": 0.2632,
+      "step": 12675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.526762667542659,
+      "learning_rate": 1.0406075055732612e-06,
+      "loss": 0.239,
+      "step": 12676
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4981714523450758,
+      "learning_rate": 1.0399856258317987e-06,
+      "loss": 0.2469,
+      "step": 12677
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5519275324277573,
+      "learning_rate": 1.039363910398365e-06,
+      "loss": 0.2467,
+      "step": 12678
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6470867041201902,
+      "learning_rate": 1.0387423592987584e-06,
+      "loss": 0.2581,
+      "step": 12679
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9412357340760982,
+      "learning_rate": 1.0381209725587666e-06,
+      "loss": 0.2582,
+      "step": 12680
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.693824949661888,
+      "learning_rate": 1.0374997502041739e-06,
+      "loss": 0.2697,
+      "step": 12681
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7987068313073133,
+      "learning_rate": 1.0368786922607548e-06,
+      "loss": 0.2595,
+      "step": 12682
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.366354239649986,
+      "learning_rate": 1.0362577987542766e-06,
+      "loss": 0.2667,
+      "step": 12683
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.726829433353076,
+      "learning_rate": 1.035637069710504e-06,
+      "loss": 0.2383,
+      "step": 12684
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.799693095943145,
+      "learning_rate": 1.0350165051551897e-06,
+      "loss": 0.2378,
+      "step": 12685
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.207478286308404,
+      "learning_rate": 1.0343961051140845e-06,
+      "loss": 0.2402,
+      "step": 12686
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.626782023532939,
+      "learning_rate": 1.0337758696129285e-06,
+      "loss": 0.2538,
+      "step": 12687
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.924991570707471,
+      "learning_rate": 1.0331557986774553e-06,
+      "loss": 0.2448,
+      "step": 12688
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2163584607766262,
+      "learning_rate": 1.0325358923333956e-06,
+      "loss": 0.2855,
+      "step": 12689
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.505883828485406,
+      "learning_rate": 1.031916150606468e-06,
+      "loss": 0.2641,
+      "step": 12690
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9471321499543168,
+      "learning_rate": 1.0312965735223868e-06,
+      "loss": 0.2357,
+      "step": 12691
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.48507471862247,
+      "learning_rate": 1.0306771611068595e-06,
+      "loss": 0.2602,
+      "step": 12692
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6326021287696546,
+      "learning_rate": 1.030057913385588e-06,
+      "loss": 0.4777,
+      "step": 12693
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.45425023918084,
+      "learning_rate": 1.0294388303842661e-06,
+      "loss": 0.2637,
+      "step": 12694
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2861385267994963,
+      "learning_rate": 1.0288199121285775e-06,
+      "loss": 0.2343,
+      "step": 12695
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7641721723017243,
+      "learning_rate": 1.0282011586442064e-06,
+      "loss": 0.2537,
+      "step": 12696
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.974476382647005,
+      "learning_rate": 1.0275825699568231e-06,
+      "loss": 0.2493,
+      "step": 12697
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.123960814989126,
+      "learning_rate": 1.0269641460920954e-06,
+      "loss": 0.2348,
+      "step": 12698
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3583238287554678,
+      "learning_rate": 1.0263458870756808e-06,
+      "loss": 0.2453,
+      "step": 12699
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.018289054983028,
+      "learning_rate": 1.0257277929332332e-06,
+      "loss": 0.2637,
+      "step": 12700
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.665058189261105,
+      "learning_rate": 1.0251098636904e-06,
+      "loss": 0.2504,
+      "step": 12701
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.91998136874884,
+      "learning_rate": 1.0244920993728185e-06,
+      "loss": 0.273,
+      "step": 12702
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3023157744331306,
+      "learning_rate": 1.0238745000061201e-06,
+      "loss": 0.2412,
+      "step": 12703
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7980366962356116,
+      "learning_rate": 1.0232570656159325e-06,
+      "loss": 0.2548,
+      "step": 12704
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4788772562573218,
+      "learning_rate": 1.0226397962278723e-06,
+      "loss": 0.26,
+      "step": 12705
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5763378591316777,
+      "learning_rate": 1.0220226918675519e-06,
+      "loss": 0.2523,
+      "step": 12706
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.201895235173311,
+      "learning_rate": 1.0214057525605742e-06,
+      "loss": 0.2688,
+      "step": 12707
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.531095714465481,
+      "learning_rate": 1.0207889783325386e-06,
+      "loss": 0.2441,
+      "step": 12708
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7185099593682156,
+      "learning_rate": 1.0201723692090376e-06,
+      "loss": 0.2475,
+      "step": 12709
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.803324053474017,
+      "learning_rate": 1.019555925215654e-06,
+      "loss": 0.2596,
+      "step": 12710
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6728505484323652,
+      "learning_rate": 1.0189396463779632e-06,
+      "loss": 0.2465,
+      "step": 12711
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.611571019272076,
+      "learning_rate": 1.0183235327215396e-06,
+      "loss": 0.4697,
+      "step": 12712
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5883228438460357,
+      "learning_rate": 1.0177075842719448e-06,
+      "loss": 0.2656,
+      "step": 12713
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6129906049332171,
+      "learning_rate": 1.0170918010547355e-06,
+      "loss": 0.4433,
+      "step": 12714
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.389832644539706,
+      "learning_rate": 1.0164761830954606e-06,
+      "loss": 0.2487,
+      "step": 12715
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.426246001497093,
+      "learning_rate": 1.0158607304196643e-06,
+      "loss": 0.2529,
+      "step": 12716
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.101607057639297,
+      "learning_rate": 1.0152454430528847e-06,
+      "loss": 0.2349,
+      "step": 12717
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9409808877926633,
+      "learning_rate": 1.0146303210206487e-06,
+      "loss": 0.2657,
+      "step": 12718
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.479657028448129,
+      "learning_rate": 1.0140153643484784e-06,
+      "loss": 0.2712,
+      "step": 12719
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.708947078782897,
+      "learning_rate": 1.0134005730618922e-06,
+      "loss": 0.2384,
+      "step": 12720
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.027054384171033,
+      "learning_rate": 1.012785947186397e-06,
+      "loss": 0.2579,
+      "step": 12721
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8035305780545818,
+      "learning_rate": 1.0121714867474936e-06,
+      "loss": 0.2394,
+      "step": 12722
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.134175423304704,
+      "learning_rate": 1.0115571917706795e-06,
+      "loss": 0.2453,
+      "step": 12723
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.70589306166117,
+      "learning_rate": 1.010943062281441e-06,
+      "loss": 0.2683,
+      "step": 12724
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6795361124172253,
+      "learning_rate": 1.0103290983052611e-06,
+      "loss": 0.4873,
+      "step": 12725
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.470026123728847,
+      "learning_rate": 1.0097152998676135e-06,
+      "loss": 0.2792,
+      "step": 12726
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.051263251291704,
+      "learning_rate": 1.009101666993964e-06,
+      "loss": 0.2619,
+      "step": 12727
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9010949523591494,
+      "learning_rate": 1.0084881997097768e-06,
+      "loss": 0.247,
+      "step": 12728
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.496350425292508,
+      "learning_rate": 1.0078748980405033e-06,
+      "loss": 0.261,
+      "step": 12729
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5563006687611018,
+      "learning_rate": 1.0072617620115904e-06,
+      "loss": 0.2591,
+      "step": 12730
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3271151319139887,
+      "learning_rate": 1.0066487916484803e-06,
+      "loss": 0.2479,
+      "step": 12731
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9768100611518893,
+      "learning_rate": 1.0060359869766028e-06,
+      "loss": 0.2716,
+      "step": 12732
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7156932569004555,
+      "learning_rate": 1.0054233480213881e-06,
+      "loss": 0.2751,
+      "step": 12733
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.698599394553698,
+      "learning_rate": 1.0048108748082536e-06,
+      "loss": 0.2638,
+      "step": 12734
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1124446176601013,
+      "learning_rate": 1.0041985673626115e-06,
+      "loss": 0.2565,
+      "step": 12735
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.215011893965315,
+      "learning_rate": 1.0035864257098687e-06,
+      "loss": 0.2579,
+      "step": 12736
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0804420682947686,
+      "learning_rate": 1.0029744498754225e-06,
+      "loss": 0.2562,
+      "step": 12737
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6035858514247574,
+      "learning_rate": 1.002362639884667e-06,
+      "loss": 0.2662,
+      "step": 12738
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.602340476526753,
+      "learning_rate": 1.0017509957629868e-06,
+      "loss": 0.2575,
+      "step": 12739
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.558152836341865,
+      "learning_rate": 1.0011395175357574e-06,
+      "loss": 0.2493,
+      "step": 12740
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.4520598093403265,
+      "learning_rate": 1.0005282052283539e-06,
+      "loss": 0.2486,
+      "step": 12741
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.725741189843362,
+      "learning_rate": 9.999170588661388e-07,
+      "loss": 0.2716,
+      "step": 12742
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.360730072590659,
+      "learning_rate": 9.99306078474469e-07,
+      "loss": 0.2407,
+      "step": 12743
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.755519875333957,
+      "learning_rate": 9.986952640786972e-07,
+      "loss": 0.2392,
+      "step": 12744
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.155347866804942,
+      "learning_rate": 9.980846157041645e-07,
+      "loss": 0.2585,
+      "step": 12745
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6009202595564043,
+      "learning_rate": 9.974741333762106e-07,
+      "loss": 0.2654,
+      "step": 12746
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.64625703476425,
+      "learning_rate": 9.968638171201644e-07,
+      "loss": 0.2722,
+      "step": 12747
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4761751049718286,
+      "learning_rate": 9.96253666961347e-07,
+      "loss": 0.2639,
+      "step": 12748
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7387890736189626,
+      "learning_rate": 9.956436829250782e-07,
+      "loss": 0.2654,
+      "step": 12749
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.40965657231041,
+      "learning_rate": 9.950338650366659e-07,
+      "loss": 0.2449,
+      "step": 12750
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7228430507127512,
+      "learning_rate": 9.944242133214098e-07,
+      "loss": 0.2585,
+      "step": 12751
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0072685071796164,
+      "learning_rate": 9.938147278046083e-07,
+      "loss": 0.2517,
+      "step": 12752
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.765746681085658,
+      "learning_rate": 9.932054085115512e-07,
+      "loss": 0.2647,
+      "step": 12753
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.536878930923567,
+      "learning_rate": 9.925962554675185e-07,
+      "loss": 0.2676,
+      "step": 12754
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6169529515121037,
+      "learning_rate": 9.919872686977849e-07,
+      "loss": 0.2656,
+      "step": 12755
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7243509227023388,
+      "learning_rate": 9.913784482276167e-07,
+      "loss": 0.2635,
+      "step": 12756
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3668179612717561,
+      "learning_rate": 9.90769794082279e-07,
+      "loss": 0.2522,
+      "step": 12757
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9917937310395633,
+      "learning_rate": 9.901613062870236e-07,
+      "loss": 0.2597,
+      "step": 12758
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6579782239575436,
+      "learning_rate": 9.89552984867096e-07,
+      "loss": 0.2537,
+      "step": 12759
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3337970745103087,
+      "learning_rate": 9.889448298477388e-07,
+      "loss": 0.2597,
+      "step": 12760
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6246909970505483,
+      "learning_rate": 9.88336841254186e-07,
+      "loss": 0.488,
+      "step": 12761
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5945754365392033,
+      "learning_rate": 9.87729019111664e-07,
+      "loss": 0.2525,
+      "step": 12762
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 26.287682843088376,
+      "learning_rate": 9.871213634453908e-07,
+      "loss": 0.2531,
+      "step": 12763
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.913740426617599,
+      "learning_rate": 9.865138742805792e-07,
+      "loss": 0.2536,
+      "step": 12764
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.300508795335946,
+      "learning_rate": 9.859065516424365e-07,
+      "loss": 0.2608,
+      "step": 12765
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8217823616327986,
+      "learning_rate": 9.852993955561607e-07,
+      "loss": 0.272,
+      "step": 12766
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.057594120481516,
+      "learning_rate": 9.846924060469433e-07,
+      "loss": 0.2533,
+      "step": 12767
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.860080822210733,
+      "learning_rate": 9.840855831399694e-07,
+      "loss": 0.2516,
+      "step": 12768
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7091044263044803,
+      "learning_rate": 9.834789268604188e-07,
+      "loss": 0.2561,
+      "step": 12769
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.678264760509789,
+      "learning_rate": 9.828724372334624e-07,
+      "loss": 0.2418,
+      "step": 12770
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0376825477549967,
+      "learning_rate": 9.822661142842621e-07,
+      "loss": 0.2461,
+      "step": 12771
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6357489493609827,
+      "learning_rate": 9.816599580379783e-07,
+      "loss": 0.2875,
+      "step": 12772
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4977430951989361,
+      "learning_rate": 9.8105396851976e-07,
+      "loss": 0.2596,
+      "step": 12773
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.921853057194947,
+      "learning_rate": 9.8044814575475e-07,
+      "loss": 0.2389,
+      "step": 12774
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7556084874755709,
+      "learning_rate": 9.798424897680876e-07,
+      "loss": 0.2353,
+      "step": 12775
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7325476307067564,
+      "learning_rate": 9.79237000584899e-07,
+      "loss": 0.2584,
+      "step": 12776
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9852158607627954,
+      "learning_rate": 9.786316782303108e-07,
+      "loss": 0.2637,
+      "step": 12777
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.209972925321309,
+      "learning_rate": 9.78026522729436e-07,
+      "loss": 0.2469,
+      "step": 12778
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8084338101734356,
+      "learning_rate": 9.774215341073844e-07,
+      "loss": 0.2574,
+      "step": 12779
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7451656808315519,
+      "learning_rate": 9.76816712389259e-07,
+      "loss": 0.2579,
+      "step": 12780
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6661527500741184,
+      "learning_rate": 9.762120576001543e-07,
+      "loss": 0.2327,
+      "step": 12781
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.656309263103782,
+      "learning_rate": 9.756075697651573e-07,
+      "loss": 0.2496,
+      "step": 12782
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8303826124579643,
+      "learning_rate": 9.750032489093514e-07,
+      "loss": 0.2729,
+      "step": 12783
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.530801960260885,
+      "learning_rate": 9.743990950578087e-07,
+      "loss": 0.2436,
+      "step": 12784
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7175450632254168,
+      "learning_rate": 9.73795108235599e-07,
+      "loss": 0.2557,
+      "step": 12785
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8929034367106248,
+      "learning_rate": 9.731912884677814e-07,
+      "loss": 0.2563,
+      "step": 12786
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5587418258075241,
+      "learning_rate": 9.72587635779409e-07,
+      "loss": 0.4351,
+      "step": 12787
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.842594908282135,
+      "learning_rate": 9.719841501955296e-07,
+      "loss": 0.2666,
+      "step": 12788
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.128309565322895,
+      "learning_rate": 9.713808317411815e-07,
+      "loss": 0.2747,
+      "step": 12789
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7326631363657472,
+      "learning_rate": 9.707776804414e-07,
+      "loss": 0.2498,
+      "step": 12790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.05237976920696,
+      "learning_rate": 9.701746963212084e-07,
+      "loss": 0.2568,
+      "step": 12791
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.393489461025843,
+      "learning_rate": 9.695718794056258e-07,
+      "loss": 0.2575,
+      "step": 12792
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.3888296321583455,
+      "learning_rate": 9.689692297196657e-07,
+      "loss": 0.2639,
+      "step": 12793
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5725615619980446,
+      "learning_rate": 9.683667472883324e-07,
+      "loss": 0.2482,
+      "step": 12794
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.647815195484356,
+      "learning_rate": 9.677644321366226e-07,
+      "loss": 0.2362,
+      "step": 12795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.174720726874843,
+      "learning_rate": 9.671622842895294e-07,
+      "loss": 0.2453,
+      "step": 12796
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4792700718431802,
+      "learning_rate": 9.66560303772035e-07,
+      "loss": 0.2445,
+      "step": 12797
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5752754867579595,
+      "learning_rate": 9.65958490609119e-07,
+      "loss": 0.2413,
+      "step": 12798
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.7160303960197085,
+      "learning_rate": 9.653568448257504e-07,
+      "loss": 0.2587,
+      "step": 12799
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0086538354064927,
+      "learning_rate": 9.647553664468918e-07,
+      "loss": 0.2394,
+      "step": 12800
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9664307217398085,
+      "learning_rate": 9.641540554975015e-07,
+      "loss": 0.2393,
+      "step": 12801
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.615912294402693,
+      "learning_rate": 9.635529120025282e-07,
+      "loss": 0.2488,
+      "step": 12802
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.917978411956469,
+      "learning_rate": 9.629519359869121e-07,
+      "loss": 0.2617,
+      "step": 12803
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1349444477823143,
+      "learning_rate": 9.623511274755914e-07,
+      "loss": 0.2725,
+      "step": 12804
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2862184235505376,
+      "learning_rate": 9.617504864934956e-07,
+      "loss": 0.2523,
+      "step": 12805
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6156777265235944,
+      "learning_rate": 9.611500130655443e-07,
+      "loss": 0.2493,
+      "step": 12806
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8710067540421056,
+      "learning_rate": 9.60549707216653e-07,
+      "loss": 0.2344,
+      "step": 12807
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0947206449001587,
+      "learning_rate": 9.599495689717276e-07,
+      "loss": 0.248,
+      "step": 12808
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.667282840868198,
+      "learning_rate": 9.593495983556723e-07,
+      "loss": 0.2886,
+      "step": 12809
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5302462702039108,
+      "learning_rate": 9.58749795393379e-07,
+      "loss": 0.2506,
+      "step": 12810
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.076053237357009,
+      "learning_rate": 9.581501601097332e-07,
+      "loss": 0.2556,
+      "step": 12811
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.006810597267588,
+      "learning_rate": 9.57550692529617e-07,
+      "loss": 0.2599,
+      "step": 12812
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.172004393315114,
+      "learning_rate": 9.569513926779033e-07,
+      "loss": 0.2536,
+      "step": 12813
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.770538705483302,
+      "learning_rate": 9.563522605794579e-07,
+      "loss": 0.2527,
+      "step": 12814
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.051092054586464,
+      "learning_rate": 9.557532962591398e-07,
+      "loss": 0.2669,
+      "step": 12815
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1166252046240914,
+      "learning_rate": 9.551544997417995e-07,
+      "loss": 0.2507,
+      "step": 12816
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1193775209546764,
+      "learning_rate": 9.545558710522844e-07,
+      "loss": 0.2492,
+      "step": 12817
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.575697964565946,
+      "learning_rate": 9.53957410215432e-07,
+      "loss": 0.2512,
+      "step": 12818
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6749766983147105,
+      "learning_rate": 9.533591172560714e-07,
+      "loss": 0.2675,
+      "step": 12819
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2145755436199703,
+      "learning_rate": 9.527609921990294e-07,
+      "loss": 0.2808,
+      "step": 12820
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2916494501173883,
+      "learning_rate": 9.521630350691235e-07,
+      "loss": 0.2596,
+      "step": 12821
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6229630033286424,
+      "learning_rate": 9.515652458911629e-07,
+      "loss": 0.4291,
+      "step": 12822
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3131772122071124,
+      "learning_rate": 9.509676246899513e-07,
+      "loss": 0.2415,
+      "step": 12823
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8486509994036227,
+      "learning_rate": 9.503701714902836e-07,
+      "loss": 0.2713,
+      "step": 12824
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7387775178104907,
+      "learning_rate": 9.497728863169514e-07,
+      "loss": 0.2508,
+      "step": 12825
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8702734825778233,
+      "learning_rate": 9.491757691947367e-07,
+      "loss": 0.2491,
+      "step": 12826
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7961841237914369,
+      "learning_rate": 9.485788201484125e-07,
+      "loss": 0.2639,
+      "step": 12827
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.576440342154987,
+      "learning_rate": 9.479820392027494e-07,
+      "loss": 0.4522,
+      "step": 12828
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.179496433329002,
+      "learning_rate": 9.473854263825105e-07,
+      "loss": 0.2623,
+      "step": 12829
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3879735818632433,
+      "learning_rate": 9.467889817124481e-07,
+      "loss": 0.2404,
+      "step": 12830
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.22086572123504,
+      "learning_rate": 9.461927052173097e-07,
+      "loss": 0.2447,
+      "step": 12831
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.394871369895212,
+      "learning_rate": 9.455965969218356e-07,
+      "loss": 0.2446,
+      "step": 12832
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.12420794196283,
+      "learning_rate": 9.450006568507614e-07,
+      "loss": 0.255,
+      "step": 12833
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0344644144945154,
+      "learning_rate": 9.444048850288112e-07,
+      "loss": 0.2411,
+      "step": 12834
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.399867326488132,
+      "learning_rate": 9.43809281480707e-07,
+      "loss": 0.2623,
+      "step": 12835
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8563141513055275,
+      "learning_rate": 9.432138462311591e-07,
+      "loss": 0.273,
+      "step": 12836
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.431892728930356,
+      "learning_rate": 9.426185793048754e-07,
+      "loss": 0.2535,
+      "step": 12837
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.930204766985941,
+      "learning_rate": 9.420234807265538e-07,
+      "loss": 0.2507,
+      "step": 12838
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1713764203900605,
+      "learning_rate": 9.414285505208859e-07,
+      "loss": 0.2457,
+      "step": 12839
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9472027670717913,
+      "learning_rate": 9.408337887125546e-07,
+      "loss": 0.2546,
+      "step": 12840
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.644738162379368,
+      "learning_rate": 9.402391953262396e-07,
+      "loss": 0.2392,
+      "step": 12841
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6318367312835584,
+      "learning_rate": 9.396447703866124e-07,
+      "loss": 0.2492,
+      "step": 12842
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7337807755938024,
+      "learning_rate": 9.390505139183359e-07,
+      "loss": 0.2597,
+      "step": 12843
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.795926326609472,
+      "learning_rate": 9.384564259460655e-07,
+      "loss": 0.2571,
+      "step": 12844
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3201580163658386,
+      "learning_rate": 9.378625064944529e-07,
+      "loss": 0.2558,
+      "step": 12845
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.997465048882375,
+      "learning_rate": 9.372687555881405e-07,
+      "loss": 0.2662,
+      "step": 12846
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0019004629313737,
+      "learning_rate": 9.366751732517632e-07,
+      "loss": 0.2517,
+      "step": 12847
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.361205174224972,
+      "learning_rate": 9.360817595099491e-07,
+      "loss": 0.2498,
+      "step": 12848
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6482585744755953,
+      "learning_rate": 9.354885143873216e-07,
+      "loss": 0.2615,
+      "step": 12849
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5524535099587276,
+      "learning_rate": 9.348954379084957e-07,
+      "loss": 0.2543,
+      "step": 12850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1563532493672573,
+      "learning_rate": 9.343025300980791e-07,
+      "loss": 0.2533,
+      "step": 12851
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5901668742846251,
+      "learning_rate": 9.337097909806703e-07,
+      "loss": 0.4666,
+      "step": 12852
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0350044462128816,
+      "learning_rate": 9.331172205808659e-07,
+      "loss": 0.2605,
+      "step": 12853
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8035053011714821,
+      "learning_rate": 9.325248189232521e-07,
+      "loss": 0.2636,
+      "step": 12854
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3109480430631197,
+      "learning_rate": 9.319325860324069e-07,
+      "loss": 0.2425,
+      "step": 12855
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.790154375725121,
+      "learning_rate": 9.313405219329041e-07,
+      "loss": 0.2624,
+      "step": 12856
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9340313613188451,
+      "learning_rate": 9.307486266493109e-07,
+      "loss": 0.2546,
+      "step": 12857
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.552539089425548,
+      "learning_rate": 9.301569002061855e-07,
+      "loss": 0.2468,
+      "step": 12858
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.235652719347328,
+      "learning_rate": 9.295653426280793e-07,
+      "loss": 0.245,
+      "step": 12859
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5643757271303158,
+      "learning_rate": 9.289739539395354e-07,
+      "loss": 0.2407,
+      "step": 12860
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.573721095433635,
+      "learning_rate": 9.283827341650942e-07,
+      "loss": 0.2391,
+      "step": 12861
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8385291457680781,
+      "learning_rate": 9.277916833292855e-07,
+      "loss": 0.2621,
+      "step": 12862
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.895979292505263,
+      "learning_rate": 9.272008014566314e-07,
+      "loss": 0.2624,
+      "step": 12863
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2921231198056478,
+      "learning_rate": 9.266100885716506e-07,
+      "loss": 0.2464,
+      "step": 12864
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.461850035186532,
+      "learning_rate": 9.260195446988535e-07,
+      "loss": 0.2441,
+      "step": 12865
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6173435668297256,
+      "learning_rate": 9.254291698627416e-07,
+      "loss": 0.249,
+      "step": 12866
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.168350408546128,
+      "learning_rate": 9.248389640878109e-07,
+      "loss": 0.2275,
+      "step": 12867
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2547508996488923,
+      "learning_rate": 9.242489273985483e-07,
+      "loss": 0.2603,
+      "step": 12868
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5311450365000812,
+      "learning_rate": 9.236590598194384e-07,
+      "loss": 0.2443,
+      "step": 12869
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.687600126500834,
+      "learning_rate": 9.230693613749547e-07,
+      "loss": 0.2647,
+      "step": 12870
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3168067255507325,
+      "learning_rate": 9.224798320895634e-07,
+      "loss": 0.24,
+      "step": 12871
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.99266560903549,
+      "learning_rate": 9.218904719877258e-07,
+      "loss": 0.2475,
+      "step": 12872
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5829464029758733,
+      "learning_rate": 9.213012810938976e-07,
+      "loss": 0.2483,
+      "step": 12873
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9153073565322267,
+      "learning_rate": 9.207122594325241e-07,
+      "loss": 0.2744,
+      "step": 12874
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7244034798619865,
+      "learning_rate": 9.201234070280446e-07,
+      "loss": 0.2507,
+      "step": 12875
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0440624380199686,
+      "learning_rate": 9.195347239048902e-07,
+      "loss": 0.2725,
+      "step": 12876
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5612430259333051,
+      "learning_rate": 9.189462100874891e-07,
+      "loss": 0.4663,
+      "step": 12877
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 41.09160100313767,
+      "learning_rate": 9.183578656002584e-07,
+      "loss": 0.2568,
+      "step": 12878
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1080926671363964,
+      "learning_rate": 9.177696904676086e-07,
+      "loss": 0.279,
+      "step": 12879
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5491718013748426,
+      "learning_rate": 9.171816847139447e-07,
+      "loss": 0.2621,
+      "step": 12880
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5374963144072058,
+      "learning_rate": 9.165938483636666e-07,
+      "loss": 0.242,
+      "step": 12881
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8972797521043407,
+      "learning_rate": 9.160061814411625e-07,
+      "loss": 0.276,
+      "step": 12882
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3082568587378285,
+      "learning_rate": 9.154186839708157e-07,
+      "loss": 0.2648,
+      "step": 12883
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.782943228581823,
+      "learning_rate": 9.148313559770011e-07,
+      "loss": 0.261,
+      "step": 12884
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6283344357888913,
+      "learning_rate": 9.142441974840915e-07,
+      "loss": 0.4548,
+      "step": 12885
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5597703822869617,
+      "learning_rate": 9.136572085164458e-07,
+      "loss": 0.4444,
+      "step": 12886
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.304495916707265,
+      "learning_rate": 9.130703890984222e-07,
+      "loss": 0.2518,
+      "step": 12887
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.488735308124743,
+      "learning_rate": 9.124837392543656e-07,
+      "loss": 0.2576,
+      "step": 12888
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6975954867493765,
+      "learning_rate": 9.118972590086206e-07,
+      "loss": 0.2461,
+      "step": 12889
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4664348549370394,
+      "learning_rate": 9.113109483855193e-07,
+      "loss": 0.2495,
+      "step": 12890
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4709030237841163,
+      "learning_rate": 9.107248074093894e-07,
+      "loss": 0.2503,
+      "step": 12891
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.001523179076291,
+      "learning_rate": 9.101388361045488e-07,
+      "loss": 0.2319,
+      "step": 12892
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.130189121210011,
+      "learning_rate": 9.095530344953141e-07,
+      "loss": 0.2468,
+      "step": 12893
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3774295368187583,
+      "learning_rate": 9.08967402605988e-07,
+      "loss": 0.2412,
+      "step": 12894
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3103110629274615,
+      "learning_rate": 9.083819404608724e-07,
+      "loss": 0.2583,
+      "step": 12895
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.244186611790629,
+      "learning_rate": 9.07796648084256e-07,
+      "loss": 0.2593,
+      "step": 12896
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6191162625720816,
+      "learning_rate": 9.072115255004266e-07,
+      "loss": 0.4698,
+      "step": 12897
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.615867861107461,
+      "learning_rate": 9.066265727336604e-07,
+      "loss": 0.2664,
+      "step": 12898
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8105279253481417,
+      "learning_rate": 9.060417898082285e-07,
+      "loss": 0.2538,
+      "step": 12899
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3438830026991466,
+      "learning_rate": 9.054571767483932e-07,
+      "loss": 0.26,
+      "step": 12900
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.057403257654064,
+      "learning_rate": 9.048727335784124e-07,
+      "loss": 0.2596,
+      "step": 12901
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3587940448726448,
+      "learning_rate": 9.042884603225372e-07,
+      "loss": 0.2396,
+      "step": 12902
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7891339199649128,
+      "learning_rate": 9.03704357005008e-07,
+      "loss": 0.2504,
+      "step": 12903
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.539726229067604,
+      "learning_rate": 9.031204236500601e-07,
+      "loss": 0.2624,
+      "step": 12904
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7644042066418655,
+      "learning_rate": 9.025366602819235e-07,
+      "loss": 0.2628,
+      "step": 12905
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4714732485959336,
+      "learning_rate": 9.019530669248195e-07,
+      "loss": 0.2528,
+      "step": 12906
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.682278135449895,
+      "learning_rate": 9.013696436029607e-07,
+      "loss": 0.2503,
+      "step": 12907
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6022755323694893,
+      "learning_rate": 9.007863903405551e-07,
+      "loss": 0.2609,
+      "step": 12908
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7965323724809963,
+      "learning_rate": 9.002033071618027e-07,
+      "loss": 0.267,
+      "step": 12909
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.890080830376393,
+      "learning_rate": 8.996203940908982e-07,
+      "loss": 0.2665,
+      "step": 12910
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6391509388317074,
+      "learning_rate": 8.990376511520272e-07,
+      "loss": 0.2571,
+      "step": 12911
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.438269915323138,
+      "learning_rate": 8.984550783693663e-07,
+      "loss": 0.2434,
+      "step": 12912
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.484500330904699,
+      "learning_rate": 8.978726757670908e-07,
+      "loss": 0.2498,
+      "step": 12913
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6036576597089971,
+      "learning_rate": 8.972904433693646e-07,
+      "loss": 0.4462,
+      "step": 12914
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.653183563269335,
+      "learning_rate": 8.967083812003446e-07,
+      "loss": 0.2505,
+      "step": 12915
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.416421650311631,
+      "learning_rate": 8.961264892841798e-07,
+      "loss": 0.2437,
+      "step": 12916
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8761232962721022,
+      "learning_rate": 8.955447676450191e-07,
+      "loss": 0.2705,
+      "step": 12917
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.757939120259039,
+      "learning_rate": 8.94963216306996e-07,
+      "loss": 0.258,
+      "step": 12918
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4166521640306378,
+      "learning_rate": 8.943818352942401e-07,
+      "loss": 0.2511,
+      "step": 12919
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5323323393674855,
+      "learning_rate": 8.938006246308734e-07,
+      "loss": 0.2661,
+      "step": 12920
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.422876001062616,
+      "learning_rate": 8.932195843410135e-07,
+      "loss": 0.2556,
+      "step": 12921
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6779079539753463,
+      "learning_rate": 8.926387144487675e-07,
+      "loss": 0.2487,
+      "step": 12922
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8720471401606082,
+      "learning_rate": 8.920580149782354e-07,
+      "loss": 0.2805,
+      "step": 12923
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9861607351680166,
+      "learning_rate": 8.914774859535131e-07,
+      "loss": 0.269,
+      "step": 12924
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.306090823850083,
+      "learning_rate": 8.908971273986883e-07,
+      "loss": 0.2541,
+      "step": 12925
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7058781362913975,
+      "learning_rate": 8.903169393378403e-07,
+      "loss": 0.249,
+      "step": 12926
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.717577300853418,
+      "learning_rate": 8.897369217950424e-07,
+      "loss": 0.2565,
+      "step": 12927
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6589760285420476,
+      "learning_rate": 8.891570747943584e-07,
+      "loss": 0.2591,
+      "step": 12928
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.305319716338477,
+      "learning_rate": 8.885773983598512e-07,
+      "loss": 0.283,
+      "step": 12929
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.79531730809761,
+      "learning_rate": 8.879978925155697e-07,
+      "loss": 0.2434,
+      "step": 12930
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0361184049130663,
+      "learning_rate": 8.874185572855581e-07,
+      "loss": 0.2496,
+      "step": 12931
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.442694204864611,
+      "learning_rate": 8.868393926938557e-07,
+      "loss": 0.2603,
+      "step": 12932
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8625723943847077,
+      "learning_rate": 8.862603987644942e-07,
+      "loss": 0.2457,
+      "step": 12933
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4946677941746433,
+      "learning_rate": 8.856815755214953e-07,
+      "loss": 0.2475,
+      "step": 12934
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2963165714071727,
+      "learning_rate": 8.851029229888752e-07,
+      "loss": 0.2418,
+      "step": 12935
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.583915943443398,
+      "learning_rate": 8.845244411906429e-07,
+      "loss": 0.4955,
+      "step": 12936
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3722575964806563,
+      "learning_rate": 8.839461301508028e-07,
+      "loss": 0.2765,
+      "step": 12937
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2057989099379087,
+      "learning_rate": 8.833679898933472e-07,
+      "loss": 0.2733,
+      "step": 12938
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.690125514832156,
+      "learning_rate": 8.827900204422674e-07,
+      "loss": 0.2689,
+      "step": 12939
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8825434705305653,
+      "learning_rate": 8.822122218215406e-07,
+      "loss": 0.2667,
+      "step": 12940
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6403611439779844,
+      "learning_rate": 8.816345940551446e-07,
+      "loss": 0.2476,
+      "step": 12941
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.505775895939719,
+      "learning_rate": 8.81057137167044e-07,
+      "loss": 0.2499,
+      "step": 12942
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2747582625413525,
+      "learning_rate": 8.80479851181199e-07,
+      "loss": 0.2676,
+      "step": 12943
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8306581193431004,
+      "learning_rate": 8.799027361215607e-07,
+      "loss": 0.2467,
+      "step": 12944
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9694901453239277,
+      "learning_rate": 8.793257920120774e-07,
+      "loss": 0.272,
+      "step": 12945
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8756997540171922,
+      "learning_rate": 8.78749018876685e-07,
+      "loss": 0.2542,
+      "step": 12946
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.621652512605524,
+      "learning_rate": 8.781724167393168e-07,
+      "loss": 0.2575,
+      "step": 12947
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6949599893835643,
+      "learning_rate": 8.775959856238953e-07,
+      "loss": 0.2553,
+      "step": 12948
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.237138056523293,
+      "learning_rate": 8.770197255543395e-07,
+      "loss": 0.2743,
+      "step": 12949
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0481638052120377,
+      "learning_rate": 8.764436365545592e-07,
+      "loss": 0.2534,
+      "step": 12950
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0585499310221427,
+      "learning_rate": 8.75867718648456e-07,
+      "loss": 0.2583,
+      "step": 12951
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.588495449316286,
+      "learning_rate": 8.752919718599256e-07,
+      "loss": 0.2605,
+      "step": 12952
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.775181838047722,
+      "learning_rate": 8.747163962128574e-07,
+      "loss": 0.2455,
+      "step": 12953
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.716596625811185,
+      "learning_rate": 8.741409917311344e-07,
+      "loss": 0.2475,
+      "step": 12954
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9314409012285993,
+      "learning_rate": 8.735657584386297e-07,
+      "loss": 0.2515,
+      "step": 12955
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.929806241612521,
+      "learning_rate": 8.729906963592105e-07,
+      "loss": 0.2601,
+      "step": 12956
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.971862189547834,
+      "learning_rate": 8.724158055167386e-07,
+      "loss": 0.2582,
+      "step": 12957
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9891388691416703,
+      "learning_rate": 8.718410859350663e-07,
+      "loss": 0.2402,
+      "step": 12958
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.022258753997461,
+      "learning_rate": 8.712665376380403e-07,
+      "loss": 0.2615,
+      "step": 12959
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5026717591833973,
+      "learning_rate": 8.706921606494973e-07,
+      "loss": 0.2603,
+      "step": 12960
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.121274332580468,
+      "learning_rate": 8.701179549932709e-07,
+      "loss": 0.2524,
+      "step": 12961
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.721103342967426,
+      "learning_rate": 8.695439206931877e-07,
+      "loss": 0.26,
+      "step": 12962
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4580206751263725,
+      "learning_rate": 8.689700577730637e-07,
+      "loss": 0.2503,
+      "step": 12963
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6261407089771402,
+      "learning_rate": 8.68396366256708e-07,
+      "loss": 0.2521,
+      "step": 12964
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.315283498841684,
+      "learning_rate": 8.678228461679272e-07,
+      "loss": 0.287,
+      "step": 12965
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.870100406785817,
+      "learning_rate": 8.672494975305157e-07,
+      "loss": 0.2541,
+      "step": 12966
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8365201530208666,
+      "learning_rate": 8.666763203682637e-07,
+      "loss": 0.2448,
+      "step": 12967
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.331207549065846,
+      "learning_rate": 8.661033147049496e-07,
+      "loss": 0.2659,
+      "step": 12968
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6157615990454055,
+      "learning_rate": 8.655304805643549e-07,
+      "loss": 0.4566,
+      "step": 12969
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8008006782529928,
+      "learning_rate": 8.649578179702434e-07,
+      "loss": 0.2544,
+      "step": 12970
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.532257319195622,
+      "learning_rate": 8.64385326946377e-07,
+      "loss": 0.2559,
+      "step": 12971
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.867931380705508,
+      "learning_rate": 8.638130075165079e-07,
+      "loss": 0.273,
+      "step": 12972
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1692684286071247,
+      "learning_rate": 8.632408597043851e-07,
+      "loss": 0.2718,
+      "step": 12973
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.268772053840724,
+      "learning_rate": 8.626688835337471e-07,
+      "loss": 0.2619,
+      "step": 12974
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6911422353850587,
+      "learning_rate": 8.620970790283251e-07,
+      "loss": 0.2482,
+      "step": 12975
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.855496193606545,
+      "learning_rate": 8.615254462118427e-07,
+      "loss": 0.2643,
+      "step": 12976
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.318438943732563,
+      "learning_rate": 8.609539851080234e-07,
+      "loss": 0.2607,
+      "step": 12977
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7953197647700785,
+      "learning_rate": 8.603826957405742e-07,
+      "loss": 0.2741,
+      "step": 12978
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.143269425400992,
+      "learning_rate": 8.598115781332006e-07,
+      "loss": 0.2431,
+      "step": 12979
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.362342912598155,
+      "learning_rate": 8.592406323095964e-07,
+      "loss": 0.2635,
+      "step": 12980
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5253350567382885,
+      "learning_rate": 8.58669858293455e-07,
+      "loss": 0.422,
+      "step": 12981
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.248639521149241,
+      "learning_rate": 8.580992561084567e-07,
+      "loss": 0.2567,
+      "step": 12982
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.737356727789742,
+      "learning_rate": 8.575288257782755e-07,
+      "loss": 0.2581,
+      "step": 12983
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8994868493645913,
+      "learning_rate": 8.569585673265818e-07,
+      "loss": 0.2729,
+      "step": 12984
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7322243587602308,
+      "learning_rate": 8.563884807770373e-07,
+      "loss": 0.2507,
+      "step": 12985
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.5259055541114055,
+      "learning_rate": 8.558185661532942e-07,
+      "loss": 0.267,
+      "step": 12986
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.718643461005986,
+      "learning_rate": 8.552488234789996e-07,
+      "loss": 0.2719,
+      "step": 12987
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.129722357225122,
+      "learning_rate": 8.546792527777925e-07,
+      "loss": 0.278,
+      "step": 12988
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4463680420732525,
+      "learning_rate": 8.541098540733067e-07,
+      "loss": 0.2546,
+      "step": 12989
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.257847127807278,
+      "learning_rate": 8.535406273891678e-07,
+      "loss": 0.2536,
+      "step": 12990
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9405059302079728,
+      "learning_rate": 8.529715727489912e-07,
+      "loss": 0.2497,
+      "step": 12991
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4784141342344626,
+      "learning_rate": 8.524026901763905e-07,
+      "loss": 0.2645,
+      "step": 12992
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9415160324746514,
+      "learning_rate": 8.518339796949704e-07,
+      "loss": 0.2873,
+      "step": 12993
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8877086490397215,
+      "learning_rate": 8.512654413283261e-07,
+      "loss": 0.2462,
+      "step": 12994
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.172067567216997,
+      "learning_rate": 8.506970751000477e-07,
+      "loss": 0.2547,
+      "step": 12995
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.178190644279585,
+      "learning_rate": 8.501288810337166e-07,
+      "loss": 0.2644,
+      "step": 12996
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.999811991749933,
+      "learning_rate": 8.495608591529103e-07,
+      "loss": 0.2413,
+      "step": 12997
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7508726125273202,
+      "learning_rate": 8.489930094811949e-07,
+      "loss": 0.256,
+      "step": 12998
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.424967339738821,
+      "learning_rate": 8.484253320421337e-07,
+      "loss": 0.2648,
+      "step": 12999
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.644868330238792,
+      "learning_rate": 8.478578268592779e-07,
+      "loss": 0.2572,
+      "step": 13000
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.041706935467664,
+      "learning_rate": 8.472904939561776e-07,
+      "loss": 0.2583,
+      "step": 13001
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.989928503212847,
+      "learning_rate": 8.467233333563707e-07,
+      "loss": 0.2508,
+      "step": 13002
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.300913008024875,
+      "learning_rate": 8.461563450833898e-07,
+      "loss": 0.2443,
+      "step": 13003
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.046459539029601,
+      "learning_rate": 8.455895291607585e-07,
+      "loss": 0.2885,
+      "step": 13004
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6342500234576756,
+      "learning_rate": 8.450228856119974e-07,
+      "loss": 0.2371,
+      "step": 13005
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8703494527860083,
+      "learning_rate": 8.444564144606177e-07,
+      "loss": 0.2566,
+      "step": 13006
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3922039419797838,
+      "learning_rate": 8.438901157301221e-07,
+      "loss": 0.2384,
+      "step": 13007
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.7703336374296015,
+      "learning_rate": 8.433239894440071e-07,
+      "loss": 0.2472,
+      "step": 13008
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9150429948459586,
+      "learning_rate": 8.427580356257636e-07,
+      "loss": 0.2544,
+      "step": 13009
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5940597519353604,
+      "learning_rate": 8.421922542988736e-07,
+      "loss": 0.4712,
+      "step": 13010
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8368424429118755,
+      "learning_rate": 8.416266454868122e-07,
+      "loss": 0.2502,
+      "step": 13011
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8532035972685212,
+      "learning_rate": 8.410612092130455e-07,
+      "loss": 0.2496,
+      "step": 13012
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.388364709508336,
+      "learning_rate": 8.40495945501037e-07,
+      "loss": 0.2618,
+      "step": 13013
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.132774989207916,
+      "learning_rate": 8.399308543742408e-07,
+      "loss": 0.2689,
+      "step": 13014
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.895250884264923,
+      "learning_rate": 8.393659358561024e-07,
+      "loss": 0.2568,
+      "step": 13015
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0614659051887556,
+      "learning_rate": 8.388011899700605e-07,
+      "loss": 0.2532,
+      "step": 13016
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7982387345417716,
+      "learning_rate": 8.382366167395495e-07,
+      "loss": 0.2294,
+      "step": 13017
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8196576703216867,
+      "learning_rate": 8.37672216187993e-07,
+      "loss": 0.2392,
+      "step": 13018
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9383974662274779,
+      "learning_rate": 8.3710798833881e-07,
+      "loss": 0.2627,
+      "step": 13019
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.959002470919046,
+      "learning_rate": 8.36543933215409e-07,
+      "loss": 0.26,
+      "step": 13020
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.624990470072323,
+      "learning_rate": 8.359800508411958e-07,
+      "loss": 0.2714,
+      "step": 13021
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8743607113147993,
+      "learning_rate": 8.354163412395671e-07,
+      "loss": 0.2504,
+      "step": 13022
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5898234152130922,
+      "learning_rate": 8.348528044339122e-07,
+      "loss": 0.4621,
+      "step": 13023
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3573439550623307,
+      "learning_rate": 8.34289440447611e-07,
+      "loss": 0.2858,
+      "step": 13024
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.501002413577344,
+      "learning_rate": 8.33726249304041e-07,
+      "loss": 0.2591,
+      "step": 13025
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8653577441948457,
+      "learning_rate": 8.331632310265691e-07,
+      "loss": 0.2568,
+      "step": 13026
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.172463980478453,
+      "learning_rate": 8.326003856385561e-07,
+      "loss": 0.2819,
+      "step": 13027
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0252902553124086,
+      "learning_rate": 8.320377131633545e-07,
+      "loss": 0.2463,
+      "step": 13028
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.009323889116786,
+      "learning_rate": 8.314752136243104e-07,
+      "loss": 0.252,
+      "step": 13029
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5814904848572607,
+      "learning_rate": 8.309128870447658e-07,
+      "loss": 0.4545,
+      "step": 13030
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6858426417017702,
+      "learning_rate": 8.303507334480499e-07,
+      "loss": 0.2535,
+      "step": 13031
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.342394518394939,
+      "learning_rate": 8.297887528574877e-07,
+      "loss": 0.2461,
+      "step": 13032
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.292131029782006,
+      "learning_rate": 8.292269452963981e-07,
+      "loss": 0.2667,
+      "step": 13033
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4481962584993637,
+      "learning_rate": 8.286653107880904e-07,
+      "loss": 0.2496,
+      "step": 13034
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5152240136766413,
+      "learning_rate": 8.281038493558674e-07,
+      "loss": 0.2593,
+      "step": 13035
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.622225821286546,
+      "learning_rate": 8.275425610230269e-07,
+      "loss": 0.2631,
+      "step": 13036
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6396270837493994,
+      "learning_rate": 8.269814458128556e-07,
+      "loss": 0.2505,
+      "step": 13037
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4101603639984648,
+      "learning_rate": 8.264205037486367e-07,
+      "loss": 0.2561,
+      "step": 13038
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.516095248356477,
+      "learning_rate": 8.258597348536452e-07,
+      "loss": 0.2655,
+      "step": 13039
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.599263168900209,
+      "learning_rate": 8.252991391511455e-07,
+      "loss": 0.2586,
+      "step": 13040
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9634882322785763,
+      "learning_rate": 8.247387166644011e-07,
+      "loss": 0.2556,
+      "step": 13041
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.347372733686193,
+      "learning_rate": 8.24178467416663e-07,
+      "loss": 0.2784,
+      "step": 13042
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.183040854119145,
+      "learning_rate": 8.236183914311769e-07,
+      "loss": 0.2704,
+      "step": 13043
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.055340739105419,
+      "learning_rate": 8.230584887311826e-07,
+      "loss": 0.2661,
+      "step": 13044
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8134922061435854,
+      "learning_rate": 8.224987593399098e-07,
+      "loss": 0.2686,
+      "step": 13045
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.811205474303526,
+      "learning_rate": 8.219392032805846e-07,
+      "loss": 0.2595,
+      "step": 13046
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.203829807712511,
+      "learning_rate": 8.213798205764228e-07,
+      "loss": 0.2692,
+      "step": 13047
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.807074794904907,
+      "learning_rate": 8.208206112506329e-07,
+      "loss": 0.2538,
+      "step": 13048
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5836032842498233,
+      "learning_rate": 8.202615753264204e-07,
+      "loss": 0.3067,
+      "step": 13049
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.563150443366848,
+      "learning_rate": 8.19702712826978e-07,
+      "loss": 0.2932,
+      "step": 13050
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.188312959201578,
+      "learning_rate": 8.191440237754961e-07,
+      "loss": 0.2533,
+      "step": 13051
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3865614737569423,
+      "learning_rate": 8.185855081951538e-07,
+      "loss": 0.2511,
+      "step": 13052
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8887032677106688,
+      "learning_rate": 8.180271661091266e-07,
+      "loss": 0.252,
+      "step": 13053
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.8215593645014145,
+      "learning_rate": 8.174689975405809e-07,
+      "loss": 0.241,
+      "step": 13054
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4799180184625094,
+      "learning_rate": 8.169110025126747e-07,
+      "loss": 0.2423,
+      "step": 13055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0505103696824696,
+      "learning_rate": 8.163531810485603e-07,
+      "loss": 0.2702,
+      "step": 13056
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.633765709039043,
+      "learning_rate": 8.157955331713846e-07,
+      "loss": 0.4643,
+      "step": 13057
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6442694501102397,
+      "learning_rate": 8.15238058904283e-07,
+      "loss": 0.2624,
+      "step": 13058
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4342128978916855,
+      "learning_rate": 8.146807582703886e-07,
+      "loss": 0.2529,
+      "step": 13059
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.573319810450124,
+      "learning_rate": 8.141236312928219e-07,
+      "loss": 0.2547,
+      "step": 13060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7221318023916532,
+      "learning_rate": 8.135666779947027e-07,
+      "loss": 0.2413,
+      "step": 13061
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.087675320801948,
+      "learning_rate": 8.13009898399137e-07,
+      "loss": 0.2439,
+      "step": 13062
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1914829503325612,
+      "learning_rate": 8.124532925292283e-07,
+      "loss": 0.2403,
+      "step": 13063
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3682928785855832,
+      "learning_rate": 8.118968604080684e-07,
+      "loss": 0.261,
+      "step": 13064
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.031624591214938,
+      "learning_rate": 8.113406020587472e-07,
+      "loss": 0.2685,
+      "step": 13065
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.358590074475766,
+      "learning_rate": 8.107845175043455e-07,
+      "loss": 0.272,
+      "step": 13066
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4451503825110303,
+      "learning_rate": 8.102286067679354e-07,
+      "loss": 0.2553,
+      "step": 13067
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.508759541027896,
+      "learning_rate": 8.096728698725803e-07,
+      "loss": 0.2408,
+      "step": 13068
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.233493219583581,
+      "learning_rate": 8.091173068413427e-07,
+      "loss": 0.2453,
+      "step": 13069
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7521175580643524,
+      "learning_rate": 8.085619176972715e-07,
+      "loss": 0.2529,
+      "step": 13070
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9641078041129796,
+      "learning_rate": 8.080067024634109e-07,
+      "loss": 0.2446,
+      "step": 13071
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6226189982102704,
+      "learning_rate": 8.07451661162797e-07,
+      "loss": 0.469,
+      "step": 13072
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6846918995664295,
+      "learning_rate": 8.068967938184608e-07,
+      "loss": 0.2421,
+      "step": 13073
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.890165279115881,
+      "learning_rate": 8.063421004534256e-07,
+      "loss": 0.2569,
+      "step": 13074
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6052195605657091,
+      "learning_rate": 8.057875810907051e-07,
+      "loss": 0.2359,
+      "step": 13075
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.615479030906575,
+      "learning_rate": 8.052332357533066e-07,
+      "loss": 0.2479,
+      "step": 13076
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.929071088557361,
+      "learning_rate": 8.046790644642327e-07,
+      "loss": 0.249,
+      "step": 13077
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3559093829198097,
+      "learning_rate": 8.041250672464768e-07,
+      "loss": 0.2535,
+      "step": 13078
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4031842626611275,
+      "learning_rate": 8.035712441230237e-07,
+      "loss": 0.2788,
+      "step": 13079
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2772086345128029,
+      "learning_rate": 8.030175951168528e-07,
+      "loss": 0.2559,
+      "step": 13080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6154076525591489,
+      "learning_rate": 8.024641202509365e-07,
+      "loss": 0.4888,
+      "step": 13081
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9684861077151803,
+      "learning_rate": 8.019108195482406e-07,
+      "loss": 0.2489,
+      "step": 13082
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1455532813566687,
+      "learning_rate": 8.013576930317218e-07,
+      "loss": 0.2679,
+      "step": 13083
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9666094734323385,
+      "learning_rate": 8.008047407243285e-07,
+      "loss": 0.2671,
+      "step": 13084
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.4877253216716495,
+      "learning_rate": 8.002519626490063e-07,
+      "loss": 0.2542,
+      "step": 13085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1984824133815075,
+      "learning_rate": 7.996993588286894e-07,
+      "loss": 0.251,
+      "step": 13086
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.715523572971109,
+      "learning_rate": 7.991469292863058e-07,
+      "loss": 0.2515,
+      "step": 13087
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.76257404545504,
+      "learning_rate": 7.985946740447792e-07,
+      "loss": 0.2559,
+      "step": 13088
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7938664839444995,
+      "learning_rate": 7.980425931270203e-07,
+      "loss": 0.2837,
+      "step": 13089
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3795535571582775,
+      "learning_rate": 7.974906865559396e-07,
+      "loss": 0.2578,
+      "step": 13090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5783046593675997,
+      "learning_rate": 7.969389543544343e-07,
+      "loss": 0.4643,
+      "step": 13091
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0700533399711567,
+      "learning_rate": 7.963873965453961e-07,
+      "loss": 0.259,
+      "step": 13092
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.932886429506239,
+      "learning_rate": 7.958360131517123e-07,
+      "loss": 0.2496,
+      "step": 13093
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9467870566710295,
+      "learning_rate": 7.952848041962601e-07,
+      "loss": 0.2563,
+      "step": 13094
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5904858129912447,
+      "learning_rate": 7.947337697019087e-07,
+      "loss": 0.2476,
+      "step": 13095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7050276569152154,
+      "learning_rate": 7.941829096915237e-07,
+      "loss": 0.2524,
+      "step": 13096
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9404042332245237,
+      "learning_rate": 7.936322241879585e-07,
+      "loss": 0.2556,
+      "step": 13097
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4866373699453335,
+      "learning_rate": 7.930817132140656e-07,
+      "loss": 0.2457,
+      "step": 13098
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7794633701255003,
+      "learning_rate": 7.92531376792684e-07,
+      "loss": 0.2714,
+      "step": 13099
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.492381490675387,
+      "learning_rate": 7.919812149466483e-07,
+      "loss": 0.2514,
+      "step": 13100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3275529548383482,
+      "learning_rate": 7.914312276987873e-07,
+      "loss": 0.2514,
+      "step": 13101
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.607591987618546,
+      "learning_rate": 7.908814150719185e-07,
+      "loss": 0.2561,
+      "step": 13102
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.048912563356653,
+      "learning_rate": 7.903317770888574e-07,
+      "loss": 0.2599,
+      "step": 13103
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.764661385919167,
+      "learning_rate": 7.897823137724081e-07,
+      "loss": 0.2659,
+      "step": 13104
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.161276809576672,
+      "learning_rate": 7.892330251453672e-07,
+      "loss": 0.2393,
+      "step": 13105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6182133375116967,
+      "learning_rate": 7.886839112305288e-07,
+      "loss": 0.4546,
+      "step": 13106
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9272856699175505,
+      "learning_rate": 7.881349720506754e-07,
+      "loss": 0.2671,
+      "step": 13107
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.285158660255171,
+      "learning_rate": 7.875862076285812e-07,
+      "loss": 0.287,
+      "step": 13108
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6366069699449706,
+      "learning_rate": 7.870376179870187e-07,
+      "loss": 0.238,
+      "step": 13109
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 23.590969575871426,
+      "learning_rate": 7.864892031487476e-07,
+      "loss": 0.2653,
+      "step": 13110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.009571809747192,
+      "learning_rate": 7.859409631365245e-07,
+      "loss": 0.2476,
+      "step": 13111
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.366996876037401,
+      "learning_rate": 7.853928979730962e-07,
+      "loss": 0.2549,
+      "step": 13112
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0365944811854693,
+      "learning_rate": 7.848450076812008e-07,
+      "loss": 0.2719,
+      "step": 13113
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.413728089958753,
+      "learning_rate": 7.842972922835745e-07,
+      "loss": 0.2596,
+      "step": 13114
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4686232985740932,
+      "learning_rate": 7.837497518029419e-07,
+      "loss": 0.2396,
+      "step": 13115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.712527937302349,
+      "learning_rate": 7.832023862620192e-07,
+      "loss": 0.2501,
+      "step": 13116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7211641572292184,
+      "learning_rate": 7.826551956835193e-07,
+      "loss": 0.2667,
+      "step": 13117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8893581009564415,
+      "learning_rate": 7.821081800901475e-07,
+      "loss": 0.2509,
+      "step": 13118
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6012426350371194,
+      "learning_rate": 7.815613395045996e-07,
+      "loss": 0.2522,
+      "step": 13119
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0291694474029827,
+      "learning_rate": 7.810146739495638e-07,
+      "loss": 0.2445,
+      "step": 13120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7227236330250217,
+      "learning_rate": 7.804681834477223e-07,
+      "loss": 0.2576,
+      "step": 13121
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5701982959971131,
+      "learning_rate": 7.799218680217513e-07,
+      "loss": 0.4314,
+      "step": 13122
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.602939963796262,
+      "learning_rate": 7.793757276943181e-07,
+      "loss": 0.2579,
+      "step": 13123
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.744144377902834,
+      "learning_rate": 7.788297624880814e-07,
+      "loss": 0.2362,
+      "step": 13124
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3946643028272416,
+      "learning_rate": 7.782839724256952e-07,
+      "loss": 0.2598,
+      "step": 13125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4460012789736063,
+      "learning_rate": 7.777383575298069e-07,
+      "loss": 0.2554,
+      "step": 13126
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.449953303242268,
+      "learning_rate": 7.771929178230542e-07,
+      "loss": 0.2568,
+      "step": 13127
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4017882233989867,
+      "learning_rate": 7.766476533280659e-07,
+      "loss": 0.273,
+      "step": 13128
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3111915189622882,
+      "learning_rate": 7.761025640674696e-07,
+      "loss": 0.2588,
+      "step": 13129
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.916331245414365,
+      "learning_rate": 7.755576500638806e-07,
+      "loss": 0.273,
+      "step": 13130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.460799444511176,
+      "learning_rate": 7.75012911339908e-07,
+      "loss": 0.2489,
+      "step": 13131
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0456496495781993,
+      "learning_rate": 7.74468347918153e-07,
+      "loss": 0.2764,
+      "step": 13132
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6935012493677806,
+      "learning_rate": 7.739239598212111e-07,
+      "loss": 0.2417,
+      "step": 13133
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1028167821480643,
+      "learning_rate": 7.733797470716725e-07,
+      "loss": 0.2647,
+      "step": 13134
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4120097639929552,
+      "learning_rate": 7.728357096921152e-07,
+      "loss": 0.2461,
+      "step": 13135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8307900217480282,
+      "learning_rate": 7.722918477051112e-07,
+      "loss": 0.2509,
+      "step": 13136
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.174653470847996,
+      "learning_rate": 7.717481611332289e-07,
+      "loss": 0.2601,
+      "step": 13137
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6613766491342608,
+      "learning_rate": 7.712046499990255e-07,
+      "loss": 0.2677,
+      "step": 13138
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.22315979333206,
+      "learning_rate": 7.706613143250524e-07,
+      "loss": 0.2511,
+      "step": 13139
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6642103270753872,
+      "learning_rate": 7.701181541338526e-07,
+      "loss": 0.259,
+      "step": 13140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.5889709885115195,
+      "learning_rate": 7.695751694479636e-07,
+      "loss": 0.2491,
+      "step": 13141
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.247138147154396,
+      "learning_rate": 7.690323602899163e-07,
+      "loss": 0.2498,
+      "step": 13142
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5433199558546358,
+      "learning_rate": 7.684897266822311e-07,
+      "loss": 0.2634,
+      "step": 13143
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9310444110559235,
+      "learning_rate": 7.67947268647422e-07,
+      "loss": 0.2526,
+      "step": 13144
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2473337510819267,
+      "learning_rate": 7.67404986207999e-07,
+      "loss": 0.2774,
+      "step": 13145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7584321078236875,
+      "learning_rate": 7.668628793864607e-07,
+      "loss": 0.2491,
+      "step": 13146
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3861916248282093,
+      "learning_rate": 7.663209482052997e-07,
+      "loss": 0.2442,
+      "step": 13147
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.506798592293803,
+      "learning_rate": 7.657791926870034e-07,
+      "loss": 0.2767,
+      "step": 13148
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.35585626017092,
+      "learning_rate": 7.652376128540478e-07,
+      "loss": 0.2345,
+      "step": 13149
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4655352478852601,
+      "learning_rate": 7.646962087289073e-07,
+      "loss": 0.2452,
+      "step": 13150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5919262894747134,
+      "learning_rate": 7.641549803340431e-07,
+      "loss": 0.5069,
+      "step": 13151
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4528374372651491,
+      "learning_rate": 7.636139276919119e-07,
+      "loss": 0.2492,
+      "step": 13152
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2757614043041645,
+      "learning_rate": 7.630730508249639e-07,
+      "loss": 0.2659,
+      "step": 13153
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.702400540208499,
+      "learning_rate": 7.625323497556414e-07,
+      "loss": 0.2468,
+      "step": 13154
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.078352471027836,
+      "learning_rate": 7.619918245063768e-07,
+      "loss": 0.2396,
+      "step": 13155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5552855032131907,
+      "learning_rate": 7.614514750996005e-07,
+      "loss": 0.2879,
+      "step": 13156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6837207980809006,
+      "learning_rate": 7.609113015577291e-07,
+      "loss": 0.2469,
+      "step": 13157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.518276060545688,
+      "learning_rate": 7.60371303903179e-07,
+      "loss": 0.2697,
+      "step": 13158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.810011764493329,
+      "learning_rate": 7.598314821583542e-07,
+      "loss": 0.26,
+      "step": 13159
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5868077906950226,
+      "learning_rate": 7.592918363456509e-07,
+      "loss": 0.2648,
+      "step": 13160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.504520961715231,
+      "learning_rate": 7.58752366487463e-07,
+      "loss": 0.2546,
+      "step": 13161
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1614706095236045,
+      "learning_rate": 7.58213072606171e-07,
+      "loss": 0.2611,
+      "step": 13162
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6102353747156255,
+      "learning_rate": 7.576739547241546e-07,
+      "loss": 0.2494,
+      "step": 13163
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.548698086483542,
+      "learning_rate": 7.571350128637811e-07,
+      "loss": 0.2656,
+      "step": 13164
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.954585837576112,
+      "learning_rate": 7.565962470474109e-07,
+      "loss": 0.2697,
+      "step": 13165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8921314004904175,
+      "learning_rate": 7.560576572974004e-07,
+      "loss": 0.2626,
+      "step": 13166
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.425673724714399,
+      "learning_rate": 7.555192436360958e-07,
+      "loss": 0.2373,
+      "step": 13167
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1670183969758345,
+      "learning_rate": 7.549810060858359e-07,
+      "loss": 0.2578,
+      "step": 13168
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4958604366379131,
+      "learning_rate": 7.544429446689544e-07,
+      "loss": 0.2669,
+      "step": 13169
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2188583707750205,
+      "learning_rate": 7.539050594077768e-07,
+      "loss": 0.2569,
+      "step": 13170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5344431709143387,
+      "learning_rate": 7.533673503246203e-07,
+      "loss": 0.2821,
+      "step": 13171
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3546647669187095,
+      "learning_rate": 7.528298174417953e-07,
+      "loss": 0.2604,
+      "step": 13172
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2383655837220653,
+      "learning_rate": 7.522924607816034e-07,
+      "loss": 0.2359,
+      "step": 13173
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.26240150075405,
+      "learning_rate": 7.517552803663441e-07,
+      "loss": 0.2444,
+      "step": 13174
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8469803590004534,
+      "learning_rate": 7.512182762183035e-07,
+      "loss": 0.2438,
+      "step": 13175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8829531334783365,
+      "learning_rate": 7.506814483597619e-07,
+      "loss": 0.2579,
+      "step": 13176
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7580614740710847,
+      "learning_rate": 7.501447968129949e-07,
+      "loss": 0.2526,
+      "step": 13177
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3951182972511615,
+      "learning_rate": 7.496083216002703e-07,
+      "loss": 0.2467,
+      "step": 13178
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5016522555526541,
+      "learning_rate": 7.490720227438453e-07,
+      "loss": 0.2339,
+      "step": 13179
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.336968063304557,
+      "learning_rate": 7.485359002659731e-07,
+      "loss": 0.2534,
+      "step": 13180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7470124960417608,
+      "learning_rate": 7.479999541888966e-07,
+      "loss": 0.2551,
+      "step": 13181
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.100544671494195,
+      "learning_rate": 7.474641845348557e-07,
+      "loss": 0.2671,
+      "step": 13182
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7244243275115387,
+      "learning_rate": 7.469285913260787e-07,
+      "loss": 0.2454,
+      "step": 13183
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3512624149281267,
+      "learning_rate": 7.46393174584788e-07,
+      "loss": 0.2696,
+      "step": 13184
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.814366230180223,
+      "learning_rate": 7.458579343331996e-07,
+      "loss": 0.261,
+      "step": 13185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7616757227815985,
+      "learning_rate": 7.453228705935228e-07,
+      "loss": 0.2757,
+      "step": 13186
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1225086198030043,
+      "learning_rate": 7.447879833879574e-07,
+      "loss": 0.2778,
+      "step": 13187
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.559266608611239,
+      "learning_rate": 7.442532727386964e-07,
+      "loss": 0.2539,
+      "step": 13188
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.19378700486509,
+      "learning_rate": 7.437187386679252e-07,
+      "loss": 0.2413,
+      "step": 13189
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0478588030052065,
+      "learning_rate": 7.431843811978246e-07,
+      "loss": 0.253,
+      "step": 13190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.267341315514969,
+      "learning_rate": 7.426502003505653e-07,
+      "loss": 0.2662,
+      "step": 13191
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3898595182236768,
+      "learning_rate": 7.421161961483098e-07,
+      "loss": 0.2508,
+      "step": 13192
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.727880332474302,
+      "learning_rate": 7.415823686132162e-07,
+      "loss": 0.2531,
+      "step": 13193
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6029757269574015,
+      "learning_rate": 7.410487177674347e-07,
+      "loss": 0.4658,
+      "step": 13194
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4805363682229404,
+      "learning_rate": 7.40515243633107e-07,
+      "loss": 0.2562,
+      "step": 13195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.652345291347373,
+      "learning_rate": 7.399819462323677e-07,
+      "loss": 0.2593,
+      "step": 13196
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.792579656590822,
+      "learning_rate": 7.394488255873422e-07,
+      "loss": 0.2517,
+      "step": 13197
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.647559192520862,
+      "learning_rate": 7.389158817201541e-07,
+      "loss": 0.2488,
+      "step": 13198
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3104969639439124,
+      "learning_rate": 7.383831146529136e-07,
+      "loss": 0.241,
+      "step": 13199
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7519835733076743,
+      "learning_rate": 7.378505244077283e-07,
+      "loss": 0.261,
+      "step": 13200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9419281616050887,
+      "learning_rate": 7.373181110066941e-07,
+      "loss": 0.2629,
+      "step": 13201
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.056614347365575,
+      "learning_rate": 7.367858744719036e-07,
+      "loss": 0.2662,
+      "step": 13202
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.19721152563661,
+      "learning_rate": 7.3625381482544e-07,
+      "loss": 0.263,
+      "step": 13203
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3233526406143272,
+      "learning_rate": 7.357219320893789e-07,
+      "loss": 0.2509,
+      "step": 13204
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.899407420166879,
+      "learning_rate": 7.351902262857874e-07,
+      "loss": 0.265,
+      "step": 13205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4408411128453946,
+      "learning_rate": 7.346586974367298e-07,
+      "loss": 0.2658,
+      "step": 13206
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8495275953807058,
+      "learning_rate": 7.341273455642577e-07,
+      "loss": 0.2699,
+      "step": 13207
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9123545368181403,
+      "learning_rate": 7.33596170690421e-07,
+      "loss": 0.2567,
+      "step": 13208
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2028353397593383,
+      "learning_rate": 7.330651728372551e-07,
+      "loss": 0.264,
+      "step": 13209
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7664593579300796,
+      "learning_rate": 7.325343520267957e-07,
+      "loss": 0.2734,
+      "step": 13210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.070524073509291,
+      "learning_rate": 7.32003708281066e-07,
+      "loss": 0.2561,
+      "step": 13211
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.088478926229157,
+      "learning_rate": 7.314732416220821e-07,
+      "loss": 0.2613,
+      "step": 13212
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8787847196137988,
+      "learning_rate": 7.309429520718558e-07,
+      "loss": 0.2474,
+      "step": 13213
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.271709088883474,
+      "learning_rate": 7.304128396523885e-07,
+      "loss": 0.2565,
+      "step": 13214
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2318000708085362,
+      "learning_rate": 7.29882904385677e-07,
+      "loss": 0.281,
+      "step": 13215
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.775665621712501,
+      "learning_rate": 7.293531462937087e-07,
+      "loss": 0.2556,
+      "step": 13216
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.339480552195945,
+      "learning_rate": 7.288235653984621e-07,
+      "loss": 0.2503,
+      "step": 13217
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.643151832930828,
+      "learning_rate": 7.282941617219142e-07,
+      "loss": 0.2349,
+      "step": 13218
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.345094487632265,
+      "learning_rate": 7.277649352860283e-07,
+      "loss": 0.2469,
+      "step": 13219
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1513817184507706,
+      "learning_rate": 7.272358861127626e-07,
+      "loss": 0.2731,
+      "step": 13220
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6633674018639215,
+      "learning_rate": 7.267070142240706e-07,
+      "loss": 0.2499,
+      "step": 13221
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9642687538246135,
+      "learning_rate": 7.261783196418937e-07,
+      "loss": 0.2829,
+      "step": 13222
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9624608906456469,
+      "learning_rate": 7.256498023881703e-07,
+      "loss": 0.2467,
+      "step": 13223
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.484682183252951,
+      "learning_rate": 7.251214624848285e-07,
+      "loss": 0.2586,
+      "step": 13224
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9693426805083505,
+      "learning_rate": 7.245932999537897e-07,
+      "loss": 0.2455,
+      "step": 13225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.792972748711006,
+      "learning_rate": 7.240653148169696e-07,
+      "loss": 0.2732,
+      "step": 13226
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.354445537378943,
+      "learning_rate": 7.235375070962747e-07,
+      "loss": 0.2671,
+      "step": 13227
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.647517482193596,
+      "learning_rate": 7.230098768136035e-07,
+      "loss": 0.2676,
+      "step": 13228
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.466791352935791,
+      "learning_rate": 7.22482423990849e-07,
+      "loss": 0.2462,
+      "step": 13229
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4521212964525203,
+      "learning_rate": 7.219551486498976e-07,
+      "loss": 0.2588,
+      "step": 13230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8064537812046055,
+      "learning_rate": 7.214280508126259e-07,
+      "loss": 0.2543,
+      "step": 13231
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.374520532959189,
+      "learning_rate": 7.20901130500904e-07,
+      "loss": 0.2537,
+      "step": 13232
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0395352338624595,
+      "learning_rate": 7.203743877365938e-07,
+      "loss": 0.2387,
+      "step": 13233
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.892245557213895,
+      "learning_rate": 7.198478225415522e-07,
+      "loss": 0.2541,
+      "step": 13234
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7168205865884385,
+      "learning_rate": 7.193214349376276e-07,
+      "loss": 0.2518,
+      "step": 13235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6872993447475668,
+      "learning_rate": 7.187952249466585e-07,
+      "loss": 0.261,
+      "step": 13236
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.214984353252086,
+      "learning_rate": 7.182691925904795e-07,
+      "loss": 0.2493,
+      "step": 13237
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7230628565198083,
+      "learning_rate": 7.177433378909183e-07,
+      "loss": 0.2305,
+      "step": 13238
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2796634006638157,
+      "learning_rate": 7.172176608697923e-07,
+      "loss": 0.2675,
+      "step": 13239
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.098432726002284,
+      "learning_rate": 7.166921615489125e-07,
+      "loss": 0.2395,
+      "step": 13240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9166388897722202,
+      "learning_rate": 7.16166839950082e-07,
+      "loss": 0.248,
+      "step": 13241
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1520564944718763,
+      "learning_rate": 7.156416960950991e-07,
+      "loss": 0.2768,
+      "step": 13242
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8871636579136695,
+      "learning_rate": 7.15116730005752e-07,
+      "loss": 0.2446,
+      "step": 13243
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.098539015704749,
+      "learning_rate": 7.145919417038216e-07,
+      "loss": 0.2442,
+      "step": 13244
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5020532192567475,
+      "learning_rate": 7.140673312110835e-07,
+      "loss": 0.2506,
+      "step": 13245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.376122086608251,
+      "learning_rate": 7.135428985493053e-07,
+      "loss": 0.2538,
+      "step": 13246
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8465195022234036,
+      "learning_rate": 7.130186437402464e-07,
+      "loss": 0.2473,
+      "step": 13247
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.438491086285166,
+      "learning_rate": 7.124945668056577e-07,
+      "loss": 0.2552,
+      "step": 13248
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.513353316282746,
+      "learning_rate": 7.119706677672844e-07,
+      "loss": 0.2412,
+      "step": 13249
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 17.7935000376769,
+      "learning_rate": 7.114469466468654e-07,
+      "loss": 0.2554,
+      "step": 13250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6543348206489981,
+      "learning_rate": 7.109234034661288e-07,
+      "loss": 0.474,
+      "step": 13251
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.700080862271829,
+      "learning_rate": 7.104000382467996e-07,
+      "loss": 0.2575,
+      "step": 13252
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3586833265131317,
+      "learning_rate": 7.098768510105908e-07,
+      "loss": 0.253,
+      "step": 13253
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6050639256172698,
+      "learning_rate": 7.093538417792128e-07,
+      "loss": 0.2639,
+      "step": 13254
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.064275598744791,
+      "learning_rate": 7.088310105743645e-07,
+      "loss": 0.2479,
+      "step": 13255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6008283946217221,
+      "learning_rate": 7.083083574177402e-07,
+      "loss": 0.4699,
+      "step": 13256
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.864594186193576,
+      "learning_rate": 7.077858823310235e-07,
+      "loss": 0.2373,
+      "step": 13257
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4337804645080292,
+      "learning_rate": 7.072635853358956e-07,
+      "loss": 0.227,
+      "step": 13258
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7580940825618323,
+      "learning_rate": 7.067414664540251e-07,
+      "loss": 0.252,
+      "step": 13259
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.630406533128059,
+      "learning_rate": 7.062195257070781e-07,
+      "loss": 0.245,
+      "step": 13260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5015015359620418,
+      "learning_rate": 7.056977631167083e-07,
+      "loss": 0.2495,
+      "step": 13261
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3854198484063311,
+      "learning_rate": 7.051761787045674e-07,
+      "loss": 0.2632,
+      "step": 13262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8801030117422544,
+      "learning_rate": 7.046547724922953e-07,
+      "loss": 0.2759,
+      "step": 13263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8163154524820007,
+      "learning_rate": 7.041335445015258e-07,
+      "loss": 0.2415,
+      "step": 13264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.217397588848267,
+      "learning_rate": 7.036124947538847e-07,
+      "loss": 0.2631,
+      "step": 13265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6033136688166005,
+      "learning_rate": 7.030916232709922e-07,
+      "loss": 0.2685,
+      "step": 13266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6105248645404295,
+      "learning_rate": 7.025709300744621e-07,
+      "loss": 0.488,
+      "step": 13267
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.035529685884794,
+      "learning_rate": 7.020504151858965e-07,
+      "loss": 0.2497,
+      "step": 13268
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5992868409870489,
+      "learning_rate": 7.015300786268925e-07,
+      "loss": 0.2444,
+      "step": 13269
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.169102854967177,
+      "learning_rate": 7.010099204190418e-07,
+      "loss": 0.2435,
+      "step": 13270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.972304788053702,
+      "learning_rate": 7.00489940583925e-07,
+      "loss": 0.2737,
+      "step": 13271
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5843583323031228,
+      "learning_rate": 6.999701391431174e-07,
+      "loss": 0.4534,
+      "step": 13272
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4708934859929377,
+      "learning_rate": 6.994505161181859e-07,
+      "loss": 0.2625,
+      "step": 13273
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.672414264833315,
+      "learning_rate": 6.989310715306902e-07,
+      "loss": 0.2716,
+      "step": 13274
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8924993545687268,
+      "learning_rate": 6.984118054021854e-07,
+      "loss": 0.2454,
+      "step": 13275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.939114888197796,
+      "learning_rate": 6.978927177542156e-07,
+      "loss": 0.2576,
+      "step": 13276
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.333441468598719,
+      "learning_rate": 6.973738086083171e-07,
+      "loss": 0.2646,
+      "step": 13277
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4711775409493217,
+      "learning_rate": 6.968550779860222e-07,
+      "loss": 0.2591,
+      "step": 13278
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5540539054006046,
+      "learning_rate": 6.963365259088539e-07,
+      "loss": 0.2614,
+      "step": 13279
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0696361553775606,
+      "learning_rate": 6.958181523983276e-07,
+      "loss": 0.23,
+      "step": 13280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8755504916308086,
+      "learning_rate": 6.952999574759483e-07,
+      "loss": 0.2683,
+      "step": 13281
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.721954281596013,
+      "learning_rate": 6.947819411632223e-07,
+      "loss": 0.2467,
+      "step": 13282
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.834870733049077,
+      "learning_rate": 6.942641034816405e-07,
+      "loss": 0.2611,
+      "step": 13283
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.995436936193978,
+      "learning_rate": 6.937464444526887e-07,
+      "loss": 0.2544,
+      "step": 13284
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8579390339059376,
+      "learning_rate": 6.932289640978446e-07,
+      "loss": 0.2307,
+      "step": 13285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5764521190366008,
+      "learning_rate": 6.927116624385821e-07,
+      "loss": 0.2841,
+      "step": 13286
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2031118651098205,
+      "learning_rate": 6.92194539496363e-07,
+      "loss": 0.2812,
+      "step": 13287
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8542181699164357,
+      "learning_rate": 6.916775952926425e-07,
+      "loss": 0.2473,
+      "step": 13288
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.453543301183245,
+      "learning_rate": 6.911608298488715e-07,
+      "loss": 0.2425,
+      "step": 13289
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2483628857390707,
+      "learning_rate": 6.906442431864919e-07,
+      "loss": 0.2574,
+      "step": 13290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.43022301675465,
+      "learning_rate": 6.90127835326937e-07,
+      "loss": 0.2297,
+      "step": 13291
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1328799968395646,
+      "learning_rate": 6.896116062916335e-07,
+      "loss": 0.2452,
+      "step": 13292
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2169989321064327,
+      "learning_rate": 6.89095556101999e-07,
+      "loss": 0.2653,
+      "step": 13293
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4787048257489264,
+      "learning_rate": 6.885796847794485e-07,
+      "loss": 0.2352,
+      "step": 13294
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5187295195723474,
+      "learning_rate": 6.880639923453846e-07,
+      "loss": 0.2444,
+      "step": 13295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0944255461533845,
+      "learning_rate": 6.875484788212039e-07,
+      "loss": 0.2459,
+      "step": 13296
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7562433286035743,
+      "learning_rate": 6.870331442282957e-07,
+      "loss": 0.2441,
+      "step": 13297
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1571936644591587,
+      "learning_rate": 6.865179885880446e-07,
+      "loss": 0.2726,
+      "step": 13298
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8982898554061316,
+      "learning_rate": 6.860030119218241e-07,
+      "loss": 0.2473,
+      "step": 13299
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7195865458687813,
+      "learning_rate": 6.854882142510011e-07,
+      "loss": 0.2874,
+      "step": 13300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.073445083949944,
+      "learning_rate": 6.849735955969339e-07,
+      "loss": 0.2666,
+      "step": 13301
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4489188422992814,
+      "learning_rate": 6.844591559809777e-07,
+      "loss": 0.2475,
+      "step": 13302
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3863288274383887,
+      "learning_rate": 6.83944895424477e-07,
+      "loss": 0.2506,
+      "step": 13303
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4542813608180722,
+      "learning_rate": 6.834308139487672e-07,
+      "loss": 0.2434,
+      "step": 13304
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.339508487499196,
+      "learning_rate": 6.829169115751799e-07,
+      "loss": 0.2617,
+      "step": 13305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.19686953944788,
+      "learning_rate": 6.824031883250393e-07,
+      "loss": 0.2518,
+      "step": 13306
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.828381161059583,
+      "learning_rate": 6.818896442196593e-07,
+      "loss": 0.2425,
+      "step": 13307
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.455156198998655,
+      "learning_rate": 6.81376279280348e-07,
+      "loss": 0.2705,
+      "step": 13308
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6372669096726111,
+      "learning_rate": 6.808630935284039e-07,
+      "loss": 0.2509,
+      "step": 13309
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.09744505580353,
+      "learning_rate": 6.803500869851232e-07,
+      "loss": 0.2365,
+      "step": 13310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.894054835084033,
+      "learning_rate": 6.798372596717889e-07,
+      "loss": 0.2611,
+      "step": 13311
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6162670830008417,
+      "learning_rate": 6.793246116096813e-07,
+      "loss": 0.4855,
+      "step": 13312
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4248765821721117,
+      "learning_rate": 6.788121428200689e-07,
+      "loss": 0.2712,
+      "step": 13313
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9485942370631357,
+      "learning_rate": 6.782998533242169e-07,
+      "loss": 0.2617,
+      "step": 13314
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.025037599325337,
+      "learning_rate": 6.777877431433799e-07,
+      "loss": 0.2631,
+      "step": 13315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.608005797982861,
+      "learning_rate": 6.772758122988072e-07,
+      "loss": 0.2475,
+      "step": 13316
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.677616574605857,
+      "learning_rate": 6.767640608117376e-07,
+      "loss": 0.2854,
+      "step": 13317
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4455361472405373,
+      "learning_rate": 6.762524887034072e-07,
+      "loss": 0.2669,
+      "step": 13318
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.889383126616042,
+      "learning_rate": 6.757410959950395e-07,
+      "loss": 0.2522,
+      "step": 13319
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.433859600496642,
+      "learning_rate": 6.752298827078563e-07,
+      "loss": 0.2429,
+      "step": 13320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.962131400877815,
+      "learning_rate": 6.747188488630651e-07,
+      "loss": 0.2784,
+      "step": 13321
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.928368177247108,
+      "learning_rate": 6.74207994481873e-07,
+      "loss": 0.2576,
+      "step": 13322
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.2226999805099,
+      "learning_rate": 6.736973195854741e-07,
+      "loss": 0.2595,
+      "step": 13323
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0145970102896236,
+      "learning_rate": 6.731868241950585e-07,
+      "loss": 0.2345,
+      "step": 13324
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.79198770341662,
+      "learning_rate": 6.726765083318054e-07,
+      "loss": 0.2609,
+      "step": 13325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8091921624436693,
+      "learning_rate": 6.721663720168897e-07,
+      "loss": 0.25,
+      "step": 13326
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5577873315549133,
+      "learning_rate": 6.716564152714799e-07,
+      "loss": 0.2506,
+      "step": 13327
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.750512885142994,
+      "learning_rate": 6.711466381167336e-07,
+      "loss": 0.2323,
+      "step": 13328
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3412387824989707,
+      "learning_rate": 6.706370405738011e-07,
+      "loss": 0.2749,
+      "step": 13329
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4847828746570872,
+      "learning_rate": 6.701276226638287e-07,
+      "loss": 0.2463,
+      "step": 13330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.790574376609553,
+      "learning_rate": 6.696183844079513e-07,
+      "loss": 0.2551,
+      "step": 13331
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0287865241314864,
+      "learning_rate": 6.691093258272995e-07,
+      "loss": 0.2726,
+      "step": 13332
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8746573160134075,
+      "learning_rate": 6.686004469429919e-07,
+      "loss": 0.252,
+      "step": 13333
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.972510693307056,
+      "learning_rate": 6.680917477761472e-07,
+      "loss": 0.2859,
+      "step": 13334
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8567151185780817,
+      "learning_rate": 6.675832283478701e-07,
+      "loss": 0.2478,
+      "step": 13335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6380011904104417,
+      "learning_rate": 6.670748886792605e-07,
+      "loss": 0.247,
+      "step": 13336
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3114982050939092,
+      "learning_rate": 6.665667287914085e-07,
+      "loss": 0.2753,
+      "step": 13337
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.475359041361067,
+      "learning_rate": 6.660587487054016e-07,
+      "loss": 0.2607,
+      "step": 13338
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6237312823262653,
+      "learning_rate": 6.655509484423145e-07,
+      "loss": 0.4768,
+      "step": 13339
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.687907174511587,
+      "learning_rate": 6.650433280232177e-07,
+      "loss": 0.2569,
+      "step": 13340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5854181964847236,
+      "learning_rate": 6.64535887469171e-07,
+      "loss": 0.252,
+      "step": 13341
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.270724207695215,
+      "learning_rate": 6.640286268012325e-07,
+      "loss": 0.2527,
+      "step": 13342
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6781150362675301,
+      "learning_rate": 6.635215460404482e-07,
+      "loss": 0.2488,
+      "step": 13343
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.107836793691485,
+      "learning_rate": 6.630146452078579e-07,
+      "loss": 0.2637,
+      "step": 13344
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5307032776601286,
+      "learning_rate": 6.625079243244914e-07,
+      "loss": 0.2651,
+      "step": 13345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.159166297456615,
+      "learning_rate": 6.62001383411377e-07,
+      "loss": 0.2504,
+      "step": 13346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4547549816024652,
+      "learning_rate": 6.6149502248953e-07,
+      "loss": 0.2554,
+      "step": 13347
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8706399694321145,
+      "learning_rate": 6.609888415799598e-07,
+      "loss": 0.25,
+      "step": 13348
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6056446752643904,
+      "learning_rate": 6.604828407036695e-07,
+      "loss": 0.4846,
+      "step": 13349
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.340569330601539,
+      "learning_rate": 6.599770198816547e-07,
+      "loss": 0.2835,
+      "step": 13350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.345268295357784,
+      "learning_rate": 6.594713791349022e-07,
+      "loss": 0.2702,
+      "step": 13351
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0795552831337734,
+      "learning_rate": 6.589659184843922e-07,
+      "loss": 0.2562,
+      "step": 13352
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5328508534892715,
+      "learning_rate": 6.584606379510955e-07,
+      "loss": 0.2458,
+      "step": 13353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.522539358597986,
+      "learning_rate": 6.579555375559793e-07,
+      "loss": 0.2622,
+      "step": 13354
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9604813076177836,
+      "learning_rate": 6.574506173200008e-07,
+      "loss": 0.2598,
+      "step": 13355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.385096274681475,
+      "learning_rate": 6.569458772641074e-07,
+      "loss": 0.2707,
+      "step": 13356
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.670842291996775,
+      "learning_rate": 6.564413174092443e-07,
+      "loss": 0.279,
+      "step": 13357
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.42196166335391,
+      "learning_rate": 6.55936937776347e-07,
+      "loss": 0.2447,
+      "step": 13358
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.81297513134383,
+      "learning_rate": 6.554327383863418e-07,
+      "loss": 0.2646,
+      "step": 13359
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9493446450335274,
+      "learning_rate": 6.549287192601494e-07,
+      "loss": 0.2555,
+      "step": 13360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8715200415549311,
+      "learning_rate": 6.54424880418681e-07,
+      "loss": 0.238,
+      "step": 13361
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.266689418164463,
+      "learning_rate": 6.539212218828439e-07,
+      "loss": 0.2708,
+      "step": 13362
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5689967724985774,
+      "learning_rate": 6.53417743673534e-07,
+      "loss": 0.2632,
+      "step": 13363
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.131417965387195,
+      "learning_rate": 6.529144458116432e-07,
+      "loss": 0.2539,
+      "step": 13364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.492466989934758,
+      "learning_rate": 6.524113283180522e-07,
+      "loss": 0.2598,
+      "step": 13365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6808072474302356,
+      "learning_rate": 6.519083912136381e-07,
+      "loss": 0.2658,
+      "step": 13366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5444482608135894,
+      "learning_rate": 6.514056345192682e-07,
+      "loss": 0.2466,
+      "step": 13367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.100620416313012,
+      "learning_rate": 6.50903058255803e-07,
+      "loss": 0.2548,
+      "step": 13368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6240214069296524,
+      "learning_rate": 6.504006624440934e-07,
+      "loss": 0.2517,
+      "step": 13369
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.831067554539817,
+      "learning_rate": 6.498984471049869e-07,
+      "loss": 0.2366,
+      "step": 13370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.425528581014166,
+      "learning_rate": 6.493964122593193e-07,
+      "loss": 0.276,
+      "step": 13371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7891392092614682,
+      "learning_rate": 6.488945579279237e-07,
+      "loss": 0.2584,
+      "step": 13372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0941617305695686,
+      "learning_rate": 6.483928841316201e-07,
+      "loss": 0.2413,
+      "step": 13373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8179463764117592,
+      "learning_rate": 6.47891390891226e-07,
+      "loss": 0.2757,
+      "step": 13374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5556348084502554,
+      "learning_rate": 6.473900782275482e-07,
+      "loss": 0.2557,
+      "step": 13375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2595594744221827,
+      "learning_rate": 6.468889461613875e-07,
+      "loss": 0.2805,
+      "step": 13376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4269481979911813,
+      "learning_rate": 6.463879947135349e-07,
+      "loss": 0.2396,
+      "step": 13377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5734877924953476,
+      "learning_rate": 6.458872239047776e-07,
+      "loss": 0.2814,
+      "step": 13378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.752253659863759,
+      "learning_rate": 6.453866337558939e-07,
+      "loss": 0.2464,
+      "step": 13379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.141801705001186,
+      "learning_rate": 6.448862242876536e-07,
+      "loss": 0.2844,
+      "step": 13380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0042754335064585,
+      "learning_rate": 6.443859955208187e-07,
+      "loss": 0.2642,
+      "step": 13381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.36639946167784,
+      "learning_rate": 6.438859474761461e-07,
+      "loss": 0.261,
+      "step": 13382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7727691878110658,
+      "learning_rate": 6.43386080174383e-07,
+      "loss": 0.2633,
+      "step": 13383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5624324691703557,
+      "learning_rate": 6.428863936362694e-07,
+      "loss": 0.2822,
+      "step": 13384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.512964644654413,
+      "learning_rate": 6.423868878825373e-07,
+      "loss": 0.2765,
+      "step": 13385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.037510962546739,
+      "learning_rate": 6.418875629339133e-07,
+      "loss": 0.2756,
+      "step": 13386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0939592328857977,
+      "learning_rate": 6.413884188111163e-07,
+      "loss": 0.2613,
+      "step": 13387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3063107985055353,
+      "learning_rate": 6.408894555348555e-07,
+      "loss": 0.2566,
+      "step": 13388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2158554344475223,
+      "learning_rate": 6.40390673125833e-07,
+      "loss": 0.2719,
+      "step": 13389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8427488080983283,
+      "learning_rate": 6.398920716047458e-07,
+      "loss": 0.2431,
+      "step": 13390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3407186830334374,
+      "learning_rate": 6.39393650992281e-07,
+      "loss": 0.2597,
+      "step": 13391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.373527650414928,
+      "learning_rate": 6.388954113091195e-07,
+      "loss": 0.2386,
+      "step": 13392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4077006452566116,
+      "learning_rate": 6.383973525759318e-07,
+      "loss": 0.2544,
+      "step": 13393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.117246292700643,
+      "learning_rate": 6.378994748133855e-07,
+      "loss": 0.2587,
+      "step": 13394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6371280794912155,
+      "learning_rate": 6.374017780421387e-07,
+      "loss": 0.2621,
+      "step": 13395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.445796724466462,
+      "learning_rate": 6.369042622828408e-07,
+      "loss": 0.2597,
+      "step": 13396
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.233675456331351,
+      "learning_rate": 6.364069275561341e-07,
+      "loss": 0.2545,
+      "step": 13397
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.886902055526752,
+      "learning_rate": 6.359097738826559e-07,
+      "loss": 0.2439,
+      "step": 13398
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4013864313638993,
+      "learning_rate": 6.354128012830319e-07,
+      "loss": 0.2656,
+      "step": 13399
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2310331489065025,
+      "learning_rate": 6.349160097778839e-07,
+      "loss": 0.2474,
+      "step": 13400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.737654862163965,
+      "learning_rate": 6.344193993878223e-07,
+      "loss": 0.247,
+      "step": 13401
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5846140988984052,
+      "learning_rate": 6.339229701334543e-07,
+      "loss": 0.4467,
+      "step": 13402
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4099723567048166,
+      "learning_rate": 6.334267220353779e-07,
+      "loss": 0.2469,
+      "step": 13403
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6596960463805375,
+      "learning_rate": 6.329306551141834e-07,
+      "loss": 0.2722,
+      "step": 13404
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.121620019677974,
+      "learning_rate": 6.324347693904515e-07,
+      "loss": 0.2378,
+      "step": 13405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9842353800341574,
+      "learning_rate": 6.319390648847596e-07,
+      "loss": 0.2748,
+      "step": 13406
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.082599947725112,
+      "learning_rate": 6.314435416176745e-07,
+      "loss": 0.2491,
+      "step": 13407
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.825142349968624,
+      "learning_rate": 6.30948199609755e-07,
+      "loss": 0.2587,
+      "step": 13408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5584881233554497,
+      "learning_rate": 6.304530388815555e-07,
+      "loss": 0.2518,
+      "step": 13409
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3550523837795236,
+      "learning_rate": 6.299580594536214e-07,
+      "loss": 0.2402,
+      "step": 13410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.775748833236635,
+      "learning_rate": 6.294632613464891e-07,
+      "loss": 0.2489,
+      "step": 13411
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4795996961937647,
+      "learning_rate": 6.289686445806897e-07,
+      "loss": 0.2497,
+      "step": 13412
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.737286113311703,
+      "learning_rate": 6.284742091767437e-07,
+      "loss": 0.2534,
+      "step": 13413
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0572373881665493,
+      "learning_rate": 6.279799551551685e-07,
+      "loss": 0.258,
+      "step": 13414
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7420740158125616,
+      "learning_rate": 6.274858825364693e-07,
+      "loss": 0.2472,
+      "step": 13415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.535841739020183,
+      "learning_rate": 6.269919913411487e-07,
+      "loss": 0.2548,
+      "step": 13416
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.37352567868235,
+      "learning_rate": 6.264982815896964e-07,
+      "loss": 0.2687,
+      "step": 13417
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0267893266036907,
+      "learning_rate": 6.260047533025998e-07,
+      "loss": 0.2646,
+      "step": 13418
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.328546667924095,
+      "learning_rate": 6.255114065003353e-07,
+      "loss": 0.2637,
+      "step": 13419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.644197023306434,
+      "learning_rate": 6.250182412033723e-07,
+      "loss": 0.2681,
+      "step": 13420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.605194812598993,
+      "learning_rate": 6.245252574321719e-07,
+      "loss": 0.2687,
+      "step": 13421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.342267842668547,
+      "learning_rate": 6.24032455207192e-07,
+      "loss": 0.2654,
+      "step": 13422
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5008454522030399,
+      "learning_rate": 6.235398345488769e-07,
+      "loss": 0.2371,
+      "step": 13423
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.216123174695491,
+      "learning_rate": 6.230473954776683e-07,
+      "loss": 0.262,
+      "step": 13424
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1194285837111244,
+      "learning_rate": 6.225551380139966e-07,
+      "loss": 0.241,
+      "step": 13425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9680174260662269,
+      "learning_rate": 6.220630621782886e-07,
+      "loss": 0.2508,
+      "step": 13426
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.765124437463708,
+      "learning_rate": 6.215711679909603e-07,
+      "loss": 0.2565,
+      "step": 13427
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0869895718564586,
+      "learning_rate": 6.210794554724209e-07,
+      "loss": 0.2593,
+      "step": 13428
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.613307153027159,
+      "learning_rate": 6.205879246430718e-07,
+      "loss": 0.2364,
+      "step": 13429
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7907274267962165,
+      "learning_rate": 6.200965755233085e-07,
+      "loss": 0.2368,
+      "step": 13430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0093124444776462,
+      "learning_rate": 6.19605408133519e-07,
+      "loss": 0.2706,
+      "step": 13431
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8182818504541047,
+      "learning_rate": 6.191144224940815e-07,
+      "loss": 0.2419,
+      "step": 13432
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4251340973267377,
+      "learning_rate": 6.186236186253669e-07,
+      "loss": 0.2584,
+      "step": 13433
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4680766014110684,
+      "learning_rate": 6.181329965477417e-07,
+      "loss": 0.2584,
+      "step": 13434
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.852248030799743,
+      "learning_rate": 6.176425562815613e-07,
+      "loss": 0.273,
+      "step": 13435
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2213768022193507,
+      "learning_rate": 6.171522978471755e-07,
+      "loss": 0.257,
+      "step": 13436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7927775350305657,
+      "learning_rate": 6.166622212649248e-07,
+      "loss": 0.2699,
+      "step": 13437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7780035068961886,
+      "learning_rate": 6.161723265551439e-07,
+      "loss": 0.2439,
+      "step": 13438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3467121746151032,
+      "learning_rate": 6.156826137381605e-07,
+      "loss": 0.2427,
+      "step": 13439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8487489334041787,
+      "learning_rate": 6.151930828342933e-07,
+      "loss": 0.2558,
+      "step": 13440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1103522994585706,
+      "learning_rate": 6.147037338638523e-07,
+      "loss": 0.2488,
+      "step": 13441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5836701096911896,
+      "learning_rate": 6.142145668471434e-07,
+      "loss": 0.248,
+      "step": 13442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.656884641141751,
+      "learning_rate": 6.137255818044618e-07,
+      "loss": 0.2581,
+      "step": 13443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5778648933304922,
+      "learning_rate": 6.132367787560972e-07,
+      "loss": 0.2572,
+      "step": 13444
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.228521511300527,
+      "learning_rate": 6.127481577223293e-07,
+      "loss": 0.2729,
+      "step": 13445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9416217940728462,
+      "learning_rate": 6.12259718723433e-07,
+      "loss": 0.2716,
+      "step": 13446
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.189065981942754,
+      "learning_rate": 6.117714617796755e-07,
+      "loss": 0.2664,
+      "step": 13447
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5858810560868786,
+      "learning_rate": 6.11283386911315e-07,
+      "loss": 0.2734,
+      "step": 13448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8034508169949752,
+      "learning_rate": 6.107954941386002e-07,
+      "loss": 0.2564,
+      "step": 13449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4459616840143066,
+      "learning_rate": 6.103077834817778e-07,
+      "loss": 0.2415,
+      "step": 13450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5788569586631858,
+      "learning_rate": 6.098202549610821e-07,
+      "loss": 0.2687,
+      "step": 13451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6741883371995105,
+      "learning_rate": 6.09332908596742e-07,
+      "loss": 0.2449,
+      "step": 13452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.280979706528602,
+      "learning_rate": 6.088457444089774e-07,
+      "loss": 0.2589,
+      "step": 13453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0334725165814875,
+      "learning_rate": 6.083587624180021e-07,
+      "loss": 0.2751,
+      "step": 13454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0825827042994396,
+      "learning_rate": 6.078719626440238e-07,
+      "loss": 0.265,
+      "step": 13455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2214091746545925,
+      "learning_rate": 6.073853451072387e-07,
+      "loss": 0.253,
+      "step": 13456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.350869610333454,
+      "learning_rate": 6.068989098278366e-07,
+      "loss": 0.2475,
+      "step": 13457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.5054463094168895,
+      "learning_rate": 6.064126568260026e-07,
+      "loss": 0.2722,
+      "step": 13458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.375200479008572,
+      "learning_rate": 6.059265861219122e-07,
+      "loss": 0.2624,
+      "step": 13459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.491099614455404,
+      "learning_rate": 6.054406977357308e-07,
+      "loss": 0.2601,
+      "step": 13460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.080453525909033,
+      "learning_rate": 6.049549916876213e-07,
+      "loss": 0.2557,
+      "step": 13461
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7852621489247482,
+      "learning_rate": 6.044694679977353e-07,
+      "loss": 0.2426,
+      "step": 13462
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.50377367865537,
+      "learning_rate": 6.03984126686219e-07,
+      "loss": 0.2959,
+      "step": 13463
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8221389980922518,
+      "learning_rate": 6.034989677732095e-07,
+      "loss": 0.2645,
+      "step": 13464
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.782062025100652,
+      "learning_rate": 6.030139912788363e-07,
+      "loss": 0.2419,
+      "step": 13465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7677736292569717,
+      "learning_rate": 6.025291972232233e-07,
+      "loss": 0.2638,
+      "step": 13466
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0981586975400246,
+      "learning_rate": 6.020445856264845e-07,
+      "loss": 0.2484,
+      "step": 13467
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.568847040983956,
+      "learning_rate": 6.015601565087265e-07,
+      "loss": 0.2353,
+      "step": 13468
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.535192413258961,
+      "learning_rate": 6.010759098900514e-07,
+      "loss": 0.2656,
+      "step": 13469
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9884449494052558,
+      "learning_rate": 6.005918457905491e-07,
+      "loss": 0.2595,
+      "step": 13470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.727137316725863,
+      "learning_rate": 6.001079642303059e-07,
+      "loss": 0.2539,
+      "step": 13471
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5961615519099264,
+      "learning_rate": 5.996242652293987e-07,
+      "loss": 0.2397,
+      "step": 13472
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.032893371999254,
+      "learning_rate": 5.99140748807896e-07,
+      "loss": 0.2435,
+      "step": 13473
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7176294075536171,
+      "learning_rate": 5.98657414985861e-07,
+      "loss": 0.2473,
+      "step": 13474
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4133225690296671,
+      "learning_rate": 5.981742637833465e-07,
+      "loss": 0.2331,
+      "step": 13475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8912674668183675,
+      "learning_rate": 5.976912952204017e-07,
+      "loss": 0.2321,
+      "step": 13476
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7216631369538566,
+      "learning_rate": 5.972085093170637e-07,
+      "loss": 0.2466,
+      "step": 13477
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.1920283523365836,
+      "learning_rate": 5.967259060933644e-07,
+      "loss": 0.2559,
+      "step": 13478
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6549034677481446,
+      "learning_rate": 5.962434855693295e-07,
+      "loss": 0.4658,
+      "step": 13479
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6649424133775486,
+      "learning_rate": 5.957612477649743e-07,
+      "loss": 0.2529,
+      "step": 13480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3857659518307384,
+      "learning_rate": 5.952791927003066e-07,
+      "loss": 0.2601,
+      "step": 13481
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6784070898508838,
+      "learning_rate": 5.9479732039533e-07,
+      "loss": 0.2544,
+      "step": 13482
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.187974414621806,
+      "learning_rate": 5.94315630870036e-07,
+      "loss": 0.2726,
+      "step": 13483
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5241617959174945,
+      "learning_rate": 5.938341241444123e-07,
+      "loss": 0.2262,
+      "step": 13484
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8372876965840628,
+      "learning_rate": 5.93352800238437e-07,
+      "loss": 0.2463,
+      "step": 13485
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8398225505536796,
+      "learning_rate": 5.928716591720813e-07,
+      "loss": 0.2664,
+      "step": 13486
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4749406718398648,
+      "learning_rate": 5.92390700965309e-07,
+      "loss": 0.2532,
+      "step": 13487
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5737951844579836,
+      "learning_rate": 5.91909925638075e-07,
+      "loss": 0.2412,
+      "step": 13488
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.508596175108206,
+      "learning_rate": 5.914293332103266e-07,
+      "loss": 0.2619,
+      "step": 13489
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.189303442915712,
+      "learning_rate": 5.90948923702006e-07,
+      "loss": 0.2576,
+      "step": 13490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.539680287597827,
+      "learning_rate": 5.90468697133047e-07,
+      "loss": 0.2418,
+      "step": 13491
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6590691911248443,
+      "learning_rate": 5.89988653523374e-07,
+      "loss": 0.2508,
+      "step": 13492
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7274892066635184,
+      "learning_rate": 5.895087928929033e-07,
+      "loss": 0.2595,
+      "step": 13493
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7782888705814661,
+      "learning_rate": 5.890291152615479e-07,
+      "loss": 0.2515,
+      "step": 13494
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8518785472865065,
+      "learning_rate": 5.885496206492097e-07,
+      "loss": 0.2582,
+      "step": 13495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.550453497656156,
+      "learning_rate": 5.880703090757833e-07,
+      "loss": 0.2589,
+      "step": 13496
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3147100621447727,
+      "learning_rate": 5.875911805611551e-07,
+      "loss": 0.2595,
+      "step": 13497
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3720167509519614,
+      "learning_rate": 5.871122351252056e-07,
+      "loss": 0.2464,
+      "step": 13498
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.546748493072562,
+      "learning_rate": 5.866334727878093e-07,
+      "loss": 0.2449,
+      "step": 13499
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3805248661240959,
+      "learning_rate": 5.861548935688288e-07,
+      "loss": 0.2417,
+      "step": 13500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.565451613317308,
+      "learning_rate": 5.856764974881213e-07,
+      "loss": 0.261,
+      "step": 13501
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.48745688257381,
+      "learning_rate": 5.851982845655368e-07,
+      "loss": 0.2569,
+      "step": 13502
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2245977594989972,
+      "learning_rate": 5.847202548209174e-07,
+      "loss": 0.2456,
+      "step": 13503
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8972157072057165,
+      "learning_rate": 5.842424082740972e-07,
+      "loss": 0.2516,
+      "step": 13504
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8840926956516535,
+      "learning_rate": 5.837647449449019e-07,
+      "loss": 0.241,
+      "step": 13505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6658739249500758,
+      "learning_rate": 5.832872648531512e-07,
+      "loss": 0.2783,
+      "step": 13506
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0168126895491363,
+      "learning_rate": 5.828099680186577e-07,
+      "loss": 0.2419,
+      "step": 13507
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9599077828272837,
+      "learning_rate": 5.823328544612245e-07,
+      "loss": 0.2429,
+      "step": 13508
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7095248415156747,
+      "learning_rate": 5.818559242006472e-07,
+      "loss": 0.2533,
+      "step": 13509
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7421267957726232,
+      "learning_rate": 5.813791772567157e-07,
+      "loss": 0.2464,
+      "step": 13510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.800904870901792,
+      "learning_rate": 5.809026136492107e-07,
+      "loss": 0.2547,
+      "step": 13511
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.881181483130517,
+      "learning_rate": 5.804262333979044e-07,
+      "loss": 0.2457,
+      "step": 13512
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.552193320850763,
+      "learning_rate": 5.799500365225647e-07,
+      "loss": 0.2456,
+      "step": 13513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.77824722130231,
+      "learning_rate": 5.794740230429475e-07,
+      "loss": 0.257,
+      "step": 13514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.251834361956646,
+      "learning_rate": 5.789981929788063e-07,
+      "loss": 0.2541,
+      "step": 13515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7435382526825194,
+      "learning_rate": 5.785225463498828e-07,
+      "loss": 0.2603,
+      "step": 13516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6986746749677004,
+      "learning_rate": 5.780470831759111e-07,
+      "loss": 0.2572,
+      "step": 13517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.969212550955034,
+      "learning_rate": 5.775718034766209e-07,
+      "loss": 0.2665,
+      "step": 13518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6106812182464696,
+      "learning_rate": 5.77096707271732e-07,
+      "loss": 0.267,
+      "step": 13519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7315584862627755,
+      "learning_rate": 5.766217945809554e-07,
+      "loss": 0.2478,
+      "step": 13520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6835443075396217,
+      "learning_rate": 5.761470654239987e-07,
+      "loss": 0.2552,
+      "step": 13521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.980618816403487,
+      "learning_rate": 5.756725198205565e-07,
+      "loss": 0.2636,
+      "step": 13522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3422479895103936,
+      "learning_rate": 5.751981577903216e-07,
+      "loss": 0.2599,
+      "step": 13523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5699725414163723,
+      "learning_rate": 5.747239793529735e-07,
+      "loss": 0.2315,
+      "step": 13524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.614291219139042,
+      "learning_rate": 5.742499845281874e-07,
+      "loss": 0.2545,
+      "step": 13525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.364417162361661,
+      "learning_rate": 5.737761733356312e-07,
+      "loss": 0.2654,
+      "step": 13526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5999843211238662,
+      "learning_rate": 5.733025457949625e-07,
+      "loss": 0.231,
+      "step": 13527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.106616621026424,
+      "learning_rate": 5.728291019258347e-07,
+      "loss": 0.263,
+      "step": 13528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9067278643917807,
+      "learning_rate": 5.723558417478914e-07,
+      "loss": 0.2597,
+      "step": 13529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8953330005585292,
+      "learning_rate": 5.718827652807673e-07,
+      "loss": 0.256,
+      "step": 13530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.104440011041646,
+      "learning_rate": 5.714098725440936e-07,
+      "loss": 0.2536,
+      "step": 13531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.749262134426467,
+      "learning_rate": 5.709371635574907e-07,
+      "loss": 0.2563,
+      "step": 13532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5855358582502723,
+      "learning_rate": 5.704646383405698e-07,
+      "loss": 0.4516,
+      "step": 13533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4983159479219632,
+      "learning_rate": 5.699922969129406e-07,
+      "loss": 0.2454,
+      "step": 13534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.589545383496322,
+      "learning_rate": 5.695201392941985e-07,
+      "loss": 0.2669,
+      "step": 13535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.378932170923166,
+      "learning_rate": 5.690481655039359e-07,
+      "loss": 0.2542,
+      "step": 13536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2701012800953952,
+      "learning_rate": 5.685763755617357e-07,
+      "loss": 0.2683,
+      "step": 13537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.00697994797012,
+      "learning_rate": 5.68104769487171e-07,
+      "loss": 0.2499,
+      "step": 13538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9762545860281695,
+      "learning_rate": 5.676333472998131e-07,
+      "loss": 0.2611,
+      "step": 13539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0532568180829864,
+      "learning_rate": 5.671621090192203e-07,
+      "loss": 0.2431,
+      "step": 13540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.670252802575042,
+      "learning_rate": 5.66691054664944e-07,
+      "loss": 0.2341,
+      "step": 13541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7785526737823396,
+      "learning_rate": 5.662201842565301e-07,
+      "loss": 0.2701,
+      "step": 13542
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.335948586720108,
+      "learning_rate": 5.657494978135169e-07,
+      "loss": 0.2658,
+      "step": 13543
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6821751756813734,
+      "learning_rate": 5.652789953554338e-07,
+      "loss": 0.2459,
+      "step": 13544
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7165543974639266,
+      "learning_rate": 5.648086769018019e-07,
+      "loss": 0.2508,
+      "step": 13545
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3152126272537419,
+      "learning_rate": 5.643385424721342e-07,
+      "loss": 0.243,
+      "step": 13546
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6376253550899063,
+      "learning_rate": 5.63868592085941e-07,
+      "loss": 0.2722,
+      "step": 13547
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2887753226860474,
+      "learning_rate": 5.633988257627187e-07,
+      "loss": 0.2433,
+      "step": 13548
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.577455232390544,
+      "learning_rate": 5.629292435219586e-07,
+      "loss": 0.4639,
+      "step": 13549
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8079642268829752,
+      "learning_rate": 5.624598453831453e-07,
+      "loss": 0.2437,
+      "step": 13550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.853634505212085,
+      "learning_rate": 5.619906313657558e-07,
+      "loss": 0.2513,
+      "step": 13551
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0104315005559203,
+      "learning_rate": 5.615216014892577e-07,
+      "loss": 0.2434,
+      "step": 13552
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7080766982438698,
+      "learning_rate": 5.610527557731126e-07,
+      "loss": 0.2691,
+      "step": 13553
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.636742689216997,
+      "learning_rate": 5.605840942367713e-07,
+      "loss": 0.2523,
+      "step": 13554
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7398274197926993,
+      "learning_rate": 5.60115616899683e-07,
+      "loss": 0.2946,
+      "step": 13555
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.470485277610429,
+      "learning_rate": 5.596473237812833e-07,
+      "loss": 0.2481,
+      "step": 13556
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2016074473377323,
+      "learning_rate": 5.591792149010022e-07,
+      "loss": 0.2522,
+      "step": 13557
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0406811642020126,
+      "learning_rate": 5.587112902782638e-07,
+      "loss": 0.2464,
+      "step": 13558
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7116847885826194,
+      "learning_rate": 5.582435499324829e-07,
+      "loss": 0.2581,
+      "step": 13559
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.439569701263958,
+      "learning_rate": 5.577759938830668e-07,
+      "loss": 0.277,
+      "step": 13560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3058979220050366,
+      "learning_rate": 5.57308622149415e-07,
+      "loss": 0.2702,
+      "step": 13561
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5615495482226585,
+      "learning_rate": 5.568414347509188e-07,
+      "loss": 0.2566,
+      "step": 13562
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.491320604001727,
+      "learning_rate": 5.563744317069642e-07,
+      "loss": 0.2455,
+      "step": 13563
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9199517882555668,
+      "learning_rate": 5.559076130369273e-07,
+      "loss": 0.2255,
+      "step": 13564
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3006377515034426,
+      "learning_rate": 5.554409787601755e-07,
+      "loss": 0.242,
+      "step": 13565
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6010076661097696,
+      "learning_rate": 5.549745288960729e-07,
+      "loss": 0.2485,
+      "step": 13566
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.113651352580854,
+      "learning_rate": 5.545082634639726e-07,
+      "loss": 0.2385,
+      "step": 13567
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8076755945116019,
+      "learning_rate": 5.54042182483221e-07,
+      "loss": 0.2542,
+      "step": 13568
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5008753258119847,
+      "learning_rate": 5.535762859731547e-07,
+      "loss": 0.2428,
+      "step": 13569
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2858870591617326,
+      "learning_rate": 5.531105739531073e-07,
+      "loss": 0.2642,
+      "step": 13570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.312244975118548,
+      "learning_rate": 5.52645046442401e-07,
+      "loss": 0.2473,
+      "step": 13571
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3237576285969577,
+      "learning_rate": 5.521797034603499e-07,
+      "loss": 0.2434,
+      "step": 13572
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5538232390013953,
+      "learning_rate": 5.517145450262639e-07,
+      "loss": 0.2557,
+      "step": 13573
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0129384304103244,
+      "learning_rate": 5.512495711594418e-07,
+      "loss": 0.2396,
+      "step": 13574
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6514149107581721,
+      "learning_rate": 5.507847818791778e-07,
+      "loss": 0.2658,
+      "step": 13575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6918203015153468,
+      "learning_rate": 5.503201772047556e-07,
+      "loss": 0.2524,
+      "step": 13576
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6636673205522431,
+      "learning_rate": 5.498557571554519e-07,
+      "loss": 0.4798,
+      "step": 13577
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.565588593550987,
+      "learning_rate": 5.493915217505386e-07,
+      "loss": 0.2524,
+      "step": 13578
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7679157801907177,
+      "learning_rate": 5.489274710092746e-07,
+      "loss": 0.2763,
+      "step": 13579
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4383566699951482,
+      "learning_rate": 5.484636049509173e-07,
+      "loss": 0.2352,
+      "step": 13580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.800519061403932,
+      "learning_rate": 5.47999923594712e-07,
+      "loss": 0.2356,
+      "step": 13581
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5724452368813069,
+      "learning_rate": 5.475364269598959e-07,
+      "loss": 0.4391,
+      "step": 13582
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7872398232081483,
+      "learning_rate": 5.470731150657033e-07,
+      "loss": 0.2431,
+      "step": 13583
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.934105082982598,
+      "learning_rate": 5.466099879313563e-07,
+      "loss": 0.2714,
+      "step": 13584
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8558269694509377,
+      "learning_rate": 5.461470455760698e-07,
+      "loss": 0.2742,
+      "step": 13585
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7107592585793334,
+      "learning_rate": 5.45684288019055e-07,
+      "loss": 0.2502,
+      "step": 13586
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9950131298505234,
+      "learning_rate": 5.452217152795092e-07,
+      "loss": 0.2433,
+      "step": 13587
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3131958912287827,
+      "learning_rate": 5.447593273766283e-07,
+      "loss": 0.2603,
+      "step": 13588
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.12074925694779,
+      "learning_rate": 5.442971243295964e-07,
+      "loss": 0.248,
+      "step": 13589
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9615825929841024,
+      "learning_rate": 5.438351061575897e-07,
+      "loss": 0.2313,
+      "step": 13590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.175260952878419,
+      "learning_rate": 5.433732728797808e-07,
+      "loss": 0.2481,
+      "step": 13591
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0754082213129466,
+      "learning_rate": 5.429116245153304e-07,
+      "loss": 0.2584,
+      "step": 13592
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.691298911844114,
+      "learning_rate": 5.424501610833921e-07,
+      "loss": 0.2483,
+      "step": 13593
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6745785689146537,
+      "learning_rate": 5.41988882603115e-07,
+      "loss": 0.2606,
+      "step": 13594
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6332867781682845,
+      "learning_rate": 5.415277890936377e-07,
+      "loss": 0.2592,
+      "step": 13595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6742408535272966,
+      "learning_rate": 5.410668805740921e-07,
+      "loss": 0.2464,
+      "step": 13596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.014704593624091,
+      "learning_rate": 5.406061570636012e-07,
+      "loss": 0.3037,
+      "step": 13597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3033209178746947,
+      "learning_rate": 5.401456185812809e-07,
+      "loss": 0.2595,
+      "step": 13598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3024031502079834,
+      "learning_rate": 5.396852651462409e-07,
+      "loss": 0.2795,
+      "step": 13599
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6235627715437748,
+      "learning_rate": 5.392250967775825e-07,
+      "loss": 0.4798,
+      "step": 13600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7957041900238733,
+      "learning_rate": 5.387651134943966e-07,
+      "loss": 0.2478,
+      "step": 13601
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.086923984648585,
+      "learning_rate": 5.383053153157703e-07,
+      "loss": 0.2586,
+      "step": 13602
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.866657186222648,
+      "learning_rate": 5.378457022607819e-07,
+      "loss": 0.2716,
+      "step": 13603
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7137875550053123,
+      "learning_rate": 5.373862743485014e-07,
+      "loss": 0.2462,
+      "step": 13604
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3864030986668119,
+      "learning_rate": 5.369270315979908e-07,
+      "loss": 0.2356,
+      "step": 13605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4348947069003946,
+      "learning_rate": 5.364679740283041e-07,
+      "loss": 0.2462,
+      "step": 13606
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5469037905580496,
+      "learning_rate": 5.360091016584901e-07,
+      "loss": 0.2552,
+      "step": 13607
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.607533664012968,
+      "learning_rate": 5.355504145075874e-07,
+      "loss": 0.2375,
+      "step": 13608
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5724723128126032,
+      "learning_rate": 5.350919125946269e-07,
+      "loss": 0.2497,
+      "step": 13609
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0558780389893303,
+      "learning_rate": 5.346335959386333e-07,
+      "loss": 0.2488,
+      "step": 13610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9389550348105287,
+      "learning_rate": 5.341754645586244e-07,
+      "loss": 0.2505,
+      "step": 13611
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.871688509492094,
+      "learning_rate": 5.337175184736077e-07,
+      "loss": 0.2274,
+      "step": 13612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4524242667794094,
+      "learning_rate": 5.332597577025845e-07,
+      "loss": 0.2704,
+      "step": 13613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7026679359135917,
+      "learning_rate": 5.328021822645463e-07,
+      "loss": 0.262,
+      "step": 13614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9682632115171763,
+      "learning_rate": 5.323447921784813e-07,
+      "loss": 0.2597,
+      "step": 13615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8148672515326587,
+      "learning_rate": 5.318875874633661e-07,
+      "loss": 0.2491,
+      "step": 13616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.560120130082495,
+      "learning_rate": 5.314305681381704e-07,
+      "loss": 0.2768,
+      "step": 13617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.424553683391421,
+      "learning_rate": 5.309737342218574e-07,
+      "loss": 0.2412,
+      "step": 13618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.572884717040157,
+      "learning_rate": 5.30517085733383e-07,
+      "loss": 0.2529,
+      "step": 13619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.6180434518983775,
+      "learning_rate": 5.300606226916938e-07,
+      "loss": 0.2694,
+      "step": 13620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0940017996641846,
+      "learning_rate": 5.296043451157285e-07,
+      "loss": 0.2543,
+      "step": 13621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3697832821973646,
+      "learning_rate": 5.29148253024418e-07,
+      "loss": 0.2648,
+      "step": 13622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9218459094353313,
+      "learning_rate": 5.286923464366889e-07,
+      "loss": 0.2452,
+      "step": 13623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.248390007409386,
+      "learning_rate": 5.282366253714555e-07,
+      "loss": 0.2567,
+      "step": 13624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.1861708252662804,
+      "learning_rate": 5.277810898476283e-07,
+      "loss": 0.2561,
+      "step": 13625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.343048640813765,
+      "learning_rate": 5.273257398841059e-07,
+      "loss": 0.2496,
+      "step": 13626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.602411140422076,
+      "learning_rate": 5.268705754997838e-07,
+      "loss": 0.2558,
+      "step": 13627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6285792398465866,
+      "learning_rate": 5.26415596713547e-07,
+      "loss": 0.2552,
+      "step": 13628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.755575220971988,
+      "learning_rate": 5.259608035442726e-07,
+      "loss": 0.2617,
+      "step": 13629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0699700530282445,
+      "learning_rate": 5.255061960108309e-07,
+      "loss": 0.2441,
+      "step": 13630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.030379181313497,
+      "learning_rate": 5.250517741320854e-07,
+      "loss": 0.2528,
+      "step": 13631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.351254439335697,
+      "learning_rate": 5.245975379268892e-07,
+      "loss": 0.247,
+      "step": 13632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2964371653384898,
+      "learning_rate": 5.241434874140916e-07,
+      "loss": 0.2484,
+      "step": 13633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1064082182356647,
+      "learning_rate": 5.236896226125293e-07,
+      "loss": 0.2584,
+      "step": 13634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.374465623231935,
+      "learning_rate": 5.232359435410366e-07,
+      "loss": 0.2646,
+      "step": 13635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9347914803804405,
+      "learning_rate": 5.227824502184364e-07,
+      "loss": 0.2734,
+      "step": 13636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2171314538639177,
+      "learning_rate": 5.223291426635447e-07,
+      "loss": 0.2626,
+      "step": 13637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2351019414444853,
+      "learning_rate": 5.21876020895169e-07,
+      "loss": 0.2675,
+      "step": 13638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2136681821585618,
+      "learning_rate": 5.214230849321111e-07,
+      "loss": 0.2506,
+      "step": 13639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.83766489256189,
+      "learning_rate": 5.209703347931655e-07,
+      "loss": 0.2735,
+      "step": 13640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.174020625700449,
+      "learning_rate": 5.205177704971159e-07,
+      "loss": 0.2671,
+      "step": 13641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.430670413441794,
+      "learning_rate": 5.200653920627402e-07,
+      "loss": 0.2969,
+      "step": 13642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3900125754278547,
+      "learning_rate": 5.196131995088089e-07,
+      "loss": 0.2399,
+      "step": 13643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2168712897411376,
+      "learning_rate": 5.191611928540846e-07,
+      "loss": 0.2571,
+      "step": 13644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.418268250393664,
+      "learning_rate": 5.187093721173198e-07,
+      "loss": 0.2576,
+      "step": 13645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0594401712808112,
+      "learning_rate": 5.182577373172642e-07,
+      "loss": 0.2432,
+      "step": 13646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3299983009554044,
+      "learning_rate": 5.178062884726547e-07,
+      "loss": 0.2425,
+      "step": 13647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.743647726617732,
+      "learning_rate": 5.173550256022241e-07,
+      "loss": 0.2393,
+      "step": 13648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8736867425503996,
+      "learning_rate": 5.169039487246958e-07,
+      "loss": 0.2661,
+      "step": 13649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7689666442057652,
+      "learning_rate": 5.164530578587845e-07,
+      "loss": 0.2541,
+      "step": 13650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 26.14275002768576,
+      "learning_rate": 5.160023530232005e-07,
+      "loss": 0.2637,
+      "step": 13651
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4701743228404707,
+      "learning_rate": 5.155518342366434e-07,
+      "loss": 0.2543,
+      "step": 13652
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1312873255106126,
+      "learning_rate": 5.151015015178046e-07,
+      "loss": 0.2821,
+      "step": 13653
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.76115379036279,
+      "learning_rate": 5.14651354885371e-07,
+      "loss": 0.233,
+      "step": 13654
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5255376443447688,
+      "learning_rate": 5.142013943580205e-07,
+      "loss": 0.2645,
+      "step": 13655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.810116332603644,
+      "learning_rate": 5.137516199544218e-07,
+      "loss": 0.2718,
+      "step": 13656
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7718179422281233,
+      "learning_rate": 5.133020316932369e-07,
+      "loss": 0.2302,
+      "step": 13657
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1572912137807267,
+      "learning_rate": 5.128526295931191e-07,
+      "loss": 0.2445,
+      "step": 13658
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.02531717500136,
+      "learning_rate": 5.124034136727163e-07,
+      "loss": 0.2533,
+      "step": 13659
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6125455807078777,
+      "learning_rate": 5.119543839506668e-07,
+      "loss": 0.2394,
+      "step": 13660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.12656228148672,
+      "learning_rate": 5.115055404456009e-07,
+      "loss": 0.2765,
+      "step": 13661
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8159771758577405,
+      "learning_rate": 5.110568831761426e-07,
+      "loss": 0.24,
+      "step": 13662
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.327296188057833,
+      "learning_rate": 5.106084121609084e-07,
+      "loss": 0.2397,
+      "step": 13663
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.4702360420956335,
+      "learning_rate": 5.101601274185053e-07,
+      "loss": 0.2505,
+      "step": 13664
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1902812194953794,
+      "learning_rate": 5.097120289675333e-07,
+      "loss": 0.2713,
+      "step": 13665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.485408023468269,
+      "learning_rate": 5.092641168265839e-07,
+      "loss": 0.2488,
+      "step": 13666
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7370814254015277,
+      "learning_rate": 5.088163910142436e-07,
+      "loss": 0.26,
+      "step": 13667
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8861457859821287,
+      "learning_rate": 5.083688515490881e-07,
+      "loss": 0.2478,
+      "step": 13668
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9405465602943042,
+      "learning_rate": 5.079214984496866e-07,
+      "loss": 0.2441,
+      "step": 13669
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5700041359223784,
+      "learning_rate": 5.074743317346009e-07,
+      "loss": 0.4724,
+      "step": 13670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9281739077401292,
+      "learning_rate": 5.070273514223856e-07,
+      "loss": 0.2599,
+      "step": 13671
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.494583527548455,
+      "learning_rate": 5.065805575315863e-07,
+      "loss": 0.2602,
+      "step": 13672
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7839197396281787,
+      "learning_rate": 5.061339500807405e-07,
+      "loss": 0.2468,
+      "step": 13673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8851465665212364,
+      "learning_rate": 5.056875290883778e-07,
+      "loss": 0.2432,
+      "step": 13674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9874666537147156,
+      "learning_rate": 5.05241294573024e-07,
+      "loss": 0.2552,
+      "step": 13675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.640552388955175,
+      "learning_rate": 5.047952465531913e-07,
+      "loss": 0.2825,
+      "step": 13676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8929427231370501,
+      "learning_rate": 5.04349385047389e-07,
+      "loss": 0.2486,
+      "step": 13677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.17727300472136,
+      "learning_rate": 5.039037100741151e-07,
+      "loss": 0.2703,
+      "step": 13678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.327589387383158,
+      "learning_rate": 5.03458221651863e-07,
+      "loss": 0.2643,
+      "step": 13679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.17373665414888,
+      "learning_rate": 5.030129197991162e-07,
+      "loss": 0.2398,
+      "step": 13680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.434944858566004,
+      "learning_rate": 5.025678045343507e-07,
+      "loss": 0.248,
+      "step": 13681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5754890071326177,
+      "learning_rate": 5.021228758760344e-07,
+      "loss": 0.2445,
+      "step": 13682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7873413417206307,
+      "learning_rate": 5.016781338426302e-07,
+      "loss": 0.256,
+      "step": 13683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3658169989413795,
+      "learning_rate": 5.01233578452589e-07,
+      "loss": 0.2623,
+      "step": 13684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8189939340242294,
+      "learning_rate": 5.007892097243588e-07,
+      "loss": 0.256,
+      "step": 13685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9382916850983416,
+      "learning_rate": 5.003450276763744e-07,
+      "loss": 0.2364,
+      "step": 13686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8440640191006197,
+      "learning_rate": 4.999010323270681e-07,
+      "loss": 0.2578,
+      "step": 13687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6054567038581964,
+      "learning_rate": 4.994572236948608e-07,
+      "loss": 0.4699,
+      "step": 13688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.129324880100568,
+      "learning_rate": 4.990136017981678e-07,
+      "loss": 0.253,
+      "step": 13689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9361134202184722,
+      "learning_rate": 4.985701666553938e-07,
+      "loss": 0.2667,
+      "step": 13690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8162227472397376,
+      "learning_rate": 4.981269182849391e-07,
+      "loss": 0.2513,
+      "step": 13691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6966984178058855,
+      "learning_rate": 4.976838567051956e-07,
+      "loss": 0.2838,
+      "step": 13692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6477148137784117,
+      "learning_rate": 4.972409819345458e-07,
+      "loss": 0.2677,
+      "step": 13693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4327947666479184,
+      "learning_rate": 4.967982939913646e-07,
+      "loss": 0.2505,
+      "step": 13694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.744999869600108,
+      "learning_rate": 4.963557928940215e-07,
+      "loss": 0.2514,
+      "step": 13695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.366089183714872,
+      "learning_rate": 4.959134786608766e-07,
+      "loss": 0.2442,
+      "step": 13696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.657426892235389,
+      "learning_rate": 4.95471351310281e-07,
+      "loss": 0.2579,
+      "step": 13697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.26204346002552,
+      "learning_rate": 4.950294108605791e-07,
+      "loss": 0.2478,
+      "step": 13698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7537968602009564,
+      "learning_rate": 4.94587657330109e-07,
+      "loss": 0.2577,
+      "step": 13699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5656007558597471,
+      "learning_rate": 4.941460907372004e-07,
+      "loss": 0.2605,
+      "step": 13700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9827331043292564,
+      "learning_rate": 4.937047111001736e-07,
+      "loss": 0.248,
+      "step": 13701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.591557031894639,
+      "learning_rate": 4.932635184373413e-07,
+      "loss": 0.2339,
+      "step": 13702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7381995198494065,
+      "learning_rate": 4.928225127670123e-07,
+      "loss": 0.2604,
+      "step": 13703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.173427432588498,
+      "learning_rate": 4.923816941074822e-07,
+      "loss": 0.2622,
+      "step": 13704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.787580573787161,
+      "learning_rate": 4.919410624770421e-07,
+      "loss": 0.265,
+      "step": 13705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5860671434993519,
+      "learning_rate": 4.915006178939724e-07,
+      "loss": 0.2413,
+      "step": 13706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9961316065912431,
+      "learning_rate": 4.910603603765523e-07,
+      "loss": 0.2378,
+      "step": 13707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.846316850977396,
+      "learning_rate": 4.906202899430463e-07,
+      "loss": 0.2671,
+      "step": 13708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5538424167891645,
+      "learning_rate": 4.901804066117144e-07,
+      "loss": 0.2652,
+      "step": 13709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.961948405895061,
+      "learning_rate": 4.897407104008067e-07,
+      "loss": 0.2477,
+      "step": 13710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6083660588937367,
+      "learning_rate": 4.893012013285692e-07,
+      "loss": 0.2593,
+      "step": 13711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1479071891206654,
+      "learning_rate": 4.88861879413236e-07,
+      "loss": 0.2587,
+      "step": 13712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5699512695917646,
+      "learning_rate": 4.884227446730372e-07,
+      "loss": 0.2552,
+      "step": 13713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.493464694099462,
+      "learning_rate": 4.879837971261897e-07,
+      "loss": 0.2419,
+      "step": 13714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.30024647646415,
+      "learning_rate": 4.875450367909101e-07,
+      "loss": 0.245,
+      "step": 13715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6888980037888235,
+      "learning_rate": 4.871064636854029e-07,
+      "loss": 0.255,
+      "step": 13716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7814713071352997,
+      "learning_rate": 4.866680778278637e-07,
+      "loss": 0.2575,
+      "step": 13717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6932877666833304,
+      "learning_rate": 4.862298792364817e-07,
+      "loss": 0.2323,
+      "step": 13718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8334929928287969,
+      "learning_rate": 4.857918679294405e-07,
+      "loss": 0.2474,
+      "step": 13719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.446489705834415,
+      "learning_rate": 4.853540439249127e-07,
+      "loss": 0.2401,
+      "step": 13720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7818780712405435,
+      "learning_rate": 4.849164072410639e-07,
+      "loss": 0.2392,
+      "step": 13721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.837794344465931,
+      "learning_rate": 4.844789578960524e-07,
+      "loss": 0.248,
+      "step": 13722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.40695280700196,
+      "learning_rate": 4.840416959080307e-07,
+      "loss": 0.2362,
+      "step": 13723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8239567390662956,
+      "learning_rate": 4.836046212951406e-07,
+      "loss": 0.2471,
+      "step": 13724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.673927413885572,
+      "learning_rate": 4.831677340755164e-07,
+      "loss": 0.255,
+      "step": 13725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.480471311827824,
+      "learning_rate": 4.827310342672847e-07,
+      "loss": 0.2653,
+      "step": 13726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8703191755513764,
+      "learning_rate": 4.822945218885672e-07,
+      "loss": 0.2748,
+      "step": 13727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.017521689606188,
+      "learning_rate": 4.818581969574743e-07,
+      "loss": 0.2774,
+      "step": 13728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.560156684088177,
+      "learning_rate": 4.814220594921087e-07,
+      "loss": 0.2772,
+      "step": 13729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2924727580181514,
+      "learning_rate": 4.809861095105678e-07,
+      "loss": 0.2722,
+      "step": 13730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.2608226204571915,
+      "learning_rate": 4.805503470309408e-07,
+      "loss": 0.2402,
+      "step": 13731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.364101019053066,
+      "learning_rate": 4.801147720713079e-07,
+      "loss": 0.2558,
+      "step": 13732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.025503460901223,
+      "learning_rate": 4.796793846497411e-07,
+      "loss": 0.2509,
+      "step": 13733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9664366367185016,
+      "learning_rate": 4.792441847843043e-07,
+      "loss": 0.2557,
+      "step": 13734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5237687523827645,
+      "learning_rate": 4.788091724930571e-07,
+      "loss": 0.2648,
+      "step": 13735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3612160673713032,
+      "learning_rate": 4.783743477940473e-07,
+      "loss": 0.2591,
+      "step": 13736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.576773378582313,
+      "learning_rate": 4.779397107053174e-07,
+      "loss": 0.2589,
+      "step": 13737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8327080019559063,
+      "learning_rate": 4.775052612449005e-07,
+      "loss": 0.2464,
+      "step": 13738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5667142764524086,
+      "learning_rate": 4.770709994308242e-07,
+      "loss": 0.4484,
+      "step": 13739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.785820436627895,
+      "learning_rate": 4.766369252811054e-07,
+      "loss": 0.2348,
+      "step": 13740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.524841412966559,
+      "learning_rate": 4.762030388137551e-07,
+      "loss": 0.2459,
+      "step": 13741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.5118561493442915,
+      "learning_rate": 4.75769340046775e-07,
+      "loss": 0.2508,
+      "step": 13742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.615071272129694,
+      "learning_rate": 4.7533582899816133e-07,
+      "loss": 0.2495,
+      "step": 13743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.296283413077024,
+      "learning_rate": 4.7490250568590137e-07,
+      "loss": 0.2477,
+      "step": 13744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.759273953850784,
+      "learning_rate": 4.744693701279735e-07,
+      "loss": 0.2501,
+      "step": 13745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.827686209491277,
+      "learning_rate": 4.7403642234234935e-07,
+      "loss": 0.2392,
+      "step": 13746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9542298104846654,
+      "learning_rate": 4.736036623469936e-07,
+      "loss": 0.2857,
+      "step": 13747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6202546621098888,
+      "learning_rate": 4.731710901598624e-07,
+      "loss": 0.2453,
+      "step": 13748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6092639131500509,
+      "learning_rate": 4.7273870579890247e-07,
+      "loss": 0.455,
+      "step": 13749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5931816960150724,
+      "learning_rate": 4.7230650928205447e-07,
+      "loss": 0.2485,
+      "step": 13750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.50300836641801,
+      "learning_rate": 4.7187450062725126e-07,
+      "loss": 0.2433,
+      "step": 13751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9782678999273278,
+      "learning_rate": 4.7144267985241856e-07,
+      "loss": 0.2375,
+      "step": 13752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.880827791846489,
+      "learning_rate": 4.7101104697547307e-07,
+      "loss": 0.2495,
+      "step": 13753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4834969288427793,
+      "learning_rate": 4.7057960201432275e-07,
+      "loss": 0.2439,
+      "step": 13754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6071836496123604,
+      "learning_rate": 4.7014834498687045e-07,
+      "loss": 0.26,
+      "step": 13755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6627098367754756,
+      "learning_rate": 4.697172759110097e-07,
+      "loss": 0.2676,
+      "step": 13756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.042789929514817,
+      "learning_rate": 4.692863948046256e-07,
+      "loss": 0.2488,
+      "step": 13757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7472584622802503,
+      "learning_rate": 4.6885570168559493e-07,
+      "loss": 0.2474,
+      "step": 13758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9202966305175444,
+      "learning_rate": 4.6842519657179066e-07,
+      "loss": 0.2454,
+      "step": 13759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9996697367178797,
+      "learning_rate": 4.67994879481074e-07,
+      "loss": 0.249,
+      "step": 13760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6712291343920274,
+      "learning_rate": 4.6756475043130024e-07,
+      "loss": 0.2482,
+      "step": 13761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5700233439317361,
+      "learning_rate": 4.671348094403139e-07,
+      "loss": 0.2602,
+      "step": 13762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.035639140420502,
+      "learning_rate": 4.667050565259568e-07,
+      "loss": 0.2517,
+      "step": 13763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.480546374106348,
+      "learning_rate": 4.662754917060591e-07,
+      "loss": 0.2502,
+      "step": 13764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.413137043776564,
+      "learning_rate": 4.658461149984439e-07,
+      "loss": 0.259,
+      "step": 13765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9089607350914712,
+      "learning_rate": 4.654169264209246e-07,
+      "loss": 0.2489,
+      "step": 13766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7866982543067795,
+      "learning_rate": 4.649879259913137e-07,
+      "loss": 0.25,
+      "step": 13767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6881030706778166,
+      "learning_rate": 4.645591137274091e-07,
+      "loss": 0.2475,
+      "step": 13768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.746952614699426,
+      "learning_rate": 4.6413048964700224e-07,
+      "loss": 0.2604,
+      "step": 13769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.169900530410927,
+      "learning_rate": 4.637020537678771e-07,
+      "loss": 0.2567,
+      "step": 13770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6558378005064505,
+      "learning_rate": 4.6327380610781235e-07,
+      "loss": 0.2552,
+      "step": 13771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.41977707890803,
+      "learning_rate": 4.628457466845754e-07,
+      "loss": 0.269,
+      "step": 13772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9287978507545889,
+      "learning_rate": 4.6241787551592645e-07,
+      "loss": 0.2408,
+      "step": 13773
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0971421383884636,
+      "learning_rate": 4.619901926196191e-07,
+      "loss": 0.2337,
+      "step": 13774
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8631539706243492,
+      "learning_rate": 4.615626980134003e-07,
+      "loss": 0.231,
+      "step": 13775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.092877746318571,
+      "learning_rate": 4.611353917150063e-07,
+      "loss": 0.2634,
+      "step": 13776
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.893296375540724,
+      "learning_rate": 4.6070827374216697e-07,
+      "loss": 0.2471,
+      "step": 13777
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5562991316454764,
+      "learning_rate": 4.602813441126025e-07,
+      "loss": 0.2428,
+      "step": 13778
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4261447464071324,
+      "learning_rate": 4.5985460284403037e-07,
+      "loss": 0.2529,
+      "step": 13779
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9234299626647677,
+      "learning_rate": 4.5942804995415423e-07,
+      "loss": 0.2584,
+      "step": 13780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9145088635375673,
+      "learning_rate": 4.590016854606727e-07,
+      "loss": 0.2472,
+      "step": 13781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.63369464339832,
+      "learning_rate": 4.585755093812766e-07,
+      "loss": 0.2552,
+      "step": 13782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.292750135238071,
+      "learning_rate": 4.5814952173365067e-07,
+      "loss": 0.2442,
+      "step": 13783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4874279774455776,
+      "learning_rate": 4.57723722535468e-07,
+      "loss": 0.2508,
+      "step": 13784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.206289831535517,
+      "learning_rate": 4.5729811180439567e-07,
+      "loss": 0.2893,
+      "step": 13785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.633603228069572,
+      "learning_rate": 4.568726895580933e-07,
+      "loss": 0.2497,
+      "step": 13786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9702627800297179,
+      "learning_rate": 4.5644745581421293e-07,
+      "loss": 0.2472,
+      "step": 13787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9207013771987873,
+      "learning_rate": 4.560224105903971e-07,
+      "loss": 0.2396,
+      "step": 13788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1453534898516478,
+      "learning_rate": 4.5559755390428284e-07,
+      "loss": 0.2474,
+      "step": 13789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5964527090143048,
+      "learning_rate": 4.5517288577349773e-07,
+      "loss": 0.2507,
+      "step": 13790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.745744419491947,
+      "learning_rate": 4.5474840621566264e-07,
+      "loss": 0.2604,
+      "step": 13791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.319181385274683,
+      "learning_rate": 4.543241152483896e-07,
+      "loss": 0.2559,
+      "step": 13792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4470094108943803,
+      "learning_rate": 4.539000128892829e-07,
+      "loss": 0.2411,
+      "step": 13793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.677805217558302,
+      "learning_rate": 4.534760991559384e-07,
+      "loss": 0.2533,
+      "step": 13794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9837572300801964,
+      "learning_rate": 4.5305237406594705e-07,
+      "loss": 0.2482,
+      "step": 13795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.477426482139802,
+      "learning_rate": 4.526288376368887e-07,
+      "loss": 0.2587,
+      "step": 13796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5367189378808117,
+      "learning_rate": 4.5220548988633707e-07,
+      "loss": 0.2347,
+      "step": 13797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7406631956229388,
+      "learning_rate": 4.5178233083185694e-07,
+      "loss": 0.2459,
+      "step": 13798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0021300722332454,
+      "learning_rate": 4.5135936049100715e-07,
+      "loss": 0.2563,
+      "step": 13799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8486010054680313,
+      "learning_rate": 4.509365788813369e-07,
+      "loss": 0.2592,
+      "step": 13800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6365894065618019,
+      "learning_rate": 4.5051398602038775e-07,
+      "loss": 0.2453,
+      "step": 13801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6871080344047993,
+      "learning_rate": 4.5009158192569345e-07,
+      "loss": 0.2438,
+      "step": 13802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6626935079153413,
+      "learning_rate": 4.496693666147811e-07,
+      "loss": 0.2542,
+      "step": 13803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5095499566522133,
+      "learning_rate": 4.4924734010517e-07,
+      "loss": 0.2382,
+      "step": 13804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3281388263377885,
+      "learning_rate": 4.488255024143695e-07,
+      "loss": 0.234,
+      "step": 13805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0891323273649456,
+      "learning_rate": 4.484038535598817e-07,
+      "loss": 0.2299,
+      "step": 13806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7026057419641507,
+      "learning_rate": 4.479823935592037e-07,
+      "loss": 0.2737,
+      "step": 13807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.185720517961521,
+      "learning_rate": 4.4756112242982153e-07,
+      "loss": 0.2582,
+      "step": 13808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.930030917707115,
+      "learning_rate": 4.471400401892145e-07,
+      "loss": 0.2391,
+      "step": 13809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 46.063136179670856,
+      "learning_rate": 4.4671914685485317e-07,
+      "loss": 0.2607,
+      "step": 13810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6535333465397057,
+      "learning_rate": 4.4629844244420126e-07,
+      "loss": 0.2566,
+      "step": 13811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7253326486467209,
+      "learning_rate": 4.458779269747171e-07,
+      "loss": 0.2543,
+      "step": 13812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7635589860368603,
+      "learning_rate": 4.4545760046384614e-07,
+      "loss": 0.2454,
+      "step": 13813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5949430510246059,
+      "learning_rate": 4.4503746292902825e-07,
+      "loss": 0.5078,
+      "step": 13814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6246812372877706,
+      "learning_rate": 4.4461751438769794e-07,
+      "loss": 0.2505,
+      "step": 13815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6625477921143004,
+      "learning_rate": 4.441977548572779e-07,
+      "loss": 0.2535,
+      "step": 13816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0977268379278056,
+      "learning_rate": 4.4377818435518474e-07,
+      "loss": 0.2446,
+      "step": 13817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5827314666332564,
+      "learning_rate": 4.433588028988267e-07,
+      "loss": 0.2744,
+      "step": 13818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8016090532650186,
+      "learning_rate": 4.4293961050560607e-07,
+      "loss": 0.2382,
+      "step": 13819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.132460261241813,
+      "learning_rate": 4.4252060719291556e-07,
+      "loss": 0.2425,
+      "step": 13820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5789131835596515,
+      "learning_rate": 4.4210179297814016e-07,
+      "loss": 0.2624,
+      "step": 13821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2596280978187036,
+      "learning_rate": 4.4168316787865594e-07,
+      "loss": 0.2693,
+      "step": 13822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.649634793274662,
+      "learning_rate": 4.412647319118346e-07,
+      "loss": 0.2507,
+      "step": 13823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.694741837403251,
+      "learning_rate": 4.4084648509503724e-07,
+      "loss": 0.2551,
+      "step": 13824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.948856281137073,
+      "learning_rate": 4.404284274456161e-07,
+      "loss": 0.2437,
+      "step": 13825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0648872061383656,
+      "learning_rate": 4.4001055898091894e-07,
+      "loss": 0.2397,
+      "step": 13826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.172524952943479,
+      "learning_rate": 4.395928797182819e-07,
+      "loss": 0.2837,
+      "step": 13827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4186274080291301,
+      "learning_rate": 4.391753896750378e-07,
+      "loss": 0.2512,
+      "step": 13828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.491560574259554,
+      "learning_rate": 4.3875808886850777e-07,
+      "loss": 0.226,
+      "step": 13829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.410133617998824,
+      "learning_rate": 4.383409773160052e-07,
+      "loss": 0.2469,
+      "step": 13830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6049657604459995,
+      "learning_rate": 4.3792405503483903e-07,
+      "loss": 0.2737,
+      "step": 13831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0627550030741637,
+      "learning_rate": 4.375073220423065e-07,
+      "loss": 0.2654,
+      "step": 13832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2355520813069347,
+      "learning_rate": 4.370907783556988e-07,
+      "loss": 0.2291,
+      "step": 13833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.11886330585489,
+      "learning_rate": 4.3667442399229985e-07,
+      "loss": 0.242,
+      "step": 13834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6402330554041666,
+      "learning_rate": 4.362582589693837e-07,
+      "loss": 0.2468,
+      "step": 13835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.214726740768506,
+      "learning_rate": 4.358422833042192e-07,
+      "loss": 0.2672,
+      "step": 13836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6252222120775173,
+      "learning_rate": 4.354264970140654e-07,
+      "loss": 0.259,
+      "step": 13837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9010466084452147,
+      "learning_rate": 4.3501090011617286e-07,
+      "loss": 0.2659,
+      "step": 13838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9746578949649074,
+      "learning_rate": 4.3459549262778736e-07,
+      "loss": 0.2697,
+      "step": 13839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6810623032065262,
+      "learning_rate": 4.3418027456614277e-07,
+      "loss": 0.2365,
+      "step": 13840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.972024926647831,
+      "learning_rate": 4.337652459484698e-07,
+      "loss": 0.2651,
+      "step": 13841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0992942064946996,
+      "learning_rate": 4.333504067919858e-07,
+      "loss": 0.2449,
+      "step": 13842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4261767185343013,
+      "learning_rate": 4.329357571139059e-07,
+      "loss": 0.266,
+      "step": 13843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6206963025094863,
+      "learning_rate": 4.3252129693143353e-07,
+      "loss": 0.2383,
+      "step": 13844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9267780552117144,
+      "learning_rate": 4.321070262617655e-07,
+      "loss": 0.2376,
+      "step": 13845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.13778457696387,
+      "learning_rate": 4.3169294512208917e-07,
+      "loss": 0.2397,
+      "step": 13846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8575974450737254,
+      "learning_rate": 4.3127905352958796e-07,
+      "loss": 0.2455,
+      "step": 13847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2793197489275094,
+      "learning_rate": 4.308653515014327e-07,
+      "loss": 0.2686,
+      "step": 13848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2701318796487207,
+      "learning_rate": 4.304518390547907e-07,
+      "loss": 0.2414,
+      "step": 13849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.41990295399939,
+      "learning_rate": 4.300385162068177e-07,
+      "loss": 0.2486,
+      "step": 13850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0833381921166843,
+      "learning_rate": 4.296253829746644e-07,
+      "loss": 0.2442,
+      "step": 13851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.911491157143949,
+      "learning_rate": 4.292124393754715e-07,
+      "loss": 0.2502,
+      "step": 13852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8997245751700704,
+      "learning_rate": 4.287996854263737e-07,
+      "loss": 0.245,
+      "step": 13853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.068283235664751,
+      "learning_rate": 4.283871211444951e-07,
+      "loss": 0.2574,
+      "step": 13854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7681632505693208,
+      "learning_rate": 4.2797474654695527e-07,
+      "loss": 0.2595,
+      "step": 13855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.108444660808345,
+      "learning_rate": 4.2756256165086443e-07,
+      "loss": 0.2533,
+      "step": 13856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.354750050889478,
+      "learning_rate": 4.27150566473325e-07,
+      "loss": 0.2423,
+      "step": 13857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4171742333104715,
+      "learning_rate": 4.267387610314294e-07,
+      "loss": 0.2518,
+      "step": 13858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.537594496454954,
+      "learning_rate": 4.2632714534226685e-07,
+      "loss": 0.2562,
+      "step": 13859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4358005144862205,
+      "learning_rate": 4.259157194229152e-07,
+      "loss": 0.2552,
+      "step": 13860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.960417309318742,
+      "learning_rate": 4.2550448329044426e-07,
+      "loss": 0.2631,
+      "step": 13861
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0160656549023637,
+      "learning_rate": 4.2509343696191695e-07,
+      "loss": 0.2341,
+      "step": 13862
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1833622724811916,
+      "learning_rate": 4.246825804543886e-07,
+      "loss": 0.2584,
+      "step": 13863
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.0390217373261255,
+      "learning_rate": 4.242719137849077e-07,
+      "loss": 0.2477,
+      "step": 13864
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1054123638811757,
+      "learning_rate": 4.238614369705124e-07,
+      "loss": 0.2627,
+      "step": 13865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.906825018692233,
+      "learning_rate": 4.2345115002823345e-07,
+      "loss": 0.2481,
+      "step": 13866
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.700215689055635,
+      "learning_rate": 4.2304105297509614e-07,
+      "loss": 0.2509,
+      "step": 13867
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6226385486277936,
+      "learning_rate": 4.2263114582811515e-07,
+      "loss": 0.4667,
+      "step": 13868
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.782780375876242,
+      "learning_rate": 4.2222142860429805e-07,
+      "loss": 0.2603,
+      "step": 13869
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.428520394096701,
+      "learning_rate": 4.2181190132064396e-07,
+      "loss": 0.2606,
+      "step": 13870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8714541166991,
+      "learning_rate": 4.2140256399414603e-07,
+      "loss": 0.25,
+      "step": 13871
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.749954869223884,
+      "learning_rate": 4.2099341664178894e-07,
+      "loss": 0.2418,
+      "step": 13872
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5808420850754918,
+      "learning_rate": 4.205844592805486e-07,
+      "loss": 0.2579,
+      "step": 13873
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6077489134122638,
+      "learning_rate": 4.2017569192739195e-07,
+      "loss": 0.2445,
+      "step": 13874
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2296919846073795,
+      "learning_rate": 4.1976711459928096e-07,
+      "loss": 0.2458,
+      "step": 13875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3033601982153282,
+      "learning_rate": 4.193587273131683e-07,
+      "loss": 0.2425,
+      "step": 13876
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5669571185389406,
+      "learning_rate": 4.189505300859981e-07,
+      "loss": 0.2415,
+      "step": 13877
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.371627591062167,
+      "learning_rate": 4.185425229347062e-07,
+      "loss": 0.2718,
+      "step": 13878
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5515246372638904,
+      "learning_rate": 4.1813470587622305e-07,
+      "loss": 0.2468,
+      "step": 13879
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5876116038820525,
+      "learning_rate": 4.1772707892746954e-07,
+      "loss": 0.2466,
+      "step": 13880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.609968141072775,
+      "learning_rate": 4.173196421053588e-07,
+      "loss": 0.2574,
+      "step": 13881
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7243799581860655,
+      "learning_rate": 4.1691239542679507e-07,
+      "loss": 0.2411,
+      "step": 13882
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.670650289852566,
+      "learning_rate": 4.165053389086776e-07,
+      "loss": 0.2344,
+      "step": 13883
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3328112575453748,
+      "learning_rate": 4.160984725678946e-07,
+      "loss": 0.2386,
+      "step": 13884
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.812917492437089,
+      "learning_rate": 4.1569179642132694e-07,
+      "loss": 0.2498,
+      "step": 13885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.894752647902837,
+      "learning_rate": 4.152853104858506e-07,
+      "loss": 0.2851,
+      "step": 13886
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7591454201584609,
+      "learning_rate": 4.148790147783288e-07,
+      "loss": 0.2484,
+      "step": 13887
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6124279603834912,
+      "learning_rate": 4.1447290931562136e-07,
+      "loss": 0.2714,
+      "step": 13888
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6797118342691032,
+      "learning_rate": 4.1406699411457804e-07,
+      "loss": 0.4849,
+      "step": 13889
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7244922133291847,
+      "learning_rate": 4.1366126919203996e-07,
+      "loss": 0.2346,
+      "step": 13890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.31265028502726,
+      "learning_rate": 4.1325573456484304e-07,
+      "loss": 0.2473,
+      "step": 13891
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.915228006423384,
+      "learning_rate": 4.1285039024981265e-07,
+      "loss": 0.2521,
+      "step": 13892
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.970610547757536,
+      "learning_rate": 4.12445236263766e-07,
+      "loss": 0.2559,
+      "step": 13893
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.088654997788465,
+      "learning_rate": 4.120402726235156e-07,
+      "loss": 0.2501,
+      "step": 13894
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5273454258558474,
+      "learning_rate": 4.116354993458632e-07,
+      "loss": 0.247,
+      "step": 13895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7837724964174653,
+      "learning_rate": 4.1123091644760413e-07,
+      "loss": 0.2514,
+      "step": 13896
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.43589432026048,
+      "learning_rate": 4.108265239455256e-07,
+      "loss": 0.2583,
+      "step": 13897
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7085514184862256,
+      "learning_rate": 4.104223218564046e-07,
+      "loss": 0.2497,
+      "step": 13898
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.785846022395972,
+      "learning_rate": 4.10018310197014e-07,
+      "loss": 0.2431,
+      "step": 13899
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3647852333575528,
+      "learning_rate": 4.096144889841158e-07,
+      "loss": 0.2541,
+      "step": 13900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.52349487523597,
+      "learning_rate": 4.092108582344673e-07,
+      "loss": 0.2454,
+      "step": 13901
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5695289241864465,
+      "learning_rate": 4.088074179648138e-07,
+      "loss": 0.2677,
+      "step": 13902
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3692451951898694,
+      "learning_rate": 4.084041681918949e-07,
+      "loss": 0.2382,
+      "step": 13903
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.023437322970993,
+      "learning_rate": 4.0800110893244314e-07,
+      "loss": 0.2872,
+      "step": 13904
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9305002568167204,
+      "learning_rate": 4.0759824020318197e-07,
+      "loss": 0.2657,
+      "step": 13905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9176846662156515,
+      "learning_rate": 4.0719556202082567e-07,
+      "loss": 0.277,
+      "step": 13906
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 44.749710952022866,
+      "learning_rate": 4.067930744020837e-07,
+      "loss": 0.2478,
+      "step": 13907
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5910365872232302,
+      "learning_rate": 4.0639077736365606e-07,
+      "loss": 0.448,
+      "step": 13908
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9684827752178375,
+      "learning_rate": 4.059886709222344e-07,
+      "loss": 0.2292,
+      "step": 13909
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9454508858254342,
+      "learning_rate": 4.055867550945025e-07,
+      "loss": 0.2514,
+      "step": 13910
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.549682611788474,
+      "learning_rate": 4.05185029897136e-07,
+      "loss": 0.2571,
+      "step": 13911
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9142156366389025,
+      "learning_rate": 4.0478349534680496e-07,
+      "loss": 0.2688,
+      "step": 13912
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6895084210184605,
+      "learning_rate": 4.043821514601681e-07,
+      "loss": 0.231,
+      "step": 13913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7273889713895971,
+      "learning_rate": 4.039809982538784e-07,
+      "loss": 0.2617,
+      "step": 13914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6850846852629544,
+      "learning_rate": 4.0358003574458017e-07,
+      "loss": 0.2704,
+      "step": 13915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.836940898758,
+      "learning_rate": 4.031792639489113e-07,
+      "loss": 0.2617,
+      "step": 13916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7979533425324887,
+      "learning_rate": 4.027786828834995e-07,
+      "loss": 0.2469,
+      "step": 13917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8854678790418142,
+      "learning_rate": 4.023782925649661e-07,
+      "loss": 0.2533,
+      "step": 13918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.0842772759433394,
+      "learning_rate": 4.0197809300992206e-07,
+      "loss": 0.2403,
+      "step": 13919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.74645143191841,
+      "learning_rate": 4.0157808423497537e-07,
+      "loss": 0.242,
+      "step": 13920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9685761744317802,
+      "learning_rate": 4.011782662567215e-07,
+      "loss": 0.2506,
+      "step": 13921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.701193911893507,
+      "learning_rate": 4.007786390917495e-07,
+      "loss": 0.2859,
+      "step": 13922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0616095332696727,
+      "learning_rate": 4.0037920275664e-07,
+      "loss": 0.2662,
+      "step": 13923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.16385087060391,
+      "learning_rate": 3.999799572679686e-07,
+      "loss": 0.2591,
+      "step": 13924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4668387854806826,
+      "learning_rate": 3.9958090264229923e-07,
+      "loss": 0.262,
+      "step": 13925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.007863733481297,
+      "learning_rate": 3.991820388961887e-07,
+      "loss": 0.2543,
+      "step": 13926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.130776382220719,
+      "learning_rate": 3.987833660461882e-07,
+      "loss": 0.2571,
+      "step": 13927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5468586335375507,
+      "learning_rate": 3.9838488410883837e-07,
+      "loss": 0.248,
+      "step": 13928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4178728606148994,
+      "learning_rate": 3.979865931006732e-07,
+      "loss": 0.2422,
+      "step": 13929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9995809761593755,
+      "learning_rate": 3.975884930382173e-07,
+      "loss": 0.2523,
+      "step": 13930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.640397388681757,
+      "learning_rate": 3.971905839379903e-07,
+      "loss": 0.2557,
+      "step": 13931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5824689224284922,
+      "learning_rate": 3.9679286581650155e-07,
+      "loss": 0.4839,
+      "step": 13932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1923699861803274,
+      "learning_rate": 3.963953386902536e-07,
+      "loss": 0.2433,
+      "step": 13933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6409875478485179,
+      "learning_rate": 3.9599800257573873e-07,
+      "loss": 0.2524,
+      "step": 13934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5029582875661958,
+      "learning_rate": 3.9560085748944545e-07,
+      "loss": 0.2452,
+      "step": 13935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4388595656727534,
+      "learning_rate": 3.9520390344785053e-07,
+      "loss": 0.2566,
+      "step": 13936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8138724186168833,
+      "learning_rate": 3.948071404674242e-07,
+      "loss": 0.2362,
+      "step": 13937
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.1889370186957295,
+      "learning_rate": 3.944105685646299e-07,
+      "loss": 0.2666,
+      "step": 13938
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7682123636206566,
+      "learning_rate": 3.940141877559212e-07,
+      "loss": 0.2374,
+      "step": 13939
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6072873298661396,
+      "learning_rate": 3.9361799805774535e-07,
+      "loss": 0.2722,
+      "step": 13940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.566699737378959,
+      "learning_rate": 3.9322199948654104e-07,
+      "loss": 0.2591,
+      "step": 13941
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9252607057894588,
+      "learning_rate": 3.928261920587373e-07,
+      "loss": 0.2651,
+      "step": 13942
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3426039067240128,
+      "learning_rate": 3.9243057579075925e-07,
+      "loss": 0.2664,
+      "step": 13943
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7662689443890587,
+      "learning_rate": 3.9203515069902055e-07,
+      "loss": 0.2601,
+      "step": 13944
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.124532112595285,
+      "learning_rate": 3.9163991679992684e-07,
+      "loss": 0.2439,
+      "step": 13945
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6089708090188581,
+      "learning_rate": 3.9124487410987955e-07,
+      "loss": 0.2543,
+      "step": 13946
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.599972590825963,
+      "learning_rate": 3.9085002264526775e-07,
+      "loss": 0.2804,
+      "step": 13947
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4550784722978682,
+      "learning_rate": 3.904553624224761e-07,
+      "loss": 0.2425,
+      "step": 13948
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7191182616936205,
+      "learning_rate": 3.900608934578787e-07,
+      "loss": 0.2789,
+      "step": 13949
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2779646523062587,
+      "learning_rate": 3.896666157678425e-07,
+      "loss": 0.2508,
+      "step": 13950
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2814600805195453,
+      "learning_rate": 3.8927252936872774e-07,
+      "loss": 0.2458,
+      "step": 13951
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8321334127552613,
+      "learning_rate": 3.8887863427688463e-07,
+      "loss": 0.2674,
+      "step": 13952
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8102398638850765,
+      "learning_rate": 3.884849305086585e-07,
+      "loss": 0.2571,
+      "step": 13953
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.041175786355012,
+      "learning_rate": 3.88091418080383e-07,
+      "loss": 0.2591,
+      "step": 13954
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.72017722148624,
+      "learning_rate": 3.8769809700838546e-07,
+      "loss": 0.2519,
+      "step": 13955
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3002952112703197,
+      "learning_rate": 3.8730496730898735e-07,
+      "loss": 0.2631,
+      "step": 13956
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9066585536648852,
+      "learning_rate": 3.86912028998499e-07,
+      "loss": 0.2697,
+      "step": 13957
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.880663232172344,
+      "learning_rate": 3.865192820932234e-07,
+      "loss": 0.2556,
+      "step": 13958
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5605055531928658,
+      "learning_rate": 3.861267266094587e-07,
+      "loss": 0.2604,
+      "step": 13959
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4718300748092226,
+      "learning_rate": 3.857343625634896e-07,
+      "loss": 0.2539,
+      "step": 13960
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3417541065616234,
+      "learning_rate": 3.8534218997159923e-07,
+      "loss": 0.2506,
+      "step": 13961
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.137743805169988,
+      "learning_rate": 3.8495020885005784e-07,
+      "loss": 0.2604,
+      "step": 13962
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5818669127051017,
+      "learning_rate": 3.8455841921512803e-07,
+      "loss": 0.261,
+      "step": 13963
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9518111291864557,
+      "learning_rate": 3.8416682108306904e-07,
+      "loss": 0.2433,
+      "step": 13964
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0509379586656284,
+      "learning_rate": 3.837754144701267e-07,
+      "loss": 0.2475,
+      "step": 13965
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8171415430236428,
+      "learning_rate": 3.8338419939254135e-07,
+      "loss": 0.2511,
+      "step": 13966
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6052968104197836,
+      "learning_rate": 3.8299317586654503e-07,
+      "loss": 0.2385,
+      "step": 13967
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8037691832569482,
+      "learning_rate": 3.826023439083637e-07,
+      "loss": 0.2419,
+      "step": 13968
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9415721273173152,
+      "learning_rate": 3.822117035342127e-07,
+      "loss": 0.2567,
+      "step": 13969
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.835155764947781,
+      "learning_rate": 3.8182125476030007e-07,
+      "loss": 0.2549,
+      "step": 13970
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.448693023832294,
+      "learning_rate": 3.8143099760282574e-07,
+      "loss": 0.2523,
+      "step": 13971
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7763422546295775,
+      "learning_rate": 3.8104093207798343e-07,
+      "loss": 0.2626,
+      "step": 13972
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.806042382047809,
+      "learning_rate": 3.8065105820195736e-07,
+      "loss": 0.2539,
+      "step": 13973
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.449936481099475,
+      "learning_rate": 3.802613759909224e-07,
+      "loss": 0.2609,
+      "step": 13974
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4197453166483975,
+      "learning_rate": 3.798718854610489e-07,
+      "loss": 0.2458,
+      "step": 13975
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3431706323038983,
+      "learning_rate": 3.794825866284979e-07,
+      "loss": 0.2467,
+      "step": 13976
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.621411461075429,
+      "learning_rate": 3.7909347950942145e-07,
+      "loss": 0.2604,
+      "step": 13977
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9454255595883063,
+      "learning_rate": 3.787045641199638e-07,
+      "loss": 0.2576,
+      "step": 13978
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.528482977675164,
+      "learning_rate": 3.783158404762616e-07,
+      "loss": 0.2829,
+      "step": 13979
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0598896829424405,
+      "learning_rate": 3.7792730859444515e-07,
+      "loss": 0.244,
+      "step": 13980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5935719863410593,
+      "learning_rate": 3.775389684906344e-07,
+      "loss": 0.2599,
+      "step": 13981
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.389542266877843,
+      "learning_rate": 3.771508201809415e-07,
+      "loss": 0.2557,
+      "step": 13982
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6116713768741266,
+      "learning_rate": 3.7676286368147184e-07,
+      "loss": 0.2321,
+      "step": 13983
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6815315489178015,
+      "learning_rate": 3.763750990083237e-07,
+      "loss": 0.2549,
+      "step": 13984
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3457852196671523,
+      "learning_rate": 3.7598752617758527e-07,
+      "loss": 0.2725,
+      "step": 13985
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2386337262106424,
+      "learning_rate": 3.756001452053376e-07,
+      "loss": 0.2878,
+      "step": 13986
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9042725851865951,
+      "learning_rate": 3.7521295610765287e-07,
+      "loss": 0.246,
+      "step": 13987
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2466391317913295,
+      "learning_rate": 3.748259589005976e-07,
+      "loss": 0.265,
+      "step": 13988
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.748354744579556,
+      "learning_rate": 3.744391536002279e-07,
+      "loss": 0.2495,
+      "step": 13989
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.063570841130431,
+      "learning_rate": 3.7405254022259475e-07,
+      "loss": 0.248,
+      "step": 13990
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4251580723790165,
+      "learning_rate": 3.73666118783737e-07,
+      "loss": 0.2545,
+      "step": 13991
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8153410444853335,
+      "learning_rate": 3.7327988929969083e-07,
+      "loss": 0.2422,
+      "step": 13992
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5122610022769543,
+      "learning_rate": 3.728938517864794e-07,
+      "loss": 0.2398,
+      "step": 13993
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.528630755370866,
+      "learning_rate": 3.725080062601211e-07,
+      "loss": 0.2474,
+      "step": 13994
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.520792065023425,
+      "learning_rate": 3.721223527366241e-07,
+      "loss": 0.2541,
+      "step": 13995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.040619967891763,
+      "learning_rate": 3.7173689123199133e-07,
+      "loss": 0.2395,
+      "step": 13996
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4496151423932235,
+      "learning_rate": 3.7135162176221483e-07,
+      "loss": 0.2449,
+      "step": 13997
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6070179068284256,
+      "learning_rate": 3.7096654434328197e-07,
+      "loss": 0.2529,
+      "step": 13998
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5364656885845123,
+      "learning_rate": 3.7058165899116816e-07,
+      "loss": 0.2596,
+      "step": 13999
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9744628668453355,
+      "learning_rate": 3.7019696572184515e-07,
+      "loss": 0.2331,
+      "step": 14000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.073741734694081,
+      "learning_rate": 3.698124645512735e-07,
+      "loss": 0.2638,
+      "step": 14001
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6839561431804393,
+      "learning_rate": 3.69428155495406e-07,
+      "loss": 0.2586,
+      "step": 14002
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.968946990777144,
+      "learning_rate": 3.690440385701899e-07,
+      "loss": 0.2401,
+      "step": 14003
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4699091569746803,
+      "learning_rate": 3.686601137915613e-07,
+      "loss": 0.2474,
+      "step": 14004
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.703743291144797,
+      "learning_rate": 3.6827638117545193e-07,
+      "loss": 0.2546,
+      "step": 14005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3149306029380574,
+      "learning_rate": 3.6789284073778187e-07,
+      "loss": 0.24,
+      "step": 14006
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.82716596559018,
+      "learning_rate": 3.67509492494465e-07,
+      "loss": 0.253,
+      "step": 14007
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.057829474780221,
+      "learning_rate": 3.6712633646140805e-07,
+      "loss": 0.251,
+      "step": 14008
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9292991903661747,
+      "learning_rate": 3.667433726545083e-07,
+      "loss": 0.2547,
+      "step": 14009
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4976106902865847,
+      "learning_rate": 3.6636060108965533e-07,
+      "loss": 0.2447,
+      "step": 14010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3557776302475415,
+      "learning_rate": 3.6597802178273136e-07,
+      "loss": 0.2606,
+      "step": 14011
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4506621013509584,
+      "learning_rate": 3.655956347496098e-07,
+      "loss": 0.2412,
+      "step": 14012
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.949391643633007,
+      "learning_rate": 3.6521344000615746e-07,
+      "loss": 0.2685,
+      "step": 14013
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8876606327230736,
+      "learning_rate": 3.6483143756823215e-07,
+      "loss": 0.2476,
+      "step": 14014
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.563638569520903,
+      "learning_rate": 3.644496274516829e-07,
+      "loss": 0.2509,
+      "step": 14015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.964949734790135,
+      "learning_rate": 3.6406800967235255e-07,
+      "loss": 0.2532,
+      "step": 14016
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4108558848561747,
+      "learning_rate": 3.636865842460752e-07,
+      "loss": 0.2566,
+      "step": 14017
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.420700808848416,
+      "learning_rate": 3.633053511886753e-07,
+      "loss": 0.2523,
+      "step": 14018
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.425152784819349,
+      "learning_rate": 3.6292431051597244e-07,
+      "loss": 0.2413,
+      "step": 14019
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4925792531002458,
+      "learning_rate": 3.625434622437768e-07,
+      "loss": 0.2517,
+      "step": 14020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7139325459117283,
+      "learning_rate": 3.6216280638789013e-07,
+      "loss": 0.2476,
+      "step": 14021
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5914139311807765,
+      "learning_rate": 3.6178234296410595e-07,
+      "loss": 0.267,
+      "step": 14022
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.605044548152336,
+      "learning_rate": 3.614020719882105e-07,
+      "loss": 0.2739,
+      "step": 14023
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5833532287491568,
+      "learning_rate": 3.610219934759829e-07,
+      "loss": 0.4876,
+      "step": 14024
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5081303045273848,
+      "learning_rate": 3.606421074431926e-07,
+      "loss": 0.2441,
+      "step": 14025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.640048597494044,
+      "learning_rate": 3.6026241390560045e-07,
+      "loss": 0.263,
+      "step": 14026
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6926999195341494,
+      "learning_rate": 3.5988291287896216e-07,
+      "loss": 0.2586,
+      "step": 14027
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.026754301064587,
+      "learning_rate": 3.5950360437902454e-07,
+      "loss": 0.2507,
+      "step": 14028
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4190361609887163,
+      "learning_rate": 3.591244884215245e-07,
+      "loss": 0.2418,
+      "step": 14029
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8392270202784946,
+      "learning_rate": 3.587455650221927e-07,
+      "loss": 0.2717,
+      "step": 14030
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9051917167279349,
+      "learning_rate": 3.5836683419675056e-07,
+      "loss": 0.2445,
+      "step": 14031
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.905415648496474,
+      "learning_rate": 3.5798829596091325e-07,
+      "loss": 0.265,
+      "step": 14032
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.982409193120526,
+      "learning_rate": 3.57609950330387e-07,
+      "loss": 0.2652,
+      "step": 14033
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6893216064066778,
+      "learning_rate": 3.572317973208689e-07,
+      "loss": 0.2448,
+      "step": 14034
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6312148138750535,
+      "learning_rate": 3.568538369480501e-07,
+      "loss": 0.2357,
+      "step": 14035
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.7249586999456525,
+      "learning_rate": 3.564760692276137e-07,
+      "loss": 0.252,
+      "step": 14036
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.150511611161764,
+      "learning_rate": 3.560984941752327e-07,
+      "loss": 0.2446,
+      "step": 14037
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.2946408833896355,
+      "learning_rate": 3.5572111180657396e-07,
+      "loss": 0.2455,
+      "step": 14038
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7397380751360467,
+      "learning_rate": 3.553439221372945e-07,
+      "loss": 0.2622,
+      "step": 14039
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.069791112226758,
+      "learning_rate": 3.549669251830462e-07,
+      "loss": 0.2431,
+      "step": 14040
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3324155647873717,
+      "learning_rate": 3.5459012095947097e-07,
+      "loss": 0.2493,
+      "step": 14041
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3983670371869428,
+      "learning_rate": 3.542135094822019e-07,
+      "loss": 0.2521,
+      "step": 14042
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.900478955509874,
+      "learning_rate": 3.538370907668659e-07,
+      "loss": 0.2554,
+      "step": 14043
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3513096505059168,
+      "learning_rate": 3.5346086482908273e-07,
+      "loss": 0.2529,
+      "step": 14044
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.418736693991127,
+      "learning_rate": 3.5308483168446094e-07,
+      "loss": 0.2541,
+      "step": 14045
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6215455821904665,
+      "learning_rate": 3.527089913486037e-07,
+      "loss": 0.2604,
+      "step": 14046
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.033883167194892,
+      "learning_rate": 3.5233334383710406e-07,
+      "loss": 0.2517,
+      "step": 14047
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6772044399704664,
+      "learning_rate": 3.5195788916555005e-07,
+      "loss": 0.2459,
+      "step": 14048
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2660530849734006,
+      "learning_rate": 3.515826273495182e-07,
+      "loss": 0.2619,
+      "step": 14049
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.069942588734494,
+      "learning_rate": 3.5120755840458043e-07,
+      "loss": 0.271,
+      "step": 14050
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.523335034358723,
+      "learning_rate": 3.5083268234629763e-07,
+      "loss": 0.2791,
+      "step": 14051
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7345082687962807,
+      "learning_rate": 3.5045799919022515e-07,
+      "loss": 0.2497,
+      "step": 14052
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2259196000159602,
+      "learning_rate": 3.500835089519089e-07,
+      "loss": 0.249,
+      "step": 14053
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.157667260164587,
+      "learning_rate": 3.4970921164688755e-07,
+      "loss": 0.269,
+      "step": 14054
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.957867738843563,
+      "learning_rate": 3.4933510729068976e-07,
+      "loss": 0.2485,
+      "step": 14055
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1714570268075017,
+      "learning_rate": 3.4896119589883925e-07,
+      "loss": 0.287,
+      "step": 14056
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9125151174917374,
+      "learning_rate": 3.4858747748684916e-07,
+      "loss": 0.2367,
+      "step": 14057
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7969448507382682,
+      "learning_rate": 3.4821395207022767e-07,
+      "loss": 0.2535,
+      "step": 14058
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0609274176620787,
+      "learning_rate": 3.4784061966447124e-07,
+      "loss": 0.2704,
+      "step": 14059
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9087048246241596,
+      "learning_rate": 3.4746748028507084e-07,
+      "loss": 0.245,
+      "step": 14060
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.714360568446815,
+      "learning_rate": 3.4709453394750847e-07,
+      "loss": 0.2378,
+      "step": 14061
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5383303949841536,
+      "learning_rate": 3.46721780667259e-07,
+      "loss": 0.2581,
+      "step": 14062
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6783334254549231,
+      "learning_rate": 3.463492204597868e-07,
+      "loss": 0.2299,
+      "step": 14063
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.326612215247788,
+      "learning_rate": 3.459768533405511e-07,
+      "loss": 0.2436,
+      "step": 14064
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.981796445493719,
+      "learning_rate": 3.456046793250034e-07,
+      "loss": 0.2461,
+      "step": 14065
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5639625510084794,
+      "learning_rate": 3.4523269842858477e-07,
+      "loss": 0.2501,
+      "step": 14066
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1504799983494314,
+      "learning_rate": 3.448609106667283e-07,
+      "loss": 0.2593,
+      "step": 14067
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.439883229790237,
+      "learning_rate": 3.444893160548618e-07,
+      "loss": 0.2311,
+      "step": 14068
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7139580722442127,
+      "learning_rate": 3.4411791460840285e-07,
+      "loss": 0.2502,
+      "step": 14069
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.773717588972613,
+      "learning_rate": 3.4374670634276073e-07,
+      "loss": 0.2537,
+      "step": 14070
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7377046459612524,
+      "learning_rate": 3.4337569127333767e-07,
+      "loss": 0.2453,
+      "step": 14071
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3101557933414867,
+      "learning_rate": 3.4300486941552913e-07,
+      "loss": 0.2499,
+      "step": 14072
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5894446259716395,
+      "learning_rate": 3.426342407847205e-07,
+      "loss": 0.2514,
+      "step": 14073
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3154537038250578,
+      "learning_rate": 3.422638053962896e-07,
+      "loss": 0.2577,
+      "step": 14074
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.073380209818981,
+      "learning_rate": 3.418935632656062e-07,
+      "loss": 0.2504,
+      "step": 14075
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.567365734048435,
+      "learning_rate": 3.4152351440803254e-07,
+      "loss": 0.2563,
+      "step": 14076
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.903767169799387,
+      "learning_rate": 3.41153658838923e-07,
+      "loss": 0.252,
+      "step": 14077
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5611119479604987,
+      "learning_rate": 3.407839965736237e-07,
+      "loss": 0.2565,
+      "step": 14078
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1750747738002634,
+      "learning_rate": 3.4041452762747007e-07,
+      "loss": 0.2578,
+      "step": 14079
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7385006166764045,
+      "learning_rate": 3.4004525201579543e-07,
+      "loss": 0.2648,
+      "step": 14080
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6157085561575384,
+      "learning_rate": 3.3967616975392026e-07,
+      "loss": 0.2698,
+      "step": 14081
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6148333518349004,
+      "learning_rate": 3.3930728085715903e-07,
+      "loss": 0.4513,
+      "step": 14082
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.023096653748799,
+      "learning_rate": 3.389385853408156e-07,
+      "loss": 0.2526,
+      "step": 14083
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7164957939293168,
+      "learning_rate": 3.385700832201905e-07,
+      "loss": 0.2379,
+      "step": 14084
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4329485094320742,
+      "learning_rate": 3.38201774510572e-07,
+      "loss": 0.2549,
+      "step": 14085
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9916654475637625,
+      "learning_rate": 3.378336592272419e-07,
+      "loss": 0.2469,
+      "step": 14086
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6054290856093585,
+      "learning_rate": 3.374657373854734e-07,
+      "loss": 0.2778,
+      "step": 14087
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7978478650048377,
+      "learning_rate": 3.3709800900053437e-07,
+      "loss": 0.2581,
+      "step": 14088
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5952629654325567,
+      "learning_rate": 3.367304740876809e-07,
+      "loss": 0.2488,
+      "step": 14089
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.246410194685025,
+      "learning_rate": 3.3636313266216304e-07,
+      "loss": 0.2753,
+      "step": 14090
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7010692136279622,
+      "learning_rate": 3.359959847392219e-07,
+      "loss": 0.2496,
+      "step": 14091
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4884789518705417,
+      "learning_rate": 3.35629030334092e-07,
+      "loss": 0.267,
+      "step": 14092
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9281994986521709,
+      "learning_rate": 3.352622694619989e-07,
+      "loss": 0.267,
+      "step": 14093
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.592981451803473,
+      "learning_rate": 3.3489570213815826e-07,
+      "loss": 0.2518,
+      "step": 14094
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.579066271172909,
+      "learning_rate": 3.3452932837778174e-07,
+      "loss": 0.2388,
+      "step": 14095
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.780184275461993,
+      "learning_rate": 3.3416314819607056e-07,
+      "loss": 0.2543,
+      "step": 14096
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.710008204718657,
+      "learning_rate": 3.3379716160821805e-07,
+      "loss": 0.2681,
+      "step": 14097
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8055841737255092,
+      "learning_rate": 3.334313686294094e-07,
+      "loss": 0.267,
+      "step": 14098
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.550078482054344,
+      "learning_rate": 3.330657692748213e-07,
+      "loss": 0.4701,
+      "step": 14099
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.820093325681551,
+      "learning_rate": 3.327003635596243e-07,
+      "loss": 0.2367,
+      "step": 14100
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.451465411921623,
+      "learning_rate": 3.3233515149897865e-07,
+      "loss": 0.2731,
+      "step": 14101
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1110474763313096,
+      "learning_rate": 3.319701331080394e-07,
+      "loss": 0.2602,
+      "step": 14102
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5133777270564037,
+      "learning_rate": 3.316053084019494e-07,
+      "loss": 0.2365,
+      "step": 14103
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.214333969472905,
+      "learning_rate": 3.312406773958482e-07,
+      "loss": 0.237,
+      "step": 14104
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.24721399983154,
+      "learning_rate": 3.3087624010486377e-07,
+      "loss": 0.2385,
+      "step": 14105
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.214415640869386,
+      "learning_rate": 3.305119965441178e-07,
+      "loss": 0.2538,
+      "step": 14106
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7871951376817026,
+      "learning_rate": 3.3014794672872165e-07,
+      "loss": 0.2494,
+      "step": 14107
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7727323497798866,
+      "learning_rate": 3.2978409067378315e-07,
+      "loss": 0.2539,
+      "step": 14108
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1630773104422993,
+      "learning_rate": 3.294204283943969e-07,
+      "loss": 0.2707,
+      "step": 14109
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4628341334900339,
+      "learning_rate": 3.2905695990565365e-07,
+      "loss": 0.2406,
+      "step": 14110
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.44670128812514,
+      "learning_rate": 3.28693685222633e-07,
+      "loss": 0.2572,
+      "step": 14111
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3998173995318388,
+      "learning_rate": 3.2833060436040955e-07,
+      "loss": 0.2351,
+      "step": 14112
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.009462336526005,
+      "learning_rate": 3.2796771733404673e-07,
+      "loss": 0.273,
+      "step": 14113
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.518736069927823,
+      "learning_rate": 3.2760502415860206e-07,
+      "loss": 0.2618,
+      "step": 14114
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2074491028281913,
+      "learning_rate": 3.2724252484912343e-07,
+      "loss": 0.2364,
+      "step": 14115
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4589832200098267,
+      "learning_rate": 3.268802194206516e-07,
+      "loss": 0.2741,
+      "step": 14116
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6906481773450206,
+      "learning_rate": 3.2651810788822125e-07,
+      "loss": 0.2707,
+      "step": 14117
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.642285376411281,
+      "learning_rate": 3.261561902668553e-07,
+      "loss": 0.2485,
+      "step": 14118
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.427392316989657,
+      "learning_rate": 3.257944665715701e-07,
+      "loss": 0.2606,
+      "step": 14119
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9736124977420078,
+      "learning_rate": 3.254329368173753e-07,
+      "loss": 0.2429,
+      "step": 14120
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.099895252274889,
+      "learning_rate": 3.2507160101927113e-07,
+      "loss": 0.2521,
+      "step": 14121
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.658880185049216,
+      "learning_rate": 3.247104591922495e-07,
+      "loss": 0.2555,
+      "step": 14122
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.07870804946324,
+      "learning_rate": 3.243495113512951e-07,
+      "loss": 0.2532,
+      "step": 14123
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3801652884066087,
+      "learning_rate": 3.239887575113837e-07,
+      "loss": 0.2649,
+      "step": 14124
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5853525265651602,
+      "learning_rate": 3.236281976874855e-07,
+      "loss": 0.2461,
+      "step": 14125
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4957718282050858,
+      "learning_rate": 3.2326783189455925e-07,
+      "loss": 0.2521,
+      "step": 14126
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9464437640196859,
+      "learning_rate": 3.229076601475567e-07,
+      "loss": 0.241,
+      "step": 14127
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9489841920359445,
+      "learning_rate": 3.225476824614238e-07,
+      "loss": 0.2653,
+      "step": 14128
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3831255692450444,
+      "learning_rate": 3.221878988510957e-07,
+      "loss": 0.2459,
+      "step": 14129
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.194250706861133,
+      "learning_rate": 3.2182830933149997e-07,
+      "loss": 0.2692,
+      "step": 14130
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6009420196293114,
+      "learning_rate": 3.2146891391755586e-07,
+      "loss": 0.2399,
+      "step": 14131
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0690534754572023,
+      "learning_rate": 3.21109712624178e-07,
+      "loss": 0.2596,
+      "step": 14132
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7260369369536,
+      "learning_rate": 3.207507054662684e-07,
+      "loss": 0.2451,
+      "step": 14133
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6769983034758233,
+      "learning_rate": 3.2039189245872404e-07,
+      "loss": 0.2526,
+      "step": 14134
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7042673355599904,
+      "learning_rate": 3.2003327361643077e-07,
+      "loss": 0.241,
+      "step": 14135
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.756754665829947,
+      "learning_rate": 3.196748489542706e-07,
+      "loss": 0.2406,
+      "step": 14136
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3749427363919344,
+      "learning_rate": 3.193166184871138e-07,
+      "loss": 0.2585,
+      "step": 14137
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0212558009921273,
+      "learning_rate": 3.189585822298241e-07,
+      "loss": 0.2545,
+      "step": 14138
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5790218799240625,
+      "learning_rate": 3.186007401972563e-07,
+      "loss": 0.2491,
+      "step": 14139
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.00351015696923,
+      "learning_rate": 3.182430924042601e-07,
+      "loss": 0.2524,
+      "step": 14140
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7627894836248061,
+      "learning_rate": 3.178856388656737e-07,
+      "loss": 0.2461,
+      "step": 14141
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0022439223692308,
+      "learning_rate": 3.175283795963291e-07,
+      "loss": 0.2889,
+      "step": 14142
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8741973059886514,
+      "learning_rate": 3.171713146110478e-07,
+      "loss": 0.2544,
+      "step": 14143
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.751331496634442,
+      "learning_rate": 3.168144439246468e-07,
+      "loss": 0.2607,
+      "step": 14144
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5805750487119463,
+      "learning_rate": 3.1645776755193314e-07,
+      "loss": 0.4583,
+      "step": 14145
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5341586703803891,
+      "learning_rate": 3.16101285507705e-07,
+      "loss": 0.2458,
+      "step": 14146
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.264454355592782,
+      "learning_rate": 3.1574499780675395e-07,
+      "loss": 0.2426,
+      "step": 14147
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5053034981439968,
+      "learning_rate": 3.1538890446386363e-07,
+      "loss": 0.2665,
+      "step": 14148
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6998832531170613,
+      "learning_rate": 3.1503300549380833e-07,
+      "loss": 0.2437,
+      "step": 14149
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5599972917303324,
+      "learning_rate": 3.1467730091135574e-07,
+      "loss": 0.2438,
+      "step": 14150
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7217109211200416,
+      "learning_rate": 3.143217907312629e-07,
+      "loss": 0.2439,
+      "step": 14151
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.288253222037634,
+      "learning_rate": 3.1396647496828245e-07,
+      "loss": 0.2578,
+      "step": 14152
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1278280755127885,
+      "learning_rate": 3.1361135363715544e-07,
+      "loss": 0.2558,
+      "step": 14153
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.190756776753083,
+      "learning_rate": 3.132564267526178e-07,
+      "loss": 0.2413,
+      "step": 14154
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.800167930523466,
+      "learning_rate": 3.1290169432939556e-07,
+      "loss": 0.2439,
+      "step": 14155
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8003421648981754,
+      "learning_rate": 3.1254715638220745e-07,
+      "loss": 0.2555,
+      "step": 14156
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6455429528094807,
+      "learning_rate": 3.12192812925764e-07,
+      "loss": 0.2428,
+      "step": 14157
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.964151007411487,
+      "learning_rate": 3.118386639747667e-07,
+      "loss": 0.2515,
+      "step": 14158
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7635693224675286,
+      "learning_rate": 3.1148470954391e-07,
+      "loss": 0.2377,
+      "step": 14159
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.644987155396305,
+      "learning_rate": 3.1113094964788095e-07,
+      "loss": 0.2636,
+      "step": 14160
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8162035050659775,
+      "learning_rate": 3.107773843013567e-07,
+      "loss": 0.2484,
+      "step": 14161
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0751770294342355,
+      "learning_rate": 3.104240135190084e-07,
+      "loss": 0.2448,
+      "step": 14162
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.366092576526622,
+      "learning_rate": 3.1007083731549705e-07,
+      "loss": 0.2435,
+      "step": 14163
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1585048475823325,
+      "learning_rate": 3.0971785570547696e-07,
+      "loss": 0.2738,
+      "step": 14164
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.872299106794302,
+      "learning_rate": 3.093650687035943e-07,
+      "loss": 0.2769,
+      "step": 14165
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6620275634652115,
+      "learning_rate": 3.090124763244867e-07,
+      "loss": 0.2637,
+      "step": 14166
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8573763061893935,
+      "learning_rate": 3.086600785827826e-07,
+      "loss": 0.2516,
+      "step": 14167
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.389102685457744,
+      "learning_rate": 3.0830787549310405e-07,
+      "loss": 0.2562,
+      "step": 14168
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.277507817959017,
+      "learning_rate": 3.079558670700666e-07,
+      "loss": 0.2492,
+      "step": 14169
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8974370453711717,
+      "learning_rate": 3.076040533282737e-07,
+      "loss": 0.2626,
+      "step": 14170
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.743171319336527,
+      "learning_rate": 3.07252434282323e-07,
+      "loss": 0.2562,
+      "step": 14171
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3050480895414043,
+      "learning_rate": 3.069010099468045e-07,
+      "loss": 0.2441,
+      "step": 14172
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.970789597954902,
+      "learning_rate": 3.0654978033629934e-07,
+      "loss": 0.2646,
+      "step": 14173
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5200774121937837,
+      "learning_rate": 3.0619874546537973e-07,
+      "loss": 0.2297,
+      "step": 14174
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0452896948858283,
+      "learning_rate": 3.058479053486113e-07,
+      "loss": 0.249,
+      "step": 14175
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.379949916591384,
+      "learning_rate": 3.0549726000055067e-07,
+      "loss": 0.257,
+      "step": 14176
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.079848781613622,
+      "learning_rate": 3.051468094357474e-07,
+      "loss": 0.2511,
+      "step": 14177
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3859222886999807,
+      "learning_rate": 3.047965536687425e-07,
+      "loss": 0.2462,
+      "step": 14178
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4474674964475036,
+      "learning_rate": 3.0444649271406723e-07,
+      "loss": 0.2438,
+      "step": 14179
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4229944900090208,
+      "learning_rate": 3.040966265862483e-07,
+      "loss": 0.2629,
+      "step": 14180
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9965843948030662,
+      "learning_rate": 3.037469552998007e-07,
+      "loss": 0.2595,
+      "step": 14181
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.320974293952236,
+      "learning_rate": 3.03397478869234e-07,
+      "loss": 0.2556,
+      "step": 14182
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.860521330464051,
+      "learning_rate": 3.0304819730904656e-07,
+      "loss": 0.2445,
+      "step": 14183
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.01671078890323,
+      "learning_rate": 3.0269911063373293e-07,
+      "loss": 0.248,
+      "step": 14184
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3216123341239854,
+      "learning_rate": 3.023502188577765e-07,
+      "loss": 0.2576,
+      "step": 14185
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7143623675867794,
+      "learning_rate": 3.02001521995654e-07,
+      "loss": 0.2564,
+      "step": 14186
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9182187327746425,
+      "learning_rate": 3.016530200618323e-07,
+      "loss": 0.236,
+      "step": 14187
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.512797203465373,
+      "learning_rate": 3.013047130707725e-07,
+      "loss": 0.248,
+      "step": 14188
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5181350357604373,
+      "learning_rate": 3.009566010369258e-07,
+      "loss": 0.2428,
+      "step": 14189
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5479538287869512,
+      "learning_rate": 3.006086839747363e-07,
+      "loss": 0.2248,
+      "step": 14190
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.25517044002195,
+      "learning_rate": 3.002609618986385e-07,
+      "loss": 0.244,
+      "step": 14191
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7239218655688637,
+      "learning_rate": 2.9991343482306144e-07,
+      "loss": 0.2516,
+      "step": 14192
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.15709818814808,
+      "learning_rate": 2.995661027624253e-07,
+      "loss": 0.2628,
+      "step": 14193
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.970519353518592,
+      "learning_rate": 2.9921896573114005e-07,
+      "loss": 0.2681,
+      "step": 14194
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.254104932453268,
+      "learning_rate": 2.988720237436088e-07,
+      "loss": 0.2515,
+      "step": 14195
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.833982346041316,
+      "learning_rate": 2.985252768142283e-07,
+      "loss": 0.2325,
+      "step": 14196
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8311408798011928,
+      "learning_rate": 2.981787249573842e-07,
+      "loss": 0.2432,
+      "step": 14197
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7006812690340607,
+      "learning_rate": 2.978323681874562e-07,
+      "loss": 0.2515,
+      "step": 14198
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.802188316984068,
+      "learning_rate": 2.97486206518815e-07,
+      "loss": 0.2527,
+      "step": 14199
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2877420965198394,
+      "learning_rate": 2.9714023996582407e-07,
+      "loss": 0.2373,
+      "step": 14200
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7574268777869833,
+      "learning_rate": 2.967944685428381e-07,
+      "loss": 0.2576,
+      "step": 14201
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4443036412371706,
+      "learning_rate": 2.964488922642034e-07,
+      "loss": 0.2607,
+      "step": 14202
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4874626471306946,
+      "learning_rate": 2.961035111442584e-07,
+      "loss": 0.2362,
+      "step": 14203
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2762819698663903,
+      "learning_rate": 2.9575832519733396e-07,
+      "loss": 0.2684,
+      "step": 14204
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5384853321987708,
+      "learning_rate": 2.9541333443775244e-07,
+      "loss": 0.2393,
+      "step": 14205
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8635726006801168,
+      "learning_rate": 2.950685388798269e-07,
+      "loss": 0.2597,
+      "step": 14206
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.778220642726601,
+      "learning_rate": 2.9472393853786473e-07,
+      "loss": 0.2505,
+      "step": 14207
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4520739525243302,
+      "learning_rate": 2.9437953342616453e-07,
+      "loss": 0.2337,
+      "step": 14208
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.024836258829966,
+      "learning_rate": 2.94035323559016e-07,
+      "loss": 0.2725,
+      "step": 14209
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1470186297578016,
+      "learning_rate": 2.936913089507004e-07,
+      "loss": 0.232,
+      "step": 14210
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.491248426266962,
+      "learning_rate": 2.9334748961549084e-07,
+      "loss": 0.2528,
+      "step": 14211
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1762916699011883,
+      "learning_rate": 2.9300386556765483e-07,
+      "loss": 0.2593,
+      "step": 14212
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.62343959964132,
+      "learning_rate": 2.926604368214486e-07,
+      "loss": 0.26,
+      "step": 14213
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.668567414873467,
+      "learning_rate": 2.92317203391122e-07,
+      "loss": 0.2283,
+      "step": 14214
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.203581337492975,
+      "learning_rate": 2.919741652909164e-07,
+      "loss": 0.2465,
+      "step": 14215
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.96762141276433,
+      "learning_rate": 2.916313225350653e-07,
+      "loss": 0.2685,
+      "step": 14216
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8209855745035428,
+      "learning_rate": 2.912886751377941e-07,
+      "loss": 0.2372,
+      "step": 14217
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9995706117839998,
+      "learning_rate": 2.909462231133192e-07,
+      "loss": 0.2564,
+      "step": 14218
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9604757938209916,
+      "learning_rate": 2.9060396647584867e-07,
+      "loss": 0.2531,
+      "step": 14219
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.60968839104951,
+      "learning_rate": 2.9026190523958553e-07,
+      "loss": 0.4309,
+      "step": 14220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.906944155396832,
+      "learning_rate": 2.899200394187202e-07,
+      "loss": 0.2336,
+      "step": 14221
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6227806585445994,
+      "learning_rate": 2.895783690274395e-07,
+      "loss": 0.2838,
+      "step": 14222
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1101751156505704,
+      "learning_rate": 2.892368940799184e-07,
+      "loss": 0.2715,
+      "step": 14223
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9735421721284456,
+      "learning_rate": 2.88895614590326e-07,
+      "loss": 0.2304,
+      "step": 14224
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.744707844567755,
+      "learning_rate": 2.885545305728227e-07,
+      "loss": 0.2486,
+      "step": 14225
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8419252848727798,
+      "learning_rate": 2.8821364204156045e-07,
+      "loss": 0.2437,
+      "step": 14226
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4988629913186162,
+      "learning_rate": 2.878729490106824e-07,
+      "loss": 0.2621,
+      "step": 14227
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1972532747467786,
+      "learning_rate": 2.875324514943256e-07,
+      "loss": 0.2395,
+      "step": 14228
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3631920326134166,
+      "learning_rate": 2.871921495066182e-07,
+      "loss": 0.2524,
+      "step": 14229
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.707015049201672,
+      "learning_rate": 2.8685204306168004e-07,
+      "loss": 0.2359,
+      "step": 14230
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2513505161088827,
+      "learning_rate": 2.865121321736203e-07,
+      "loss": 0.2479,
+      "step": 14231
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2149533956543466,
+      "learning_rate": 2.861724168565461e-07,
+      "loss": 0.2502,
+      "step": 14232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4707674493259502,
+      "learning_rate": 2.8583289712455e-07,
+      "loss": 0.2347,
+      "step": 14233
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7805922254398394,
+      "learning_rate": 2.8549357299172077e-07,
+      "loss": 0.2522,
+      "step": 14234
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.128419079823596,
+      "learning_rate": 2.851544444721366e-07,
+      "loss": 0.2654,
+      "step": 14235
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5927763243012352,
+      "learning_rate": 2.8481551157986896e-07,
+      "loss": 0.4553,
+      "step": 14236
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.457753305233146,
+      "learning_rate": 2.8447677432898115e-07,
+      "loss": 0.2534,
+      "step": 14237
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.431673570825116,
+      "learning_rate": 2.8413823273352793e-07,
+      "loss": 0.2714,
+      "step": 14238
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.510482169073478,
+      "learning_rate": 2.8379988680755533e-07,
+      "loss": 0.2539,
+      "step": 14239
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0339140084886718,
+      "learning_rate": 2.8346173656510266e-07,
+      "loss": 0.2578,
+      "step": 14240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0177448908096074,
+      "learning_rate": 2.8312378202020043e-07,
+      "loss": 0.2609,
+      "step": 14241
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7390656271983145,
+      "learning_rate": 2.827860231868701e-07,
+      "loss": 0.246,
+      "step": 14242
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8694647885348494,
+      "learning_rate": 2.824484600791261e-07,
+      "loss": 0.2546,
+      "step": 14243
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.422264378514909,
+      "learning_rate": 2.821110927109744e-07,
+      "loss": 0.2517,
+      "step": 14244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0384834894355954,
+      "learning_rate": 2.817739210964143e-07,
+      "loss": 0.2479,
+      "step": 14245
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.198299849321342,
+      "learning_rate": 2.814369452494348e-07,
+      "loss": 0.2527,
+      "step": 14246
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.563116367906744,
+      "learning_rate": 2.811001651840167e-07,
+      "loss": 0.2415,
+      "step": 14247
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3240036326220106,
+      "learning_rate": 2.807635809141357e-07,
+      "loss": 0.2539,
+      "step": 14248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7415728194434241,
+      "learning_rate": 2.804271924537555e-07,
+      "loss": 0.2681,
+      "step": 14249
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9932981358469144,
+      "learning_rate": 2.800909998168333e-07,
+      "loss": 0.2923,
+      "step": 14250
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.173290918256281,
+      "learning_rate": 2.797550030173196e-07,
+      "loss": 0.2571,
+      "step": 14251
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6016106518789888,
+      "learning_rate": 2.7941920206915443e-07,
+      "loss": 0.2537,
+      "step": 14252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9258956452904854,
+      "learning_rate": 2.790835969862721e-07,
+      "loss": 0.2591,
+      "step": 14253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5544338341820572,
+      "learning_rate": 2.7874818778259647e-07,
+      "loss": 0.2523,
+      "step": 14254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7179957423032626,
+      "learning_rate": 2.7841297447204374e-07,
+      "loss": 0.2558,
+      "step": 14255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6069787880469615,
+      "learning_rate": 2.780779570685238e-07,
+      "loss": 0.4768,
+      "step": 14256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9796644903922684,
+      "learning_rate": 2.7774313558593667e-07,
+      "loss": 0.2461,
+      "step": 14257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.394635213376667,
+      "learning_rate": 2.774085100381735e-07,
+      "loss": 0.2583,
+      "step": 14258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.704137124756935,
+      "learning_rate": 2.770740804391209e-07,
+      "loss": 0.2651,
+      "step": 14259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5746760808446791,
+      "learning_rate": 2.767398468026522e-07,
+      "loss": 0.2547,
+      "step": 14260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4546499789228757,
+      "learning_rate": 2.764058091426375e-07,
+      "loss": 0.2529,
+      "step": 14261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9139363708796495,
+      "learning_rate": 2.760719674729362e-07,
+      "loss": 0.2607,
+      "step": 14262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6077906748820867,
+      "learning_rate": 2.757383218073983e-07,
+      "loss": 0.2686,
+      "step": 14263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.965502298767634,
+      "learning_rate": 2.7540487215986956e-07,
+      "loss": 0.2644,
+      "step": 14264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.499242183724396,
+      "learning_rate": 2.750716185441843e-07,
+      "loss": 0.2774,
+      "step": 14265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7116661990277742,
+      "learning_rate": 2.747385609741704e-07,
+      "loss": 0.2505,
+      "step": 14266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5395731190456927,
+      "learning_rate": 2.7440569946364624e-07,
+      "loss": 0.2774,
+      "step": 14267
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7632531684727295,
+      "learning_rate": 2.7407303402642305e-07,
+      "loss": 0.2357,
+      "step": 14268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.288278037230518,
+      "learning_rate": 2.737405646763042e-07,
+      "loss": 0.2523,
+      "step": 14269
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7551981683033855,
+      "learning_rate": 2.7340829142708413e-07,
+      "loss": 0.2626,
+      "step": 14270
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.598934700154759,
+      "learning_rate": 2.730762142925492e-07,
+      "loss": 0.2698,
+      "step": 14271
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3412242311082365,
+      "learning_rate": 2.727443332864782e-07,
+      "loss": 0.2633,
+      "step": 14272
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.464517531893419,
+      "learning_rate": 2.724126484226408e-07,
+      "loss": 0.2382,
+      "step": 14273
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4759007137241236,
+      "learning_rate": 2.720811597148004e-07,
+      "loss": 0.2479,
+      "step": 14274
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 25.146667317474584,
+      "learning_rate": 2.7174986717670995e-07,
+      "loss": 0.2456,
+      "step": 14275
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3813183177187598,
+      "learning_rate": 2.7141877082211664e-07,
+      "loss": 0.2673,
+      "step": 14276
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2713790345902896,
+      "learning_rate": 2.710878706647574e-07,
+      "loss": 0.2469,
+      "step": 14277
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.790607553768402,
+      "learning_rate": 2.707571667183617e-07,
+      "loss": 0.2391,
+      "step": 14278
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5958337388918039,
+      "learning_rate": 2.704266589966503e-07,
+      "loss": 0.4604,
+      "step": 14279
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.13640474544427,
+      "learning_rate": 2.700963475133378e-07,
+      "loss": 0.2527,
+      "step": 14280
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.868114083564988,
+      "learning_rate": 2.6976623228212984e-07,
+      "loss": 0.2594,
+      "step": 14281
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.599182555203552,
+      "learning_rate": 2.694363133167227e-07,
+      "loss": 0.2476,
+      "step": 14282
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4228390381837008,
+      "learning_rate": 2.6910659063080436e-07,
+      "loss": 0.2458,
+      "step": 14283
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0723920804520524,
+      "learning_rate": 2.687770642380577e-07,
+      "loss": 0.2479,
+      "step": 14284
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8751990552545472,
+      "learning_rate": 2.684477341521541e-07,
+      "loss": 0.2351,
+      "step": 14285
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.010661882498264,
+      "learning_rate": 2.681186003867581e-07,
+      "loss": 0.2362,
+      "step": 14286
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0588857871857464,
+      "learning_rate": 2.6778966295552546e-07,
+      "loss": 0.2571,
+      "step": 14287
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7254992247541738,
+      "learning_rate": 2.674609218721053e-07,
+      "loss": 0.2582,
+      "step": 14288
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4153781255581883,
+      "learning_rate": 2.6713237715013773e-07,
+      "loss": 0.2272,
+      "step": 14289
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.188195995898477,
+      "learning_rate": 2.668040288032547e-07,
+      "loss": 0.2548,
+      "step": 14290
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9173221933713778,
+      "learning_rate": 2.664758768450787e-07,
+      "loss": 0.2433,
+      "step": 14291
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.454731247284654,
+      "learning_rate": 2.6614792128922704e-07,
+      "loss": 0.2399,
+      "step": 14292
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8180687209741582,
+      "learning_rate": 2.6582016214930617e-07,
+      "loss": 0.2396,
+      "step": 14293
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.232075975795993,
+      "learning_rate": 2.6549259943891567e-07,
+      "loss": 0.264,
+      "step": 14294
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4984812271406946,
+      "learning_rate": 2.6516523317164647e-07,
+      "loss": 0.2487,
+      "step": 14295
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.888018384220596,
+      "learning_rate": 2.648380633610814e-07,
+      "loss": 0.2498,
+      "step": 14296
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2812566108764694,
+      "learning_rate": 2.645110900207959e-07,
+      "loss": 0.2475,
+      "step": 14297
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.02868864682591,
+      "learning_rate": 2.641843131643573e-07,
+      "loss": 0.2654,
+      "step": 14298
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.377008929284228,
+      "learning_rate": 2.638577328053221e-07,
+      "loss": 0.2389,
+      "step": 14299
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6874707178334174,
+      "learning_rate": 2.6353134895724273e-07,
+      "loss": 0.2376,
+      "step": 14300
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5864689451606748,
+      "learning_rate": 2.632051616336606e-07,
+      "loss": 0.4357,
+      "step": 14301
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7089429809272876,
+      "learning_rate": 2.628791708481099e-07,
+      "loss": 0.2554,
+      "step": 14302
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9622773307672783,
+      "learning_rate": 2.625533766141153e-07,
+      "loss": 0.2371,
+      "step": 14303
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5524344654306752,
+      "learning_rate": 2.622277789451966e-07,
+      "loss": 0.4532,
+      "step": 14304
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.304183526786705,
+      "learning_rate": 2.6190237785486237e-07,
+      "loss": 0.2519,
+      "step": 14305
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.68409037718728,
+      "learning_rate": 2.615771733566147e-07,
+      "loss": 0.2468,
+      "step": 14306
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.051235304864181,
+      "learning_rate": 2.6125216546394605e-07,
+      "loss": 0.2468,
+      "step": 14307
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5563011857453501,
+      "learning_rate": 2.609273541903423e-07,
+      "loss": 0.2477,
+      "step": 14308
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5047866240943875,
+      "learning_rate": 2.606027395492805e-07,
+      "loss": 0.2472,
+      "step": 14309
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3888584474943344,
+      "learning_rate": 2.6027832155422816e-07,
+      "loss": 0.2334,
+      "step": 14310
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 19.57147551750055,
+      "learning_rate": 2.599541002186479e-07,
+      "loss": 0.2357,
+      "step": 14311
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9039100663689588,
+      "learning_rate": 2.5963007555599053e-07,
+      "loss": 0.2385,
+      "step": 14312
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9219853419890818,
+      "learning_rate": 2.5930624757970205e-07,
+      "loss": 0.2683,
+      "step": 14313
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6234271727171482,
+      "learning_rate": 2.5898261630321717e-07,
+      "loss": 0.4833,
+      "step": 14314
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.272651886494399,
+      "learning_rate": 2.5865918173996476e-07,
+      "loss": 0.2535,
+      "step": 14315
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1642482305498887,
+      "learning_rate": 2.5833594390336447e-07,
+      "loss": 0.2562,
+      "step": 14316
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4491858440788827,
+      "learning_rate": 2.580129028068273e-07,
+      "loss": 0.2455,
+      "step": 14317
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0427806683094007,
+      "learning_rate": 2.576900584637582e-07,
+      "loss": 0.2562,
+      "step": 14318
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8198197369485298,
+      "learning_rate": 2.5736741088755237e-07,
+      "loss": 0.2501,
+      "step": 14319
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7798821511568765,
+      "learning_rate": 2.570449600915953e-07,
+      "loss": 0.2625,
+      "step": 14320
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.810146008158133,
+      "learning_rate": 2.5672270608926855e-07,
+      "loss": 0.2701,
+      "step": 14321
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6614336260490514,
+      "learning_rate": 2.5640064889394133e-07,
+      "loss": 0.2521,
+      "step": 14322
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0944552441167548,
+      "learning_rate": 2.560787885189758e-07,
+      "loss": 0.2378,
+      "step": 14323
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.766888339721811,
+      "learning_rate": 2.5575712497772855e-07,
+      "loss": 0.238,
+      "step": 14324
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3644201234443156,
+      "learning_rate": 2.554356582835443e-07,
+      "loss": 0.2379,
+      "step": 14325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6866521254325533,
+      "learning_rate": 2.55114388449762e-07,
+      "loss": 0.2405,
+      "step": 14326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5681223064819383,
+      "learning_rate": 2.5479331548971197e-07,
+      "loss": 0.2653,
+      "step": 14327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.453224986863318,
+      "learning_rate": 2.544724394167153e-07,
+      "loss": 0.2489,
+      "step": 14328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2150220301972587,
+      "learning_rate": 2.5415176024408685e-07,
+      "loss": 0.2578,
+      "step": 14329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.820358215394252,
+      "learning_rate": 2.5383127798513094e-07,
+      "loss": 0.2462,
+      "step": 14330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4161023122403655,
+      "learning_rate": 2.535109926531448e-07,
+      "loss": 0.2375,
+      "step": 14331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6473279939021952,
+      "learning_rate": 2.531909042614189e-07,
+      "loss": 0.2449,
+      "step": 14332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0198870517579244,
+      "learning_rate": 2.528710128232337e-07,
+      "loss": 0.2579,
+      "step": 14333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.056495398028475,
+      "learning_rate": 2.525513183518624e-07,
+      "loss": 0.2614,
+      "step": 14334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9809781313122357,
+      "learning_rate": 2.5223182086056897e-07,
+      "loss": 0.2776,
+      "step": 14335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0434178773472356,
+      "learning_rate": 2.5191252036260995e-07,
+      "loss": 0.2606,
+      "step": 14336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 24.372143294614947,
+      "learning_rate": 2.5159341687123416e-07,
+      "loss": 0.2412,
+      "step": 14337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2169340053483104,
+      "learning_rate": 2.512745103996822e-07,
+      "loss": 0.2494,
+      "step": 14338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.239186597790997,
+      "learning_rate": 2.5095580096118454e-07,
+      "loss": 0.2634,
+      "step": 14339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.440272853098814,
+      "learning_rate": 2.506372885689662e-07,
+      "loss": 0.2594,
+      "step": 14340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.489070722551848,
+      "learning_rate": 2.503189732362432e-07,
+      "loss": 0.2652,
+      "step": 14341
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5590398551133493,
+      "learning_rate": 2.5000085497622227e-07,
+      "loss": 0.255,
+      "step": 14342
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.299914823370509,
+      "learning_rate": 2.496829338021028e-07,
+      "loss": 0.2448,
+      "step": 14343
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.067053193087143,
+      "learning_rate": 2.4936520972707487e-07,
+      "loss": 0.2686,
+      "step": 14344
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.384261052662673,
+      "learning_rate": 2.49047682764324e-07,
+      "loss": 0.2452,
+      "step": 14345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5748765177490964,
+      "learning_rate": 2.4873035292702243e-07,
+      "loss": 0.2394,
+      "step": 14346
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.994306853704753,
+      "learning_rate": 2.484132202283379e-07,
+      "loss": 0.235,
+      "step": 14347
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.551821758825019,
+      "learning_rate": 2.4809628468142834e-07,
+      "loss": 0.2566,
+      "step": 14348
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.728590827808792,
+      "learning_rate": 2.477795462994448e-07,
+      "loss": 0.2514,
+      "step": 14349
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7651862151945834,
+      "learning_rate": 2.47463005095529e-07,
+      "loss": 0.2599,
+      "step": 14350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5845281034009686,
+      "learning_rate": 2.4714666108281436e-07,
+      "loss": 0.2386,
+      "step": 14351
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3639805165174725,
+      "learning_rate": 2.468305142744259e-07,
+      "loss": 0.2458,
+      "step": 14352
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3358090310402124,
+      "learning_rate": 2.465145646834832e-07,
+      "loss": 0.2839,
+      "step": 14353
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8326084014107749,
+      "learning_rate": 2.4619881232309405e-07,
+      "loss": 0.2583,
+      "step": 14354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.716078901090518,
+      "learning_rate": 2.458832572063591e-07,
+      "loss": 0.2569,
+      "step": 14355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.665638152776214,
+      "learning_rate": 2.4556789934637226e-07,
+      "loss": 0.2643,
+      "step": 14356
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3691234592486214,
+      "learning_rate": 2.452527387562187e-07,
+      "loss": 0.264,
+      "step": 14357
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6303826087677693,
+      "learning_rate": 2.44937775448974e-07,
+      "loss": 0.248,
+      "step": 14358
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0313384949244813,
+      "learning_rate": 2.446230094377067e-07,
+      "loss": 0.2543,
+      "step": 14359
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.520181498866394,
+      "learning_rate": 2.4430844073547786e-07,
+      "loss": 0.2543,
+      "step": 14360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.48213903885956,
+      "learning_rate": 2.439940693553389e-07,
+      "loss": 0.2608,
+      "step": 14361
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6595244579834318,
+      "learning_rate": 2.436798953103331e-07,
+      "loss": 0.2354,
+      "step": 14362
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.821776732101652,
+      "learning_rate": 2.4336591861349734e-07,
+      "loss": 0.2395,
+      "step": 14363
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6243041882966267,
+      "learning_rate": 2.430521392778573e-07,
+      "loss": 0.2387,
+      "step": 14364
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.224976937315625,
+      "learning_rate": 2.4273855731643427e-07,
+      "loss": 0.265,
+      "step": 14365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.428468718138498,
+      "learning_rate": 2.4242517274223776e-07,
+      "loss": 0.2701,
+      "step": 14366
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.651647854495142,
+      "learning_rate": 2.421119855682713e-07,
+      "loss": 0.2388,
+      "step": 14367
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1431795516245846,
+      "learning_rate": 2.417989958075295e-07,
+      "loss": 0.2473,
+      "step": 14368
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.736890072614024,
+      "learning_rate": 2.414862034729998e-07,
+      "loss": 0.2267,
+      "step": 14369
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6582103219946127,
+      "learning_rate": 2.4117360857765836e-07,
+      "loss": 0.2711,
+      "step": 14370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9213955854946294,
+      "learning_rate": 2.408612111344771e-07,
+      "loss": 0.2475,
+      "step": 14371
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.453657492157642,
+      "learning_rate": 2.4054901115641684e-07,
+      "loss": 0.2423,
+      "step": 14372
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2477633989523333,
+      "learning_rate": 2.402370086564326e-07,
+      "loss": 0.2734,
+      "step": 14373
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1075849917982463,
+      "learning_rate": 2.399252036474686e-07,
+      "loss": 0.2374,
+      "step": 14374
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5765891430791905,
+      "learning_rate": 2.396135961424628e-07,
+      "loss": 0.4522,
+      "step": 14375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3969949856910713,
+      "learning_rate": 2.393021861543449e-07,
+      "loss": 0.2587,
+      "step": 14376
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2012000712684663,
+      "learning_rate": 2.3899097369603385e-07,
+      "loss": 0.2437,
+      "step": 14377
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3542205146342408,
+      "learning_rate": 2.38679958780445e-07,
+      "loss": 0.2485,
+      "step": 14378
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7431888350531586,
+      "learning_rate": 2.3836914142048194e-07,
+      "loss": 0.2611,
+      "step": 14379
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.878420594068702,
+      "learning_rate": 2.3805852162903987e-07,
+      "loss": 0.2539,
+      "step": 14380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0501687454298296,
+      "learning_rate": 2.3774809941900844e-07,
+      "loss": 0.2462,
+      "step": 14381
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6814288297169528,
+      "learning_rate": 2.3743787480326742e-07,
+      "loss": 0.2493,
+      "step": 14382
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4408187485143142,
+      "learning_rate": 2.3712784779468756e-07,
+      "loss": 0.2543,
+      "step": 14383
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.189193046287574,
+      "learning_rate": 2.3681801840613362e-07,
+      "loss": 0.2395,
+      "step": 14384
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.229798587352424,
+      "learning_rate": 2.3650838665045972e-07,
+      "loss": 0.2501,
+      "step": 14385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5913219911759955,
+      "learning_rate": 2.361989525405145e-07,
+      "loss": 0.5088,
+      "step": 14386
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6949106012743453,
+      "learning_rate": 2.3588971608913604e-07,
+      "loss": 0.2478,
+      "step": 14387
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6218761311965364,
+      "learning_rate": 2.3558067730915513e-07,
+      "loss": 0.265,
+      "step": 14388
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5670724941813509,
+      "learning_rate": 2.3527183621339434e-07,
+      "loss": 0.4422,
+      "step": 14389
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.608523501003697,
+      "learning_rate": 2.3496319281466895e-07,
+      "loss": 0.242,
+      "step": 14390
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.733549344985571,
+      "learning_rate": 2.346547471257832e-07,
+      "loss": 0.2607,
+      "step": 14391
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.095912393550003,
+      "learning_rate": 2.3434649915953623e-07,
+      "loss": 0.2513,
+      "step": 14392
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9514309466854387,
+      "learning_rate": 2.3403844892871896e-07,
+      "loss": 0.2403,
+      "step": 14393
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.280382700689735,
+      "learning_rate": 2.337305964461112e-07,
+      "loss": 0.2472,
+      "step": 14394
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.861369097382789,
+      "learning_rate": 2.334229417244871e-07,
+      "loss": 0.2915,
+      "step": 14395
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7780582043153892,
+      "learning_rate": 2.3311548477661038e-07,
+      "loss": 0.2412,
+      "step": 14396
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3023809363689742,
+      "learning_rate": 2.3280822561524031e-07,
+      "loss": 0.2448,
+      "step": 14397
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3497213326356283,
+      "learning_rate": 2.325011642531244e-07,
+      "loss": 0.2253,
+      "step": 14398
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0955837113376727,
+      "learning_rate": 2.3219430070300254e-07,
+      "loss": 0.2589,
+      "step": 14399
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2183426439622114,
+      "learning_rate": 2.3188763497760725e-07,
+      "loss": 0.2486,
+      "step": 14400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.473385434235075,
+      "learning_rate": 2.3158116708966448e-07,
+      "loss": 0.2436,
+      "step": 14401
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.461460164221812,
+      "learning_rate": 2.31274897051888e-07,
+      "loss": 0.2448,
+      "step": 14402
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.590503190490564,
+      "learning_rate": 2.3096882487698703e-07,
+      "loss": 0.2574,
+      "step": 14403
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8828166767004388,
+      "learning_rate": 2.306629505776592e-07,
+      "loss": 0.2563,
+      "step": 14404
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.951661123735288,
+      "learning_rate": 2.3035727416659769e-07,
+      "loss": 0.2545,
+      "step": 14405
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.580776276640741,
+      "learning_rate": 2.300517956564846e-07,
+      "loss": 0.2396,
+      "step": 14406
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6372326130218704,
+      "learning_rate": 2.2974651505999425e-07,
+      "loss": 0.4515,
+      "step": 14407
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6914910927303681,
+      "learning_rate": 2.2944143238979366e-07,
+      "loss": 0.2527,
+      "step": 14408
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5940336644161897,
+      "learning_rate": 2.291365476585422e-07,
+      "loss": 0.2505,
+      "step": 14409
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0332868183136243,
+      "learning_rate": 2.2883186087888977e-07,
+      "loss": 0.2595,
+      "step": 14410
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3147909450333746,
+      "learning_rate": 2.2852737206347786e-07,
+      "loss": 0.2482,
+      "step": 14411
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6875666958481723,
+      "learning_rate": 2.2822308122493976e-07,
+      "loss": 0.2426,
+      "step": 14412
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4211114418589434,
+      "learning_rate": 2.2791898837590197e-07,
+      "loss": 0.2424,
+      "step": 14413
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3004597385252072,
+      "learning_rate": 2.2761509352898114e-07,
+      "loss": 0.2405,
+      "step": 14414
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.79546572072472,
+      "learning_rate": 2.2731139669678714e-07,
+      "loss": 0.2537,
+      "step": 14415
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8991778276486342,
+      "learning_rate": 2.2700789789192047e-07,
+      "loss": 0.2639,
+      "step": 14416
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1232139901719056,
+      "learning_rate": 2.267045971269738e-07,
+      "loss": 0.2575,
+      "step": 14417
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.485955926120402,
+      "learning_rate": 2.2640149441453208e-07,
+      "loss": 0.2552,
+      "step": 14418
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.736427901752236,
+      "learning_rate": 2.2609858976717136e-07,
+      "loss": 0.2514,
+      "step": 14419
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1880579531991087,
+      "learning_rate": 2.2579588319745883e-07,
+      "loss": 0.232,
+      "step": 14420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.815118180142573,
+      "learning_rate": 2.2549337471795553e-07,
+      "loss": 0.2424,
+      "step": 14421
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3752905844804904,
+      "learning_rate": 2.2519106434121252e-07,
+      "loss": 0.2527,
+      "step": 14422
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.792289956779686,
+      "learning_rate": 2.2488895207977312e-07,
+      "loss": 0.2513,
+      "step": 14423
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.562866945134141,
+      "learning_rate": 2.245870379461723e-07,
+      "loss": 0.2743,
+      "step": 14424
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6285545836109259,
+      "learning_rate": 2.242853219529384e-07,
+      "loss": 0.2343,
+      "step": 14425
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.25526385611361,
+      "learning_rate": 2.2398380411258858e-07,
+      "loss": 0.247,
+      "step": 14426
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3273318668415506,
+      "learning_rate": 2.2368248443763451e-07,
+      "loss": 0.2348,
+      "step": 14427
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2994044788467543,
+      "learning_rate": 2.2338136294057677e-07,
+      "loss": 0.2448,
+      "step": 14428
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.657924935810576,
+      "learning_rate": 2.2308043963391034e-07,
+      "loss": 0.2749,
+      "step": 14429
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9342738710559804,
+      "learning_rate": 2.2277971453012193e-07,
+      "loss": 0.2678,
+      "step": 14430
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.58915421386696,
+      "learning_rate": 2.2247918764168874e-07,
+      "loss": 0.2363,
+      "step": 14431
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3492375979172468,
+      "learning_rate": 2.2217885898107915e-07,
+      "loss": 0.2566,
+      "step": 14432
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0956570689213967,
+      "learning_rate": 2.2187872856075544e-07,
+      "loss": 0.2487,
+      "step": 14433
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3599289344086705,
+      "learning_rate": 2.2157879639317038e-07,
+      "loss": 0.2549,
+      "step": 14434
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4080477810531007,
+      "learning_rate": 2.2127906249076903e-07,
+      "loss": 0.2407,
+      "step": 14435
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.772172873548713,
+      "learning_rate": 2.209795268659859e-07,
+      "loss": 0.2739,
+      "step": 14436
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7411833702255235,
+      "learning_rate": 2.2068018953125103e-07,
+      "loss": 0.2594,
+      "step": 14437
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.220970600829254,
+      "learning_rate": 2.203810504989845e-07,
+      "loss": 0.2555,
+      "step": 14438
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5872756190965056,
+      "learning_rate": 2.2008210978159806e-07,
+      "loss": 0.465,
+      "step": 14439
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.851639766650723,
+      "learning_rate": 2.1978336739149454e-07,
+      "loss": 0.2621,
+      "step": 14440
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6739715102465174,
+      "learning_rate": 2.1948482334106957e-07,
+      "loss": 0.2453,
+      "step": 14441
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.050088892825613,
+      "learning_rate": 2.1918647764271105e-07,
+      "loss": 0.2378,
+      "step": 14442
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8791823839520525,
+      "learning_rate": 2.1888833030879685e-07,
+      "loss": 0.2463,
+      "step": 14443
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.755452865345263,
+      "learning_rate": 2.1859038135169764e-07,
+      "loss": 0.2743,
+      "step": 14444
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.249296166630754,
+      "learning_rate": 2.1829263078377683e-07,
+      "loss": 0.259,
+      "step": 14445
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.645143216423926,
+      "learning_rate": 2.179950786173879e-07,
+      "loss": 0.2462,
+      "step": 14446
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.779318669753915,
+      "learning_rate": 2.1769772486487705e-07,
+      "loss": 0.2586,
+      "step": 14447
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.837384937670963,
+      "learning_rate": 2.174005695385817e-07,
+      "loss": 0.2718,
+      "step": 14448
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.052738596929044,
+      "learning_rate": 2.171036126508319e-07,
+      "loss": 0.2602,
+      "step": 14449
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7928776179684216,
+      "learning_rate": 2.1680685421394842e-07,
+      "loss": 0.2711,
+      "step": 14450
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4483639431258877,
+      "learning_rate": 2.1651029424024417e-07,
+      "loss": 0.2412,
+      "step": 14451
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.2787994474514255,
+      "learning_rate": 2.1621393274202429e-07,
+      "loss": 0.259,
+      "step": 14452
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6261215902194026,
+      "learning_rate": 2.1591776973158564e-07,
+      "loss": 0.2349,
+      "step": 14453
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6680651222154,
+      "learning_rate": 2.1562180522121613e-07,
+      "loss": 0.2498,
+      "step": 14454
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8700943683892384,
+      "learning_rate": 2.153260392231965e-07,
+      "loss": 0.2511,
+      "step": 14455
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.567964666506948,
+      "learning_rate": 2.1503047174979695e-07,
+      "loss": 0.2715,
+      "step": 14456
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.614727062017292,
+      "learning_rate": 2.147351028132827e-07,
+      "loss": 0.2762,
+      "step": 14457
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8235129338382885,
+      "learning_rate": 2.144399324259089e-07,
+      "loss": 0.2731,
+      "step": 14458
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5803840532779754,
+      "learning_rate": 2.1414496059992183e-07,
+      "loss": 0.2477,
+      "step": 14459
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0769844778878053,
+      "learning_rate": 2.138501873475607e-07,
+      "loss": 0.2418,
+      "step": 14460
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7474217544896757,
+      "learning_rate": 2.135556126810573e-07,
+      "loss": 0.241,
+      "step": 14461
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3863284081175946,
+      "learning_rate": 2.13261236612633e-07,
+      "loss": 0.2686,
+      "step": 14462
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6041412725538277,
+      "learning_rate": 2.1296705915450244e-07,
+      "loss": 0.488,
+      "step": 14463
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.34452761724846,
+      "learning_rate": 2.1267308031887036e-07,
+      "loss": 0.237,
+      "step": 14464
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4947442283899726,
+      "learning_rate": 2.123793001179364e-07,
+      "loss": 0.2694,
+      "step": 14465
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.26729705768746,
+      "learning_rate": 2.1208571856388915e-07,
+      "loss": 0.2596,
+      "step": 14466
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.775837042839868,
+      "learning_rate": 2.117923356689089e-07,
+      "loss": 0.2589,
+      "step": 14467
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5906264427923897,
+      "learning_rate": 2.114991514451692e-07,
+      "loss": 0.2517,
+      "step": 14468
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6102118575831724,
+      "learning_rate": 2.112061659048359e-07,
+      "loss": 0.4594,
+      "step": 14469
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4509361422530795,
+      "learning_rate": 2.109133790600648e-07,
+      "loss": 0.2598,
+      "step": 14470
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.615792194612049,
+      "learning_rate": 2.1062079092300402e-07,
+      "loss": 0.2468,
+      "step": 14471
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0041096957855564,
+      "learning_rate": 2.103284015057927e-07,
+      "loss": 0.2534,
+      "step": 14472
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2664870866984965,
+      "learning_rate": 2.1003621082056392e-07,
+      "loss": 0.2418,
+      "step": 14473
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3768270188270497,
+      "learning_rate": 2.097442188794402e-07,
+      "loss": 0.2837,
+      "step": 14474
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.427340659654771,
+      "learning_rate": 2.0945242569453795e-07,
+      "loss": 0.253,
+      "step": 14475
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5161216053208117,
+      "learning_rate": 2.091608312779625e-07,
+      "loss": 0.2401,
+      "step": 14476
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.675503374474496,
+      "learning_rate": 2.088694356418147e-07,
+      "loss": 0.2688,
+      "step": 14477
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8923580657315366,
+      "learning_rate": 2.0857823879818384e-07,
+      "loss": 0.2454,
+      "step": 14478
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3759402065898474,
+      "learning_rate": 2.0828724075915298e-07,
+      "loss": 0.2556,
+      "step": 14479
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.476388684288924,
+      "learning_rate": 2.079964415367941e-07,
+      "loss": 0.2338,
+      "step": 14480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4283349735457755,
+      "learning_rate": 2.0770584114317483e-07,
+      "loss": 0.2636,
+      "step": 14481
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6486641676674842,
+      "learning_rate": 2.074154395903527e-07,
+      "loss": 0.2402,
+      "step": 14482
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6213172513410736,
+      "learning_rate": 2.0712523689037645e-07,
+      "loss": 0.2556,
+      "step": 14483
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8537371799773197,
+      "learning_rate": 2.0683523305528696e-07,
+      "loss": 0.24,
+      "step": 14484
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7785473319886127,
+      "learning_rate": 2.0654542809711798e-07,
+      "loss": 0.2585,
+      "step": 14485
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7651719366311103,
+      "learning_rate": 2.0625582202789317e-07,
+      "loss": 0.2319,
+      "step": 14486
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1273793923211395,
+      "learning_rate": 2.0596641485962854e-07,
+      "loss": 0.2457,
+      "step": 14487
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.21819875878106,
+      "learning_rate": 2.056772066043322e-07,
+      "loss": 0.2698,
+      "step": 14488
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5709435104374834,
+      "learning_rate": 2.053881972740046e-07,
+      "loss": 0.2511,
+      "step": 14489
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8761287437363046,
+      "learning_rate": 2.0509938688063723e-07,
+      "loss": 0.2411,
+      "step": 14490
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9211021468018759,
+      "learning_rate": 2.0481077543621275e-07,
+      "loss": 0.2368,
+      "step": 14491
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5115448951643704,
+      "learning_rate": 2.04522362952706e-07,
+      "loss": 0.2417,
+      "step": 14492
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.088136531250573,
+      "learning_rate": 2.0423414944208464e-07,
+      "loss": 0.2441,
+      "step": 14493
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.087766755890298,
+      "learning_rate": 2.0394613491630688e-07,
+      "loss": 0.2645,
+      "step": 14494
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.107246310436187,
+      "learning_rate": 2.036583193873226e-07,
+      "loss": 0.2588,
+      "step": 14495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.204240909998728,
+      "learning_rate": 2.0337070286707283e-07,
+      "loss": 0.2476,
+      "step": 14496
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.413546846926048,
+      "learning_rate": 2.0308328536749355e-07,
+      "loss": 0.2509,
+      "step": 14497
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7561510964926772,
+      "learning_rate": 2.0279606690050856e-07,
+      "loss": 0.2469,
+      "step": 14498
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.338753213289381,
+      "learning_rate": 2.0250904747803614e-07,
+      "loss": 0.2612,
+      "step": 14499
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0824586160643634,
+      "learning_rate": 2.0222222711198392e-07,
+      "loss": 0.2393,
+      "step": 14500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.007315300107275,
+      "learning_rate": 2.0193560581425408e-07,
+      "loss": 0.2318,
+      "step": 14501
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9865525920497915,
+      "learning_rate": 2.0164918359673769e-07,
+      "loss": 0.2509,
+      "step": 14502
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7197148031101857,
+      "learning_rate": 2.013629604713202e-07,
+      "loss": 0.2382,
+      "step": 14503
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8947838501290832,
+      "learning_rate": 2.0107693644987492e-07,
+      "loss": 0.254,
+      "step": 14504
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.902731433394685,
+      "learning_rate": 2.007911115442729e-07,
+      "loss": 0.2588,
+      "step": 14505
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.289651256926037,
+      "learning_rate": 2.005054857663724e-07,
+      "loss": 0.2502,
+      "step": 14506
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.186328093635684,
+      "learning_rate": 2.0022005912802345e-07,
+      "loss": 0.2609,
+      "step": 14507
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.237839020173491,
+      "learning_rate": 1.999348316410693e-07,
+      "loss": 0.2598,
+      "step": 14508
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3844167421565585,
+      "learning_rate": 1.9964980331734552e-07,
+      "loss": 0.2406,
+      "step": 14509
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5850318063128662,
+      "learning_rate": 1.9936497416867816e-07,
+      "loss": 0.4568,
+      "step": 14510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7980759564399063,
+      "learning_rate": 1.9908034420688394e-07,
+      "loss": 0.266,
+      "step": 14511
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.731553582372513,
+      "learning_rate": 1.9879591344377335e-07,
+      "loss": 0.2584,
+      "step": 14512
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4881645457346415,
+      "learning_rate": 1.9851168189114923e-07,
+      "loss": 0.2767,
+      "step": 14513
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.590704362330037,
+      "learning_rate": 1.9822764956080375e-07,
+      "loss": 0.2573,
+      "step": 14514
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.706228859345715,
+      "learning_rate": 1.9794381646452198e-07,
+      "loss": 0.2427,
+      "step": 14515
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.93471096274332,
+      "learning_rate": 1.9766018261408005e-07,
+      "loss": 0.2557,
+      "step": 14516
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.9366506521476845,
+      "learning_rate": 1.973767480212474e-07,
+      "loss": 0.255,
+      "step": 14517
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2290112195682714,
+      "learning_rate": 1.9709351269778408e-07,
+      "loss": 0.272,
+      "step": 14518
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3340555024224874,
+      "learning_rate": 1.9681047665544127e-07,
+      "loss": 0.2726,
+      "step": 14519
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.747504705651108,
+      "learning_rate": 1.9652763990596345e-07,
+      "loss": 0.2516,
+      "step": 14520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.562454675871726,
+      "learning_rate": 1.9624500246108625e-07,
+      "loss": 0.2443,
+      "step": 14521
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5122784108971654,
+      "learning_rate": 1.9596256433253635e-07,
+      "loss": 0.2681,
+      "step": 14522
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.092076801555144,
+      "learning_rate": 1.956803255320322e-07,
+      "loss": 0.2794,
+      "step": 14523
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.704990015234847,
+      "learning_rate": 1.953982860712844e-07,
+      "loss": 0.2503,
+      "step": 14524
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6040679585377352,
+      "learning_rate": 1.9511644596199643e-07,
+      "loss": 0.4704,
+      "step": 14525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2386558262934875,
+      "learning_rate": 1.9483480521586107e-07,
+      "loss": 0.2623,
+      "step": 14526
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6204712726676649,
+      "learning_rate": 1.945533638445646e-07,
+      "loss": 0.4777,
+      "step": 14527
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6546399832572103,
+      "learning_rate": 1.9427212185978428e-07,
+      "loss": 0.2554,
+      "step": 14528
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.868847229193744,
+      "learning_rate": 1.9399107927319028e-07,
+      "loss": 0.2563,
+      "step": 14529
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.100870480623982,
+      "learning_rate": 1.9371023609644268e-07,
+      "loss": 0.2759,
+      "step": 14530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7400093479151766,
+      "learning_rate": 1.9342959234119385e-07,
+      "loss": 0.2337,
+      "step": 14531
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7265859632734524,
+      "learning_rate": 1.931491480190889e-07,
+      "loss": 0.2514,
+      "step": 14532
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2983224300409826,
+      "learning_rate": 1.9286890314176353e-07,
+      "loss": 0.2502,
+      "step": 14533
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2969022201778935,
+      "learning_rate": 1.9258885772084567e-07,
+      "loss": 0.2416,
+      "step": 14534
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.508959980294738,
+      "learning_rate": 1.9230901176795548e-07,
+      "loss": 0.2665,
+      "step": 14535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.582752042354064,
+      "learning_rate": 1.9202936529470363e-07,
+      "loss": 0.2625,
+      "step": 14536
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.081445959892226,
+      "learning_rate": 1.9174991831269362e-07,
+      "loss": 0.2679,
+      "step": 14537
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.287830484110232,
+      "learning_rate": 1.9147067083351954e-07,
+      "loss": 0.256,
+      "step": 14538
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5175800942879811,
+      "learning_rate": 1.911916228687688e-07,
+      "loss": 0.2376,
+      "step": 14539
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4282350306356104,
+      "learning_rate": 1.909127744300182e-07,
+      "loss": 0.2453,
+      "step": 14540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5090376650146013,
+      "learning_rate": 1.9063412552883852e-07,
+      "loss": 0.2629,
+      "step": 14541
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2789070229309223,
+      "learning_rate": 1.9035567617679163e-07,
+      "loss": 0.2532,
+      "step": 14542
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.200914609783236,
+      "learning_rate": 1.9007742638543104e-07,
+      "loss": 0.2634,
+      "step": 14543
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4406926563074134,
+      "learning_rate": 1.8979937616630084e-07,
+      "loss": 0.2368,
+      "step": 14544
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9354443070109852,
+      "learning_rate": 1.8952152553093906e-07,
+      "loss": 0.24,
+      "step": 14545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.382994471291222,
+      "learning_rate": 1.892438744908731e-07,
+      "loss": 0.2618,
+      "step": 14546
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4377555631615566,
+      "learning_rate": 1.889664230576238e-07,
+      "loss": 0.251,
+      "step": 14547
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6563752045766305,
+      "learning_rate": 1.8868917124270248e-07,
+      "loss": 0.2372,
+      "step": 14548
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7992839791383575,
+      "learning_rate": 1.8841211905761326e-07,
+      "loss": 0.2531,
+      "step": 14549
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.48632246323107,
+      "learning_rate": 1.881352665138525e-07,
+      "loss": 0.2559,
+      "step": 14550
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2703408625191157,
+      "learning_rate": 1.8785861362290603e-07,
+      "loss": 0.2454,
+      "step": 14551
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1449732033066944,
+      "learning_rate": 1.8758216039625243e-07,
+      "loss": 0.263,
+      "step": 14552
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.95268284637273,
+      "learning_rate": 1.873059068453631e-07,
+      "loss": 0.2396,
+      "step": 14553
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5756193425105325,
+      "learning_rate": 1.870298529817005e-07,
+      "loss": 0.2581,
+      "step": 14554
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.413862692424448,
+      "learning_rate": 1.8675399881671774e-07,
+      "loss": 0.2625,
+      "step": 14555
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.978311636808432,
+      "learning_rate": 1.8647834436186009e-07,
+      "loss": 0.2501,
+      "step": 14556
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0528327277025133,
+      "learning_rate": 1.862028896285667e-07,
+      "loss": 0.2568,
+      "step": 14557
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7813793578847024,
+      "learning_rate": 1.859276346282657e-07,
+      "loss": 0.2491,
+      "step": 14558
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5221083090039145,
+      "learning_rate": 1.856525793723779e-07,
+      "loss": 0.2474,
+      "step": 14559
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.390504163933587,
+      "learning_rate": 1.853777238723148e-07,
+      "loss": 0.2787,
+      "step": 14560
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.200196410811054,
+      "learning_rate": 1.8510306813948276e-07,
+      "loss": 0.241,
+      "step": 14561
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.498783783971337,
+      "learning_rate": 1.848286121852766e-07,
+      "loss": 0.2649,
+      "step": 14562
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.679129399596893,
+      "learning_rate": 1.8455435602108328e-07,
+      "loss": 0.2649,
+      "step": 14563
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9345832226516606,
+      "learning_rate": 1.8428029965828265e-07,
+      "loss": 0.2578,
+      "step": 14564
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.894186801501865,
+      "learning_rate": 1.8400644310824722e-07,
+      "loss": 0.2427,
+      "step": 14565
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.265496929716585,
+      "learning_rate": 1.8373278638233848e-07,
+      "loss": 0.2575,
+      "step": 14566
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.193911921550233,
+      "learning_rate": 1.834593294919107e-07,
+      "loss": 0.2396,
+      "step": 14567
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.398539465303125,
+      "learning_rate": 1.8318607244831033e-07,
+      "loss": 0.2497,
+      "step": 14568
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.081467693048924,
+      "learning_rate": 1.8291301526287609e-07,
+      "loss": 0.2617,
+      "step": 14569
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7714467938424623,
+      "learning_rate": 1.8264015794693668e-07,
+      "loss": 0.269,
+      "step": 14570
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8395494084039425,
+      "learning_rate": 1.8236750051181306e-07,
+      "loss": 0.249,
+      "step": 14571
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.3142334681720875,
+      "learning_rate": 1.8209504296881896e-07,
+      "loss": 0.2662,
+      "step": 14572
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3435171338601832,
+      "learning_rate": 1.818227853292598e-07,
+      "loss": 0.2555,
+      "step": 14573
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7046417545149053,
+      "learning_rate": 1.8155072760443149e-07,
+      "loss": 0.2416,
+      "step": 14574
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7680273985641055,
+      "learning_rate": 1.812788698056217e-07,
+      "loss": 0.2514,
+      "step": 14575
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.417788014275917,
+      "learning_rate": 1.810072119441103e-07,
+      "loss": 0.2207,
+      "step": 14576
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4467649856462055,
+      "learning_rate": 1.8073575403116938e-07,
+      "loss": 0.261,
+      "step": 14577
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5405873106651053,
+      "learning_rate": 1.804644960780616e-07,
+      "loss": 0.2691,
+      "step": 14578
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2291964329191467,
+      "learning_rate": 1.8019343809604295e-07,
+      "loss": 0.2568,
+      "step": 14579
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.00943408163364,
+      "learning_rate": 1.7992258009635942e-07,
+      "loss": 0.2352,
+      "step": 14580
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2836406128994153,
+      "learning_rate": 1.7965192209024928e-07,
+      "loss": 0.2456,
+      "step": 14581
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.832438125777982,
+      "learning_rate": 1.7938146408894296e-07,
+      "loss": 0.2412,
+      "step": 14582
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.929832837409836,
+      "learning_rate": 1.7911120610366262e-07,
+      "loss": 0.2451,
+      "step": 14583
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6513242660518177,
+      "learning_rate": 1.7884114814561983e-07,
+      "loss": 0.2472,
+      "step": 14584
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1905309233386983,
+      "learning_rate": 1.7857129022602226e-07,
+      "loss": 0.2466,
+      "step": 14585
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4813312295210759,
+      "learning_rate": 1.783016323560649e-07,
+      "loss": 0.2347,
+      "step": 14586
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.563174579858422,
+      "learning_rate": 1.7803217454693766e-07,
+      "loss": 0.2482,
+      "step": 14587
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9113023448252877,
+      "learning_rate": 1.7776291680981995e-07,
+      "loss": 0.2412,
+      "step": 14588
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2786178310528573,
+      "learning_rate": 1.7749385915588446e-07,
+      "loss": 0.2415,
+      "step": 14589
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 29.05300778970683,
+      "learning_rate": 1.7722500159629452e-07,
+      "loss": 0.2653,
+      "step": 14590
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.290002453105786,
+      "learning_rate": 1.769563441422051e-07,
+      "loss": 0.2697,
+      "step": 14591
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.092070753675626,
+      "learning_rate": 1.7668788680476334e-07,
+      "loss": 0.2562,
+      "step": 14592
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3838436164833396,
+      "learning_rate": 1.7641962959510872e-07,
+      "loss": 0.252,
+      "step": 14593
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3092892132355134,
+      "learning_rate": 1.7615157252437121e-07,
+      "loss": 0.2522,
+      "step": 14594
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8603278509510393,
+      "learning_rate": 1.7588371560367357e-07,
+      "loss": 0.253,
+      "step": 14595
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3223387568769163,
+      "learning_rate": 1.7561605884412858e-07,
+      "loss": 0.2407,
+      "step": 14596
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5872890972249871,
+      "learning_rate": 1.7534860225684292e-07,
+      "loss": 0.2713,
+      "step": 14597
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.67526615629178,
+      "learning_rate": 1.7508134585291325e-07,
+      "loss": 0.2395,
+      "step": 14598
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.986689727135611,
+      "learning_rate": 1.748142896434285e-07,
+      "loss": 0.2468,
+      "step": 14599
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0775548200314744,
+      "learning_rate": 1.7454743363946868e-07,
+      "loss": 0.249,
+      "step": 14600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.308339774678808,
+      "learning_rate": 1.7428077785210663e-07,
+      "loss": 0.2538,
+      "step": 14601
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9148666821890743,
+      "learning_rate": 1.7401432229240733e-07,
+      "loss": 0.252,
+      "step": 14602
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2472774814567003,
+      "learning_rate": 1.7374806697142588e-07,
+      "loss": 0.2507,
+      "step": 14603
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1864405904788144,
+      "learning_rate": 1.7348201190020897e-07,
+      "loss": 0.2447,
+      "step": 14604
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1737520105448525,
+      "learning_rate": 1.7321615708979666e-07,
+      "loss": 0.2357,
+      "step": 14605
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8597862814083623,
+      "learning_rate": 1.729505025512196e-07,
+      "loss": 0.2456,
+      "step": 14606
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7164992487378634,
+      "learning_rate": 1.726850482954995e-07,
+      "loss": 0.2436,
+      "step": 14607
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.306614413990062,
+      "learning_rate": 1.7241979433365086e-07,
+      "loss": 0.2708,
+      "step": 14608
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5303694058213413,
+      "learning_rate": 1.721547406766799e-07,
+      "loss": 0.2535,
+      "step": 14609
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5516930701767238,
+      "learning_rate": 1.7188988733558397e-07,
+      "loss": 0.2565,
+      "step": 14610
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.152221530150172,
+      "learning_rate": 1.7162523432135258e-07,
+      "loss": 0.2453,
+      "step": 14611
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9981888361709716,
+      "learning_rate": 1.7136078164496585e-07,
+      "loss": 0.2567,
+      "step": 14612
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4227979480329496,
+      "learning_rate": 1.7109652931739717e-07,
+      "loss": 0.2439,
+      "step": 14613
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0908760732808362,
+      "learning_rate": 1.7083247734961117e-07,
+      "loss": 0.2493,
+      "step": 14614
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.599814462824902,
+      "learning_rate": 1.705686257525624e-07,
+      "loss": 0.2726,
+      "step": 14615
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6584689323598625,
+      "learning_rate": 1.703049745371993e-07,
+      "loss": 0.2503,
+      "step": 14616
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.167687084073127,
+      "learning_rate": 1.7004152371446146e-07,
+      "loss": 0.2372,
+      "step": 14617
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3676841305083616,
+      "learning_rate": 1.6977827329528017e-07,
+      "loss": 0.2458,
+      "step": 14618
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2506013133346543,
+      "learning_rate": 1.6951522329057778e-07,
+      "loss": 0.2397,
+      "step": 14619
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.640926651918078,
+      "learning_rate": 1.6925237371126835e-07,
+      "loss": 0.2512,
+      "step": 14620
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0368443257180364,
+      "learning_rate": 1.689897245682587e-07,
+      "loss": 0.2287,
+      "step": 14621
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0133313326001674,
+      "learning_rate": 1.687272758724462e-07,
+      "loss": 0.2415,
+      "step": 14622
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4267649921275667,
+      "learning_rate": 1.6846502763471939e-07,
+      "loss": 0.2583,
+      "step": 14623
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3878008863385654,
+      "learning_rate": 1.682029798659618e-07,
+      "loss": 0.268,
+      "step": 14624
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4831645260225281,
+      "learning_rate": 1.679411325770436e-07,
+      "loss": 0.2292,
+      "step": 14625
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.735729228860837,
+      "learning_rate": 1.6767948577883109e-07,
+      "loss": 0.2554,
+      "step": 14626
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0309999751510386,
+      "learning_rate": 1.6741803948218005e-07,
+      "loss": 0.2595,
+      "step": 14627
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.356528461492822,
+      "learning_rate": 1.6715679369793737e-07,
+      "loss": 0.2595,
+      "step": 14628
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.275932866756124,
+      "learning_rate": 1.6689574843694433e-07,
+      "loss": 0.261,
+      "step": 14629
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3970316663035125,
+      "learning_rate": 1.6663490371003122e-07,
+      "loss": 0.2604,
+      "step": 14630
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.076284490200964,
+      "learning_rate": 1.663742595280199e-07,
+      "loss": 0.2542,
+      "step": 14631
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.965953426197452,
+      "learning_rate": 1.661138159017267e-07,
+      "loss": 0.2699,
+      "step": 14632
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1067068405456113,
+      "learning_rate": 1.6585357284195748e-07,
+      "loss": 0.2421,
+      "step": 14633
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3118731101592522,
+      "learning_rate": 1.6559353035950964e-07,
+      "loss": 0.2449,
+      "step": 14634
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0456770023365625,
+      "learning_rate": 1.653336884651735e-07,
+      "loss": 0.2421,
+      "step": 14635
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2109111172764475,
+      "learning_rate": 1.650740471697293e-07,
+      "loss": 0.2572,
+      "step": 14636
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8522415018301728,
+      "learning_rate": 1.648146064839512e-07,
+      "loss": 0.2899,
+      "step": 14637
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0267974383100484,
+      "learning_rate": 1.6455536641860282e-07,
+      "loss": 0.2638,
+      "step": 14638
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.106440965143541,
+      "learning_rate": 1.6429632698444165e-07,
+      "loss": 0.2435,
+      "step": 14639
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.92420689986729,
+      "learning_rate": 1.6403748819221464e-07,
+      "loss": 0.2491,
+      "step": 14640
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6832401826812362,
+      "learning_rate": 1.6377885005266214e-07,
+      "loss": 0.246,
+      "step": 14641
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5134706548275112,
+      "learning_rate": 1.6352041257651496e-07,
+      "loss": 0.2451,
+      "step": 14642
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9732603558655655,
+      "learning_rate": 1.6326217577449677e-07,
+      "loss": 0.2621,
+      "step": 14643
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5648125019183925,
+      "learning_rate": 1.6300413965732176e-07,
+      "loss": 0.2573,
+      "step": 14644
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2224410022222307,
+      "learning_rate": 1.6274630423569582e-07,
+      "loss": 0.2426,
+      "step": 14645
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.098652634617864,
+      "learning_rate": 1.6248866952031816e-07,
+      "loss": 0.2636,
+      "step": 14646
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9518050289061362,
+      "learning_rate": 1.622312355218786e-07,
+      "loss": 0.2593,
+      "step": 14647
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2932777745542174,
+      "learning_rate": 1.6197400225105686e-07,
+      "loss": 0.246,
+      "step": 14648
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.716278503850628,
+      "learning_rate": 1.617169697185278e-07,
+      "loss": 0.2448,
+      "step": 14649
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.036732537531056,
+      "learning_rate": 1.6146013793495564e-07,
+      "loss": 0.2515,
+      "step": 14650
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.430620711198442,
+      "learning_rate": 1.612035069109963e-07,
+      "loss": 0.2367,
+      "step": 14651
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.377390063577401,
+      "learning_rate": 1.6094707665729736e-07,
+      "loss": 0.2644,
+      "step": 14652
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6673442596094272,
+      "learning_rate": 1.6069084718449978e-07,
+      "loss": 0.2721,
+      "step": 14653
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7329225185422226,
+      "learning_rate": 1.6043481850323505e-07,
+      "loss": 0.2418,
+      "step": 14654
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.859340110203497,
+      "learning_rate": 1.6017899062412578e-07,
+      "loss": 0.2399,
+      "step": 14655
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.983389765371011,
+      "learning_rate": 1.5992336355778572e-07,
+      "loss": 0.2482,
+      "step": 14656
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.257335331812706,
+      "learning_rate": 1.59667937314823e-07,
+      "loss": 0.2784,
+      "step": 14657
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.519793464270759,
+      "learning_rate": 1.5941271190583473e-07,
+      "loss": 0.2484,
+      "step": 14658
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7246468564172275,
+      "learning_rate": 1.5915768734141135e-07,
+      "loss": 0.2494,
+      "step": 14659
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.079357549788692,
+      "learning_rate": 1.5890286363213326e-07,
+      "loss": 0.2515,
+      "step": 14660
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3967370792960778,
+      "learning_rate": 1.5864824078857365e-07,
+      "loss": 0.2315,
+      "step": 14661
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.742727083738558,
+      "learning_rate": 1.5839381882129856e-07,
+      "loss": 0.2474,
+      "step": 14662
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7703861623718138,
+      "learning_rate": 1.5813959774086396e-07,
+      "loss": 0.2495,
+      "step": 14663
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.21980916951687,
+      "learning_rate": 1.5788557755781642e-07,
+      "loss": 0.2424,
+      "step": 14664
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.670680642865481,
+      "learning_rate": 1.576317582826975e-07,
+      "loss": 0.244,
+      "step": 14665
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7967100125911086,
+      "learning_rate": 1.5737813992603822e-07,
+      "loss": 0.2396,
+      "step": 14666
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4662996903675816,
+      "learning_rate": 1.571247224983613e-07,
+      "loss": 0.2282,
+      "step": 14667
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.131216940287285,
+      "learning_rate": 1.5687150601018053e-07,
+      "loss": 0.246,
+      "step": 14668
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.264780672903606,
+      "learning_rate": 1.5661849047200361e-07,
+      "loss": 0.2483,
+      "step": 14669
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9630664181022202,
+      "learning_rate": 1.5636567589432939e-07,
+      "loss": 0.2606,
+      "step": 14670
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7084230757094758,
+      "learning_rate": 1.5611306228764612e-07,
+      "loss": 0.277,
+      "step": 14671
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6489836653289587,
+      "learning_rate": 1.5586064966243487e-07,
+      "loss": 0.2559,
+      "step": 14672
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3012312620610775,
+      "learning_rate": 1.5560843802917001e-07,
+      "loss": 0.2304,
+      "step": 14673
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5971566825111916,
+      "learning_rate": 1.5535642739831602e-07,
+      "loss": 0.2489,
+      "step": 14674
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.060596037769323,
+      "learning_rate": 1.5510461778032836e-07,
+      "loss": 0.2475,
+      "step": 14675
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3823634922069425,
+      "learning_rate": 1.5485300918565593e-07,
+      "loss": 0.2478,
+      "step": 14676
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1116629273618397,
+      "learning_rate": 1.5460160162473757e-07,
+      "loss": 0.2436,
+      "step": 14677
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.303742306352026,
+      "learning_rate": 1.5435039510800555e-07,
+      "loss": 0.2688,
+      "step": 14678
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5724092937533349,
+      "learning_rate": 1.5409938964588254e-07,
+      "loss": 0.4582,
+      "step": 14679
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3640143428741367,
+      "learning_rate": 1.538485852487831e-07,
+      "loss": 0.252,
+      "step": 14680
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.24412176119006,
+      "learning_rate": 1.5359798192711383e-07,
+      "loss": 0.2668,
+      "step": 14681
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9384342811547017,
+      "learning_rate": 1.5334757969127202e-07,
+      "loss": 0.2374,
+      "step": 14682
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9555963481625724,
+      "learning_rate": 1.5309737855164763e-07,
+      "loss": 0.2467,
+      "step": 14683
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1158433271097215,
+      "learning_rate": 1.5284737851862296e-07,
+      "loss": 0.2479,
+      "step": 14684
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0871617596354586,
+      "learning_rate": 1.525975796025686e-07,
+      "loss": 0.2537,
+      "step": 14685
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6673814261868143,
+      "learning_rate": 1.523479818138518e-07,
+      "loss": 0.2444,
+      "step": 14686
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9984525737896035,
+      "learning_rate": 1.520985851628276e-07,
+      "loss": 0.2594,
+      "step": 14687
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.968073224426047,
+      "learning_rate": 1.518493896598433e-07,
+      "loss": 0.2605,
+      "step": 14688
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.527886828538164,
+      "learning_rate": 1.5160039531523952e-07,
+      "loss": 0.2522,
+      "step": 14689
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.890379303935964,
+      "learning_rate": 1.5135160213934685e-07,
+      "loss": 0.2545,
+      "step": 14690
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7934664789798513,
+      "learning_rate": 1.5110301014248874e-07,
+      "loss": 0.2629,
+      "step": 14691
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.08821812985751,
+      "learning_rate": 1.508546193349797e-07,
+      "loss": 0.2369,
+      "step": 14692
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.761791256746903,
+      "learning_rate": 1.5060642972712425e-07,
+      "loss": 0.2438,
+      "step": 14693
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9194521860460347,
+      "learning_rate": 1.503584413292225e-07,
+      "loss": 0.2624,
+      "step": 14694
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.392746401289107,
+      "learning_rate": 1.501106541515629e-07,
+      "loss": 0.2517,
+      "step": 14695
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9097213975952536,
+      "learning_rate": 1.4986306820442609e-07,
+      "loss": 0.2392,
+      "step": 14696
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.627763148595492,
+      "learning_rate": 1.4961568349808607e-07,
+      "loss": 0.2517,
+      "step": 14697
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1211150282529068,
+      "learning_rate": 1.493685000428058e-07,
+      "loss": 0.258,
+      "step": 14698
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3680378321681013,
+      "learning_rate": 1.491215178488431e-07,
+      "loss": 0.2396,
+      "step": 14699
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6550561107235926,
+      "learning_rate": 1.4887473692644428e-07,
+      "loss": 0.2441,
+      "step": 14700
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.03823116777284,
+      "learning_rate": 1.4862815728584888e-07,
+      "loss": 0.254,
+      "step": 14701
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.08310411298709,
+      "learning_rate": 1.4838177893728878e-07,
+      "loss": 0.2368,
+      "step": 14702
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7705972847750595,
+      "learning_rate": 1.481356018909863e-07,
+      "loss": 0.2491,
+      "step": 14703
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.236174680489363,
+      "learning_rate": 1.4788962615715498e-07,
+      "loss": 0.2484,
+      "step": 14704
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7189632644735378,
+      "learning_rate": 1.4764385174600104e-07,
+      "loss": 0.2314,
+      "step": 14705
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.197762770919978,
+      "learning_rate": 1.4739827866772304e-07,
+      "loss": 0.2461,
+      "step": 14706
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.039026192120943,
+      "learning_rate": 1.4715290693251006e-07,
+      "loss": 0.2515,
+      "step": 14707
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.676748054690642,
+      "learning_rate": 1.4690773655054225e-07,
+      "loss": 0.2501,
+      "step": 14708
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.409012596567277,
+      "learning_rate": 1.4666276753199204e-07,
+      "loss": 0.2634,
+      "step": 14709
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.662885215623147,
+      "learning_rate": 1.4641799988702466e-07,
+      "loss": 0.2401,
+      "step": 14710
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6202459054772875,
+      "learning_rate": 1.4617343362579528e-07,
+      "loss": 0.2667,
+      "step": 14711
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.967525754697255,
+      "learning_rate": 1.4592906875845137e-07,
+      "loss": 0.2438,
+      "step": 14712
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8626807965066814,
+      "learning_rate": 1.4568490529513203e-07,
+      "loss": 0.2446,
+      "step": 14713
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0173005214412774,
+      "learning_rate": 1.454409432459686e-07,
+      "loss": 0.2634,
+      "step": 14714
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8575748264284897,
+      "learning_rate": 1.4519718262108296e-07,
+      "loss": 0.2685,
+      "step": 14715
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4461683819404567,
+      "learning_rate": 1.4495362343058872e-07,
+      "loss": 0.2551,
+      "step": 14716
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1198459156949172,
+      "learning_rate": 1.4471026568459335e-07,
+      "loss": 0.2503,
+      "step": 14717
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.660210048012057,
+      "learning_rate": 1.4446710939319263e-07,
+      "loss": 0.2433,
+      "step": 14718
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.711256064625923,
+      "learning_rate": 1.4422415456647577e-07,
+      "loss": 0.2506,
+      "step": 14719
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.548776361514621,
+      "learning_rate": 1.4398140121452297e-07,
+      "loss": 0.2609,
+      "step": 14720
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3296713450433755,
+      "learning_rate": 1.4373884934740733e-07,
+      "loss": 0.2477,
+      "step": 14721
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6352457523620478,
+      "learning_rate": 1.43496498975193e-07,
+      "loss": 0.2445,
+      "step": 14722
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5220597702437866,
+      "learning_rate": 1.432543501079353e-07,
+      "loss": 0.2491,
+      "step": 14723
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.929715711269501,
+      "learning_rate": 1.4301240275568006e-07,
+      "loss": 0.2342,
+      "step": 14724
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6164918458979822,
+      "learning_rate": 1.4277065692846815e-07,
+      "loss": 0.2332,
+      "step": 14725
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6748028845409968,
+      "learning_rate": 1.425291126363293e-07,
+      "loss": 0.2513,
+      "step": 14726
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2800775503316286,
+      "learning_rate": 1.422877698892844e-07,
+      "loss": 0.2614,
+      "step": 14727
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.630118151963712,
+      "learning_rate": 1.420466286973493e-07,
+      "loss": 0.228,
+      "step": 14728
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7986034412082135,
+      "learning_rate": 1.4180568907052717e-07,
+      "loss": 0.2692,
+      "step": 14729
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.117392210899359,
+      "learning_rate": 1.4156495101881662e-07,
+      "loss": 0.2381,
+      "step": 14730
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.937335080838224,
+      "learning_rate": 1.4132441455220637e-07,
+      "loss": 0.2561,
+      "step": 14731
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.191753514025765,
+      "learning_rate": 1.4108407968067506e-07,
+      "loss": 0.2523,
+      "step": 14732
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.810721605975489,
+      "learning_rate": 1.4084394641419641e-07,
+      "loss": 0.2494,
+      "step": 14733
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8148423028706926,
+      "learning_rate": 1.4060401476273356e-07,
+      "loss": 0.2674,
+      "step": 14734
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.572950294863925,
+      "learning_rate": 1.403642847362402e-07,
+      "loss": 0.2552,
+      "step": 14735
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.585125551063138,
+      "learning_rate": 1.401247563446656e-07,
+      "loss": 0.2495,
+      "step": 14736
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.390141227624101,
+      "learning_rate": 1.3988542959794627e-07,
+      "loss": 0.2582,
+      "step": 14737
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6615242836036774,
+      "learning_rate": 1.3964630450601314e-07,
+      "loss": 0.2558,
+      "step": 14738
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8693471718371781,
+      "learning_rate": 1.3940738107878826e-07,
+      "loss": 0.253,
+      "step": 14739
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1627254114858387,
+      "learning_rate": 1.3916865932618373e-07,
+      "loss": 0.2313,
+      "step": 14740
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4294468952066604,
+      "learning_rate": 1.389301392581055e-07,
+      "loss": 0.243,
+      "step": 14741
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5798209152898646,
+      "learning_rate": 1.386918208844501e-07,
+      "loss": 0.2698,
+      "step": 14742
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.345631241114109,
+      "learning_rate": 1.3845370421510573e-07,
+      "loss": 0.2437,
+      "step": 14743
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3598763642601126,
+      "learning_rate": 1.3821578925995282e-07,
+      "loss": 0.2586,
+      "step": 14744
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4424637281151522,
+      "learning_rate": 1.3797807602886128e-07,
+      "loss": 0.2349,
+      "step": 14745
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.6177531521316855,
+      "learning_rate": 1.3774056453169647e-07,
+      "loss": 0.249,
+      "step": 14746
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7221479334460554,
+      "learning_rate": 1.3750325477831116e-07,
+      "loss": 0.2579,
+      "step": 14747
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1935476166031673,
+      "learning_rate": 1.3726614677855243e-07,
+      "loss": 0.2647,
+      "step": 14748
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.301822660137684,
+      "learning_rate": 1.3702924054225908e-07,
+      "loss": 0.2598,
+      "step": 14749
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8087374879222473,
+      "learning_rate": 1.367925360792599e-07,
+      "loss": 0.2465,
+      "step": 14750
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.626035391338491,
+      "learning_rate": 1.3655603339937652e-07,
+      "loss": 0.2419,
+      "step": 14751
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5735250489342802,
+      "learning_rate": 1.3631973251242158e-07,
+      "loss": 0.2521,
+      "step": 14752
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5395385324028603,
+      "learning_rate": 1.3608363342819952e-07,
+      "loss": 0.2329,
+      "step": 14753
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5198408950056004,
+      "learning_rate": 1.3584773615650748e-07,
+      "loss": 0.256,
+      "step": 14754
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9608517463716955,
+      "learning_rate": 1.3561204070713264e-07,
+      "loss": 0.2576,
+      "step": 14755
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0182165284492783,
+      "learning_rate": 1.3537654708985326e-07,
+      "loss": 0.2557,
+      "step": 14756
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6962218711990547,
+      "learning_rate": 1.351412553144421e-07,
+      "loss": 0.2469,
+      "step": 14757
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8662322458064182,
+      "learning_rate": 1.3490616539066138e-07,
+      "loss": 0.2358,
+      "step": 14758
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1189782158984234,
+      "learning_rate": 1.3467127732826545e-07,
+      "loss": 0.2495,
+      "step": 14759
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8511935868137062,
+      "learning_rate": 1.3443659113699993e-07,
+      "loss": 0.2464,
+      "step": 14760
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.088782987727659,
+      "learning_rate": 1.3420210682660194e-07,
+      "loss": 0.2713,
+      "step": 14761
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.7116420973206825,
+      "learning_rate": 1.3396782440680157e-07,
+      "loss": 0.2399,
+      "step": 14762
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3910080932062456,
+      "learning_rate": 1.337337438873193e-07,
+      "loss": 0.2592,
+      "step": 14763
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8280881242317046,
+      "learning_rate": 1.3349986527786684e-07,
+      "loss": 0.2558,
+      "step": 14764
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.283105581231077,
+      "learning_rate": 1.332661885881492e-07,
+      "loss": 0.2803,
+      "step": 14765
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2888267676262974,
+      "learning_rate": 1.3303271382786199e-07,
+      "loss": 0.2386,
+      "step": 14766
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3866444593266483,
+      "learning_rate": 1.3279944100669183e-07,
+      "loss": 0.2534,
+      "step": 14767
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0197531724971145,
+      "learning_rate": 1.3256637013431827e-07,
+      "loss": 0.2561,
+      "step": 14768
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3216086750641067,
+      "learning_rate": 1.3233350122041133e-07,
+      "loss": 0.2549,
+      "step": 14769
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8146968874953524,
+      "learning_rate": 1.3210083427463326e-07,
+      "loss": 0.2606,
+      "step": 14770
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.782544295627537,
+      "learning_rate": 1.318683693066386e-07,
+      "loss": 0.2327,
+      "step": 14771
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2115745025861973,
+      "learning_rate": 1.3163610632607128e-07,
+      "loss": 0.2511,
+      "step": 14772
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.529932750224458,
+      "learning_rate": 1.3140404534256912e-07,
+      "loss": 0.2469,
+      "step": 14773
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.592638550689956,
+      "learning_rate": 1.3117218636576113e-07,
+      "loss": 0.2387,
+      "step": 14774
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4553856056557426,
+      "learning_rate": 1.3094052940526735e-07,
+      "loss": 0.2684,
+      "step": 14775
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.79622933342868,
+      "learning_rate": 1.3070907447069903e-07,
+      "loss": 0.2478,
+      "step": 14776
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.23924406557907,
+      "learning_rate": 1.3047782157166013e-07,
+      "loss": 0.274,
+      "step": 14777
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.051012920838647,
+      "learning_rate": 1.3024677071774573e-07,
+      "loss": 0.2337,
+      "step": 14778
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4814995187614595,
+      "learning_rate": 1.3001592191854261e-07,
+      "loss": 0.2735,
+      "step": 14779
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8391784390693475,
+      "learning_rate": 1.2978527518362815e-07,
+      "loss": 0.2427,
+      "step": 14780
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.613378542900168,
+      "learning_rate": 1.2955483052257355e-07,
+      "loss": 0.4245,
+      "step": 14781
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.30056942275997,
+      "learning_rate": 1.2932458794494007e-07,
+      "loss": 0.2374,
+      "step": 14782
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7433319544034398,
+      "learning_rate": 1.2909454746028062e-07,
+      "loss": 0.2659,
+      "step": 14783
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3767453602211006,
+      "learning_rate": 1.288647090781403e-07,
+      "loss": 0.2436,
+      "step": 14784
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9230309022735703,
+      "learning_rate": 1.2863507280805488e-07,
+      "loss": 0.255,
+      "step": 14785
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2803826526183115,
+      "learning_rate": 1.2840563865955335e-07,
+      "loss": 0.2485,
+      "step": 14786
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3656651932396824,
+      "learning_rate": 1.2817640664215425e-07,
+      "loss": 0.2539,
+      "step": 14787
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.059815424855098,
+      "learning_rate": 1.2794737676536993e-07,
+      "loss": 0.2443,
+      "step": 14788
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9648153798521695,
+      "learning_rate": 1.277185490387023e-07,
+      "loss": 0.2516,
+      "step": 14789
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.101183941336599,
+      "learning_rate": 1.2748992347164647e-07,
+      "loss": 0.2503,
+      "step": 14790
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9660790907228658,
+      "learning_rate": 1.272615000736882e-07,
+      "loss": 0.2535,
+      "step": 14791
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3704224188798575,
+      "learning_rate": 1.270332788543055e-07,
+      "loss": 0.2598,
+      "step": 14792
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4233464489865375,
+      "learning_rate": 1.2680525982296688e-07,
+      "loss": 0.2513,
+      "step": 14793
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5876366069095593,
+      "learning_rate": 1.265774429891342e-07,
+      "loss": 0.2374,
+      "step": 14794
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1348939968086453,
+      "learning_rate": 1.2634982836225884e-07,
+      "loss": 0.244,
+      "step": 14795
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1557460702912743,
+      "learning_rate": 1.2612241595178654e-07,
+      "loss": 0.2486,
+      "step": 14796
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.403929852882607,
+      "learning_rate": 1.2589520576715142e-07,
+      "loss": 0.2632,
+      "step": 14797
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6775150488385586,
+      "learning_rate": 1.256681978177826e-07,
+      "loss": 0.2869,
+      "step": 14798
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.891385801479705,
+      "learning_rate": 1.2544139211309757e-07,
+      "loss": 0.2361,
+      "step": 14799
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3701499166962425,
+      "learning_rate": 1.2521478866250714e-07,
+      "loss": 0.2386,
+      "step": 14800
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7063263012705459,
+      "learning_rate": 1.2498838747541374e-07,
+      "loss": 0.2405,
+      "step": 14801
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5580683549274503,
+      "learning_rate": 1.2476218856121103e-07,
+      "loss": 0.238,
+      "step": 14802
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.144137443763496,
+      "learning_rate": 1.245361919292848e-07,
+      "loss": 0.2485,
+      "step": 14803
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.354295341876992,
+      "learning_rate": 1.2431039758901198e-07,
+      "loss": 0.239,
+      "step": 14804
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.679861511521898,
+      "learning_rate": 1.2408480554976067e-07,
+      "loss": 0.2415,
+      "step": 14805
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.945784493613841,
+      "learning_rate": 1.2385941582089168e-07,
+      "loss": 0.2479,
+      "step": 14806
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.563666531930614,
+      "learning_rate": 1.2363422841175644e-07,
+      "loss": 0.2722,
+      "step": 14807
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9128993895990256,
+      "learning_rate": 1.2340924333169802e-07,
+      "loss": 0.244,
+      "step": 14808
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.144993451579595,
+      "learning_rate": 1.2318446059005173e-07,
+      "loss": 0.2555,
+      "step": 14809
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.279521249213373,
+      "learning_rate": 1.2295988019614514e-07,
+      "loss": 0.2424,
+      "step": 14810
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2654319812179002,
+      "learning_rate": 1.227355021592952e-07,
+      "loss": 0.242,
+      "step": 14811
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.042725802533616,
+      "learning_rate": 1.225113264888128e-07,
+      "loss": 0.2491,
+      "step": 14812
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8868224572855978,
+      "learning_rate": 1.222873531939983e-07,
+      "loss": 0.2765,
+      "step": 14813
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.576975161629902,
+      "learning_rate": 1.2206358228414593e-07,
+      "loss": 0.2748,
+      "step": 14814
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.581179031064587,
+      "learning_rate": 1.218400137685394e-07,
+      "loss": 0.2444,
+      "step": 14815
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.887526890406968,
+      "learning_rate": 1.2161664765645454e-07,
+      "loss": 0.2552,
+      "step": 14816
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.500305220643014,
+      "learning_rate": 1.2139348395716066e-07,
+      "loss": 0.2606,
+      "step": 14817
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.79615867270646,
+      "learning_rate": 1.2117052267991648e-07,
+      "loss": 0.2489,
+      "step": 14818
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6042147400663841,
+      "learning_rate": 1.2094776383397344e-07,
+      "loss": 0.4384,
+      "step": 14819
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5319324540550467,
+      "learning_rate": 1.2072520742857363e-07,
+      "loss": 0.2571,
+      "step": 14820
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9823282127894013,
+      "learning_rate": 1.2050285347295077e-07,
+      "loss": 0.2385,
+      "step": 14821
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0395177496762456,
+      "learning_rate": 1.2028070197633246e-07,
+      "loss": 0.2582,
+      "step": 14822
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.045153436724005,
+      "learning_rate": 1.2005875294793522e-07,
+      "loss": 0.2829,
+      "step": 14823
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.798874262521781,
+      "learning_rate": 1.1983700639696727e-07,
+      "loss": 0.243,
+      "step": 14824
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5626622951190008,
+      "learning_rate": 1.1961546233263011e-07,
+      "loss": 0.2233,
+      "step": 14825
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.677377284479592,
+      "learning_rate": 1.193941207641164e-07,
+      "loss": 0.2554,
+      "step": 14826
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9631409156833521,
+      "learning_rate": 1.1917298170060987e-07,
+      "loss": 0.273,
+      "step": 14827
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.48583609118683,
+      "learning_rate": 1.1895204515128544e-07,
+      "loss": 0.2649,
+      "step": 14828
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.277225363412385,
+      "learning_rate": 1.1873131112530967e-07,
+      "loss": 0.2658,
+      "step": 14829
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1121952784711446,
+      "learning_rate": 1.1851077963184243e-07,
+      "loss": 0.2496,
+      "step": 14830
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5936972522637975,
+      "learning_rate": 1.1829045068003364e-07,
+      "loss": 0.2751,
+      "step": 14831
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.806239788123379,
+      "learning_rate": 1.1807032427902488e-07,
+      "loss": 0.2448,
+      "step": 14832
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.668576839612423,
+      "learning_rate": 1.1785040043794882e-07,
+      "loss": 0.2567,
+      "step": 14833
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5403410715242014,
+      "learning_rate": 1.1763067916593263e-07,
+      "loss": 0.2494,
+      "step": 14834
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1104326000099825,
+      "learning_rate": 1.1741116047209067e-07,
+      "loss": 0.2458,
+      "step": 14835
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.390632507976463,
+      "learning_rate": 1.1719184436553288e-07,
+      "loss": 0.2611,
+      "step": 14836
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.223425231824431,
+      "learning_rate": 1.1697273085535755e-07,
+      "loss": 0.2453,
+      "step": 14837
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.835627269380618,
+      "learning_rate": 1.1675381995065738e-07,
+      "loss": 0.2377,
+      "step": 14838
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.252125941938038,
+      "learning_rate": 1.1653511166051457e-07,
+      "loss": 0.2505,
+      "step": 14839
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7293188119700607,
+      "learning_rate": 1.1631660599400407e-07,
+      "loss": 0.2301,
+      "step": 14840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.947457346600411,
+      "learning_rate": 1.1609830296019142e-07,
+      "loss": 0.2451,
+      "step": 14841
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.380286439795994,
+      "learning_rate": 1.1588020256813604e-07,
+      "loss": 0.2607,
+      "step": 14842
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6688985718457132,
+      "learning_rate": 1.1566230482688567e-07,
+      "loss": 0.2476,
+      "step": 14843
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.491500712169069,
+      "learning_rate": 1.1544460974548199e-07,
+      "loss": 0.2514,
+      "step": 14844
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8003044903232,
+      "learning_rate": 1.1522711733295722e-07,
+      "loss": 0.252,
+      "step": 14845
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9114935174019365,
+      "learning_rate": 1.1500982759833579e-07,
+      "loss": 0.2585,
+      "step": 14846
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4443451893432773,
+      "learning_rate": 1.1479274055063327e-07,
+      "loss": 0.2583,
+      "step": 14847
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.937288000014147,
+      "learning_rate": 1.1457585619885692e-07,
+      "loss": 0.2452,
+      "step": 14848
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.854551755026898,
+      "learning_rate": 1.1435917455200562e-07,
+      "loss": 0.2517,
+      "step": 14849
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9920062840874568,
+      "learning_rate": 1.1414269561907054e-07,
+      "loss": 0.2611,
+      "step": 14850
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.664366102243156,
+      "learning_rate": 1.1392641940903337e-07,
+      "loss": 0.2729,
+      "step": 14851
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3581522728478457,
+      "learning_rate": 1.1371034593086749e-07,
+      "loss": 0.2433,
+      "step": 14852
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.339598302772925,
+      "learning_rate": 1.1349447519353796e-07,
+      "loss": 0.2434,
+      "step": 14853
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3366893826110973,
+      "learning_rate": 1.1327880720600205e-07,
+      "loss": 0.2384,
+      "step": 14854
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6284451781576239,
+      "learning_rate": 1.1306334197720925e-07,
+      "loss": 0.4828,
+      "step": 14855
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.561425902150267,
+      "learning_rate": 1.12848079516098e-07,
+      "loss": 0.2665,
+      "step": 14856
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.023058133647233,
+      "learning_rate": 1.1263301983160002e-07,
+      "loss": 0.2497,
+      "step": 14857
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.916481387111643,
+      "learning_rate": 1.124181629326393e-07,
+      "loss": 0.2697,
+      "step": 14858
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4385159119139947,
+      "learning_rate": 1.1220350882813091e-07,
+      "loss": 0.2673,
+      "step": 14859
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.651392137228159,
+      "learning_rate": 1.1198905752697997e-07,
+      "loss": 0.2486,
+      "step": 14860
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9391530254969616,
+      "learning_rate": 1.117748090380849e-07,
+      "loss": 0.2568,
+      "step": 14861
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8450702886107342,
+      "learning_rate": 1.1156076337033583e-07,
+      "loss": 0.2385,
+      "step": 14862
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6987128208037827,
+      "learning_rate": 1.1134692053261343e-07,
+      "loss": 0.2416,
+      "step": 14863
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.674250944224833,
+      "learning_rate": 1.1113328053379113e-07,
+      "loss": 0.2634,
+      "step": 14864
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0697943603454942,
+      "learning_rate": 1.1091984338273188e-07,
+      "loss": 0.2233,
+      "step": 14865
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8118060020421574,
+      "learning_rate": 1.1070660908829245e-07,
+      "loss": 0.2265,
+      "step": 14866
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.651358541009252,
+      "learning_rate": 1.1049357765932078e-07,
+      "loss": 0.2519,
+      "step": 14867
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8171109253631073,
+      "learning_rate": 1.1028074910465481e-07,
+      "loss": 0.2572,
+      "step": 14868
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3117779403954093,
+      "learning_rate": 1.1006812343312467e-07,
+      "loss": 0.2469,
+      "step": 14869
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5892931590116841,
+      "learning_rate": 1.09855700653555e-07,
+      "loss": 0.4931,
+      "step": 14870
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5588480046248643,
+      "learning_rate": 1.0964348077475817e-07,
+      "loss": 0.2355,
+      "step": 14871
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4196301254817536,
+      "learning_rate": 1.0943146380553938e-07,
+      "loss": 0.257,
+      "step": 14872
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6697389354697143,
+      "learning_rate": 1.092196497546949e-07,
+      "loss": 0.2584,
+      "step": 14873
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2976118335943434,
+      "learning_rate": 1.0900803863101494e-07,
+      "loss": 0.2447,
+      "step": 14874
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5761961831633281,
+      "learning_rate": 1.0879663044327915e-07,
+      "loss": 0.4653,
+      "step": 14875
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8734145186457203,
+      "learning_rate": 1.0858542520025828e-07,
+      "loss": 0.2423,
+      "step": 14876
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.153025758468859,
+      "learning_rate": 1.0837442291071587e-07,
+      "loss": 0.2702,
+      "step": 14877
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.889608873689613,
+      "learning_rate": 1.0816362358340826e-07,
+      "loss": 0.2807,
+      "step": 14878
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5377182230887265,
+      "learning_rate": 1.0795302722708012e-07,
+      "loss": 0.2508,
+      "step": 14879
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1302354393816434,
+      "learning_rate": 1.0774263385047057e-07,
+      "loss": 0.2591,
+      "step": 14880
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1321161995180526,
+      "learning_rate": 1.0753244346230818e-07,
+      "loss": 0.2572,
+      "step": 14881
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.977443119313658,
+      "learning_rate": 1.0732245607131542e-07,
+      "loss": 0.264,
+      "step": 14882
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.755625879920134,
+      "learning_rate": 1.071126716862042e-07,
+      "loss": 0.2435,
+      "step": 14883
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2010017414711407,
+      "learning_rate": 1.0690309031567814e-07,
+      "loss": 0.2478,
+      "step": 14884
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1106095110096064,
+      "learning_rate": 1.066937119684347e-07,
+      "loss": 0.2456,
+      "step": 14885
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1943699278626636,
+      "learning_rate": 1.0648453665316028e-07,
+      "loss": 0.2483,
+      "step": 14886
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5251210715085795,
+      "learning_rate": 1.0627556437853459e-07,
+      "loss": 0.2629,
+      "step": 14887
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7813309320362822,
+      "learning_rate": 1.0606679515322849e-07,
+      "loss": 0.2611,
+      "step": 14888
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.591465764685345,
+      "learning_rate": 1.0585822898590281e-07,
+      "loss": 0.2479,
+      "step": 14889
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6430687983850112,
+      "learning_rate": 1.056498658852123e-07,
+      "loss": 0.2616,
+      "step": 14890
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6878018452247103,
+      "learning_rate": 1.0544170585980229e-07,
+      "loss": 0.2474,
+      "step": 14891
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3871371505976424,
+      "learning_rate": 1.0523374891830972e-07,
+      "loss": 0.237,
+      "step": 14892
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.598496951092441,
+      "learning_rate": 1.0502599506936273e-07,
+      "loss": 0.2536,
+      "step": 14893
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4978143617102027,
+      "learning_rate": 1.0481844432158162e-07,
+      "loss": 0.2616,
+      "step": 14894
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9994766603064726,
+      "learning_rate": 1.0461109668357894e-07,
+      "loss": 0.2497,
+      "step": 14895
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6608440324327385,
+      "learning_rate": 1.0440395216395616e-07,
+      "loss": 0.2401,
+      "step": 14896
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.151947723333777,
+      "learning_rate": 1.0419701077130917e-07,
+      "loss": 0.2757,
+      "step": 14897
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.112650879500353,
+      "learning_rate": 1.0399027251422444e-07,
+      "loss": 0.2534,
+      "step": 14898
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.759699635665989,
+      "learning_rate": 1.0378373740127901e-07,
+      "loss": 0.2596,
+      "step": 14899
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6249261058165243,
+      "learning_rate": 1.0357740544104323e-07,
+      "loss": 0.2532,
+      "step": 14900
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1306714110980276,
+      "learning_rate": 1.0337127664207803e-07,
+      "loss": 0.2454,
+      "step": 14901
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.653563634502309,
+      "learning_rate": 1.0316535101293601e-07,
+      "loss": 0.2685,
+      "step": 14902
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.025075574024474,
+      "learning_rate": 1.0295962856216146e-07,
+      "loss": 0.2491,
+      "step": 14903
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9884337748471876,
+      "learning_rate": 1.0275410929828977e-07,
+      "loss": 0.2661,
+      "step": 14904
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6602176710280725,
+      "learning_rate": 1.0254879322984856e-07,
+      "loss": 0.2496,
+      "step": 14905
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.579321150659445,
+      "learning_rate": 1.0234368036535658e-07,
+      "loss": 0.2698,
+      "step": 14906
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0472771201422084,
+      "learning_rate": 1.0213877071332478e-07,
+      "loss": 0.242,
+      "step": 14907
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2621855761825884,
+      "learning_rate": 1.0193406428225528e-07,
+      "loss": 0.266,
+      "step": 14908
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.860810566412694,
+      "learning_rate": 1.017295610806407e-07,
+      "loss": 0.2344,
+      "step": 14909
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.795315979060604,
+      "learning_rate": 1.0152526111696759e-07,
+      "loss": 0.2565,
+      "step": 14910
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.634323388444125,
+      "learning_rate": 1.0132116439971196e-07,
+      "loss": 0.2555,
+      "step": 14911
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5809638260400033,
+      "learning_rate": 1.0111727093734203e-07,
+      "loss": 0.4658,
+      "step": 14912
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.379199203821618,
+      "learning_rate": 1.009135807383177e-07,
+      "loss": 0.2499,
+      "step": 14913
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.57060843423349,
+      "learning_rate": 1.0071009381109053e-07,
+      "loss": 0.2687,
+      "step": 14914
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6136274412195344,
+      "learning_rate": 1.0050681016410435e-07,
+      "loss": 0.2417,
+      "step": 14915
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.465798957401808,
+      "learning_rate": 1.0030372980579294e-07,
+      "loss": 0.2354,
+      "step": 14916
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.478186727032362,
+      "learning_rate": 1.0010085274458236e-07,
+      "loss": 0.2527,
+      "step": 14917
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3420232502846567,
+      "learning_rate": 9.989817898889087e-08,
+      "loss": 0.2502,
+      "step": 14918
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5862422368537016,
+      "learning_rate": 9.969570854712785e-08,
+      "loss": 0.463,
+      "step": 14919
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.186868427191014,
+      "learning_rate": 9.949344142769323e-08,
+      "loss": 0.2473,
+      "step": 14920
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4484885809275756,
+      "learning_rate": 9.929137763897923e-08,
+      "loss": 0.2739,
+      "step": 14921
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5240728567295854,
+      "learning_rate": 9.908951718937187e-08,
+      "loss": 0.2563,
+      "step": 14922
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4450657920114627,
+      "learning_rate": 9.888786008724504e-08,
+      "loss": 0.2589,
+      "step": 14923
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.12220374274891,
+      "learning_rate": 9.868640634096649e-08,
+      "loss": 0.2486,
+      "step": 14924
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9133550174936365,
+      "learning_rate": 9.848515595889452e-08,
+      "loss": 0.2531,
+      "step": 14925
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6291496748320533,
+      "learning_rate": 9.828410894937912e-08,
+      "loss": 0.2408,
+      "step": 14926
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.610575991622655,
+      "learning_rate": 9.808326532076307e-08,
+      "loss": 0.2346,
+      "step": 14927
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1243219912170472,
+      "learning_rate": 9.788262508137913e-08,
+      "loss": 0.2532,
+      "step": 14928
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0825214835054133,
+      "learning_rate": 9.768218823955122e-08,
+      "loss": 0.2421,
+      "step": 14929
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.534624003973788,
+      "learning_rate": 9.748195480359768e-08,
+      "loss": 0.2403,
+      "step": 14930
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8788918273845305,
+      "learning_rate": 9.728192478182574e-08,
+      "loss": 0.2333,
+      "step": 14931
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2294369483115686,
+      "learning_rate": 9.708209818253489e-08,
+      "loss": 0.2545,
+      "step": 14932
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6569128406506433,
+      "learning_rate": 9.68824750140157e-08,
+      "loss": 0.2584,
+      "step": 14933
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.83855571000696,
+      "learning_rate": 9.668305528455212e-08,
+      "loss": 0.2572,
+      "step": 14934
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9125795231859963,
+      "learning_rate": 9.648383900241808e-08,
+      "loss": 0.2447,
+      "step": 14935
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.365058972026654,
+      "learning_rate": 9.628482617587864e-08,
+      "loss": 0.2636,
+      "step": 14936
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 21.5959645402046,
+      "learning_rate": 9.608601681319163e-08,
+      "loss": 0.2661,
+      "step": 14937
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2741830157771217,
+      "learning_rate": 9.588741092260656e-08,
+      "loss": 0.2293,
+      "step": 14938
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5611052605208338,
+      "learning_rate": 9.568900851236351e-08,
+      "loss": 0.2521,
+      "step": 14939
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9518424864515562,
+      "learning_rate": 9.549080959069423e-08,
+      "loss": 0.2522,
+      "step": 14940
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5103558130100954,
+      "learning_rate": 9.52928141658227e-08,
+      "loss": 0.2702,
+      "step": 14941
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3833230777335555,
+      "learning_rate": 9.509502224596401e-08,
+      "loss": 0.2482,
+      "step": 14942
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6141775631828845,
+      "learning_rate": 9.489743383932548e-08,
+      "loss": 0.2636,
+      "step": 14943
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.225754419827806,
+      "learning_rate": 9.470004895410445e-08,
+      "loss": 0.2495,
+      "step": 14944
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.493882899914474,
+      "learning_rate": 9.450286759849048e-08,
+      "loss": 0.2477,
+      "step": 14945
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.010916841228309,
+      "learning_rate": 9.430588978066647e-08,
+      "loss": 0.2546,
+      "step": 14946
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8776254788072027,
+      "learning_rate": 9.410911550880474e-08,
+      "loss": 0.2467,
+      "step": 14947
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6419471230118083,
+      "learning_rate": 9.391254479106993e-08,
+      "loss": 0.2477,
+      "step": 14948
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.398983459069012,
+      "learning_rate": 9.371617763561658e-08,
+      "loss": 0.2386,
+      "step": 14949
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8036445531365404,
+      "learning_rate": 9.352001405059486e-08,
+      "loss": 0.2445,
+      "step": 14950
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.566218181636836,
+      "learning_rate": 9.332405404414158e-08,
+      "loss": 0.2581,
+      "step": 14951
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9469322935504314,
+      "learning_rate": 9.312829762438914e-08,
+      "loss": 0.2408,
+      "step": 14952
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2739847687461685,
+      "learning_rate": 9.29327447994588e-08,
+      "loss": 0.2665,
+      "step": 14953
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6358838442443655,
+      "learning_rate": 9.27373955774652e-08,
+      "loss": 0.2404,
+      "step": 14954
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.082394203426563,
+      "learning_rate": 9.254224996651351e-08,
+      "loss": 0.2456,
+      "step": 14955
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9432220621634133,
+      "learning_rate": 9.23473079747006e-08,
+      "loss": 0.2391,
+      "step": 14956
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6062936362069737,
+      "learning_rate": 9.215256961011443e-08,
+      "loss": 0.2485,
+      "step": 14957
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.917760650477404,
+      "learning_rate": 9.195803488083521e-08,
+      "loss": 0.248,
+      "step": 14958
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.693826435999172,
+      "learning_rate": 9.176370379493482e-08,
+      "loss": 0.2592,
+      "step": 14959
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.468578559414902,
+      "learning_rate": 9.15695763604768e-08,
+      "loss": 0.2753,
+      "step": 14960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.576035149210496,
+      "learning_rate": 9.137565258551529e-08,
+      "loss": 0.247,
+      "step": 14961
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4686507036846277,
+      "learning_rate": 9.11819324780966e-08,
+      "loss": 0.2434,
+      "step": 14962
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.0674562135055385,
+      "learning_rate": 9.098841604625874e-08,
+      "loss": 0.2457,
+      "step": 14963
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1382600137892642,
+      "learning_rate": 9.079510329803087e-08,
+      "loss": 0.2476,
+      "step": 14964
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5833971663826814,
+      "learning_rate": 9.060199424143378e-08,
+      "loss": 0.27,
+      "step": 14965
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9141649054444287,
+      "learning_rate": 9.040908888447941e-08,
+      "loss": 0.2503,
+      "step": 14966
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.8754388846088315,
+      "learning_rate": 9.021638723517357e-08,
+      "loss": 0.2495,
+      "step": 14967
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5964389953001117,
+      "learning_rate": 9.002388930150984e-08,
+      "loss": 0.2387,
+      "step": 14968
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7661147240402528,
+      "learning_rate": 8.983159509147577e-08,
+      "loss": 0.2673,
+      "step": 14969
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7161758825257591,
+      "learning_rate": 8.963950461305104e-08,
+      "loss": 0.2499,
+      "step": 14970
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.12529174164548,
+      "learning_rate": 8.944761787420486e-08,
+      "loss": 0.253,
+      "step": 14971
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4962954036350211,
+      "learning_rate": 8.925593488289918e-08,
+      "loss": 0.2443,
+      "step": 14972
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9436273201767413,
+      "learning_rate": 8.906445564708655e-08,
+      "loss": 0.2393,
+      "step": 14973
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7900452883307112,
+      "learning_rate": 8.88731801747128e-08,
+      "loss": 0.2557,
+      "step": 14974
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5962253752787288,
+      "learning_rate": 8.868210847371384e-08,
+      "loss": 0.2517,
+      "step": 14975
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5808888653862876,
+      "learning_rate": 8.84912405520183e-08,
+      "loss": 0.2523,
+      "step": 14976
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7288831022952715,
+      "learning_rate": 8.83005764175443e-08,
+      "loss": 0.2666,
+      "step": 14977
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3162567983125717,
+      "learning_rate": 8.811011607820386e-08,
+      "loss": 0.247,
+      "step": 14978
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.445641448881334,
+      "learning_rate": 8.791985954189952e-08,
+      "loss": 0.2445,
+      "step": 14979
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1815497088289924,
+      "learning_rate": 8.772980681652444e-08,
+      "loss": 0.2432,
+      "step": 14980
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.421640283671424,
+      "learning_rate": 8.753995790996505e-08,
+      "loss": 0.2469,
+      "step": 14981
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9510725059112817,
+      "learning_rate": 8.735031283009842e-08,
+      "loss": 0.2497,
+      "step": 14982
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.912487167504048,
+      "learning_rate": 8.716087158479324e-08,
+      "loss": 0.264,
+      "step": 14983
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.834676258213733,
+      "learning_rate": 8.697163418190935e-08,
+      "loss": 0.2354,
+      "step": 14984
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7326168534625863,
+      "learning_rate": 8.678260062929877e-08,
+      "loss": 0.243,
+      "step": 14985
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2672894190913695,
+      "learning_rate": 8.659377093480581e-08,
+      "loss": 0.2508,
+      "step": 14986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.359989194874794,
+      "learning_rate": 8.640514510626363e-08,
+      "loss": 0.2445,
+      "step": 14987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5369905580351089,
+      "learning_rate": 8.621672315149987e-08,
+      "loss": 0.2529,
+      "step": 14988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.060738960519203,
+      "learning_rate": 8.602850507833161e-08,
+      "loss": 0.2556,
+      "step": 14989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2529995206791567,
+      "learning_rate": 8.584049089456981e-08,
+      "loss": 0.2445,
+      "step": 14990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6858711108570388,
+      "learning_rate": 8.56526806080138e-08,
+      "loss": 0.2659,
+      "step": 14991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2813124187359204,
+      "learning_rate": 8.546507422645734e-08,
+      "loss": 0.265,
+      "step": 14992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.848471024723462,
+      "learning_rate": 8.52776717576842e-08,
+      "loss": 0.264,
+      "step": 14993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4011796314883185,
+      "learning_rate": 8.509047320946984e-08,
+      "loss": 0.25,
+      "step": 14994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0682365920212433,
+      "learning_rate": 8.490347858958192e-08,
+      "loss": 0.265,
+      "step": 14995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7556126978556454,
+      "learning_rate": 8.471668790577814e-08,
+      "loss": 0.23,
+      "step": 14996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0838923037676325,
+      "learning_rate": 8.45301011658095e-08,
+      "loss": 0.2506,
+      "step": 14997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.085841390214276,
+      "learning_rate": 8.434371837741817e-08,
+      "loss": 0.2425,
+      "step": 14998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8883906705242588,
+      "learning_rate": 8.415753954833738e-08,
+      "loss": 0.2346,
+      "step": 14999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.727605274229623,
+      "learning_rate": 8.397156468629209e-08,
+      "loss": 0.2544,
+      "step": 15000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.83609080121886,
+      "learning_rate": 8.378579379899776e-08,
+      "loss": 0.2489,
+      "step": 15001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3191386930002,
+      "learning_rate": 8.360022689416324e-08,
+      "loss": 0.2397,
+      "step": 15002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.977769675272214,
+      "learning_rate": 8.34148639794874e-08,
+      "loss": 0.273,
+      "step": 15003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.456322072507468,
+      "learning_rate": 8.322970506266237e-08,
+      "loss": 0.2705,
+      "step": 15004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5705162597152489,
+      "learning_rate": 8.304475015136925e-08,
+      "loss": 0.2507,
+      "step": 15005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7703958453252453,
+      "learning_rate": 8.285999925328359e-08,
+      "loss": 0.2599,
+      "step": 15006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6048938773161612,
+      "learning_rate": 8.267545237607034e-08,
+      "loss": 0.2806,
+      "step": 15007
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9367012150118517,
+      "learning_rate": 8.249110952738726e-08,
+      "loss": 0.2694,
+      "step": 15008
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.542691414723443,
+      "learning_rate": 8.230697071488159e-08,
+      "loss": 0.2444,
+      "step": 15009
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8070296743870038,
+      "learning_rate": 8.212303594619497e-08,
+      "loss": 0.2357,
+      "step": 15010
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6064423374607273,
+      "learning_rate": 8.193930522895799e-08,
+      "loss": 0.2422,
+      "step": 15011
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.226075116398095,
+      "learning_rate": 8.175577857079565e-08,
+      "loss": 0.2658,
+      "step": 15012
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4496437824003148,
+      "learning_rate": 8.157245597932129e-08,
+      "loss": 0.2333,
+      "step": 15013
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6669729270492224,
+      "learning_rate": 8.138933746214218e-08,
+      "loss": 0.2417,
+      "step": 15014
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7373856422281864,
+      "learning_rate": 8.120642302685556e-08,
+      "loss": 0.2746,
+      "step": 15015
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5336644560674157,
+      "learning_rate": 8.102371268105147e-08,
+      "loss": 0.2556,
+      "step": 15016
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0907280259399346,
+      "learning_rate": 8.084120643231052e-08,
+      "loss": 0.2781,
+      "step": 15017
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.130195687230351,
+      "learning_rate": 8.0658904288205e-08,
+      "loss": 0.2447,
+      "step": 15018
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9566567457339217,
+      "learning_rate": 8.047680625629994e-08,
+      "loss": 0.2598,
+      "step": 15019
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0714324441971175,
+      "learning_rate": 8.029491234414987e-08,
+      "loss": 0.245,
+      "step": 15020
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.337980440495657,
+      "learning_rate": 8.01132225593021e-08,
+      "loss": 0.2596,
+      "step": 15021
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.88405739578724,
+      "learning_rate": 7.993173690929556e-08,
+      "loss": 0.2569,
+      "step": 15022
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.187463976959984,
+      "learning_rate": 7.975045540166038e-08,
+      "loss": 0.2596,
+      "step": 15023
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6669701272689876,
+      "learning_rate": 7.956937804391829e-08,
+      "loss": 0.2441,
+      "step": 15024
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.487160049490356,
+      "learning_rate": 7.938850484358219e-08,
+      "loss": 0.2497,
+      "step": 15025
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.416276728830592,
+      "learning_rate": 7.920783580815661e-08,
+      "loss": 0.2573,
+      "step": 15026
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4075389471818622,
+      "learning_rate": 7.90273709451389e-08,
+      "loss": 0.2511,
+      "step": 15027
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9727910733199612,
+      "learning_rate": 7.884711026201586e-08,
+      "loss": 0.2477,
+      "step": 15028
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5693306121869144,
+      "learning_rate": 7.866705376626704e-08,
+      "loss": 0.4655,
+      "step": 15029
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0371422961176253,
+      "learning_rate": 7.848720146536426e-08,
+      "loss": 0.2606,
+      "step": 15030
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.406852070741884,
+      "learning_rate": 7.830755336676821e-08,
+      "loss": 0.2479,
+      "step": 15031
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5372067319191896,
+      "learning_rate": 7.812810947793403e-08,
+      "loss": 0.2406,
+      "step": 15032
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.855173786389777,
+      "learning_rate": 7.79488698063069e-08,
+      "loss": 0.2345,
+      "step": 15033
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7803893474498058,
+      "learning_rate": 7.776983435932306e-08,
+      "loss": 0.2698,
+      "step": 15034
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.087583867651988,
+      "learning_rate": 7.759100314441215e-08,
+      "loss": 0.2364,
+      "step": 15035
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5924903559618606,
+      "learning_rate": 7.741237616899377e-08,
+      "loss": 0.2318,
+      "step": 15036
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.902985871795474,
+      "learning_rate": 7.723395344047924e-08,
+      "loss": 0.2538,
+      "step": 15037
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4096437595077678,
+      "learning_rate": 7.705573496627205e-08,
+      "loss": 0.2553,
+      "step": 15038
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9456729915431172,
+      "learning_rate": 7.687772075376632e-08,
+      "loss": 0.2485,
+      "step": 15039
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4682587504945,
+      "learning_rate": 7.669991081034889e-08,
+      "loss": 0.256,
+      "step": 15040
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8428692360405576,
+      "learning_rate": 7.65223051433961e-08,
+      "loss": 0.2554,
+      "step": 15041
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.502670488676359,
+      "learning_rate": 7.634490376027814e-08,
+      "loss": 0.2487,
+      "step": 15042
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9273363329139828,
+      "learning_rate": 7.616770666835582e-08,
+      "loss": 0.2646,
+      "step": 15043
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7949329714879687,
+      "learning_rate": 7.599071387498102e-08,
+      "loss": 0.2502,
+      "step": 15044
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5112986784603015,
+      "learning_rate": 7.581392538749732e-08,
+      "loss": 0.2472,
+      "step": 15045
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.149502030469735,
+      "learning_rate": 7.563734121324052e-08,
+      "loss": 0.2416,
+      "step": 15046
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9709106388728097,
+      "learning_rate": 7.546096135953696e-08,
+      "loss": 0.2388,
+      "step": 15047
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7272547408062873,
+      "learning_rate": 7.528478583370469e-08,
+      "loss": 0.2577,
+      "step": 15048
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3685895048468675,
+      "learning_rate": 7.510881464305452e-08,
+      "loss": 0.2716,
+      "step": 15049
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.436988767256087,
+      "learning_rate": 7.493304779488675e-08,
+      "loss": 0.2371,
+      "step": 15050
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9474019351298308,
+      "learning_rate": 7.475748529649496e-08,
+      "loss": 0.2501,
+      "step": 15051
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.414570049926212,
+      "learning_rate": 7.458212715516389e-08,
+      "loss": 0.2675,
+      "step": 15052
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5213522155912707,
+      "learning_rate": 7.440697337816771e-08,
+      "loss": 0.2414,
+      "step": 15053
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6456077863037066,
+      "learning_rate": 7.423202397277618e-08,
+      "loss": 0.2575,
+      "step": 15054
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9318247935661166,
+      "learning_rate": 7.405727894624626e-08,
+      "loss": 0.2592,
+      "step": 15055
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.316707003497598,
+      "learning_rate": 7.388273830582937e-08,
+      "loss": 0.2544,
+      "step": 15056
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.678109446238628,
+      "learning_rate": 7.370840205876806e-08,
+      "loss": 0.2528,
+      "step": 15057
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0675570350701333,
+      "learning_rate": 7.353427021229486e-08,
+      "loss": 0.2776,
+      "step": 15058
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1464382725465234,
+      "learning_rate": 7.336034277363513e-08,
+      "loss": 0.2456,
+      "step": 15059
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.433912470881727,
+      "learning_rate": 7.318661975000585e-08,
+      "loss": 0.2528,
+      "step": 15060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9775996362674466,
+      "learning_rate": 7.301310114861404e-08,
+      "loss": 0.2734,
+      "step": 15061
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.356976812157187,
+      "learning_rate": 7.283978697666005e-08,
+      "loss": 0.2637,
+      "step": 15062
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4432642208060746,
+      "learning_rate": 7.266667724133536e-08,
+      "loss": 0.239,
+      "step": 15063
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.875363497132506,
+      "learning_rate": 7.249377194982199e-08,
+      "loss": 0.2493,
+      "step": 15064
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.750832913721003,
+      "learning_rate": 7.23210711092942e-08,
+      "loss": 0.2551,
+      "step": 15065
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1965919730419743,
+      "learning_rate": 7.214857472691794e-08,
+      "loss": 0.2591,
+      "step": 15066
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4181288946124706,
+      "learning_rate": 7.19762828098497e-08,
+      "loss": 0.2371,
+      "step": 15067
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9364263772404247,
+      "learning_rate": 7.18041953652393e-08,
+      "loss": 0.2429,
+      "step": 15068
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.27039854486909,
+      "learning_rate": 7.16323124002255e-08,
+      "loss": 0.2847,
+      "step": 15069
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.423011681705568,
+      "learning_rate": 7.146063392194147e-08,
+      "loss": 0.2654,
+      "step": 15070
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.917807672109599,
+      "learning_rate": 7.128915993750929e-08,
+      "loss": 0.2587,
+      "step": 15071
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7531678404609035,
+      "learning_rate": 7.111789045404493e-08,
+      "loss": 0.2276,
+      "step": 15072
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6049922607264762,
+      "learning_rate": 7.094682547865328e-08,
+      "loss": 0.2468,
+      "step": 15073
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1576768292165927,
+      "learning_rate": 7.07759650184331e-08,
+      "loss": 0.2592,
+      "step": 15074
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.952215411700562,
+      "learning_rate": 7.06053090804737e-08,
+      "loss": 0.251,
+      "step": 15075
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.157583088277711,
+      "learning_rate": 7.043485767185554e-08,
+      "loss": 0.2449,
+      "step": 15076
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5970967109403809,
+      "learning_rate": 7.026461079965075e-08,
+      "loss": 0.4728,
+      "step": 15077
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.564200936615172,
+      "learning_rate": 7.009456847092311e-08,
+      "loss": 0.2579,
+      "step": 15078
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3959489720807112,
+      "learning_rate": 6.99247306927292e-08,
+      "loss": 0.2551,
+      "step": 15079
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7185456841383286,
+      "learning_rate": 6.975509747211451e-08,
+      "loss": 0.2431,
+      "step": 15080
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.557992125173991,
+      "learning_rate": 6.958566881611783e-08,
+      "loss": 0.2555,
+      "step": 15081
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.473151134879203,
+      "learning_rate": 6.941644473176967e-08,
+      "loss": 0.2649,
+      "step": 15082
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8596363940244411,
+      "learning_rate": 6.924742522609052e-08,
+      "loss": 0.2565,
+      "step": 15083
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3633102982141923,
+      "learning_rate": 6.90786103060942e-08,
+      "loss": 0.2463,
+      "step": 15084
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0047519931535174,
+      "learning_rate": 6.890999997878345e-08,
+      "loss": 0.2507,
+      "step": 15085
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9008710161498206,
+      "learning_rate": 6.8741594251156e-08,
+      "loss": 0.2304,
+      "step": 15086
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4535875985794044,
+      "learning_rate": 6.857339313019906e-08,
+      "loss": 0.2404,
+      "step": 15087
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2091228671246594,
+      "learning_rate": 6.840539662289148e-08,
+      "loss": 0.2338,
+      "step": 15088
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5789767588666381,
+      "learning_rate": 6.823760473620267e-08,
+      "loss": 0.2537,
+      "step": 15089
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5449648442695354,
+      "learning_rate": 6.807001747709596e-08,
+      "loss": 0.2328,
+      "step": 15090
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4428670856828945,
+      "learning_rate": 6.790263485252413e-08,
+      "loss": 0.2286,
+      "step": 15091
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9390355575658578,
+      "learning_rate": 6.773545686943272e-08,
+      "loss": 0.2468,
+      "step": 15092
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6580568443351344,
+      "learning_rate": 6.756848353475732e-08,
+      "loss": 0.2672,
+      "step": 15093
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6013723138859637,
+      "learning_rate": 6.740171485542624e-08,
+      "loss": 0.2356,
+      "step": 15094
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8487835934831445,
+      "learning_rate": 6.723515083836008e-08,
+      "loss": 0.2636,
+      "step": 15095
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0637371279345085,
+      "learning_rate": 6.706879149046886e-08,
+      "loss": 0.2614,
+      "step": 15096
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.01843306386544,
+      "learning_rate": 6.690263681865539e-08,
+      "loss": 0.2572,
+      "step": 15097
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2889947065763074,
+      "learning_rate": 6.67366868298136e-08,
+      "loss": 0.2775,
+      "step": 15098
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7265614047933293,
+      "learning_rate": 6.657094153082911e-08,
+      "loss": 0.2431,
+      "step": 15099
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9326881289441615,
+      "learning_rate": 6.640540092857916e-08,
+      "loss": 0.2275,
+      "step": 15100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4671067198548577,
+      "learning_rate": 6.624006502993219e-08,
+      "loss": 0.2304,
+      "step": 15101
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.3049156344853685,
+      "learning_rate": 6.607493384174768e-08,
+      "loss": 0.2589,
+      "step": 15102
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7234941969486317,
+      "learning_rate": 6.59100073708785e-08,
+      "loss": 0.2327,
+      "step": 15103
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9464370795901975,
+      "learning_rate": 6.574528562416693e-08,
+      "loss": 0.2623,
+      "step": 15104
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2161176030459635,
+      "learning_rate": 6.558076860844752e-08,
+      "loss": 0.2487,
+      "step": 15105
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4946278049032118,
+      "learning_rate": 6.54164563305465e-08,
+      "loss": 0.2314,
+      "step": 15106
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9899022284806125,
+      "learning_rate": 6.525234879728171e-08,
+      "loss": 0.2554,
+      "step": 15107
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.34642547748058,
+      "learning_rate": 6.508844601546216e-08,
+      "loss": 0.2299,
+      "step": 15108
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7554524323469507,
+      "learning_rate": 6.492474799188797e-08,
+      "loss": 0.2528,
+      "step": 15109
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8134892649807535,
+      "learning_rate": 6.476125473335149e-08,
+      "loss": 0.2243,
+      "step": 15110
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.700990344186364,
+      "learning_rate": 6.459796624663728e-08,
+      "loss": 0.2436,
+      "step": 15111
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1122002978515972,
+      "learning_rate": 6.443488253851937e-08,
+      "loss": 0.2323,
+      "step": 15112
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8460731893107822,
+      "learning_rate": 6.427200361576458e-08,
+      "loss": 0.2474,
+      "step": 15113
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.770180126530419,
+      "learning_rate": 6.410932948513138e-08,
+      "loss": 0.2674,
+      "step": 15114
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.761875834139763,
+      "learning_rate": 6.394686015336882e-08,
+      "loss": 0.2701,
+      "step": 15115
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8154205062658497,
+      "learning_rate": 6.378459562721928e-08,
+      "loss": 0.2609,
+      "step": 15116
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0372390233772855,
+      "learning_rate": 6.362253591341405e-08,
+      "loss": 0.2741,
+      "step": 15117
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.583921479661432,
+      "learning_rate": 6.346068101867775e-08,
+      "loss": 0.2517,
+      "step": 15118
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5572213060348525,
+      "learning_rate": 6.329903094972612e-08,
+      "loss": 0.2502,
+      "step": 15119
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6531779228784358,
+      "learning_rate": 6.313758571326656e-08,
+      "loss": 0.2426,
+      "step": 15120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9409091805365677,
+      "learning_rate": 6.29763453159965e-08,
+      "loss": 0.2331,
+      "step": 15121
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0001766903916027,
+      "learning_rate": 6.281530976460781e-08,
+      "loss": 0.2695,
+      "step": 15122
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1533895809246753,
+      "learning_rate": 6.265447906578126e-08,
+      "loss": 0.282,
+      "step": 15123
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6730403870285562,
+      "learning_rate": 6.24938532261904e-08,
+      "loss": 0.2532,
+      "step": 15124
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7155203670651034,
+      "learning_rate": 6.233343225249933e-08,
+      "loss": 0.2575,
+      "step": 15125
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.999608808053576,
+      "learning_rate": 6.217321615136385e-08,
+      "loss": 0.2409,
+      "step": 15126
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5850726638459498,
+      "learning_rate": 6.201320492943252e-08,
+      "loss": 0.4774,
+      "step": 15127
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6008261510209832,
+      "learning_rate": 6.18533985933445e-08,
+      "loss": 0.465,
+      "step": 15128
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5886253994562793,
+      "learning_rate": 6.169379714973e-08,
+      "loss": 0.2581,
+      "step": 15129
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8513448530907524,
+      "learning_rate": 6.153440060521099e-08,
+      "loss": 0.2394,
+      "step": 15130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9469107701208122,
+      "learning_rate": 6.137520896640159e-08,
+      "loss": 0.2506,
+      "step": 15131
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7405142829403057,
+      "learning_rate": 6.121622223990708e-08,
+      "loss": 0.2295,
+      "step": 15132
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4737681393194144,
+      "learning_rate": 6.105744043232331e-08,
+      "loss": 0.2386,
+      "step": 15133
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.862263669110029,
+      "learning_rate": 6.089886355023889e-08,
+      "loss": 0.2727,
+      "step": 15134
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8948451670601327,
+      "learning_rate": 6.074049160023355e-08,
+      "loss": 0.2458,
+      "step": 15135
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8907923683215584,
+      "learning_rate": 6.05823245888787e-08,
+      "loss": 0.2381,
+      "step": 15136
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.933079028711817,
+      "learning_rate": 6.042436252273631e-08,
+      "loss": 0.2426,
+      "step": 15137
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6434428434070673,
+      "learning_rate": 6.026660540836005e-08,
+      "loss": 0.4853,
+      "step": 15138
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6234215158027452,
+      "learning_rate": 6.010905325229744e-08,
+      "loss": 0.4676,
+      "step": 15139
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0697498060359614,
+      "learning_rate": 5.99517060610838e-08,
+      "loss": 0.2556,
+      "step": 15140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6493974884538214,
+      "learning_rate": 5.979456384124893e-08,
+      "loss": 0.269,
+      "step": 15141
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8341882391238067,
+      "learning_rate": 5.963762659931205e-08,
+      "loss": 0.2673,
+      "step": 15142
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.287705961590037,
+      "learning_rate": 5.948089434178517e-08,
+      "loss": 0.249,
+      "step": 15143
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.897937430213338,
+      "learning_rate": 5.932436707517142e-08,
+      "loss": 0.2577,
+      "step": 15144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8812675656392879,
+      "learning_rate": 5.91680448059645e-08,
+      "loss": 0.247,
+      "step": 15145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3695730808633724,
+      "learning_rate": 5.901192754065199e-08,
+      "loss": 0.2478,
+      "step": 15146
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3134578969140613,
+      "learning_rate": 5.8856015285710946e-08,
+      "loss": 0.2592,
+      "step": 15147
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.519528924904941,
+      "learning_rate": 5.870030804761007e-08,
+      "loss": 0.2606,
+      "step": 15148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8495245079254352,
+      "learning_rate": 5.8544805832810304e-08,
+      "loss": 0.2696,
+      "step": 15149
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6019247537774153,
+      "learning_rate": 5.838950864776316e-08,
+      "loss": 0.27,
+      "step": 15150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1327114632354314,
+      "learning_rate": 5.8234416498912925e-08,
+      "loss": 0.2542,
+      "step": 15151
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.571365519697078,
+      "learning_rate": 5.8079529392694453e-08,
+      "loss": 0.2636,
+      "step": 15152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.16703759712683,
+      "learning_rate": 5.792484733553372e-08,
+      "loss": 0.2447,
+      "step": 15153
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.010961220939667,
+      "learning_rate": 5.7770370333849466e-08,
+      "loss": 0.2569,
+      "step": 15154
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.640572411925826,
+      "learning_rate": 5.7616098394051026e-08,
+      "loss": 0.262,
+      "step": 15155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5472826722278206,
+      "learning_rate": 5.746203152253937e-08,
+      "loss": 0.2489,
+      "step": 15156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4749328699426574,
+      "learning_rate": 5.730816972570663e-08,
+      "loss": 0.249,
+      "step": 15157
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.423856974609935,
+      "learning_rate": 5.715451300993713e-08,
+      "loss": 0.2591,
+      "step": 15158
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.009584840737862,
+      "learning_rate": 5.700106138160688e-08,
+      "loss": 0.24,
+      "step": 15159
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3581814776620722,
+      "learning_rate": 5.684781484708135e-08,
+      "loss": 0.2549,
+      "step": 15160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9425870284310038,
+      "learning_rate": 5.6694773412721e-08,
+      "loss": 0.2465,
+      "step": 15161
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2810184054673877,
+      "learning_rate": 5.65419370848741e-08,
+      "loss": 0.2414,
+      "step": 15162
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.703815186506678,
+      "learning_rate": 5.638930586988334e-08,
+      "loss": 0.2448,
+      "step": 15163
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6213095471075559,
+      "learning_rate": 5.623687977408088e-08,
+      "loss": 0.4262,
+      "step": 15164
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.682341354058051,
+      "learning_rate": 5.6084658803791656e-08,
+      "loss": 0.2496,
+      "step": 15165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.215689411723757,
+      "learning_rate": 5.5932642965331184e-08,
+      "loss": 0.2569,
+      "step": 15166
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6587726584601461,
+      "learning_rate": 5.578083226500663e-08,
+      "loss": 0.4948,
+      "step": 15167
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.168351878895283,
+      "learning_rate": 5.562922670911741e-08,
+      "loss": 0.2371,
+      "step": 15168
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8955329916455563,
+      "learning_rate": 5.547782630395404e-08,
+      "loss": 0.2467,
+      "step": 15169
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.588851101880383,
+      "learning_rate": 5.532663105579761e-08,
+      "loss": 0.2449,
+      "step": 15170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6489272704683102,
+      "learning_rate": 5.517564097092254e-08,
+      "loss": 0.2478,
+      "step": 15171
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.90788333060237,
+      "learning_rate": 5.5024856055593266e-08,
+      "loss": 0.2429,
+      "step": 15172
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.28757202297265,
+      "learning_rate": 5.487427631606534e-08,
+      "loss": 0.2411,
+      "step": 15173
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8575776825344417,
+      "learning_rate": 5.472390175858766e-08,
+      "loss": 0.2436,
+      "step": 15174
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2881137453485976,
+      "learning_rate": 5.457373238939856e-08,
+      "loss": 0.2565,
+      "step": 15175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8198144124812337,
+      "learning_rate": 5.4423768214730274e-08,
+      "loss": 0.2694,
+      "step": 15176
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8031232580146708,
+      "learning_rate": 5.427400924080395e-08,
+      "loss": 0.267,
+      "step": 15177
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6278084081181734,
+      "learning_rate": 5.4124455473832935e-08,
+      "loss": 0.2529,
+      "step": 15178
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.021537135167698,
+      "learning_rate": 5.397510692002339e-08,
+      "loss": 0.2541,
+      "step": 15179
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.402138276385479,
+      "learning_rate": 5.3825963585572015e-08,
+      "loss": 0.2371,
+      "step": 15180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8101820977372205,
+      "learning_rate": 5.367702547666664e-08,
+      "loss": 0.2647,
+      "step": 15181
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8894330996454136,
+      "learning_rate": 5.3528292599486776e-08,
+      "loss": 0.2579,
+      "step": 15182
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8951110145995576,
+      "learning_rate": 5.337976496020469e-08,
+      "loss": 0.2417,
+      "step": 15183
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9426591456784061,
+      "learning_rate": 5.323144256498214e-08,
+      "loss": 0.2446,
+      "step": 15184
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4369025705594134,
+      "learning_rate": 5.308332541997363e-08,
+      "loss": 0.2347,
+      "step": 15185
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8318883764582616,
+      "learning_rate": 5.293541353132425e-08,
+      "loss": 0.2488,
+      "step": 15186
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6685713685598524,
+      "learning_rate": 5.2787706905171874e-08,
+      "loss": 0.2597,
+      "step": 15187
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5352402352444785,
+      "learning_rate": 5.264020554764438e-08,
+      "loss": 0.2425,
+      "step": 15188
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7706277661782548,
+      "learning_rate": 5.249290946486241e-08,
+      "loss": 0.2435,
+      "step": 15189
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.736725762980753,
+      "learning_rate": 5.23458186629372e-08,
+      "loss": 0.2516,
+      "step": 15190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.764797371212197,
+      "learning_rate": 5.2198933147972194e-08,
+      "loss": 0.2485,
+      "step": 15191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0022105372767873,
+      "learning_rate": 5.205225292606198e-08,
+      "loss": 0.2567,
+      "step": 15192
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.594500950542812,
+      "learning_rate": 5.1905778003292216e-08,
+      "loss": 0.2422,
+      "step": 15193
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9204515577690455,
+      "learning_rate": 5.175950838574029e-08,
+      "loss": 0.2449,
+      "step": 15194
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5331104549298378,
+      "learning_rate": 5.1613444079475774e-08,
+      "loss": 0.2333,
+      "step": 15195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.47501339219459,
+      "learning_rate": 5.1467585090558825e-08,
+      "loss": 0.2581,
+      "step": 15196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.684539632185198,
+      "learning_rate": 5.132193142504072e-08,
+      "loss": 0.247,
+      "step": 15197
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.027822068369087,
+      "learning_rate": 5.1176483088966054e-08,
+      "loss": 0.265,
+      "step": 15198
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6415789247623427,
+      "learning_rate": 5.103124008836891e-08,
+      "loss": 0.2477,
+      "step": 15199
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.164109607072536,
+      "learning_rate": 5.088620242927666e-08,
+      "loss": 0.2474,
+      "step": 15200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.494603751571222,
+      "learning_rate": 5.0741370117705634e-08,
+      "loss": 0.2491,
+      "step": 15201
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0621501866695215,
+      "learning_rate": 5.0596743159666565e-08,
+      "loss": 0.2587,
+      "step": 15202
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7005264148884391,
+      "learning_rate": 5.0452321561159646e-08,
+      "loss": 0.2458,
+      "step": 15203
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2018407435742287,
+      "learning_rate": 5.030810532817732e-08,
+      "loss": 0.256,
+      "step": 15204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7323948238513522,
+      "learning_rate": 5.016409446670312e-08,
+      "loss": 0.2473,
+      "step": 15205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.363654318878327,
+      "learning_rate": 5.002028898271227e-08,
+      "loss": 0.2645,
+      "step": 15206
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.236699145240007,
+      "learning_rate": 4.987668888217223e-08,
+      "loss": 0.2473,
+      "step": 15207
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.194310880329028,
+      "learning_rate": 4.9733294171041e-08,
+      "loss": 0.2478,
+      "step": 15208
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8973970223237857,
+      "learning_rate": 4.959010485526772e-08,
+      "loss": 0.2526,
+      "step": 15209
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6451938731953772,
+      "learning_rate": 4.9447120940793735e-08,
+      "loss": 0.466,
+      "step": 15210
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.048792251373977,
+      "learning_rate": 4.9304342433552086e-08,
+      "loss": 0.2539,
+      "step": 15211
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.019522967877576,
+      "learning_rate": 4.9161769339466926e-08,
+      "loss": 0.236,
+      "step": 15212
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.906157398199811,
+      "learning_rate": 4.9019401664453516e-08,
+      "loss": 0.2635,
+      "step": 15213
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.303859144721179,
+      "learning_rate": 4.887723941441824e-08,
+      "loss": 0.2594,
+      "step": 15214
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3740232477896925,
+      "learning_rate": 4.8735282595261393e-08,
+      "loss": 0.2337,
+      "step": 15215
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.811416062202169,
+      "learning_rate": 4.8593531212872136e-08,
+      "loss": 0.268,
+      "step": 15216
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.247098452250411,
+      "learning_rate": 4.845198527313188e-08,
+      "loss": 0.2526,
+      "step": 15217
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3133988129658465,
+      "learning_rate": 4.831064478191316e-08,
+      "loss": 0.2411,
+      "step": 15218
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.008405293056618,
+      "learning_rate": 4.8169509745081275e-08,
+      "loss": 0.2535,
+      "step": 15219
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.710309663320064,
+      "learning_rate": 4.80285801684921e-08,
+      "loss": 0.2553,
+      "step": 15220
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9966221546370129,
+      "learning_rate": 4.7887856057993175e-08,
+      "loss": 0.2604,
+      "step": 15221
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1092487647744567,
+      "learning_rate": 4.774733741942206e-08,
+      "loss": 0.2524,
+      "step": 15222
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5856665390191502,
+      "learning_rate": 4.7607024258610744e-08,
+      "loss": 0.4571,
+      "step": 15223
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4573738058872443,
+      "learning_rate": 4.746691658138014e-08,
+      "loss": 0.2579,
+      "step": 15224
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.677805411856043,
+      "learning_rate": 4.732701439354448e-08,
+      "loss": 0.2383,
+      "step": 15225
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5062308680197356,
+      "learning_rate": 4.7187317700906896e-08,
+      "loss": 0.257,
+      "step": 15226
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9946536866791273,
+      "learning_rate": 4.704782650926554e-08,
+      "loss": 0.249,
+      "step": 15227
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8148324274248189,
+      "learning_rate": 4.6908540824406876e-08,
+      "loss": 0.2462,
+      "step": 15228
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0039248350527212,
+      "learning_rate": 4.676946065211074e-08,
+      "loss": 0.236,
+      "step": 15229
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.078730454449582,
+      "learning_rate": 4.6630585998147515e-08,
+      "loss": 0.2339,
+      "step": 15230
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.510118986408492,
+      "learning_rate": 4.649191686827925e-08,
+      "loss": 0.2611,
+      "step": 15231
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.525831984473719,
+      "learning_rate": 4.635345326826024e-08,
+      "loss": 0.2518,
+      "step": 15232
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.9945673532516714,
+      "learning_rate": 4.621519520383477e-08,
+      "loss": 0.2547,
+      "step": 15233
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5247985117279472,
+      "learning_rate": 4.607714268073937e-08,
+      "loss": 0.2339,
+      "step": 15234
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0937951549587672,
+      "learning_rate": 4.5939295704703344e-08,
+      "loss": 0.2394,
+      "step": 15235
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5020797471347267,
+      "learning_rate": 4.58016542814449e-08,
+      "loss": 0.2435,
+      "step": 15236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6930752115180783,
+      "learning_rate": 4.5664218416675565e-08,
+      "loss": 0.2256,
+      "step": 15237
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.582958312268396,
+      "learning_rate": 4.5526988116097457e-08,
+      "loss": 0.2464,
+      "step": 15238
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2690202315888017,
+      "learning_rate": 4.5389963385405467e-08,
+      "loss": 0.2457,
+      "step": 15239
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.265770290941597,
+      "learning_rate": 4.525314423028393e-08,
+      "loss": 0.2421,
+      "step": 15240
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7212025690661907,
+      "learning_rate": 4.5116530656409415e-08,
+      "loss": 0.2394,
+      "step": 15241
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6488513883421396,
+      "learning_rate": 4.498012266945129e-08,
+      "loss": 0.2444,
+      "step": 15242
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.300497615099552,
+      "learning_rate": 4.4843920275068896e-08,
+      "loss": 0.2392,
+      "step": 15243
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5748133550149743,
+      "learning_rate": 4.470792347891384e-08,
+      "loss": 0.2573,
+      "step": 15244
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.558385985174884,
+      "learning_rate": 4.457213228662882e-08,
+      "loss": 0.2549,
+      "step": 15245
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.0750824986953065,
+      "learning_rate": 4.443654670384767e-08,
+      "loss": 0.2583,
+      "step": 15246
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5004155359426379,
+      "learning_rate": 4.430116673619589e-08,
+      "loss": 0.2519,
+      "step": 15247
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.177245468921564,
+      "learning_rate": 4.4165992389291754e-08,
+      "loss": 0.2343,
+      "step": 15248
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.55994081459802,
+      "learning_rate": 4.4031023668742454e-08,
+      "loss": 0.2513,
+      "step": 15249
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.236676502308055,
+      "learning_rate": 4.3896260580149067e-08,
+      "loss": 0.2675,
+      "step": 15250
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4437197521405114,
+      "learning_rate": 4.3761703129102664e-08,
+      "loss": 0.2501,
+      "step": 15251
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.823777221525274,
+      "learning_rate": 4.362735132118656e-08,
+      "loss": 0.2368,
+      "step": 15252
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5703363684189728,
+      "learning_rate": 4.349320516197575e-08,
+      "loss": 0.2443,
+      "step": 15253
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0679800829988095,
+      "learning_rate": 4.335926465703466e-08,
+      "loss": 0.2592,
+      "step": 15254
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.356343132527862,
+      "learning_rate": 4.3225529811922186e-08,
+      "loss": 0.2567,
+      "step": 15255
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9619603710102285,
+      "learning_rate": 4.309200063218666e-08,
+      "loss": 0.2355,
+      "step": 15256
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7831380507861792,
+      "learning_rate": 4.29586771233681e-08,
+      "loss": 0.2513,
+      "step": 15257
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.860489360122903,
+      "learning_rate": 4.28255592909993e-08,
+      "loss": 0.248,
+      "step": 15258
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.005892768479395,
+      "learning_rate": 4.2692647140602507e-08,
+      "loss": 0.2482,
+      "step": 15259
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.023296402407743,
+      "learning_rate": 4.255994067769331e-08,
+      "loss": 0.2537,
+      "step": 15260
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.220633330102731,
+      "learning_rate": 4.2427439907777866e-08,
+      "loss": 0.2471,
+      "step": 15261
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.869521845471784,
+      "learning_rate": 4.229514483635288e-08,
+      "loss": 0.2561,
+      "step": 15262
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8901609326295237,
+      "learning_rate": 4.216305546890842e-08,
+      "loss": 0.2535,
+      "step": 15263
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.390020115903151,
+      "learning_rate": 4.2031171810925083e-08,
+      "loss": 0.2356,
+      "step": 15264
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5936812413296529,
+      "learning_rate": 4.189949386787462e-08,
+      "loss": 0.4859,
+      "step": 15265
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.337940223191999,
+      "learning_rate": 4.176802164522042e-08,
+      "loss": 0.2596,
+      "step": 15266
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4417081749550698,
+      "learning_rate": 4.163675514841814e-08,
+      "loss": 0.2438,
+      "step": 15267
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.660431688649492,
+      "learning_rate": 4.1505694382913965e-08,
+      "loss": 0.2727,
+      "step": 15268
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.53405796262389,
+      "learning_rate": 4.1374839354146325e-08,
+      "loss": 0.2781,
+      "step": 15269
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.377763272319495,
+      "learning_rate": 4.12441900675431e-08,
+      "loss": 0.2716,
+      "step": 15270
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9765658395738166,
+      "learning_rate": 4.111374652852662e-08,
+      "loss": 0.2513,
+      "step": 15271
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5273943077283598,
+      "learning_rate": 4.098350874250867e-08,
+      "loss": 0.2389,
+      "step": 15272
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.407040531772963,
+      "learning_rate": 4.085347671489382e-08,
+      "loss": 0.2365,
+      "step": 15273
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3047366264035998,
+      "learning_rate": 4.072365045107551e-08,
+      "loss": 0.2713,
+      "step": 15274
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8629879975745975,
+      "learning_rate": 4.059402995644224e-08,
+      "loss": 0.2543,
+      "step": 15275
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.072581575390451,
+      "learning_rate": 4.046461523637191e-08,
+      "loss": 0.254,
+      "step": 15276
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3297485353315848,
+      "learning_rate": 4.033540629623356e-08,
+      "loss": 0.2598,
+      "step": 15277
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.317567473137255,
+      "learning_rate": 4.020640314138846e-08,
+      "loss": 0.242,
+      "step": 15278
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.165336164376101,
+      "learning_rate": 4.0077605777189e-08,
+      "loss": 0.2333,
+      "step": 15279
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.028975643073122,
+      "learning_rate": 3.9949014208979784e-08,
+      "loss": 0.2373,
+      "step": 15280
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.164789527296874,
+      "learning_rate": 3.9820628442096e-08,
+      "loss": 0.2557,
+      "step": 15281
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1806254702067998,
+      "learning_rate": 3.969244848186449e-08,
+      "loss": 0.267,
+      "step": 15282
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0176171263476332,
+      "learning_rate": 3.9564474333603776e-08,
+      "loss": 0.2469,
+      "step": 15283
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6824030486712855,
+      "learning_rate": 3.943670600262406e-08,
+      "loss": 0.2523,
+      "step": 15284
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7226523653560994,
+      "learning_rate": 3.9309143494226097e-08,
+      "loss": 0.2443,
+      "step": 15285
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.582261031254253,
+      "learning_rate": 3.918178681370288e-08,
+      "loss": 0.2482,
+      "step": 15286
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6870161425798145,
+      "learning_rate": 3.905463596633852e-08,
+      "loss": 0.2538,
+      "step": 15287
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.18981167878124,
+      "learning_rate": 3.8927690957409893e-08,
+      "loss": 0.2541,
+      "step": 15288
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.774093789964343,
+      "learning_rate": 3.880095179218224e-08,
+      "loss": 0.2418,
+      "step": 15289
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8224984339852053,
+      "learning_rate": 3.86744184759158e-08,
+      "loss": 0.2511,
+      "step": 15290
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8674508338644273,
+      "learning_rate": 3.8548091013859704e-08,
+      "loss": 0.2294,
+      "step": 15291
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.49393034334191,
+      "learning_rate": 3.842196941125587e-08,
+      "loss": 0.2419,
+      "step": 15292
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.091091727312069,
+      "learning_rate": 3.8296053673337884e-08,
+      "loss": 0.2439,
+      "step": 15293
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.896800212058695,
+      "learning_rate": 3.8170343805328245e-08,
+      "loss": 0.2341,
+      "step": 15294
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.168275542320641,
+      "learning_rate": 3.8044839812445554e-08,
+      "loss": 0.2693,
+      "step": 15295
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7257924184085172,
+      "learning_rate": 3.791954169989509e-08,
+      "loss": 0.2476,
+      "step": 15296
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.648661235010644,
+      "learning_rate": 3.7794449472877135e-08,
+      "loss": 0.2383,
+      "step": 15297
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5738737607136402,
+      "learning_rate": 3.7669563136580875e-08,
+      "loss": 0.2597,
+      "step": 15298
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.802143903144524,
+      "learning_rate": 3.754488269618883e-08,
+      "loss": 0.2622,
+      "step": 15299
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3231835161972927,
+      "learning_rate": 3.742040815687353e-08,
+      "loss": 0.2636,
+      "step": 15300
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4970093285623225,
+      "learning_rate": 3.729613952380029e-08,
+      "loss": 0.2484,
+      "step": 15301
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.997487054648361,
+      "learning_rate": 3.7172076802124426e-08,
+      "loss": 0.2698,
+      "step": 15302
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5618913184894834,
+      "learning_rate": 3.704821999699515e-08,
+      "loss": 0.2391,
+      "step": 15303
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5674667541409975,
+      "learning_rate": 3.692456911354947e-08,
+      "loss": 0.246,
+      "step": 15304
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9085429118159454,
+      "learning_rate": 3.6801124156919385e-08,
+      "loss": 0.2421,
+      "step": 15305
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8696135249059707,
+      "learning_rate": 3.667788513222581e-08,
+      "loss": 0.2791,
+      "step": 15306
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8700458518702057,
+      "learning_rate": 3.655485204458353e-08,
+      "loss": 0.2482,
+      "step": 15307
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9142051065697887,
+      "learning_rate": 3.6432024899095694e-08,
+      "loss": 0.2413,
+      "step": 15308
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4659636876350683,
+      "learning_rate": 3.630940370085934e-08,
+      "loss": 0.2555,
+      "step": 15309
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.777845902709994,
+      "learning_rate": 3.61869884549626e-08,
+      "loss": 0.2485,
+      "step": 15310
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8942817230681803,
+      "learning_rate": 3.606477916648477e-08,
+      "loss": 0.2484,
+      "step": 15311
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.51286077522307,
+      "learning_rate": 3.594277584049566e-08,
+      "loss": 0.2529,
+      "step": 15312
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.048395491240227,
+      "learning_rate": 3.5820978482058454e-08,
+      "loss": 0.2435,
+      "step": 15313
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.345472689640056,
+      "learning_rate": 3.569938709622578e-08,
+      "loss": 0.2331,
+      "step": 15314
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6320413597453367,
+      "learning_rate": 3.557800168804359e-08,
+      "loss": 0.2619,
+      "step": 15315
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.702694496363717,
+      "learning_rate": 3.545682226254732e-08,
+      "loss": 0.2455,
+      "step": 15316
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.936633360459089,
+      "learning_rate": 3.5335848824765706e-08,
+      "loss": 0.235,
+      "step": 15317
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.488881124586551,
+      "learning_rate": 3.521508137971807e-08,
+      "loss": 0.25,
+      "step": 15318
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.250987039169261,
+      "learning_rate": 3.5094519932415417e-08,
+      "loss": 0.2662,
+      "step": 15319
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0940118980209528,
+      "learning_rate": 3.4974164487859285e-08,
+      "loss": 0.2497,
+      "step": 15320
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0866958738940644,
+      "learning_rate": 3.485401505104458e-08,
+      "loss": 0.276,
+      "step": 15321
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.139820748030344,
+      "learning_rate": 3.473407162695508e-08,
+      "loss": 0.2411,
+      "step": 15322
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1695480514946923,
+      "learning_rate": 3.4614334220569036e-08,
+      "loss": 0.2628,
+      "step": 15323
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8046521207051538,
+      "learning_rate": 3.449480283685302e-08,
+      "loss": 0.2346,
+      "step": 15324
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5955712789364611,
+      "learning_rate": 3.4375477480768084e-08,
+      "loss": 0.2492,
+      "step": 15325
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2857057519472268,
+      "learning_rate": 3.4256358157264135e-08,
+      "loss": 0.2481,
+      "step": 15326
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2482746178734794,
+      "learning_rate": 3.41374448712839e-08,
+      "loss": 0.2562,
+      "step": 15327
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.036424485527265,
+      "learning_rate": 3.4018737627761754e-08,
+      "loss": 0.2627,
+      "step": 15328
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8016908421713853,
+      "learning_rate": 3.390023643162266e-08,
+      "loss": 0.2429,
+      "step": 15329
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7336543605057326,
+      "learning_rate": 3.3781941287783224e-08,
+      "loss": 0.2595,
+      "step": 15330
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2087324847340573,
+      "learning_rate": 3.3663852201152314e-08,
+      "loss": 0.2451,
+      "step": 15331
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0713500706815813,
+      "learning_rate": 3.354596917662989e-08,
+      "loss": 0.2574,
+      "step": 15332
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0011230273975733,
+      "learning_rate": 3.342829221910593e-08,
+      "loss": 0.2434,
+      "step": 15333
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4172602084071526,
+      "learning_rate": 3.331082133346375e-08,
+      "loss": 0.2604,
+      "step": 15334
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5952833301376854,
+      "learning_rate": 3.3193556524577784e-08,
+      "loss": 0.2522,
+      "step": 15335
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3358224769108644,
+      "learning_rate": 3.3076497797313034e-08,
+      "loss": 0.2514,
+      "step": 15336
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7516779276850174,
+      "learning_rate": 3.295964515652672e-08,
+      "loss": 0.2582,
+      "step": 15337
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5643769409130932,
+      "learning_rate": 3.284299860706719e-08,
+      "loss": 0.2499,
+      "step": 15338
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.434310465857878,
+      "learning_rate": 3.27265581537739e-08,
+      "loss": 0.2326,
+      "step": 15339
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.992712023337565,
+      "learning_rate": 3.2610323801479104e-08,
+      "loss": 0.2592,
+      "step": 15340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.135565201470293,
+      "learning_rate": 3.249429555500505e-08,
+      "loss": 0.2453,
+      "step": 15341
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8042016918879589,
+      "learning_rate": 3.2378473419165665e-08,
+      "loss": 0.2573,
+      "step": 15342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7511426969549913,
+      "learning_rate": 3.226285739876711e-08,
+      "loss": 0.2443,
+      "step": 15343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8925565070592203,
+      "learning_rate": 3.21474474986061e-08,
+      "loss": 0.2407,
+      "step": 15344
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.81522924280968,
+      "learning_rate": 3.203224372347158e-08,
+      "loss": 0.2434,
+      "step": 15345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4283249443979504,
+      "learning_rate": 3.191724607814306e-08,
+      "loss": 0.2528,
+      "step": 15346
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6450169322172354,
+      "learning_rate": 3.1802454567392837e-08,
+      "loss": 0.2834,
+      "step": 15347
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.363494925275413,
+      "learning_rate": 3.1687869195983215e-08,
+      "loss": 0.2407,
+      "step": 15348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6525994502119377,
+      "learning_rate": 3.157348996866816e-08,
+      "loss": 0.2457,
+      "step": 15349
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.843293513615338,
+      "learning_rate": 3.145931689019388e-08,
+      "loss": 0.2526,
+      "step": 15350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4776435375146455,
+      "learning_rate": 3.134534996529826e-08,
+      "loss": 0.2527,
+      "step": 15351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.082787644659762,
+      "learning_rate": 3.123158919870917e-08,
+      "loss": 0.2484,
+      "step": 15352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3803864607492664,
+      "learning_rate": 3.111803459514673e-08,
+      "loss": 0.2596,
+      "step": 15353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4063893010098014,
+      "learning_rate": 3.1004686159322726e-08,
+      "loss": 0.237,
+      "step": 15354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6274120290660266,
+      "learning_rate": 3.089154389594062e-08,
+      "loss": 0.2482,
+      "step": 15355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7085256371492532,
+      "learning_rate": 3.0778607809694436e-08,
+      "loss": 0.2512,
+      "step": 15356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8195484725010698,
+      "learning_rate": 3.066587790526987e-08,
+      "loss": 0.241,
+      "step": 15357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7289982967678537,
+      "learning_rate": 3.055335418734429e-08,
+      "loss": 0.2586,
+      "step": 15358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.061965627953282,
+      "learning_rate": 3.0441036660587284e-08,
+      "loss": 0.2397,
+      "step": 15359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.354734299441215,
+      "learning_rate": 3.032892532965848e-08,
+      "loss": 0.2589,
+      "step": 15360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.004312778668439,
+      "learning_rate": 3.021702019920969e-08,
+      "loss": 0.2737,
+      "step": 15361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0144702909820316,
+      "learning_rate": 3.010532127388388e-08,
+      "loss": 0.2508,
+      "step": 15362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4394008718255376,
+      "learning_rate": 2.999382855831623e-08,
+      "loss": 0.2509,
+      "step": 15363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8584716934521521,
+      "learning_rate": 2.9882542057131926e-08,
+      "loss": 0.2515,
+      "step": 15364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0465589848756323,
+      "learning_rate": 2.9771461774949494e-08,
+      "loss": 0.2526,
+      "step": 15365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3813740686620206,
+      "learning_rate": 2.9660587716376366e-08,
+      "loss": 0.252,
+      "step": 15366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.745844013136599,
+      "learning_rate": 2.954991988601441e-08,
+      "loss": 0.2466,
+      "step": 15367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.638236706039188,
+      "learning_rate": 2.9439458288454404e-08,
+      "loss": 0.2594,
+      "step": 15368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 25.038448563217997,
+      "learning_rate": 2.9329202928280452e-08,
+      "loss": 0.2502,
+      "step": 15369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5932460614647606,
+      "learning_rate": 2.9219153810066124e-08,
+      "loss": 0.2472,
+      "step": 15370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.268113488709931,
+      "learning_rate": 2.9109310938378875e-08,
+      "loss": 0.2559,
+      "step": 15371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.813926907059919,
+      "learning_rate": 2.899967431777506e-08,
+      "loss": 0.2567,
+      "step": 15372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.452551507658095,
+      "learning_rate": 2.889024395280493e-08,
+      "loss": 0.2387,
+      "step": 15373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3181105248950122,
+      "learning_rate": 2.8781019848007628e-08,
+      "loss": 0.2358,
+      "step": 15374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6590910298493007,
+      "learning_rate": 2.8672002007915646e-08,
+      "loss": 0.4294,
+      "step": 15375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.022896727496877,
+      "learning_rate": 2.856319043705258e-08,
+      "loss": 0.2814,
+      "step": 15376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3888693843039,
+      "learning_rate": 2.8454585139933154e-08,
+      "loss": 0.2563,
+      "step": 15377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4575689950990918,
+      "learning_rate": 2.8346186121063214e-08,
+      "loss": 0.2508,
+      "step": 15378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.189791923584707,
+      "learning_rate": 2.8237993384940822e-08,
+      "loss": 0.2488,
+      "step": 15379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.899432846434351,
+      "learning_rate": 2.8130006936055167e-08,
+      "loss": 0.2619,
+      "step": 15380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.512986310092415,
+      "learning_rate": 2.8022226778885997e-08,
+      "loss": 0.2454,
+      "step": 15381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.269107239754572,
+      "learning_rate": 2.7914652917906405e-08,
+      "loss": 0.2504,
+      "step": 15382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9599260607219626,
+      "learning_rate": 2.7807285357578374e-08,
+      "loss": 0.2604,
+      "step": 15383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6436723989716344,
+      "learning_rate": 2.7700124102358896e-08,
+      "loss": 0.2492,
+      "step": 15384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8879250478746306,
+      "learning_rate": 2.759316915669219e-08,
+      "loss": 0.2588,
+      "step": 15385
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.293526808767585,
+      "learning_rate": 2.7486420525017487e-08,
+      "loss": 0.2644,
+      "step": 15386
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6808194686052513,
+      "learning_rate": 2.737987821176291e-08,
+      "loss": 0.2605,
+      "step": 15387
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4008204885221236,
+      "learning_rate": 2.7273542221349925e-08,
+      "loss": 0.2593,
+      "step": 15388
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9107654947024302,
+      "learning_rate": 2.7167412558189997e-08,
+      "loss": 0.2473,
+      "step": 15389
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9418409289716292,
+      "learning_rate": 2.7061489226686832e-08,
+      "loss": 0.2492,
+      "step": 15390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.665281881741754,
+      "learning_rate": 2.6955772231235243e-08,
+      "loss": 0.2605,
+      "step": 15391
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6116081555458277,
+      "learning_rate": 2.685026157622228e-08,
+      "loss": 0.231,
+      "step": 15392
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.538967965499641,
+      "learning_rate": 2.6744957266024996e-08,
+      "loss": 0.2465,
+      "step": 15393
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.183306124935992,
+      "learning_rate": 2.663985930501267e-08,
+      "loss": 0.2425,
+      "step": 15394
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2725031333267514,
+      "learning_rate": 2.653496769754682e-08,
+      "loss": 0.2721,
+      "step": 15395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0530465258193353,
+      "learning_rate": 2.6430282447978404e-08,
+      "loss": 0.2602,
+      "step": 15396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6272586979980177,
+      "learning_rate": 2.6325803560652286e-08,
+      "loss": 0.4753,
+      "step": 15397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4854977102106606,
+      "learning_rate": 2.6221531039902214e-08,
+      "loss": 0.2478,
+      "step": 15398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.603740926459719,
+      "learning_rate": 2.611746489005529e-08,
+      "loss": 0.2354,
+      "step": 15399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.184837014693623,
+      "learning_rate": 2.601360511542972e-08,
+      "loss": 0.2535,
+      "step": 15400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5317966287391347,
+      "learning_rate": 2.5909951720334837e-08,
+      "loss": 0.2505,
+      "step": 15401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6660374121280324,
+      "learning_rate": 2.580650470906998e-08,
+      "loss": 0.2465,
+      "step": 15402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8738114424906243,
+      "learning_rate": 2.570326408592949e-08,
+      "loss": 0.2437,
+      "step": 15403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.308996330877488,
+      "learning_rate": 2.56002298551955e-08,
+      "loss": 0.2437,
+      "step": 15404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.550698509333058,
+      "learning_rate": 2.549740202114348e-08,
+      "loss": 0.2572,
+      "step": 15405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6730352932360026,
+      "learning_rate": 2.5394780588039458e-08,
+      "loss": 0.2435,
+      "step": 15406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4473176371846725,
+      "learning_rate": 2.5292365560142252e-08,
+      "loss": 0.2696,
+      "step": 15407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.624639056178406,
+      "learning_rate": 2.5190156941700684e-08,
+      "loss": 0.2465,
+      "step": 15408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3593923238769758,
+      "learning_rate": 2.508815473695636e-08,
+      "loss": 0.25,
+      "step": 15409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7386045395773486,
+      "learning_rate": 2.4986358950140343e-08,
+      "loss": 0.2516,
+      "step": 15410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6639875402424946,
+      "learning_rate": 2.488476958547703e-08,
+      "loss": 0.2334,
+      "step": 15411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0738530249733325,
+      "learning_rate": 2.478338664718194e-08,
+      "loss": 0.2721,
+      "step": 15412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.110522461368836,
+      "learning_rate": 2.4682210139460593e-08,
+      "loss": 0.2526,
+      "step": 15413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3880459034330768,
+      "learning_rate": 2.4581240066511304e-08,
+      "loss": 0.2446,
+      "step": 15414
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.31725079189721,
+      "learning_rate": 2.4480476432524048e-08,
+      "loss": 0.2545,
+      "step": 15415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5664437641217384,
+      "learning_rate": 2.4379919241679373e-08,
+      "loss": 0.2427,
+      "step": 15416
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2831622809714314,
+      "learning_rate": 2.4279568498149497e-08,
+      "loss": 0.24,
+      "step": 15417
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.498901627489701,
+      "learning_rate": 2.4179424206098314e-08,
+      "loss": 0.2312,
+      "step": 15418
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7619439479261354,
+      "learning_rate": 2.4079486369680826e-08,
+      "loss": 0.2488,
+      "step": 15419
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.457464055928428,
+      "learning_rate": 2.3979754993043724e-08,
+      "loss": 0.2428,
+      "step": 15420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5065920712687704,
+      "learning_rate": 2.3880230080324806e-08,
+      "loss": 0.2288,
+      "step": 15421
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4139226223849604,
+      "learning_rate": 2.3780911635653547e-08,
+      "loss": 0.2548,
+      "step": 15422
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3718508502763997,
+      "learning_rate": 2.3681799663151096e-08,
+      "loss": 0.237,
+      "step": 15423
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7419795838586325,
+      "learning_rate": 2.358289416693027e-08,
+      "loss": 0.2354,
+      "step": 15424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1453250036548543,
+      "learning_rate": 2.3484195151093902e-08,
+      "loss": 0.2387,
+      "step": 15425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6081404995631711,
+      "learning_rate": 2.338570261973705e-08,
+      "loss": 0.4764,
+      "step": 15426
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 44.83175261527692,
+      "learning_rate": 2.328741657694755e-08,
+      "loss": 0.2558,
+      "step": 15427
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0605645560487424,
+      "learning_rate": 2.3189337026802705e-08,
+      "loss": 0.2572,
+      "step": 15428
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4817074203972622,
+      "learning_rate": 2.309146397337203e-08,
+      "loss": 0.2215,
+      "step": 15429
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8491849328290098,
+      "learning_rate": 2.2993797420716168e-08,
+      "loss": 0.2571,
+      "step": 15430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.38646516516461,
+      "learning_rate": 2.2896337372887988e-08,
+      "loss": 0.2439,
+      "step": 15431
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3661684537440792,
+      "learning_rate": 2.279908383393148e-08,
+      "loss": 0.2512,
+      "step": 15432
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6656297796028037,
+      "learning_rate": 2.270203680788119e-08,
+      "loss": 0.2546,
+      "step": 15433
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.933725777283793,
+      "learning_rate": 2.26051962987639e-08,
+      "loss": 0.2518,
+      "step": 15434
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.072465880182918,
+      "learning_rate": 2.2508562310598057e-08,
+      "loss": 0.2501,
+      "step": 15435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1916570925011354,
+      "learning_rate": 2.2412134847392687e-08,
+      "loss": 0.2322,
+      "step": 15436
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8154144355138124,
+      "learning_rate": 2.2315913913149578e-08,
+      "loss": 0.2652,
+      "step": 15437
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5564233721356238,
+      "learning_rate": 2.221989951185999e-08,
+      "loss": 0.2607,
+      "step": 15438
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6670904655104286,
+      "learning_rate": 2.2124091647509062e-08,
+      "loss": 0.2332,
+      "step": 15439
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.606807819661613,
+      "learning_rate": 2.202849032407084e-08,
+      "loss": 0.236,
+      "step": 15440
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.765907588009598,
+      "learning_rate": 2.193309554551215e-08,
+      "loss": 0.2437,
+      "step": 15441
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.44992618005198,
+      "learning_rate": 2.1837907315791495e-08,
+      "loss": 0.2725,
+      "step": 15442
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.149696000434322,
+      "learning_rate": 2.174292563885849e-08,
+      "loss": 0.2739,
+      "step": 15443
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8933918783675505,
+      "learning_rate": 2.1648150518653875e-08,
+      "loss": 0.2779,
+      "step": 15444
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5328495552060035,
+      "learning_rate": 2.1553581959110058e-08,
+      "loss": 0.2617,
+      "step": 15445
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.205367458624773,
+      "learning_rate": 2.145921996415057e-08,
+      "loss": 0.2679,
+      "step": 15446
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5079999903802914,
+      "learning_rate": 2.1365064537691162e-08,
+      "loss": 0.2599,
+      "step": 15447
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5755684981923537,
+      "learning_rate": 2.1271115683638154e-08,
+      "loss": 0.2687,
+      "step": 15448
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.141764026252246,
+      "learning_rate": 2.11773734058901e-08,
+      "loss": 0.2464,
+      "step": 15449
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.341899232552611,
+      "learning_rate": 2.1083837708335554e-08,
+      "loss": 0.2536,
+      "step": 15450
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1419034502819514,
+      "learning_rate": 2.0990508594856407e-08,
+      "loss": 0.2622,
+      "step": 15451
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9159169909098033,
+      "learning_rate": 2.089738606932512e-08,
+      "loss": 0.2618,
+      "step": 15452
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9294412910823078,
+      "learning_rate": 2.0804470135604714e-08,
+      "loss": 0.2548,
+      "step": 15453
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9508149838605389,
+      "learning_rate": 2.0711760797550996e-08,
+      "loss": 0.2355,
+      "step": 15454
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5438058980307416,
+      "learning_rate": 2.0619258059010883e-08,
+      "loss": 0.2515,
+      "step": 15455
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6200725801548135,
+      "learning_rate": 2.0526961923821864e-08,
+      "loss": 0.4887,
+      "step": 15456
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3413472880789359,
+      "learning_rate": 2.0434872395813655e-08,
+      "loss": 0.2345,
+      "step": 15457
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2186162078452494,
+      "learning_rate": 2.034298947880764e-08,
+      "loss": 0.2521,
+      "step": 15458
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.486717289994227,
+      "learning_rate": 2.0251313176615218e-08,
+      "loss": 0.2518,
+      "step": 15459
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9950808963470017,
+      "learning_rate": 2.015984349304112e-08,
+      "loss": 0.2483,
+      "step": 15460
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.413461480828005,
+      "learning_rate": 2.0068580431880647e-08,
+      "loss": 0.2463,
+      "step": 15461
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2821608242697047,
+      "learning_rate": 1.9977523996919658e-08,
+      "loss": 0.2445,
+      "step": 15462
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.851568948622649,
+      "learning_rate": 1.9886674191937348e-08,
+      "loss": 0.2472,
+      "step": 15463
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0410138494571632,
+      "learning_rate": 1.9796031020702376e-08,
+      "loss": 0.2572,
+      "step": 15464
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6359560002023097,
+      "learning_rate": 1.970559448697562e-08,
+      "loss": 0.2428,
+      "step": 15465
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4668214706477742,
+      "learning_rate": 1.9615364594509633e-08,
+      "loss": 0.2549,
+      "step": 15466
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6060142031172588,
+      "learning_rate": 1.9525341347048643e-08,
+      "loss": 0.4653,
+      "step": 15467
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.766276291352814,
+      "learning_rate": 1.943552474832744e-08,
+      "loss": 0.257,
+      "step": 15468
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4656088738430961,
+      "learning_rate": 1.934591480207304e-08,
+      "loss": 0.2327,
+      "step": 15469
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7355086601017375,
+      "learning_rate": 1.925651151200303e-08,
+      "loss": 0.2464,
+      "step": 15470
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.817621729816515,
+      "learning_rate": 1.9167314881827214e-08,
+      "loss": 0.2401,
+      "step": 15471
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0191311831408743,
+      "learning_rate": 1.9078324915246527e-08,
+      "loss": 0.2651,
+      "step": 15472
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.793188698041432,
+      "learning_rate": 1.898954161595301e-08,
+      "loss": 0.2711,
+      "step": 15473
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.369588591877338,
+      "learning_rate": 1.8900964987630388e-08,
+      "loss": 0.2505,
+      "step": 15474
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.923550585192585,
+      "learning_rate": 1.8812595033954607e-08,
+      "loss": 0.2578,
+      "step": 15475
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6308461250264963,
+      "learning_rate": 1.8724431758592177e-08,
+      "loss": 0.2697,
+      "step": 15476
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9505678082335662,
+      "learning_rate": 1.8636475165200176e-08,
+      "loss": 0.2364,
+      "step": 15477
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0326938089135087,
+      "learning_rate": 1.8548725257429014e-08,
+      "loss": 0.2496,
+      "step": 15478
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.625344297906875,
+      "learning_rate": 1.8461182038919666e-08,
+      "loss": 0.2496,
+      "step": 15479
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.096852475475712,
+      "learning_rate": 1.8373845513303124e-08,
+      "loss": 0.2666,
+      "step": 15480
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6312141642097142,
+      "learning_rate": 1.8286715684204815e-08,
+      "loss": 0.2385,
+      "step": 15481
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.692134022410049,
+      "learning_rate": 1.819979255523907e-08,
+      "loss": 0.2397,
+      "step": 15482
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.99791696158461,
+      "learning_rate": 1.8113076130012453e-08,
+      "loss": 0.2468,
+      "step": 15483
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.399654529619578,
+      "learning_rate": 1.802656641212375e-08,
+      "loss": 0.2539,
+      "step": 15484
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4177192326430204,
+      "learning_rate": 1.7940263405161195e-08,
+      "loss": 0.2839,
+      "step": 15485
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2935373024116066,
+      "learning_rate": 1.785416711270638e-08,
+      "loss": 0.2504,
+      "step": 15486
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9160151498153903,
+      "learning_rate": 1.7768277538331435e-08,
+      "loss": 0.2612,
+      "step": 15487
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9762390016418894,
+      "learning_rate": 1.7682594685600184e-08,
+      "loss": 0.2539,
+      "step": 15488
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7756054644360466,
+      "learning_rate": 1.759711855806756e-08,
+      "loss": 0.2508,
+      "step": 15489
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6438743333723528,
+      "learning_rate": 1.751184915928017e-08,
+      "loss": 0.4626,
+      "step": 15490
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0046281064284197,
+      "learning_rate": 1.742678649277574e-08,
+      "loss": 0.2559,
+      "step": 15491
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5446583872634272,
+      "learning_rate": 1.7341930562084775e-08,
+      "loss": 0.2403,
+      "step": 15492
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.942016087724381,
+      "learning_rate": 1.7257281370726688e-08,
+      "loss": 0.2788,
+      "step": 15493
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.612550085794354,
+      "learning_rate": 1.7172838922214773e-08,
+      "loss": 0.2519,
+      "step": 15494
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6222136585881686,
+      "learning_rate": 1.7088603220051792e-08,
+      "loss": 0.2407,
+      "step": 15495
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5933577644267796,
+      "learning_rate": 1.7004574267733832e-08,
+      "loss": 0.248,
+      "step": 15496
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5081318471711267,
+      "learning_rate": 1.6920752068746438e-08,
+      "loss": 0.2567,
+      "step": 15497
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3722598268780937,
+      "learning_rate": 1.6837136626568496e-08,
+      "loss": 0.2556,
+      "step": 15498
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1696781911728378,
+      "learning_rate": 1.6753727944668342e-08,
+      "loss": 0.2415,
+      "step": 15499
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2465984209971315,
+      "learning_rate": 1.667052602650765e-08,
+      "loss": 0.2682,
+      "step": 15500
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.073112480747819,
+      "learning_rate": 1.6587530875538106e-08,
+      "loss": 0.2451,
+      "step": 15501
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0039550855240713,
+      "learning_rate": 1.6504742495203064e-08,
+      "loss": 0.2682,
+      "step": 15502
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.716537895168068,
+      "learning_rate": 1.642216088893811e-08,
+      "loss": 0.2615,
+      "step": 15503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6264490824986113,
+      "learning_rate": 1.6339786060169393e-08,
+      "loss": 0.2498,
+      "step": 15504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9944649389374496,
+      "learning_rate": 1.625761801231529e-08,
+      "loss": 0.2611,
+      "step": 15505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9685160961845805,
+      "learning_rate": 1.617565674878474e-08,
+      "loss": 0.2435,
+      "step": 15506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9792377528694678,
+      "learning_rate": 1.6093902272978357e-08,
+      "loss": 0.2532,
+      "step": 15507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5625692901850257,
+      "learning_rate": 1.6012354588288425e-08,
+      "loss": 0.2504,
+      "step": 15508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.872568214140006,
+      "learning_rate": 1.5931013698098353e-08,
+      "loss": 0.2818,
+      "step": 15509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.658827039770151,
+      "learning_rate": 1.5849879605783214e-08,
+      "loss": 0.2456,
+      "step": 15510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7288425240765597,
+      "learning_rate": 1.5768952314709763e-08,
+      "loss": 0.2483,
+      "step": 15511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8807100627466924,
+      "learning_rate": 1.5688231828234757e-08,
+      "loss": 0.248,
+      "step": 15512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8400376668637661,
+      "learning_rate": 1.560771814970885e-08,
+      "loss": 0.2626,
+      "step": 15513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9019812541798318,
+      "learning_rate": 1.5527411282471594e-08,
+      "loss": 0.2478,
+      "step": 15514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0171208771970015,
+      "learning_rate": 1.5447311229855876e-08,
+      "loss": 0.2468,
+      "step": 15515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9905667456664624,
+      "learning_rate": 1.5367417995184597e-08,
+      "loss": 0.2474,
+      "step": 15516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.473877934475984,
+      "learning_rate": 1.5287731581772326e-08,
+      "loss": 0.2532,
+      "step": 15517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4390865518188125,
+      "learning_rate": 1.5208251992926415e-08,
+      "loss": 0.2654,
+      "step": 15518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4060498089965208,
+      "learning_rate": 1.512897923194423e-08,
+      "loss": 0.2412,
+      "step": 15519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5454154221178362,
+      "learning_rate": 1.504991330211425e-08,
+      "loss": 0.2609,
+      "step": 15520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.583097433815207,
+      "learning_rate": 1.4971054206718294e-08,
+      "loss": 0.2747,
+      "step": 15521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.73913176597066,
+      "learning_rate": 1.489240194902708e-08,
+      "loss": 0.23,
+      "step": 15522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3986844251397104,
+      "learning_rate": 1.4813956532305218e-08,
+      "loss": 0.2508,
+      "step": 15523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.873586140642064,
+      "learning_rate": 1.4735717959806773e-08,
+      "loss": 0.2375,
+      "step": 15524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.141798204858941,
+      "learning_rate": 1.4657686234778035e-08,
+      "loss": 0.2457,
+      "step": 15525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5805240595153993,
+      "learning_rate": 1.4579861360457525e-08,
+      "loss": 0.2656,
+      "step": 15526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7488438587677326,
+      "learning_rate": 1.4502243340072663e-08,
+      "loss": 0.2449,
+      "step": 15527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.712316081614816,
+      "learning_rate": 1.4424832176845871e-08,
+      "loss": 0.2637,
+      "step": 15528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9890059149896477,
+      "learning_rate": 1.4347627873987912e-08,
+      "loss": 0.2353,
+      "step": 15529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5888310231399467,
+      "learning_rate": 1.4270630434701782e-08,
+      "loss": 0.2559,
+      "step": 15530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5157996681402393,
+      "learning_rate": 1.4193839862183812e-08,
+      "loss": 0.2622,
+      "step": 15531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2218685234480606,
+      "learning_rate": 1.4117256159618676e-08,
+      "loss": 0.2466,
+      "step": 15532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.744680873333697,
+      "learning_rate": 1.4040879330184387e-08,
+      "loss": 0.2638,
+      "step": 15533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.730767422762499,
+      "learning_rate": 1.3964709377050079e-08,
+      "loss": 0.2698,
+      "step": 15534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.182953404008009,
+      "learning_rate": 1.3888746303376554e-08,
+      "loss": 0.242,
+      "step": 15535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8149028322694094,
+      "learning_rate": 1.3812990112315184e-08,
+      "loss": 0.2658,
+      "step": 15536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.093215393823958,
+      "learning_rate": 1.3737440807009006e-08,
+      "loss": 0.2417,
+      "step": 15537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.060166318007312,
+      "learning_rate": 1.3662098390593292e-08,
+      "loss": 0.2604,
+      "step": 15538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.620758001831998,
+      "learning_rate": 1.3586962866193875e-08,
+      "loss": 0.244,
+      "step": 15539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.927748783609269,
+      "learning_rate": 1.3512034236927706e-08,
+      "loss": 0.2574,
+      "step": 15540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2031237302777953,
+      "learning_rate": 1.3437312505905075e-08,
+      "loss": 0.2363,
+      "step": 15541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.546532599190121,
+      "learning_rate": 1.3362797676224614e-08,
+      "loss": 0.239,
+      "step": 15542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5388820068781249,
+      "learning_rate": 1.3288489750979406e-08,
+      "loss": 0.4509,
+      "step": 15543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9298914157208156,
+      "learning_rate": 1.3214388733252536e-08,
+      "loss": 0.2424,
+      "step": 15544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7125518414190737,
+      "learning_rate": 1.3140494626117661e-08,
+      "loss": 0.2542,
+      "step": 15545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8646247302854344,
+      "learning_rate": 1.3066807432641216e-08,
+      "loss": 0.254,
+      "step": 15546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.445369105515142,
+      "learning_rate": 1.2993327155880753e-08,
+      "loss": 0.2486,
+      "step": 15547
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2019131649084476,
+      "learning_rate": 1.2920053798885501e-08,
+      "loss": 0.2489,
+      "step": 15548
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1827807345721975,
+      "learning_rate": 1.2846987364695252e-08,
+      "loss": 0.2577,
+      "step": 15549
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.213672803846767,
+      "learning_rate": 1.2774127856341467e-08,
+      "loss": 0.27,
+      "step": 15550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7643336230298274,
+      "learning_rate": 1.2701475276847286e-08,
+      "loss": 0.2403,
+      "step": 15551
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.206378727249492,
+      "learning_rate": 1.2629029629227518e-08,
+      "loss": 0.2517,
+      "step": 15552
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9358869169210136,
+      "learning_rate": 1.2556790916488093e-08,
+      "loss": 0.2446,
+      "step": 15553
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8228967276809194,
+      "learning_rate": 1.2484759141625502e-08,
+      "loss": 0.2459,
+      "step": 15554
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0295069711263496,
+      "learning_rate": 1.2412934307629575e-08,
+      "loss": 0.2421,
+      "step": 15555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7658988365955928,
+      "learning_rate": 1.2341316417479598e-08,
+      "loss": 0.2474,
+      "step": 15556
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.233087249645574,
+      "learning_rate": 1.2269905474147636e-08,
+      "loss": 0.2555,
+      "step": 15557
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5075996025079546,
+      "learning_rate": 1.2198701480596875e-08,
+      "loss": 0.2483,
+      "step": 15558
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6876271772076141,
+      "learning_rate": 1.2127704439781062e-08,
+      "loss": 0.2457,
+      "step": 15559
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.793953200667405,
+      "learning_rate": 1.2056914354646176e-08,
+      "loss": 0.2621,
+      "step": 15560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0639247294016263,
+      "learning_rate": 1.1986331228129311e-08,
+      "loss": 0.2438,
+      "step": 15561
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5793979500542205,
+      "learning_rate": 1.1915955063159235e-08,
+      "loss": 0.4536,
+      "step": 15562
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7061081259812236,
+      "learning_rate": 1.1845785862656389e-08,
+      "loss": 0.2573,
+      "step": 15563
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.576161420494555,
+      "learning_rate": 1.177582362953178e-08,
+      "loss": 0.2634,
+      "step": 15564
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9030760447820265,
+      "learning_rate": 1.1706068366688083e-08,
+      "loss": 0.252,
+      "step": 15565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0726290133180094,
+      "learning_rate": 1.1636520077020207e-08,
+      "loss": 0.2577,
+      "step": 15566
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5106248481346085,
+      "learning_rate": 1.1567178763413068e-08,
+      "loss": 0.2475,
+      "step": 15567
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9558351080447594,
+      "learning_rate": 1.149804442874436e-08,
+      "loss": 0.2494,
+      "step": 15568
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.334714522306353,
+      "learning_rate": 1.1429117075882345e-08,
+      "loss": 0.2684,
+      "step": 15569
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.570248109556965,
+      "learning_rate": 1.1360396707686961e-08,
+      "loss": 0.2425,
+      "step": 15570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8808737821337238,
+      "learning_rate": 1.1291883327009257e-08,
+      "loss": 0.2564,
+      "step": 15571
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4292775300354967,
+      "learning_rate": 1.1223576936692515e-08,
+      "loss": 0.2506,
+      "step": 15572
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.519205841181352,
+      "learning_rate": 1.115547753957058e-08,
+      "loss": 0.2727,
+      "step": 15573
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.46784944440388,
+      "learning_rate": 1.1087585138469525e-08,
+      "loss": 0.2453,
+      "step": 15574
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6034317401022334,
+      "learning_rate": 1.101989973620543e-08,
+      "loss": 0.4681,
+      "step": 15575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.070420111968326,
+      "learning_rate": 1.095242133558716e-08,
+      "loss": 0.2451,
+      "step": 15576
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5708556517898078,
+      "learning_rate": 1.0885149939414141e-08,
+      "loss": 0.4573,
+      "step": 15577
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.197566537185328,
+      "learning_rate": 1.0818085550478585e-08,
+      "loss": 0.2465,
+      "step": 15578
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6830089399389538,
+      "learning_rate": 1.0751228171561601e-08,
+      "loss": 0.2567,
+      "step": 15579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.2262230473066715,
+      "learning_rate": 1.0684577805438744e-08,
+      "loss": 0.2353,
+      "step": 15580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2733926244862537,
+      "learning_rate": 1.0618134454874473e-08,
+      "loss": 0.2528,
+      "step": 15581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0396419553233973,
+      "learning_rate": 1.0551898122626025e-08,
+      "loss": 0.2399,
+      "step": 15582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.229861437779296,
+      "learning_rate": 1.0485868811441757e-08,
+      "loss": 0.2535,
+      "step": 15583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7751439642267388,
+      "learning_rate": 1.0420046524061145e-08,
+      "loss": 0.2878,
+      "step": 15584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6262032897060499,
+      "learning_rate": 1.0354431263214782e-08,
+      "loss": 0.2643,
+      "step": 15585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.040780135198672,
+      "learning_rate": 1.0289023031626044e-08,
+      "loss": 0.2429,
+      "step": 15586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.965122746676793,
+      "learning_rate": 1.0223821832008873e-08,
+      "loss": 0.267,
+      "step": 15587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4362883851264039,
+      "learning_rate": 1.0158827667067772e-08,
+      "loss": 0.2502,
+      "step": 15588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.511057216270994,
+      "learning_rate": 1.0094040539499473e-08,
+      "loss": 0.2568,
+      "step": 15589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.246104436570195,
+      "learning_rate": 1.0029460451992933e-08,
+      "loss": 0.2413,
+      "step": 15590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6716867312882933,
+      "learning_rate": 9.965087407227125e-09,
+      "loss": 0.2654,
+      "step": 15591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7472619600651458,
+      "learning_rate": 9.900921407873243e-09,
+      "loss": 0.2385,
+      "step": 15592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.725106848403077,
+      "learning_rate": 9.836962456593602e-09,
+      "loss": 0.2644,
+      "step": 15593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6500384360411164,
+      "learning_rate": 9.77321055604219e-09,
+      "loss": 0.2547,
+      "step": 15594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.632168563033859,
+      "learning_rate": 9.709665708863558e-09,
+      "loss": 0.2495,
+      "step": 15595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.006364411613694,
+      "learning_rate": 9.646327917694486e-09,
+      "loss": 0.2405,
+      "step": 15596
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9932137846393272,
+      "learning_rate": 9.58319718516343e-09,
+      "loss": 0.2555,
+      "step": 15597
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9316758936017333,
+      "learning_rate": 9.5202735138894e-09,
+      "loss": 0.2499,
+      "step": 15598
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5850876663634432,
+      "learning_rate": 9.45755690648309e-09,
+      "loss": 0.4632,
+      "step": 15599
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.855408833305095,
+      "learning_rate": 9.395047365547416e-09,
+      "loss": 0.2405,
+      "step": 15600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.236241917557413,
+      "learning_rate": 9.332744893675306e-09,
+      "loss": 0.2761,
+      "step": 15601
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6952567859126504,
+      "learning_rate": 9.27064949345191e-09,
+      "loss": 0.25,
+      "step": 15602
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9630476757870707,
+      "learning_rate": 9.208761167453507e-09,
+      "loss": 0.2625,
+      "step": 15603
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.962904644754545,
+      "learning_rate": 9.147079918249146e-09,
+      "loss": 0.2604,
+      "step": 15604
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1188646288857007,
+      "learning_rate": 9.085605748396231e-09,
+      "loss": 0.242,
+      "step": 15605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4531888245836753,
+      "learning_rate": 9.024338660447162e-09,
+      "loss": 0.2361,
+      "step": 15606
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7107240539871507,
+      "learning_rate": 8.963278656942687e-09,
+      "loss": 0.2487,
+      "step": 15607
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.318751386047623,
+      "learning_rate": 8.902425740416886e-09,
+      "loss": 0.2462,
+      "step": 15608
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5774320024795008,
+      "learning_rate": 8.841779913394966e-09,
+      "loss": 0.2671,
+      "step": 15609
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6641220630708113,
+      "learning_rate": 8.781341178393244e-09,
+      "loss": 0.2442,
+      "step": 15610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9925184886437746,
+      "learning_rate": 8.72110953791805e-09,
+      "loss": 0.2538,
+      "step": 15611
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4335687958758183,
+      "learning_rate": 8.66108499447016e-09,
+      "loss": 0.2492,
+      "step": 15612
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5373189443861768,
+      "learning_rate": 8.601267550539805e-09,
+      "loss": 0.2437,
+      "step": 15613
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5720932307561466,
+      "learning_rate": 8.541657208607778e-09,
+      "loss": 0.4494,
+      "step": 15614
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7043081794435382,
+      "learning_rate": 8.482253971148768e-09,
+      "loss": 0.266,
+      "step": 15615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6149659013118141,
+      "learning_rate": 8.423057840626914e-09,
+      "loss": 0.4963,
+      "step": 15616
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.613573159584638,
+      "learning_rate": 8.364068819498029e-09,
+      "loss": 0.2672,
+      "step": 15617
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.908704026341994,
+      "learning_rate": 8.305286910210709e-09,
+      "loss": 0.2619,
+      "step": 15618
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9634236224880925,
+      "learning_rate": 8.246712115203004e-09,
+      "loss": 0.2433,
+      "step": 15619
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8058417900140946,
+      "learning_rate": 8.188344436905193e-09,
+      "loss": 0.2485,
+      "step": 15620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0499520966327007,
+      "learning_rate": 8.130183877739784e-09,
+      "loss": 0.2506,
+      "step": 15621
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.464304071789921,
+      "learning_rate": 8.072230440119843e-09,
+      "loss": 0.2453,
+      "step": 15622
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9687961653155635,
+      "learning_rate": 8.014484126449008e-09,
+      "loss": 0.2397,
+      "step": 15623
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0285807789622403,
+      "learning_rate": 7.9569449391248e-09,
+      "loss": 0.273,
+      "step": 15624
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.591317288318226,
+      "learning_rate": 7.899612880533092e-09,
+      "loss": 0.2368,
+      "step": 15625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6196789453325597,
+      "learning_rate": 7.842487953054202e-09,
+      "loss": 0.2361,
+      "step": 15626
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2419885920789207,
+      "learning_rate": 7.785570159056788e-09,
+      "loss": 0.2387,
+      "step": 15627
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.905001739989543,
+      "learning_rate": 7.728859500903408e-09,
+      "loss": 0.2557,
+      "step": 15628
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7865140658066843,
+      "learning_rate": 7.672355980947177e-09,
+      "loss": 0.2647,
+      "step": 15629
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1961707316110086,
+      "learning_rate": 7.616059601532333e-09,
+      "loss": 0.2467,
+      "step": 15630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7961826128246297,
+      "learning_rate": 7.55997036499423e-09,
+      "loss": 0.2543,
+      "step": 15631
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9538604690082362,
+      "learning_rate": 7.504088273661003e-09,
+      "loss": 0.2887,
+      "step": 15632
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2481173055607213,
+      "learning_rate": 7.448413329850801e-09,
+      "loss": 0.2598,
+      "step": 15633
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9807230966892162,
+      "learning_rate": 7.392945535873441e-09,
+      "loss": 0.2485,
+      "step": 15634
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.754303864248791,
+      "learning_rate": 7.337684894030417e-09,
+      "loss": 0.2374,
+      "step": 15635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.07331941428063,
+      "learning_rate": 7.282631406615448e-09,
+      "loss": 0.2354,
+      "step": 15636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7190981757398527,
+      "learning_rate": 7.2277850759117085e-09,
+      "loss": 0.2255,
+      "step": 15637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5759893447618776,
+      "learning_rate": 7.173145904195156e-09,
+      "loss": 0.2581,
+      "step": 15638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8790706297505653,
+      "learning_rate": 7.11871389373342e-09,
+      "loss": 0.2496,
+      "step": 15639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.577593385186562,
+      "learning_rate": 7.064489046784695e-09,
+      "loss": 0.2642,
+      "step": 15640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6178861605317176,
+      "learning_rate": 7.010471365598292e-09,
+      "loss": 0.2423,
+      "step": 15641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.979658809139444,
+      "learning_rate": 6.956660852416308e-09,
+      "loss": 0.2475,
+      "step": 15642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.265061644184496,
+      "learning_rate": 6.903057509470845e-09,
+      "loss": 0.2513,
+      "step": 15643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.183084640261362,
+      "learning_rate": 6.8496613389867905e-09,
+      "loss": 0.2348,
+      "step": 15644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0015993779853,
+      "learning_rate": 6.796472343178484e-09,
+      "loss": 0.2701,
+      "step": 15645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.768687284571175,
+      "learning_rate": 6.743490524254159e-09,
+      "loss": 0.2748,
+      "step": 15646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.518585898499266,
+      "learning_rate": 6.690715884410947e-09,
+      "loss": 0.2454,
+      "step": 15647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 32.006304581087356,
+      "learning_rate": 6.638148425838764e-09,
+      "loss": 0.2526,
+      "step": 15648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.733578765111345,
+      "learning_rate": 6.585788150719752e-09,
+      "loss": 0.2453,
+      "step": 15649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.082609978964282,
+      "learning_rate": 6.533635061225507e-09,
+      "loss": 0.2773,
+      "step": 15650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1356784380508858,
+      "learning_rate": 6.481689159519855e-09,
+      "loss": 0.2572,
+      "step": 15651
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.234922177587976,
+      "learning_rate": 6.429950447758293e-09,
+      "loss": 0.2598,
+      "step": 15652
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.653254671064186,
+      "learning_rate": 6.378418928087993e-09,
+      "loss": 0.2567,
+      "step": 15653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8209219972317714,
+      "learning_rate": 6.327094602646688e-09,
+      "loss": 0.2629,
+      "step": 15654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.331626509852034,
+      "learning_rate": 6.275977473563788e-09,
+      "loss": 0.2584,
+      "step": 15655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9876333163916338,
+      "learning_rate": 6.2250675429609274e-09,
+      "loss": 0.2544,
+      "step": 15656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5051829218987804,
+      "learning_rate": 6.174364812949752e-09,
+      "loss": 0.2595,
+      "step": 15657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0610047525807573,
+      "learning_rate": 6.123869285634132e-09,
+      "loss": 0.2503,
+      "step": 15658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.385699698392544,
+      "learning_rate": 6.073580963109061e-09,
+      "loss": 0.2527,
+      "step": 15659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.234972326993217,
+      "learning_rate": 6.023499847461201e-09,
+      "loss": 0.2408,
+      "step": 15660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8396623098913352,
+      "learning_rate": 5.973625940769445e-09,
+      "loss": 0.2506,
+      "step": 15661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.006516227653823,
+      "learning_rate": 5.9239592451015845e-09,
+      "loss": 0.2293,
+      "step": 15662
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.629313662584817,
+      "learning_rate": 5.874499762519303e-09,
+      "loss": 0.2426,
+      "step": 15663
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5684966662259532,
+      "learning_rate": 5.825247495074849e-09,
+      "loss": 0.2422,
+      "step": 15664
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6761734588470107,
+      "learning_rate": 5.776202444811585e-09,
+      "loss": 0.2435,
+      "step": 15665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.046085493563638,
+      "learning_rate": 5.727364613763997e-09,
+      "loss": 0.2548,
+      "step": 15666
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0059501740205063,
+      "learning_rate": 5.678734003958797e-09,
+      "loss": 0.2685,
+      "step": 15667
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.98830537254205,
+      "learning_rate": 5.63031061741437e-09,
+      "loss": 0.2554,
+      "step": 15668
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0175163559699087,
+      "learning_rate": 5.582094456139109e-09,
+      "loss": 0.2514,
+      "step": 15669
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5522079285470918,
+      "learning_rate": 5.534085522133637e-09,
+      "loss": 0.2492,
+      "step": 15670
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.112999635984669,
+      "learning_rate": 5.486283817390248e-09,
+      "loss": 0.2524,
+      "step": 15671
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.520334575559302,
+      "learning_rate": 5.438689343892356e-09,
+      "loss": 0.2453,
+      "step": 15672
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1018062447739836,
+      "learning_rate": 5.391302103615048e-09,
+      "loss": 0.2641,
+      "step": 15673
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4175783931073094,
+      "learning_rate": 5.344122098523419e-09,
+      "loss": 0.2365,
+      "step": 15674
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.016727443018837,
+      "learning_rate": 5.2971493305753465e-09,
+      "loss": 0.2684,
+      "step": 15675
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2345848687961936,
+      "learning_rate": 5.2503838017203824e-09,
+      "loss": 0.258,
+      "step": 15676
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.011374306576539,
+      "learning_rate": 5.2038255138991965e-09,
+      "loss": 0.2425,
+      "step": 15677
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.356641889740356,
+      "learning_rate": 5.157474469042467e-09,
+      "loss": 0.231,
+      "step": 15678
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0070010729827286,
+      "learning_rate": 5.111330669074211e-09,
+      "loss": 0.2491,
+      "step": 15679
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.776880313302346,
+      "learning_rate": 5.065394115909006e-09,
+      "loss": 0.2489,
+      "step": 15680
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8691306274525914,
+      "learning_rate": 5.019664811451997e-09,
+      "loss": 0.264,
+      "step": 15681
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.12316657401478,
+      "learning_rate": 4.974142757601663e-09,
+      "loss": 0.2579,
+      "step": 15682
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6520729617270807,
+      "learning_rate": 4.92882795624594e-09,
+      "loss": 0.4264,
+      "step": 15683
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5793191284545007,
+      "learning_rate": 4.8837204092655465e-09,
+      "loss": 0.2545,
+      "step": 15684
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.038925760909837,
+      "learning_rate": 4.838820118532317e-09,
+      "loss": 0.2648,
+      "step": 15685
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8814905684080334,
+      "learning_rate": 4.7941270859086505e-09,
+      "loss": 0.2638,
+      "step": 15686
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7749038360763734,
+      "learning_rate": 4.749641313248621e-09,
+      "loss": 0.2488,
+      "step": 15687
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.117625713689163,
+      "learning_rate": 4.705362802399082e-09,
+      "loss": 0.273,
+      "step": 15688
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2739046925445483,
+      "learning_rate": 4.661291555196346e-09,
+      "loss": 0.2455,
+      "step": 15689
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.335449567235336,
+      "learning_rate": 4.617427573469502e-09,
+      "loss": 0.2577,
+      "step": 15690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9443615278019875,
+      "learning_rate": 4.573770859038207e-09,
+      "loss": 0.236,
+      "step": 15691
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.722616558266238,
+      "learning_rate": 4.530321413714345e-09,
+      "loss": 0.2628,
+      "step": 15692
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.163307092856542,
+      "learning_rate": 4.487079239300363e-09,
+      "loss": 0.2504,
+      "step": 15693
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5143142547497772,
+      "learning_rate": 4.444044337590381e-09,
+      "loss": 0.2548,
+      "step": 15694
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.357737687257555,
+      "learning_rate": 4.401216710370193e-09,
+      "loss": 0.2414,
+      "step": 15695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.296475992183728,
+      "learning_rate": 4.358596359416712e-09,
+      "loss": 0.2516,
+      "step": 15696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3018184705700144,
+      "learning_rate": 4.316183286497966e-09,
+      "loss": 0.2452,
+      "step": 15697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0956838160957654,
+      "learning_rate": 4.273977493374215e-09,
+      "loss": 0.2639,
+      "step": 15698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8657023384057065,
+      "learning_rate": 4.231978981796836e-09,
+      "loss": 0.2547,
+      "step": 15699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5622571930646347,
+      "learning_rate": 4.190187753507769e-09,
+      "loss": 0.2471,
+      "step": 15700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8443428570988152,
+      "learning_rate": 4.148603810241736e-09,
+      "loss": 0.2454,
+      "step": 15701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6598271444935904,
+      "learning_rate": 4.107227153724025e-09,
+      "loss": 0.2499,
+      "step": 15702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7876087746085303,
+      "learning_rate": 4.066057785670486e-09,
+      "loss": 0.2591,
+      "step": 15703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.151683819391787,
+      "learning_rate": 4.0250957077903055e-09,
+      "loss": 0.239,
+      "step": 15704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6222148335428344,
+      "learning_rate": 3.984340921782681e-09,
+      "loss": 0.2608,
+      "step": 15705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1699803301021197,
+      "learning_rate": 3.9437934293384824e-09,
+      "loss": 0.2648,
+      "step": 15706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.633422960341346,
+      "learning_rate": 3.903453232140808e-09,
+      "loss": 0.2743,
+      "step": 15707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.548380016945056,
+      "learning_rate": 3.863320331862763e-09,
+      "loss": 0.2577,
+      "step": 15708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6850794141643164,
+      "learning_rate": 3.823394730169683e-09,
+      "loss": 0.2489,
+      "step": 15709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8572656559590008,
+      "learning_rate": 3.78367642871802e-09,
+      "loss": 0.2517,
+      "step": 15710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5703480181271305,
+      "learning_rate": 3.744165429155344e-09,
+      "loss": 0.2683,
+      "step": 15711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6371975870181545,
+      "learning_rate": 3.7048617331225667e-09,
+      "loss": 0.2547,
+      "step": 15712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.740872662529989,
+      "learning_rate": 3.6657653422489392e-09,
+      "loss": 0.2464,
+      "step": 15713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9950757202341172,
+      "learning_rate": 3.626876258157608e-09,
+      "loss": 0.2741,
+      "step": 15714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7426453294714634,
+      "learning_rate": 3.5881944824611716e-09,
+      "loss": 0.2569,
+      "step": 15715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5666520416987435,
+      "learning_rate": 3.5497200167655676e-09,
+      "loss": 0.2771,
+      "step": 15716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1091677599967005,
+      "learning_rate": 3.511452862666742e-09,
+      "loss": 0.2572,
+      "step": 15717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7261772790606957,
+      "learning_rate": 3.473393021752314e-09,
+      "loss": 0.2442,
+      "step": 15718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.517246950564713,
+      "learning_rate": 3.4355404956021297e-09,
+      "loss": 0.226,
+      "step": 15719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6825080849977034,
+      "learning_rate": 3.3978952857854907e-09,
+      "loss": 0.2485,
+      "step": 15720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.79111575766408,
+      "learning_rate": 3.3604573938655903e-09,
+      "loss": 0.2495,
+      "step": 15721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6244642837968324,
+      "learning_rate": 3.323226821395631e-09,
+      "loss": 0.2741,
+      "step": 15722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8370531258603937,
+      "learning_rate": 3.2862035699199323e-09,
+      "loss": 0.2484,
+      "step": 15723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.320558568930975,
+      "learning_rate": 3.2493876409744886e-09,
+      "loss": 0.2359,
+      "step": 15724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3111944569878298,
+      "learning_rate": 3.212779036087521e-09,
+      "loss": 0.258,
+      "step": 15725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.748964897584076,
+      "learning_rate": 3.17637775677726e-09,
+      "loss": 0.2621,
+      "step": 15726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8969812154128607,
+      "learning_rate": 3.1401838045547197e-09,
+      "loss": 0.2515,
+      "step": 15727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7619772801912825,
+      "learning_rate": 3.1041971809209204e-09,
+      "loss": 0.2615,
+      "step": 15728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.075026575288025,
+      "learning_rate": 3.068417887370223e-09,
+      "loss": 0.2475,
+      "step": 15729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6039535760746912,
+      "learning_rate": 3.0328459253858854e-09,
+      "loss": 0.2509,
+      "step": 15730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.519921143423708,
+      "learning_rate": 2.9974812964445042e-09,
+      "loss": 0.2557,
+      "step": 15731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1964090774125284,
+      "learning_rate": 2.9623240020132395e-09,
+      "loss": 0.2567,
+      "step": 15732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.817719585498989,
+      "learning_rate": 2.9273740435514787e-09,
+      "loss": 0.2522,
+      "step": 15733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5741963012685687,
+      "learning_rate": 2.892631422508063e-09,
+      "loss": 0.2473,
+      "step": 15734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.812412797870793,
+      "learning_rate": 2.8580961403251726e-09,
+      "loss": 0.2561,
+      "step": 15735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9472104741231293,
+      "learning_rate": 2.8237681984361054e-09,
+      "loss": 0.2661,
+      "step": 15736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.120596604382238,
+      "learning_rate": 2.789647598264722e-09,
+      "loss": 0.2563,
+      "step": 15737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0668650554990875,
+      "learning_rate": 2.755734341227112e-09,
+      "loss": 0.2436,
+      "step": 15738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5249412288243576,
+      "learning_rate": 2.722028428730483e-09,
+      "loss": 0.2396,
+      "step": 15739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.724048197980926,
+      "learning_rate": 2.6885298621726062e-09,
+      "loss": 0.2658,
+      "step": 15740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8083641336883494,
+      "learning_rate": 2.6552386429434806e-09,
+      "loss": 0.2402,
+      "step": 15741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.129485889147207,
+      "learning_rate": 2.6221547724253337e-09,
+      "loss": 0.2671,
+      "step": 15742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6534557948417292,
+      "learning_rate": 2.5892782519904015e-09,
+      "loss": 0.2695,
+      "step": 15743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.642079432730648,
+      "learning_rate": 2.5566090830025924e-09,
+      "loss": 0.2415,
+      "step": 15744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7533545002534725,
+      "learning_rate": 2.5241472668174894e-09,
+      "loss": 0.2553,
+      "step": 15745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.678989742258969,
+      "learning_rate": 2.4918928047817924e-09,
+      "loss": 0.2423,
+      "step": 15746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.397889746513507,
+      "learning_rate": 2.459845698234431e-09,
+      "loss": 0.2508,
+      "step": 15747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2909800370640574,
+      "learning_rate": 2.4280059485043418e-09,
+      "loss": 0.2471,
+      "step": 15748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0282399774213684,
+      "learning_rate": 2.396373556913245e-09,
+      "loss": 0.2549,
+      "step": 15749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6786566203014583,
+      "learning_rate": 2.364948524773425e-09,
+      "loss": 0.258,
+      "step": 15750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.319857187054752,
+      "learning_rate": 2.333730853388283e-09,
+      "loss": 0.2609,
+      "step": 15751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.984921089124113,
+      "learning_rate": 2.3027205440540047e-09,
+      "loss": 0.2581,
+      "step": 15752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7630771807481402,
+      "learning_rate": 2.271917598056228e-09,
+      "loss": 0.2774,
+      "step": 15753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.303271519897911,
+      "learning_rate": 2.24132201667393e-09,
+      "loss": 0.2545,
+      "step": 15754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4398670295687306,
+      "learning_rate": 2.210933801176096e-09,
+      "loss": 0.2438,
+      "step": 15755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.558926362763449,
+      "learning_rate": 2.1807529528239392e-09,
+      "loss": 0.2329,
+      "step": 15756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.820919925346816,
+      "learning_rate": 2.1507794728692354e-09,
+      "loss": 0.2495,
+      "step": 15757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6281178551012414,
+      "learning_rate": 2.121013362555435e-09,
+      "loss": 0.2455,
+      "step": 15758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.12444545364323,
+      "learning_rate": 2.0914546231187717e-09,
+      "loss": 0.284,
+      "step": 15759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.310457589552341,
+      "learning_rate": 2.0621032557843755e-09,
+      "loss": 0.2672,
+      "step": 15760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2508547205728386,
+      "learning_rate": 2.0329592617712724e-09,
+      "loss": 0.2461,
+      "step": 15761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8216047821695156,
+      "learning_rate": 2.0040226422873844e-09,
+      "loss": 0.2507,
+      "step": 15762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.340007702846668,
+      "learning_rate": 1.975293398534528e-09,
+      "loss": 0.262,
+      "step": 15763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0180235361840233,
+      "learning_rate": 1.9467715317039725e-09,
+      "loss": 0.2574,
+      "step": 15764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4677583375723975,
+      "learning_rate": 1.918457042979216e-09,
+      "loss": 0.2313,
+      "step": 15765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9576456015077983,
+      "learning_rate": 1.8903499335359842e-09,
+      "loss": 0.2334,
+      "step": 15766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.617471321302201,
+      "learning_rate": 1.8624502045389015e-09,
+      "loss": 0.2472,
+      "step": 15767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.239156455603878,
+      "learning_rate": 1.8347578571470404e-09,
+      "loss": 0.2591,
+      "step": 15768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2377567823075415,
+      "learning_rate": 1.8072728925089266e-09,
+      "loss": 0.2524,
+      "step": 15769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.370827325043801,
+      "learning_rate": 1.7799953117642045e-09,
+      "loss": 0.2456,
+      "step": 15770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1988984807116836,
+      "learning_rate": 1.752925116045856e-09,
+      "loss": 0.269,
+      "step": 15771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7036850099691692,
+      "learning_rate": 1.7260623064763172e-09,
+      "loss": 0.2628,
+      "step": 15772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.498788690928352,
+      "learning_rate": 1.6994068841702515e-09,
+      "loss": 0.2555,
+      "step": 15773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6238314332509672,
+      "learning_rate": 1.6729588502339966e-09,
+      "loss": 0.4649,
+      "step": 15774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5360329641170327,
+      "learning_rate": 1.6467182057650077e-09,
+      "loss": 0.2423,
+      "step": 15775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8678257484980374,
+      "learning_rate": 1.6206849518513036e-09,
+      "loss": 0.265,
+      "step": 15776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4887046116583083,
+      "learning_rate": 1.5948590895736859e-09,
+      "loss": 0.2588,
+      "step": 15777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4301680936463346,
+      "learning_rate": 1.5692406200035204e-09,
+      "loss": 0.2352,
+      "step": 15778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3356872740475663,
+      "learning_rate": 1.5438295442032902e-09,
+      "loss": 0.2508,
+      "step": 15779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7309428404312084,
+      "learning_rate": 1.5186258632282625e-09,
+      "loss": 0.2546,
+      "step": 15780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6027635420829873,
+      "learning_rate": 1.493629578123712e-09,
+      "loss": 0.2466,
+      "step": 15781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5931391141911617,
+      "learning_rate": 1.4688406899271424e-09,
+      "loss": 0.4778,
+      "step": 15782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.626739295954805,
+      "learning_rate": 1.44425919966662e-09,
+      "loss": 0.263,
+      "step": 15783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.968552764813573,
+      "learning_rate": 1.4198851083618847e-09,
+      "loss": 0.2673,
+      "step": 15784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.866082105062887,
+      "learning_rate": 1.395718417024905e-09,
+      "loss": 0.2523,
+      "step": 15785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0104393193516383,
+      "learning_rate": 1.3717591266576569e-09,
+      "loss": 0.2483,
+      "step": 15786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2977593434636496,
+      "learning_rate": 1.3480072382549003e-09,
+      "loss": 0.2622,
+      "step": 15787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8680346634061324,
+      "learning_rate": 1.3244627528019582e-09,
+      "loss": 0.242,
+      "step": 15788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6706379357274512,
+      "learning_rate": 1.3011256712758268e-09,
+      "loss": 0.245,
+      "step": 15789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.806097211194816,
+      "learning_rate": 1.2779959946446209e-09,
+      "loss": 0.2431,
+      "step": 15790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9816721896288432,
+      "learning_rate": 1.2550737238681277e-09,
+      "loss": 0.2592,
+      "step": 15791
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.773295092238538,
+      "learning_rate": 1.2323588598972536e-09,
+      "loss": 0.2515,
+      "step": 15792
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7284142664191817,
+      "learning_rate": 1.209851403675133e-09,
+      "loss": 0.2438,
+      "step": 15793
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.970835538706795,
+      "learning_rate": 1.1875513561343531e-09,
+      "loss": 0.2575,
+      "step": 15794
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4136915313654406,
+      "learning_rate": 1.1654587182013955e-09,
+      "loss": 0.2522,
+      "step": 15795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2831527404125618,
+      "learning_rate": 1.1435734907921936e-09,
+      "loss": 0.2637,
+      "step": 15796
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4595129764799224,
+      "learning_rate": 1.1218956748154652e-09,
+      "loss": 0.2572,
+      "step": 15797
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.535759149745823,
+      "learning_rate": 1.100425271170491e-09,
+      "loss": 0.2367,
+      "step": 15798
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6849277462826926,
+      "learning_rate": 1.0791622807471147e-09,
+      "loss": 0.25,
+      "step": 15799
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.577346126166828,
+      "learning_rate": 1.0581067044290738e-09,
+      "loss": 0.2434,
+      "step": 15800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5184578024039044,
+      "learning_rate": 1.0372585430890037e-09,
+      "loss": 0.2529,
+      "step": 15801
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7656763929677726,
+      "learning_rate": 1.016617797592323e-09,
+      "loss": 0.2501,
+      "step": 15802
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.271362466702832,
+      "learning_rate": 9.96184468795014e-10,
+      "loss": 0.2572,
+      "step": 15803
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4264577474246123,
+      "learning_rate": 9.75958557545842e-10,
+      "loss": 0.2431,
+      "step": 15804
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6250150979802267,
+      "learning_rate": 9.559400646830253e-10,
+      "loss": 0.2377,
+      "step": 15805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1494566019549723,
+      "learning_rate": 9.36128991037566e-10,
+      "loss": 0.2628,
+      "step": 15806
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9165050424740038,
+      "learning_rate": 9.165253374315841e-10,
+      "loss": 0.2624,
+      "step": 15807
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6692706515997264,
+      "learning_rate": 8.971291046783182e-10,
+      "loss": 0.2418,
+      "step": 15808
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.037366880622655,
+      "learning_rate": 8.779402935826797e-10,
+      "loss": 0.2546,
+      "step": 15809
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.026183330748864,
+      "learning_rate": 8.589589049406988e-10,
+      "loss": 0.2605,
+      "step": 15810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6477617273822848,
+      "learning_rate": 8.401849395395234e-10,
+      "loss": 0.4714,
+      "step": 15811
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7289607764256854,
+      "learning_rate": 8.216183981590853e-10,
+      "loss": 0.2438,
+      "step": 15812
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7450392671904176,
+      "learning_rate": 8.032592815693241e-10,
+      "loss": 0.2455,
+      "step": 15813
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.457123771551411,
+      "learning_rate": 7.851075905312977e-10,
+      "loss": 0.2626,
+      "step": 15814
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.861847714252405,
+      "learning_rate": 7.671633257994026e-10,
+      "loss": 0.2697,
+      "step": 15815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0476718483286223,
+      "learning_rate": 7.494264881174884e-10,
+      "loss": 0.2674,
+      "step": 15816
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2708692136065403,
+      "learning_rate": 7.318970782210777e-10,
+      "loss": 0.2535,
+      "step": 15817
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.011896763030752,
+      "learning_rate": 7.145750968384768e-10,
+      "loss": 0.2311,
+      "step": 15818
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.89508554758061,
+      "learning_rate": 6.97460544687445e-10,
+      "loss": 0.264,
+      "step": 15819
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.808696185442965,
+      "learning_rate": 6.805534224785248e-10,
+      "loss": 0.2623,
+      "step": 15820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8490841953622192,
+      "learning_rate": 6.638537309133775e-10,
+      "loss": 0.2665,
+      "step": 15821
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.448466173634456,
+      "learning_rate": 6.473614706847819e-10,
+      "loss": 0.2433,
+      "step": 15822
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.379150334893618,
+      "learning_rate": 6.310766424771908e-10,
+      "loss": 0.2646,
+      "step": 15823
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.778485870475655,
+      "learning_rate": 6.149992469661747e-10,
+      "loss": 0.29,
+      "step": 15824
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8603618420930625,
+      "learning_rate": 5.991292848184227e-10,
+      "loss": 0.2474,
+      "step": 15825
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0839198732403688,
+      "learning_rate": 5.83466756692852e-10,
+      "loss": 0.2466,
+      "step": 15826
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.881392580062999,
+      "learning_rate": 5.680116632389432e-10,
+      "loss": 0.2509,
+      "step": 15827
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4326814213899266,
+      "learning_rate": 5.527640050984051e-10,
+      "loss": 0.2532,
+      "step": 15828
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.111285348025681,
+      "learning_rate": 5.377237829040649e-10,
+      "loss": 0.2332,
+      "step": 15829
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.345627418125003,
+      "learning_rate": 5.228909972793129e-10,
+      "loss": 0.259,
+      "step": 15830
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.885811337884381,
+      "learning_rate": 5.082656488397675e-10,
+      "loss": 0.2576,
+      "step": 15831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.720055567255641,
+      "learning_rate": 4.938477381921658e-10,
+      "loss": 0.2491,
+      "step": 15832
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8205032582896923,
+      "learning_rate": 4.796372659354731e-10,
+      "loss": 0.2564,
+      "step": 15833
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7918494828314753,
+      "learning_rate": 4.656342326586627e-10,
+      "loss": 0.2618,
+      "step": 15834
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2529947044315977,
+      "learning_rate": 4.5183863894293633e-10,
+      "loss": 0.2503,
+      "step": 15835
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8860136646471233,
+      "learning_rate": 4.382504853600589e-10,
+      "loss": 0.2625,
+      "step": 15836
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.084821422668374,
+      "learning_rate": 4.2486977247513384e-10,
+      "loss": 0.2601,
+      "step": 15837
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4990367180345032,
+      "learning_rate": 4.116965008421625e-10,
+      "loss": 0.2525,
+      "step": 15838
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.549193529741228,
+      "learning_rate": 3.987306710079297e-10,
+      "loss": 0.2696,
+      "step": 15839
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6142007502113539,
+      "learning_rate": 3.859722835114488e-10,
+      "loss": 0.475,
+      "step": 15840
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1968732666911017,
+      "learning_rate": 3.734213388806307e-10,
+      "loss": 0.2521,
+      "step": 15841
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3236889276011943,
+      "learning_rate": 3.6107783763728033e-10,
+      "loss": 0.2618,
+      "step": 15842
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7506306941277392,
+      "learning_rate": 3.4894178029265537e-10,
+      "loss": 0.2204,
+      "step": 15843
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.660360868628995,
+      "learning_rate": 3.3701316735135213e-10,
+      "loss": 0.2597,
+      "step": 15844
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.311502716852735,
+      "learning_rate": 3.2529199930741996e-10,
+      "loss": 0.2475,
+      "step": 15845
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.693419771168762,
+      "learning_rate": 3.1377827664769156e-10,
+      "loss": 0.2484,
+      "step": 15846
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6339411603170874,
+      "learning_rate": 3.024719998495629e-10,
+      "loss": 0.4808,
+      "step": 15847
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.071155479781498,
+      "learning_rate": 2.9137316938265826e-10,
+      "loss": 0.2594,
+      "step": 15848
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9132552133377398,
+      "learning_rate": 2.804817857071651e-10,
+      "loss": 0.2572,
+      "step": 15849
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.448989462257928,
+      "learning_rate": 2.697978492749442e-10,
+      "loss": 0.2482,
+      "step": 15850
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.550668758577631,
+      "learning_rate": 2.593213605289746e-10,
+      "loss": 0.2423,
+      "step": 15851
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6655926186417245,
+      "learning_rate": 2.4905231990446367e-10,
+      "loss": 0.246,
+      "step": 15852
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.202039656965659,
+      "learning_rate": 2.3899072782773703e-10,
+      "loss": 0.2553,
+      "step": 15853
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9603793889571737,
+      "learning_rate": 2.2913658471512834e-10,
+      "loss": 0.2544,
+      "step": 15854
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.060598463139521,
+      "learning_rate": 2.1948989097686503e-10,
+      "loss": 0.2634,
+      "step": 15855
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0960565068295582,
+      "learning_rate": 2.1005064701262733e-10,
+      "loss": 0.2478,
+      "step": 15856
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.82967641871937,
+      "learning_rate": 2.008188532132138e-10,
+      "loss": 0.2424,
+      "step": 15857
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9338115148910051,
+      "learning_rate": 1.9179450996331673e-10,
+      "loss": 0.2551,
+      "step": 15858
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.993639365186874,
+      "learning_rate": 1.82977617635971e-10,
+      "loss": 0.2549,
+      "step": 15859
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7870016346903912,
+      "learning_rate": 1.7436817659810534e-10,
+      "loss": 0.26,
+      "step": 15860
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.873555944531384,
+      "learning_rate": 1.6596618720610135e-10,
+      "loss": 0.2539,
+      "step": 15861
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8403489382086395,
+      "learning_rate": 1.5777164980856908e-10,
+      "loss": 0.2622,
+      "step": 15862
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8220043644736423,
+      "learning_rate": 1.4978456474634694e-10,
+      "loss": 0.2485,
+      "step": 15863
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5134512089914693,
+      "learning_rate": 1.4200493234972633e-10,
+      "loss": 0.2714,
+      "step": 15864
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8601289860607761,
+      "learning_rate": 1.3443275294233727e-10,
+      "loss": 0.2703,
+      "step": 15865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.55878366768738,
+      "learning_rate": 1.2706802683837282e-10,
+      "loss": 0.2404,
+      "step": 15866
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2219665945766827,
+      "learning_rate": 1.1991075434258924e-10,
+      "loss": 0.255,
+      "step": 15867
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5993892094123481,
+      "learning_rate": 1.1296093575308142e-10,
+      "loss": 0.4483,
+      "step": 15868
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4817994356802835,
+      "learning_rate": 1.062185713573971e-10,
+      "loss": 0.2407,
+      "step": 15869
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5873082189193888,
+      "learning_rate": 9.96836614353125e-11,
+      "loss": 0.2353,
+      "step": 15870
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6658670506839874,
+      "learning_rate": 9.335620625827712e-11,
+      "loss": 0.2402,
+      "step": 15871
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5357579254195182,
+      "learning_rate": 8.723620608885875e-11,
+      "loss": 0.2492,
+      "step": 15872
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9062473174883259,
+      "learning_rate": 8.13236611807433e-11,
+      "loss": 0.2477,
+      "step": 15873
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 29.69040100379666,
+      "learning_rate": 7.56185717792901e-11,
+      "loss": 0.2662,
+      "step": 15874
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.6464851171485275,
+      "learning_rate": 7.012093812153175e-11,
+      "loss": 0.2362,
+      "step": 15875
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.762516786693178,
+      "learning_rate": 6.483076043506398e-11,
+      "loss": 0.2303,
+      "step": 15876
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.826107944026365,
+      "learning_rate": 5.974803893971093e-11,
+      "loss": 0.2523,
+      "step": 15877
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.22165720210061,
+      "learning_rate": 5.4872773846414984e-11,
+      "loss": 0.2604,
+      "step": 15878
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.016177664600205,
+      "learning_rate": 5.0204965357236735e-11,
+      "loss": 0.2421,
+      "step": 15879
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.134437829279411,
+      "learning_rate": 4.5744613665910096e-11,
+      "loss": 0.2467,
+      "step": 15880
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2137617734187565,
+      "learning_rate": 4.149171895784232e-11,
+      "loss": 0.2435,
+      "step": 15881
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1415668129522456,
+      "learning_rate": 3.744628140900375e-11,
+      "loss": 0.2699,
+      "step": 15882
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4250313758395325,
+      "learning_rate": 3.360830118759317e-11,
+      "loss": 0.2462,
+      "step": 15883
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.481129533467969,
+      "learning_rate": 2.997777845237249e-11,
+      "loss": 0.2671,
+      "step": 15884
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5735979939100921,
+      "learning_rate": 2.655471335433202e-11,
+      "loss": 0.2434,
+      "step": 15885
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.087847386570869,
+      "learning_rate": 2.3339106036135428e-11,
+      "loss": 0.2548,
+      "step": 15886
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5564851815198115,
+      "learning_rate": 2.0330956629899256e-11,
+      "loss": 0.239,
+      "step": 15887
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.614140017072922,
+      "learning_rate": 1.7530265261633817e-11,
+      "loss": 0.2525,
+      "step": 15888
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4842722954885819,
+      "learning_rate": 1.493703204624719e-11,
+      "loss": 0.2512,
+      "step": 15889
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3875573976995257,
+      "learning_rate": 1.2551257092541236e-11,
+      "loss": 0.2483,
+      "step": 15890
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1632674817088717,
+      "learning_rate": 1.0372940499325801e-11,
+      "loss": 0.2434,
+      "step": 15891
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6108113224693525,
+      "learning_rate": 8.402082356528951e-12,
+      "loss": 0.2375,
+      "step": 15892
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0574607295809564,
+      "learning_rate": 6.638682745752078e-12,
+      "loss": 0.2501,
+      "step": 15893
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9223745668222756,
+      "learning_rate": 5.082741740825015e-12,
+      "loss": 0.2481,
+      "step": 15894
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.911128079235186,
+      "learning_rate": 3.734259406140694e-12,
+      "loss": 0.2495,
+      "step": 15895
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6302976214860023,
+      "learning_rate": 2.5932357972102695e-12,
+      "loss": 0.4538,
+      "step": 15896
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5654230457121145,
+      "learning_rate": 1.6596709617733298e-12,
+      "loss": 0.2426,
+      "step": 15897
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.76785231886163,
+      "learning_rate": 9.33564938687681e-13,
+      "loss": 0.2583,
+      "step": 15898
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8857348369708409,
+      "learning_rate": 4.149177579293451e-13,
+      "loss": 0.2469,
+      "step": 15899
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9988417197412622,
+      "learning_rate": 1.037294405925593e-13,
+      "loss": 0.2535,
+      "step": 15900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.689559692396786,
+      "learning_rate": 0.0,
+      "loss": 0.2487,
+      "step": 15901
+    },
+    {
+      "epoch": 1.0,
+      "step": 15901,
+      "total_flos": 1.4822327104372736e+16,
+      "train_loss": 0.06416682437559874,
+      "train_runtime": 143299.6981,
+      "train_samples_per_second": 28.408,
+      "train_steps_per_second": 0.111
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 15901,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3000,
+  "total_flos": 1.4822327104372736e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}