diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,86914 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999395782562285,
+  "eval_steps": 500,
+  "global_step": 12412,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 111.06907213368356,
+      "learning_rate": 2.6809651474530834e-08,
+      "loss": 2.3926,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 122.57469072561354,
+      "learning_rate": 5.361930294906167e-08,
+      "loss": 2.5268,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 100.62349832355093,
+      "learning_rate": 8.04289544235925e-08,
+      "loss": 2.2562,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 90.28389686992509,
+      "learning_rate": 1.0723860589812334e-07,
+      "loss": 2.5248,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 75.80776502213655,
+      "learning_rate": 1.3404825737265417e-07,
+      "loss": 2.3516,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 78.10573773693335,
+      "learning_rate": 1.60857908847185e-07,
+      "loss": 2.5393,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 105.50692608615883,
+      "learning_rate": 1.8766756032171582e-07,
+      "loss": 2.3677,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 75.6952043936806,
+      "learning_rate": 2.1447721179624667e-07,
+      "loss": 2.463,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.190875966581483,
+      "learning_rate": 2.412868632707775e-07,
+      "loss": 0.9404,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 109.4248955165473,
+      "learning_rate": 2.6809651474530835e-07,
+      "loss": 2.3585,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 99.59380465048183,
+      "learning_rate": 2.9490616621983914e-07,
+      "loss": 2.4175,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 68.81999458016412,
+      "learning_rate": 3.2171581769437e-07,
+      "loss": 2.3206,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 85.62017057129694,
+      "learning_rate": 3.4852546916890084e-07,
+      "loss": 2.417,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 144.1487196571025,
+      "learning_rate": 3.7533512064343164e-07,
+      "loss": 2.2488,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 93.32265002552015,
+      "learning_rate": 4.021447721179625e-07,
+      "loss": 2.1031,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 182.79458820933567,
+      "learning_rate": 4.2895442359249334e-07,
+      "loss": 2.1022,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 163.0738878225148,
+      "learning_rate": 4.557640750670242e-07,
+      "loss": 2.1135,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 64.47243983256514,
+      "learning_rate": 4.82573726541555e-07,
+      "loss": 2.118,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 66.74365484559563,
+      "learning_rate": 5.093833780160858e-07,
+      "loss": 1.9807,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 71.76674495041182,
+      "learning_rate": 5.361930294906167e-07,
+      "loss": 1.8048,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 109.23880399694352,
+      "learning_rate": 5.630026809651475e-07,
+      "loss": 1.6711,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 46.65886934661648,
+      "learning_rate": 5.898123324396783e-07,
+      "loss": 1.6922,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 148.9421893121272,
+      "learning_rate": 6.166219839142092e-07,
+      "loss": 1.7409,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 58.966936003100365,
+      "learning_rate": 6.4343163538874e-07,
+      "loss": 1.6452,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 71.36941366857242,
+      "learning_rate": 6.702412868632709e-07,
+      "loss": 1.5311,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 49.064400645965556,
+      "learning_rate": 6.970509383378017e-07,
+      "loss": 1.1842,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.64077530444452,
+      "learning_rate": 7.238605898123326e-07,
+      "loss": 1.1651,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 30.752823655269193,
+      "learning_rate": 7.506702412868633e-07,
+      "loss": 0.955,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 48.937434341529446,
+      "learning_rate": 7.774798927613941e-07,
+      "loss": 1.0551,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 63.37420619465647,
+      "learning_rate": 8.04289544235925e-07,
+      "loss": 1.1285,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.27878443183434,
+      "learning_rate": 8.310991957104558e-07,
+      "loss": 1.0806,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.79650007251078,
+      "learning_rate": 8.579088471849867e-07,
+      "loss": 0.9198,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8926532202369173,
+      "learning_rate": 8.847184986595175e-07,
+      "loss": 0.728,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 39.310164985315204,
+      "learning_rate": 9.115281501340484e-07,
+      "loss": 1.0737,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 42.54679694656224,
+      "learning_rate": 9.383378016085791e-07,
+      "loss": 0.9975,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.181364442449466,
+      "learning_rate": 9.6514745308311e-07,
+      "loss": 1.0426,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.875990860725299,
+      "learning_rate": 9.919571045576408e-07,
+      "loss": 1.0418,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.873507223774514,
+      "learning_rate": 1.0187667560321716e-06,
+      "loss": 0.8424,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.153543914244185,
+      "learning_rate": 1.0455764075067026e-06,
+      "loss": 1.022,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.458052094727243,
+      "learning_rate": 1.0723860589812334e-06,
+      "loss": 0.9638,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.35496404842189,
+      "learning_rate": 1.0991957104557642e-06,
+      "loss": 1.0095,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.071577129329377,
+      "learning_rate": 1.126005361930295e-06,
+      "loss": 0.8324,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.934904322497675,
+      "learning_rate": 1.1528150134048258e-06,
+      "loss": 0.7924,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.84385069391789,
+      "learning_rate": 1.1796246648793566e-06,
+      "loss": 0.8623,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.220955275256024,
+      "learning_rate": 1.2064343163538874e-06,
+      "loss": 0.9353,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.02763926347162,
+      "learning_rate": 1.2332439678284184e-06,
+      "loss": 0.81,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.84395125276659,
+      "learning_rate": 1.2600536193029492e-06,
+      "loss": 0.9208,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3106299924520064,
+      "learning_rate": 1.28686327077748e-06,
+      "loss": 0.7365,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.667720716271171,
+      "learning_rate": 1.3136729222520108e-06,
+      "loss": 0.8416,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.02565410925133,
+      "learning_rate": 1.3404825737265418e-06,
+      "loss": 0.9394,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.366928581250008,
+      "learning_rate": 1.3672922252010726e-06,
+      "loss": 0.99,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.630458709877042,
+      "learning_rate": 1.3941018766756034e-06,
+      "loss": 0.9096,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.401327406676558,
+      "learning_rate": 1.4209115281501342e-06,
+      "loss": 0.7832,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.346630206712224,
+      "learning_rate": 1.4477211796246652e-06,
+      "loss": 1.135,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.825135706674494,
+      "learning_rate": 1.4745308310991958e-06,
+      "loss": 0.9713,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.43420167640646,
+      "learning_rate": 1.5013404825737266e-06,
+      "loss": 0.7911,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.532806480309018,
+      "learning_rate": 1.5281501340482574e-06,
+      "loss": 0.8331,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.278626484644239,
+      "learning_rate": 1.5549597855227882e-06,
+      "loss": 0.8776,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.439463443492862,
+      "learning_rate": 1.5817694369973192e-06,
+      "loss": 0.74,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.10429934785669,
+      "learning_rate": 1.60857908847185e-06,
+      "loss": 0.8675,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.345314363951537,
+      "learning_rate": 1.6353887399463808e-06,
+      "loss": 0.8439,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.1269723277515,
+      "learning_rate": 1.6621983914209116e-06,
+      "loss": 0.845,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.456338137217182,
+      "learning_rate": 1.6890080428954426e-06,
+      "loss": 0.7972,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.489158870377263,
+      "learning_rate": 1.7158176943699734e-06,
+      "loss": 0.7923,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.80424544627228,
+      "learning_rate": 1.7426273458445042e-06,
+      "loss": 0.8123,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 37.82298793701088,
+      "learning_rate": 1.769436997319035e-06,
+      "loss": 0.8784,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.249263690979113,
+      "learning_rate": 1.796246648793566e-06,
+      "loss": 0.9049,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.989586064709329,
+      "learning_rate": 1.8230563002680968e-06,
+      "loss": 0.8887,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.702841307709207,
+      "learning_rate": 1.8498659517426276e-06,
+      "loss": 0.7855,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.382513167616544,
+      "learning_rate": 1.8766756032171582e-06,
+      "loss": 0.687,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.242136408089937,
+      "learning_rate": 1.903485254691689e-06,
+      "loss": 0.741,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.349592045261627,
+      "learning_rate": 1.93029490616622e-06,
+      "loss": 0.7338,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.188201105667641,
+      "learning_rate": 1.9571045576407505e-06,
+      "loss": 0.8158,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.866357648660966,
+      "learning_rate": 1.9839142091152816e-06,
+      "loss": 0.887,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.875270769070864,
+      "learning_rate": 2.0107238605898126e-06,
+      "loss": 0.7808,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.08699977922816,
+      "learning_rate": 2.037533512064343e-06,
+      "loss": 0.7048,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.098242234307795,
+      "learning_rate": 2.064343163538874e-06,
+      "loss": 0.8221,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.458041748752553,
+      "learning_rate": 2.091152815013405e-06,
+      "loss": 0.8726,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.260028801031023,
+      "learning_rate": 2.1179624664879358e-06,
+      "loss": 0.7088,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.866965401922165,
+      "learning_rate": 2.1447721179624668e-06,
+      "loss": 0.6023,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4765225997217297,
+      "learning_rate": 2.1715817694369974e-06,
+      "loss": 0.7131,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.38138246921691,
+      "learning_rate": 2.1983914209115284e-06,
+      "loss": 0.7926,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.187421417640117,
+      "learning_rate": 2.2252010723860594e-06,
+      "loss": 0.7934,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.136754766658676,
+      "learning_rate": 2.25201072386059e-06,
+      "loss": 0.7477,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.755687978200916,
+      "learning_rate": 2.278820375335121e-06,
+      "loss": 0.7332,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.311205222380686,
+      "learning_rate": 2.3056300268096516e-06,
+      "loss": 0.7711,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 47.23629251427299,
+      "learning_rate": 2.3324396782841826e-06,
+      "loss": 0.7435,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.185466523855021,
+      "learning_rate": 2.359249329758713e-06,
+      "loss": 0.6627,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.194115492889853,
+      "learning_rate": 2.386058981233244e-06,
+      "loss": 0.6915,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.906853953124452,
+      "learning_rate": 2.4128686327077747e-06,
+      "loss": 0.6381,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.156501708348474,
+      "learning_rate": 2.4396782841823058e-06,
+      "loss": 0.8333,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.968312328171331,
+      "learning_rate": 2.4664879356568368e-06,
+      "loss": 0.8011,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.42247218203922,
+      "learning_rate": 2.4932975871313673e-06,
+      "loss": 0.7005,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.083347329061078,
+      "learning_rate": 2.5201072386058984e-06,
+      "loss": 0.7796,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.29321735877993,
+      "learning_rate": 2.5469168900804294e-06,
+      "loss": 0.6663,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.444084350825705,
+      "learning_rate": 2.57372654155496e-06,
+      "loss": 0.6594,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.740007331667963,
+      "learning_rate": 2.600536193029491e-06,
+      "loss": 0.7627,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.506739639358339,
+      "learning_rate": 2.6273458445040215e-06,
+      "loss": 0.7298,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.603218925372476,
+      "learning_rate": 2.6541554959785526e-06,
+      "loss": 0.7948,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.594828663150945,
+      "learning_rate": 2.6809651474530836e-06,
+      "loss": 0.7578,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.326060527209085,
+      "learning_rate": 2.707774798927614e-06,
+      "loss": 0.6855,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.009986821714675,
+      "learning_rate": 2.734584450402145e-06,
+      "loss": 0.6957,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.646574356647346,
+      "learning_rate": 2.761394101876676e-06,
+      "loss": 0.7444,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 12.325587448042153,
+      "learning_rate": 2.7882037533512068e-06,
+      "loss": 0.7068,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 24.069759828453563,
+      "learning_rate": 2.8150134048257378e-06,
+      "loss": 0.7758,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.922976189400241,
+      "learning_rate": 2.8418230563002683e-06,
+      "loss": 0.7331,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.830506608999528,
+      "learning_rate": 2.8686327077747994e-06,
+      "loss": 0.7465,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.7859298180436864,
+      "learning_rate": 2.8954423592493304e-06,
+      "loss": 0.6878,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.398580911734382,
+      "learning_rate": 2.9222520107238605e-06,
+      "loss": 0.8924,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.807616698425461,
+      "learning_rate": 2.9490616621983915e-06,
+      "loss": 0.6137,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2038362794527857,
+      "learning_rate": 2.975871313672922e-06,
+      "loss": 0.6327,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.340788539262372,
+      "learning_rate": 3.002680965147453e-06,
+      "loss": 0.6452,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.31421894148441,
+      "learning_rate": 3.029490616621984e-06,
+      "loss": 0.6857,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.561422842815732,
+      "learning_rate": 3.0563002680965147e-06,
+      "loss": 0.6709,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.956272073368498,
+      "learning_rate": 3.0831099195710457e-06,
+      "loss": 0.8003,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.838631017953537,
+      "learning_rate": 3.1099195710455763e-06,
+      "loss": 0.5902,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.475877282710105,
+      "learning_rate": 3.1367292225201073e-06,
+      "loss": 0.6314,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.039640285472611,
+      "learning_rate": 3.1635388739946383e-06,
+      "loss": 0.7525,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.738668529314289,
+      "learning_rate": 3.190348525469169e-06,
+      "loss": 0.6543,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 22.604838468964267,
+      "learning_rate": 3.2171581769437e-06,
+      "loss": 0.841,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.65521576681332,
+      "learning_rate": 3.243967828418231e-06,
+      "loss": 0.691,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.312841209423713,
+      "learning_rate": 3.2707774798927615e-06,
+      "loss": 0.7204,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.971287898996771,
+      "learning_rate": 3.2975871313672925e-06,
+      "loss": 0.6969,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.1214677391755,
+      "learning_rate": 3.324396782841823e-06,
+      "loss": 0.7931,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.847407829299312,
+      "learning_rate": 3.351206434316354e-06,
+      "loss": 0.8255,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.341070212436199,
+      "learning_rate": 3.378016085790885e-06,
+      "loss": 0.7762,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.124212238333406,
+      "learning_rate": 3.4048257372654157e-06,
+      "loss": 0.7003,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.631572103894618,
+      "learning_rate": 3.4316353887399467e-06,
+      "loss": 0.5839,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.950368074487955,
+      "learning_rate": 3.4584450402144778e-06,
+      "loss": 0.7791,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.613672846303096,
+      "learning_rate": 3.4852546916890083e-06,
+      "loss": 0.6245,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.619498261841813,
+      "learning_rate": 3.5120643431635393e-06,
+      "loss": 0.7681,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 37.40555563843904,
+      "learning_rate": 3.53887399463807e-06,
+      "loss": 0.6614,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.088517791966749,
+      "learning_rate": 3.565683646112601e-06,
+      "loss": 0.6845,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.984453588163199,
+      "learning_rate": 3.592493297587132e-06,
+      "loss": 0.6399,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.52046488796881,
+      "learning_rate": 3.6193029490616625e-06,
+      "loss": 0.8554,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.672637674109138,
+      "learning_rate": 3.6461126005361935e-06,
+      "loss": 0.7655,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.781878018107301,
+      "learning_rate": 3.6729222520107246e-06,
+      "loss": 0.5851,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.4701232931136685,
+      "learning_rate": 3.699731903485255e-06,
+      "loss": 0.7417,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.830469526671185,
+      "learning_rate": 3.726541554959786e-06,
+      "loss": 0.695,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.492661953015231,
+      "learning_rate": 3.7533512064343163e-06,
+      "loss": 0.7783,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 83.70279732518462,
+      "learning_rate": 3.7801608579088473e-06,
+      "loss": 0.6837,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.47605619909543,
+      "learning_rate": 3.806970509383378e-06,
+      "loss": 0.5808,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.790490998027301,
+      "learning_rate": 3.833780160857909e-06,
+      "loss": 0.5842,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.100203341503974,
+      "learning_rate": 3.86058981233244e-06,
+      "loss": 0.7716,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.606289216307983,
+      "learning_rate": 3.8873994638069705e-06,
+      "loss": 0.8139,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.486917434951024,
+      "learning_rate": 3.914209115281501e-06,
+      "loss": 0.7325,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3286804242221972,
+      "learning_rate": 3.9410187667560325e-06,
+      "loss": 0.707,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.989360776116529,
+      "learning_rate": 3.967828418230563e-06,
+      "loss": 0.6378,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.048700411875139,
+      "learning_rate": 3.994638069705094e-06,
+      "loss": 0.6641,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.840097942615614,
+      "learning_rate": 4.021447721179625e-06,
+      "loss": 0.7436,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.702358863308681,
+      "learning_rate": 4.048257372654156e-06,
+      "loss": 0.7175,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.623880792376663,
+      "learning_rate": 4.075067024128686e-06,
+      "loss": 0.5381,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.382865765654515,
+      "learning_rate": 4.101876675603218e-06,
+      "loss": 0.7426,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.943700190572278,
+      "learning_rate": 4.128686327077748e-06,
+      "loss": 0.595,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.847614293622386,
+      "learning_rate": 4.155495978552279e-06,
+      "loss": 0.7335,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.560853370592575,
+      "learning_rate": 4.18230563002681e-06,
+      "loss": 0.7155,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.730080423669779,
+      "learning_rate": 4.209115281501341e-06,
+      "loss": 0.7571,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.856450401507202,
+      "learning_rate": 4.2359249329758715e-06,
+      "loss": 0.6724,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.223801521982305,
+      "learning_rate": 4.262734584450403e-06,
+      "loss": 0.7268,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.18418339943064,
+      "learning_rate": 4.2895442359249335e-06,
+      "loss": 0.6496,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.464442751193876,
+      "learning_rate": 4.316353887399464e-06,
+      "loss": 0.6918,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.777784447853006,
+      "learning_rate": 4.343163538873995e-06,
+      "loss": 0.7863,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.897937792035265,
+      "learning_rate": 4.369973190348526e-06,
+      "loss": 0.6364,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.872942879974851,
+      "learning_rate": 4.396782841823057e-06,
+      "loss": 0.7069,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.882408151948045,
+      "learning_rate": 4.423592493297587e-06,
+      "loss": 0.6021,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.128248797266213,
+      "learning_rate": 4.450402144772119e-06,
+      "loss": 0.7614,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.946526131920074,
+      "learning_rate": 4.477211796246649e-06,
+      "loss": 0.613,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.34276773855277,
+      "learning_rate": 4.50402144772118e-06,
+      "loss": 0.5972,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.779463750874537,
+      "learning_rate": 4.530831099195711e-06,
+      "loss": 0.6905,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.609752714578294,
+      "learning_rate": 4.557640750670242e-06,
+      "loss": 0.6859,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.9881489056780035,
+      "learning_rate": 4.5844504021447725e-06,
+      "loss": 0.655,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.3763287506086215,
+      "learning_rate": 4.611260053619303e-06,
+      "loss": 0.6669,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.320565365519057,
+      "learning_rate": 4.638069705093834e-06,
+      "loss": 0.8001,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.341252690809038,
+      "learning_rate": 4.664879356568365e-06,
+      "loss": 0.7326,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.546239652204985,
+      "learning_rate": 4.691689008042896e-06,
+      "loss": 0.7558,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.323576391654177,
+      "learning_rate": 4.718498659517426e-06,
+      "loss": 0.6456,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.668584694348578,
+      "learning_rate": 4.745308310991958e-06,
+      "loss": 0.6618,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.980544984345709,
+      "learning_rate": 4.772117962466488e-06,
+      "loss": 0.6873,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.783060037805118,
+      "learning_rate": 4.798927613941019e-06,
+      "loss": 0.6822,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.62178286375343,
+      "learning_rate": 4.8257372654155495e-06,
+      "loss": 0.6271,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.177746665922566,
+      "learning_rate": 4.852546916890081e-06,
+      "loss": 0.6305,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.636223767406328,
+      "learning_rate": 4.8793565683646115e-06,
+      "loss": 0.6381,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.155446335906081,
+      "learning_rate": 4.906166219839142e-06,
+      "loss": 0.5507,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.955607973267833,
+      "learning_rate": 4.9329758713136735e-06,
+      "loss": 0.7265,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.51555391139486,
+      "learning_rate": 4.959785522788204e-06,
+      "loss": 0.5986,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.802698038699153,
+      "learning_rate": 4.986595174262735e-06,
+      "loss": 0.6181,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.987691712885555,
+      "learning_rate": 5.013404825737266e-06,
+      "loss": 0.5389,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.210445108958239,
+      "learning_rate": 5.040214477211797e-06,
+      "loss": 0.4754,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.780290072499368,
+      "learning_rate": 5.067024128686327e-06,
+      "loss": 0.6533,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.346659852625581,
+      "learning_rate": 5.093833780160859e-06,
+      "loss": 0.7583,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.919947529525403,
+      "learning_rate": 5.120643431635389e-06,
+      "loss": 0.6926,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.714100294046948,
+      "learning_rate": 5.14745308310992e-06,
+      "loss": 0.6167,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.995871768570122,
+      "learning_rate": 5.174262734584451e-06,
+      "loss": 0.6606,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.960115313014813,
+      "learning_rate": 5.201072386058982e-06,
+      "loss": 0.5961,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.172288698705337,
+      "learning_rate": 5.2278820375335125e-06,
+      "loss": 0.6542,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.570358223571498,
+      "learning_rate": 5.254691689008043e-06,
+      "loss": 0.715,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.757382037397196,
+      "learning_rate": 5.2815013404825745e-06,
+      "loss": 0.6798,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.849045025108514,
+      "learning_rate": 5.308310991957105e-06,
+      "loss": 0.69,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.108555690370628,
+      "learning_rate": 5.335120643431636e-06,
+      "loss": 0.7453,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.866678700750645,
+      "learning_rate": 5.361930294906167e-06,
+      "loss": 0.7416,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.599629117848702,
+      "learning_rate": 5.388739946380698e-06,
+      "loss": 0.7309,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.117707386104997,
+      "learning_rate": 5.415549597855228e-06,
+      "loss": 0.6115,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.5837553175339165,
+      "learning_rate": 5.44235924932976e-06,
+      "loss": 0.6819,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.614036798710173,
+      "learning_rate": 5.46916890080429e-06,
+      "loss": 0.7292,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.528638744552816,
+      "learning_rate": 5.495978552278821e-06,
+      "loss": 0.6584,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.000790359489285,
+      "learning_rate": 5.522788203753352e-06,
+      "loss": 0.6728,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.071535158275958,
+      "learning_rate": 5.549597855227883e-06,
+      "loss": 0.6738,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.677519936084339,
+      "learning_rate": 5.5764075067024135e-06,
+      "loss": 0.711,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.869560646406654,
+      "learning_rate": 5.603217158176945e-06,
+      "loss": 0.6493,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.547775805608433,
+      "learning_rate": 5.6300268096514755e-06,
+      "loss": 0.6969,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.83971451087532,
+      "learning_rate": 5.656836461126006e-06,
+      "loss": 0.5271,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.682905489184935,
+      "learning_rate": 5.683646112600537e-06,
+      "loss": 0.6446,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.630904842572068,
+      "learning_rate": 5.710455764075068e-06,
+      "loss": 0.644,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.1851823871925555,
+      "learning_rate": 5.737265415549599e-06,
+      "loss": 0.6862,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.882114378652549,
+      "learning_rate": 5.764075067024129e-06,
+      "loss": 0.6442,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.203777509738813,
+      "learning_rate": 5.790884718498661e-06,
+      "loss": 0.6043,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.393090594807507,
+      "learning_rate": 5.817694369973191e-06,
+      "loss": 0.5547,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.060893402595755,
+      "learning_rate": 5.844504021447721e-06,
+      "loss": 0.5992,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.937869569214672,
+      "learning_rate": 5.871313672922252e-06,
+      "loss": 0.6686,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.462343590239806,
+      "learning_rate": 5.898123324396783e-06,
+      "loss": 0.672,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.571907932731333,
+      "learning_rate": 5.924932975871314e-06,
+      "loss": 0.5726,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2134842948132296,
+      "learning_rate": 5.951742627345844e-06,
+      "loss": 0.7131,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.956747560744936,
+      "learning_rate": 5.978552278820376e-06,
+      "loss": 0.6505,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9476029375226598,
+      "learning_rate": 6.005361930294906e-06,
+      "loss": 0.7486,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.874356806839605,
+      "learning_rate": 6.032171581769437e-06,
+      "loss": 0.6175,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.768606555357628,
+      "learning_rate": 6.058981233243968e-06,
+      "loss": 0.7331,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.566960030267936,
+      "learning_rate": 6.085790884718499e-06,
+      "loss": 0.7321,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.052806329166555,
+      "learning_rate": 6.1126005361930295e-06,
+      "loss": 0.6674,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.582163987781495,
+      "learning_rate": 6.139410187667561e-06,
+      "loss": 0.6573,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.62099145250934,
+      "learning_rate": 6.1662198391420915e-06,
+      "loss": 0.5558,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.930559043752479,
+      "learning_rate": 6.193029490616622e-06,
+      "loss": 0.6005,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.00026510595833,
+      "learning_rate": 6.219839142091153e-06,
+      "loss": 0.5398,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.116825177455287,
+      "learning_rate": 6.246648793565684e-06,
+      "loss": 0.6816,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.3973814038637595,
+      "learning_rate": 6.273458445040215e-06,
+      "loss": 0.5372,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.831467752218199,
+      "learning_rate": 6.300268096514745e-06,
+      "loss": 0.6255,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.302030913857966,
+      "learning_rate": 6.327077747989277e-06,
+      "loss": 0.6713,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.607464962038097,
+      "learning_rate": 6.353887399463807e-06,
+      "loss": 0.5675,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.827444914792632,
+      "learning_rate": 6.380697050938338e-06,
+      "loss": 0.7888,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.220091518224656,
+      "learning_rate": 6.407506702412869e-06,
+      "loss": 0.6569,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.499799578613052,
+      "learning_rate": 6.4343163538874e-06,
+      "loss": 0.6178,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.191577508506997,
+      "learning_rate": 6.4611260053619305e-06,
+      "loss": 0.6698,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.682636092794401,
+      "learning_rate": 6.487935656836462e-06,
+      "loss": 0.6986,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.939138336810605,
+      "learning_rate": 6.5147453083109925e-06,
+      "loss": 0.5618,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.067103657243278,
+      "learning_rate": 6.541554959785523e-06,
+      "loss": 0.528,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.90252255113877,
+      "learning_rate": 6.5683646112600545e-06,
+      "loss": 0.7933,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.928996352784312,
+      "learning_rate": 6.595174262734585e-06,
+      "loss": 0.6507,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.783621283387384,
+      "learning_rate": 6.621983914209116e-06,
+      "loss": 0.6598,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.833685438804293,
+      "learning_rate": 6.648793565683646e-06,
+      "loss": 0.6444,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.535728333607127,
+      "learning_rate": 6.675603217158178e-06,
+      "loss": 0.5648,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.205243651849763,
+      "learning_rate": 6.702412868632708e-06,
+      "loss": 0.691,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.898061838511929,
+      "learning_rate": 6.729222520107239e-06,
+      "loss": 0.6423,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.30246786621298,
+      "learning_rate": 6.75603217158177e-06,
+      "loss": 0.6121,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.723679101784859,
+      "learning_rate": 6.782841823056301e-06,
+      "loss": 0.5457,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.277954764311047,
+      "learning_rate": 6.8096514745308315e-06,
+      "loss": 0.6126,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.9700703691890205,
+      "learning_rate": 6.836461126005363e-06,
+      "loss": 0.5052,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.096892438386883,
+      "learning_rate": 6.8632707774798935e-06,
+      "loss": 0.6107,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.704085867855948,
+      "learning_rate": 6.890080428954424e-06,
+      "loss": 0.7586,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.376990941554452,
+      "learning_rate": 6.9168900804289555e-06,
+      "loss": 0.5289,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.478698555670895,
+      "learning_rate": 6.943699731903486e-06,
+      "loss": 0.6046,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.973071812554681,
+      "learning_rate": 6.970509383378017e-06,
+      "loss": 0.5608,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.280611720727913,
+      "learning_rate": 6.997319034852548e-06,
+      "loss": 0.7344,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.755983496550087,
+      "learning_rate": 7.024128686327079e-06,
+      "loss": 0.6056,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.643931972713098,
+      "learning_rate": 7.050938337801609e-06,
+      "loss": 0.7485,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.488115771194938,
+      "learning_rate": 7.07774798927614e-06,
+      "loss": 0.5818,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.711639718578028,
+      "learning_rate": 7.104557640750671e-06,
+      "loss": 0.7112,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.287712231454048,
+      "learning_rate": 7.131367292225202e-06,
+      "loss": 0.6021,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.730546702597632,
+      "learning_rate": 7.1581769436997325e-06,
+      "loss": 0.56,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.337944584967096,
+      "learning_rate": 7.184986595174264e-06,
+      "loss": 0.5922,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.099238216761735,
+      "learning_rate": 7.2117962466487945e-06,
+      "loss": 0.6599,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.934961568303146,
+      "learning_rate": 7.238605898123325e-06,
+      "loss": 0.6676,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.032711716230696,
+      "learning_rate": 7.2654155495978565e-06,
+      "loss": 0.657,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.689262255823219,
+      "learning_rate": 7.292225201072387e-06,
+      "loss": 0.5862,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.785178282650161,
+      "learning_rate": 7.319034852546918e-06,
+      "loss": 0.6491,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.821788969460188,
+      "learning_rate": 7.345844504021449e-06,
+      "loss": 0.6536,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.944984428016963,
+      "learning_rate": 7.37265415549598e-06,
+      "loss": 0.7167,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.415842106268652,
+      "learning_rate": 7.39946380697051e-06,
+      "loss": 0.6494,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.900623321335587,
+      "learning_rate": 7.426273458445042e-06,
+      "loss": 0.6724,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.969809672924321,
+      "learning_rate": 7.453083109919572e-06,
+      "loss": 0.6307,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.209355802507353,
+      "learning_rate": 7.479892761394103e-06,
+      "loss": 0.6012,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.886306574143408,
+      "learning_rate": 7.506702412868633e-06,
+      "loss": 0.5593,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.224605489387802,
+      "learning_rate": 7.533512064343163e-06,
+      "loss": 0.5099,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.595006428172308,
+      "learning_rate": 7.560321715817695e-06,
+      "loss": 0.689,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.309185610509053,
+      "learning_rate": 7.587131367292225e-06,
+      "loss": 0.6023,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.655224729035256,
+      "learning_rate": 7.613941018766756e-06,
+      "loss": 0.6506,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.9514994612443415,
+      "learning_rate": 7.640750670241287e-06,
+      "loss": 0.5994,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.893982092000809,
+      "learning_rate": 7.667560321715819e-06,
+      "loss": 0.6591,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.701990817764118,
+      "learning_rate": 7.694369973190348e-06,
+      "loss": 0.6177,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.6935585994725315,
+      "learning_rate": 7.72117962466488e-06,
+      "loss": 0.627,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.099760015542169,
+      "learning_rate": 7.74798927613941e-06,
+      "loss": 0.7744,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.144719294407982,
+      "learning_rate": 7.774798927613941e-06,
+      "loss": 0.6352,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.092494071076228,
+      "learning_rate": 7.801608579088472e-06,
+      "loss": 0.6075,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.283671146189653,
+      "learning_rate": 7.828418230563002e-06,
+      "loss": 0.6502,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.829386942908994,
+      "learning_rate": 7.855227882037534e-06,
+      "loss": 0.5593,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.942041740739474,
+      "learning_rate": 7.882037533512065e-06,
+      "loss": 0.5154,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.346472025584439,
+      "learning_rate": 7.908847184986595e-06,
+      "loss": 0.5617,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.906043651221386,
+      "learning_rate": 7.935656836461126e-06,
+      "loss": 0.6432,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.257997927651868,
+      "learning_rate": 7.962466487935658e-06,
+      "loss": 0.6479,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.531588510964356,
+      "learning_rate": 7.989276139410187e-06,
+      "loss": 0.7339,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.266593317496769,
+      "learning_rate": 8.016085790884719e-06,
+      "loss": 0.5929,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.782270184496351,
+      "learning_rate": 8.04289544235925e-06,
+      "loss": 0.611,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.453066227859715,
+      "learning_rate": 8.06970509383378e-06,
+      "loss": 0.6242,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.776357115041182,
+      "learning_rate": 8.096514745308311e-06,
+      "loss": 0.7295,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.426526631147303,
+      "learning_rate": 8.123324396782843e-06,
+      "loss": 0.5753,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.521354060100845,
+      "learning_rate": 8.150134048257373e-06,
+      "loss": 0.5624,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.443044576389355,
+      "learning_rate": 8.176943699731904e-06,
+      "loss": 0.6004,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.729240330765151,
+      "learning_rate": 8.203753351206435e-06,
+      "loss": 0.6471,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.080597605108732,
+      "learning_rate": 8.230563002680965e-06,
+      "loss": 0.7951,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.082273783532623,
+      "learning_rate": 8.257372654155497e-06,
+      "loss": 0.6382,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.4761263015682,
+      "learning_rate": 8.284182305630028e-06,
+      "loss": 0.6589,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 38.822665565877905,
+      "learning_rate": 8.310991957104558e-06,
+      "loss": 0.7091,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.501721847691924,
+      "learning_rate": 8.33780160857909e-06,
+      "loss": 0.5705,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.335754551131912,
+      "learning_rate": 8.36461126005362e-06,
+      "loss": 0.6163,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.814939034106668,
+      "learning_rate": 8.39142091152815e-06,
+      "loss": 0.5948,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.123544199191704,
+      "learning_rate": 8.418230563002682e-06,
+      "loss": 0.4916,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.682665231968716,
+      "learning_rate": 8.445040214477213e-06,
+      "loss": 0.5765,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.272781192698249,
+      "learning_rate": 8.471849865951743e-06,
+      "loss": 0.6177,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.080147663828159,
+      "learning_rate": 8.498659517426274e-06,
+      "loss": 0.6806,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.285122200780735,
+      "learning_rate": 8.525469168900806e-06,
+      "loss": 0.5546,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.06057217531914,
+      "learning_rate": 8.552278820375336e-06,
+      "loss": 0.6564,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.174294061620935,
+      "learning_rate": 8.579088471849867e-06,
+      "loss": 0.6488,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.009628338325557,
+      "learning_rate": 8.605898123324398e-06,
+      "loss": 0.6305,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.7711758047406585,
+      "learning_rate": 8.632707774798928e-06,
+      "loss": 0.6179,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.737671234148584,
+      "learning_rate": 8.65951742627346e-06,
+      "loss": 0.6323,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.667617348015529,
+      "learning_rate": 8.68632707774799e-06,
+      "loss": 0.6172,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.644205868130122,
+      "learning_rate": 8.71313672922252e-06,
+      "loss": 0.4951,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.107166365624764,
+      "learning_rate": 8.739946380697052e-06,
+      "loss": 0.6634,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.359083841916988,
+      "learning_rate": 8.766756032171582e-06,
+      "loss": 0.6112,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.228021853281659,
+      "learning_rate": 8.793565683646113e-06,
+      "loss": 0.5945,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.7137110349696165,
+      "learning_rate": 8.820375335120645e-06,
+      "loss": 0.6237,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.886624583394349,
+      "learning_rate": 8.847184986595175e-06,
+      "loss": 0.6626,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.686570722659138,
+      "learning_rate": 8.873994638069706e-06,
+      "loss": 0.8064,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.707893151339013,
+      "learning_rate": 8.900804289544237e-06,
+      "loss": 0.7414,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.1969220228074136,
+      "learning_rate": 8.927613941018767e-06,
+      "loss": 0.5951,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.27343543873902,
+      "learning_rate": 8.954423592493299e-06,
+      "loss": 0.5621,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.336482198307235,
+      "learning_rate": 8.98123324396783e-06,
+      "loss": 0.7102,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.787649081395831,
+      "learning_rate": 9.00804289544236e-06,
+      "loss": 0.4602,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.081675243882314,
+      "learning_rate": 9.034852546916891e-06,
+      "loss": 0.6017,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.8915471478428065,
+      "learning_rate": 9.061662198391423e-06,
+      "loss": 0.5429,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.792854975350584,
+      "learning_rate": 9.088471849865952e-06,
+      "loss": 0.6347,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.466496646116509,
+      "learning_rate": 9.115281501340484e-06,
+      "loss": 0.6164,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.984008812827013,
+      "learning_rate": 9.142091152815015e-06,
+      "loss": 0.6224,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.159902376201595,
+      "learning_rate": 9.168900804289545e-06,
+      "loss": 0.7589,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.658039077333587,
+      "learning_rate": 9.195710455764075e-06,
+      "loss": 0.603,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.514609283934334,
+      "learning_rate": 9.222520107238606e-06,
+      "loss": 0.8015,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.357575803414787,
+      "learning_rate": 9.249329758713138e-06,
+      "loss": 0.681,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.036409049828143,
+      "learning_rate": 9.276139410187667e-06,
+      "loss": 0.6576,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1219113517589774,
+      "learning_rate": 9.302949061662199e-06,
+      "loss": 0.6888,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.414219794216217,
+      "learning_rate": 9.32975871313673e-06,
+      "loss": 0.6474,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.104169966120765,
+      "learning_rate": 9.35656836461126e-06,
+      "loss": 0.6155,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.534882754338667,
+      "learning_rate": 9.383378016085791e-06,
+      "loss": 0.5907,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.591310199372394,
+      "learning_rate": 9.410187667560323e-06,
+      "loss": 0.7675,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.31656620734291,
+      "learning_rate": 9.436997319034853e-06,
+      "loss": 0.7031,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.5793514868471155,
+      "learning_rate": 9.463806970509384e-06,
+      "loss": 0.548,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.557286358655546,
+      "learning_rate": 9.490616621983915e-06,
+      "loss": 0.6385,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.321111049086398,
+      "learning_rate": 9.517426273458445e-06,
+      "loss": 0.5393,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.742584500670051,
+      "learning_rate": 9.544235924932977e-06,
+      "loss": 0.6066,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.89503313015838,
+      "learning_rate": 9.571045576407506e-06,
+      "loss": 0.6801,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.380443592171769,
+      "learning_rate": 9.597855227882038e-06,
+      "loss": 0.6048,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.888932222543188,
+      "learning_rate": 9.62466487935657e-06,
+      "loss": 0.5796,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.79289494871947,
+      "learning_rate": 9.651474530831099e-06,
+      "loss": 0.6949,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.914785604794807,
+      "learning_rate": 9.67828418230563e-06,
+      "loss": 0.5743,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.352011455063279,
+      "learning_rate": 9.705093833780162e-06,
+      "loss": 0.602,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5009207733725414,
+      "learning_rate": 9.731903485254692e-06,
+      "loss": 0.7631,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.1167555658535395,
+      "learning_rate": 9.758713136729223e-06,
+      "loss": 0.5997,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.014390252303107,
+      "learning_rate": 9.785522788203754e-06,
+      "loss": 0.7997,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.759891368407603,
+      "learning_rate": 9.812332439678284e-06,
+      "loss": 0.7848,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0267284989468917,
+      "learning_rate": 9.839142091152816e-06,
+      "loss": 0.7327,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.197991772764897,
+      "learning_rate": 9.865951742627347e-06,
+      "loss": 0.5714,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.826860218068223,
+      "learning_rate": 9.892761394101877e-06,
+      "loss": 0.6445,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.389424527430027,
+      "learning_rate": 9.919571045576408e-06,
+      "loss": 0.5913,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.61157252107286,
+      "learning_rate": 9.94638069705094e-06,
+      "loss": 0.5576,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.8998890640728145,
+      "learning_rate": 9.97319034852547e-06,
+      "loss": 0.6046,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.57503736047468,
+      "learning_rate": 1e-05,
+      "loss": 0.5406,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.508692610966933,
+      "learning_rate": 9.999999829761054e-06,
+      "loss": 0.5516,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.630101750583523,
+      "learning_rate": 9.999999319044228e-06,
+      "loss": 0.4942,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.674969624315066,
+      "learning_rate": 9.999998467849553e-06,
+      "loss": 0.5867,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.346691080924932,
+      "learning_rate": 9.99999727617709e-06,
+      "loss": 0.5602,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.883115629231502,
+      "learning_rate": 9.999995744026921e-06,
+      "loss": 0.6671,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.116418848773912,
+      "learning_rate": 9.99999387139915e-06,
+      "loss": 0.531,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.436194283510765,
+      "learning_rate": 9.999991658293901e-06,
+      "loss": 0.577,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.464967582323264,
+      "learning_rate": 9.99998910471133e-06,
+      "loss": 0.5239,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.946438034914115,
+      "learning_rate": 9.999986210651606e-06,
+      "loss": 0.5536,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.654755365283333,
+      "learning_rate": 9.999982976114929e-06,
+      "loss": 0.615,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 14.32130171320425,
+      "learning_rate": 9.999979401101519e-06,
+      "loss": 0.5859,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.441716840116557,
+      "learning_rate": 9.999975485611618e-06,
+      "loss": 0.5586,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.612735446400066,
+      "learning_rate": 9.999971229645494e-06,
+      "loss": 0.7339,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.304343618110067,
+      "learning_rate": 9.999966633203437e-06,
+      "loss": 0.6948,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.4109404232007146,
+      "learning_rate": 9.999961696285758e-06,
+      "loss": 0.5954,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.740002099249732,
+      "learning_rate": 9.999956418892797e-06,
+      "loss": 0.5481,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.123550427285188,
+      "learning_rate": 9.999950801024908e-06,
+      "loss": 0.7017,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.5064940208286455,
+      "learning_rate": 9.999944842682479e-06,
+      "loss": 0.6738,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.35906446724027,
+      "learning_rate": 9.999938543865913e-06,
+      "loss": 0.5245,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.425287996943937,
+      "learning_rate": 9.999931904575638e-06,
+      "loss": 0.6223,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.607504298554862,
+      "learning_rate": 9.999924924812108e-06,
+      "loss": 0.6252,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.763349819871314,
+      "learning_rate": 9.999917604575797e-06,
+      "loss": 0.6508,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.056584840058331,
+      "learning_rate": 9.999909943867206e-06,
+      "loss": 0.6476,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.888682118393513,
+      "learning_rate": 9.999901942686852e-06,
+      "loss": 0.5858,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.0162929460993775,
+      "learning_rate": 9.999893601035284e-06,
+      "loss": 0.61,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.697253829264872,
+      "learning_rate": 9.999884918913068e-06,
+      "loss": 0.7124,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.456822094412279,
+      "learning_rate": 9.999875896320797e-06,
+      "loss": 0.6486,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.977495078103326,
+      "learning_rate": 9.999866533259081e-06,
+      "loss": 0.5723,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.304168392568399,
+      "learning_rate": 9.999856829728564e-06,
+      "loss": 0.6717,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.399703319137561,
+      "learning_rate": 9.999846785729903e-06,
+      "loss": 0.6301,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.099028354468247,
+      "learning_rate": 9.999836401263782e-06,
+      "loss": 0.6008,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.149293051171149,
+      "learning_rate": 9.99982567633091e-06,
+      "loss": 0.5994,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.627906580036083,
+      "learning_rate": 9.999814610932014e-06,
+      "loss": 0.7313,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.406228845741484,
+      "learning_rate": 9.99980320506785e-06,
+      "loss": 0.6869,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.827631391001245,
+      "learning_rate": 9.999791458739194e-06,
+      "loss": 0.5624,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.509184051903391,
+      "learning_rate": 9.999779371946848e-06,
+      "loss": 0.6053,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.697578348158185,
+      "learning_rate": 9.99976694469163e-06,
+      "loss": 0.5893,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.404641391369749,
+      "learning_rate": 9.99975417697439e-06,
+      "loss": 0.5777,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.725571730575014,
+      "learning_rate": 9.999741068795998e-06,
+      "loss": 0.5314,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.007245546301691,
+      "learning_rate": 9.999727620157345e-06,
+      "loss": 0.5816,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.326645689732777,
+      "learning_rate": 9.999713831059348e-06,
+      "loss": 0.5842,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.998121192246539,
+      "learning_rate": 9.999699701502943e-06,
+      "loss": 0.5849,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.663436888915598,
+      "learning_rate": 9.999685231489096e-06,
+      "loss": 0.7568,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.059592006309952,
+      "learning_rate": 9.99967042101879e-06,
+      "loss": 0.7873,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.908830838304987,
+      "learning_rate": 9.999655270093034e-06,
+      "loss": 0.8212,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.6760900060233075,
+      "learning_rate": 9.99963977871286e-06,
+      "loss": 0.6449,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.41183454770552,
+      "learning_rate": 9.999623946879322e-06,
+      "loss": 0.5486,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.806285805797526,
+      "learning_rate": 9.999607774593501e-06,
+      "loss": 0.6804,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.48624536935526,
+      "learning_rate": 9.999591261856494e-06,
+      "loss": 0.7242,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.616725355649693,
+      "learning_rate": 9.999574408669428e-06,
+      "loss": 0.5885,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.875119753036814,
+      "learning_rate": 9.999557215033451e-06,
+      "loss": 0.48,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.77407552318965,
+      "learning_rate": 9.999539680949732e-06,
+      "loss": 0.6035,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.235439854225651,
+      "learning_rate": 9.999521806419466e-06,
+      "loss": 0.6623,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.042980505259546,
+      "learning_rate": 9.99950359144387e-06,
+      "loss": 0.5613,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.1212667939115475,
+      "learning_rate": 9.999485036024186e-06,
+      "loss": 0.5336,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.314472523098836,
+      "learning_rate": 9.999466140161676e-06,
+      "loss": 0.6666,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.928620243077904,
+      "learning_rate": 9.999446903857625e-06,
+      "loss": 0.6433,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9021700746690315,
+      "learning_rate": 9.999427327113347e-06,
+      "loss": 0.7265,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.763986943023307,
+      "learning_rate": 9.99940740993017e-06,
+      "loss": 0.5564,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.77631666370379,
+      "learning_rate": 9.999387152309454e-06,
+      "loss": 0.5581,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.600486284156332,
+      "learning_rate": 9.999366554252578e-06,
+      "loss": 0.5306,
+      "step": 434
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 28.147426128019696,
+      "learning_rate": 9.999345615760944e-06,
+      "loss": 0.7387,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.75362960895393,
+      "learning_rate": 9.999324336835979e-06,
+      "loss": 0.5623,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.512037133897585,
+      "learning_rate": 9.99930271747913e-06,
+      "loss": 0.605,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.844037321191896,
+      "learning_rate": 9.99928075769187e-06,
+      "loss": 0.6372,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.126757790100472,
+      "learning_rate": 9.999258457475696e-06,
+      "loss": 0.5878,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.081805091319472,
+      "learning_rate": 9.999235816832123e-06,
+      "loss": 0.6539,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.794789929854415,
+      "learning_rate": 9.999212835762695e-06,
+      "loss": 0.571,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.6750568461153446,
+      "learning_rate": 9.999189514268977e-06,
+      "loss": 0.6135,
+      "step": 442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.554140259897694,
+      "learning_rate": 9.999165852352556e-06,
+      "loss": 0.5967,
+      "step": 443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.1578434986179875,
+      "learning_rate": 9.999141850015046e-06,
+      "loss": 0.7161,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.433736482752552,
+      "learning_rate": 9.999117507258078e-06,
+      "loss": 0.5206,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9701611019362142,
+      "learning_rate": 9.999092824083312e-06,
+      "loss": 0.7354,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.702709274401544,
+      "learning_rate": 9.999067800492427e-06,
+      "loss": 0.6557,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.101387214111046,
+      "learning_rate": 9.99904243648713e-06,
+      "loss": 0.6779,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.797911980120324,
+      "learning_rate": 9.999016732069144e-06,
+      "loss": 0.5692,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.393460464132476,
+      "learning_rate": 9.998990687240222e-06,
+      "loss": 0.5575,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.511521628075598,
+      "learning_rate": 9.998964302002137e-06,
+      "loss": 0.5755,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.799276319719473,
+      "learning_rate": 9.998937576356687e-06,
+      "loss": 0.6061,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.125316212333039,
+      "learning_rate": 9.99891051030569e-06,
+      "loss": 0.6055,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.229142301755386,
+      "learning_rate": 9.99888310385099e-06,
+      "loss": 0.644,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.467845611992761,
+      "learning_rate": 9.99885535699445e-06,
+      "loss": 0.5404,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.634969176529141,
+      "learning_rate": 9.998827269737964e-06,
+      "loss": 0.5708,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7679322227727652,
+      "learning_rate": 9.998798842083445e-06,
+      "loss": 0.6851,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.085467735069233,
+      "learning_rate": 9.998770074032827e-06,
+      "loss": 0.626,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.699838586324562,
+      "learning_rate": 9.998740965588065e-06,
+      "loss": 0.6082,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.948168783050038,
+      "learning_rate": 9.998711516751148e-06,
+      "loss": 0.6026,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.951625587396531,
+      "learning_rate": 9.998681727524077e-06,
+      "loss": 0.7167,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.593931790007032,
+      "learning_rate": 9.998651597908884e-06,
+      "loss": 0.5757,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.379987647363585,
+      "learning_rate": 9.998621127907617e-06,
+      "loss": 0.6377,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.576820068062017,
+      "learning_rate": 9.998590317522352e-06,
+      "loss": 0.6413,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.547374724498416,
+      "learning_rate": 9.998559166755187e-06,
+      "loss": 0.5492,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 26.143736096834573,
+      "learning_rate": 9.998527675608246e-06,
+      "loss": 0.639,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.380915279240922,
+      "learning_rate": 9.998495844083669e-06,
+      "loss": 0.4842,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.41224690349593,
+      "learning_rate": 9.998463672183625e-06,
+      "loss": 0.634,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6096183221806233,
+      "learning_rate": 9.998431159910307e-06,
+      "loss": 0.8317,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.988049910880573,
+      "learning_rate": 9.998398307265927e-06,
+      "loss": 0.5578,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.19220076931391,
+      "learning_rate": 9.998365114252722e-06,
+      "loss": 0.5973,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.350325057325014,
+      "learning_rate": 9.998331580872952e-06,
+      "loss": 0.5906,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.232421370511597,
+      "learning_rate": 9.998297707128904e-06,
+      "loss": 0.6014,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.766162004696768,
+      "learning_rate": 9.99826349302288e-06,
+      "loss": 0.489,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.20984008302196,
+      "learning_rate": 9.998228938557212e-06,
+      "loss": 0.615,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.619887837293652,
+      "learning_rate": 9.998194043734253e-06,
+      "loss": 0.6525,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 17.99310345193075,
+      "learning_rate": 9.998158808556378e-06,
+      "loss": 0.6017,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.500697688052427,
+      "learning_rate": 9.998123233025989e-06,
+      "loss": 0.7444,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.084365115150798,
+      "learning_rate": 9.998087317145505e-06,
+      "loss": 0.5878,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.248405440067243,
+      "learning_rate": 9.998051060917374e-06,
+      "loss": 0.648,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.108671721642725,
+      "learning_rate": 9.998014464344065e-06,
+      "loss": 0.5467,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6549040214134556,
+      "learning_rate": 9.997977527428069e-06,
+      "loss": 0.6971,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.614005861556339,
+      "learning_rate": 9.997940250171903e-06,
+      "loss": 0.5932,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.340550255323108,
+      "learning_rate": 9.997902632578103e-06,
+      "loss": 0.6271,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.407805981996987,
+      "learning_rate": 9.997864674649231e-06,
+      "loss": 0.5665,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.171318210707566,
+      "learning_rate": 9.997826376387874e-06,
+      "loss": 0.5513,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.6798027717988875,
+      "learning_rate": 9.997787737796636e-06,
+      "loss": 0.6414,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.078431935132171,
+      "learning_rate": 9.997748758878153e-06,
+      "loss": 0.6225,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.33892287879157,
+      "learning_rate": 9.997709439635076e-06,
+      "loss": 0.5822,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.622148820200381,
+      "learning_rate": 9.997669780070084e-06,
+      "loss": 0.5854,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.911999079524254,
+      "learning_rate": 9.997629780185876e-06,
+      "loss": 0.551,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.272507223573783,
+      "learning_rate": 9.997589439985178e-06,
+      "loss": 0.6966,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.726080021226044,
+      "learning_rate": 9.997548759470735e-06,
+      "loss": 0.5846,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.229597418131615,
+      "learning_rate": 9.997507738645318e-06,
+      "loss": 0.6599,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.788804825633731,
+      "learning_rate": 9.99746637751172e-06,
+      "loss": 0.7452,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.555050817961806,
+      "learning_rate": 9.997424676072758e-06,
+      "loss": 0.5881,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0272236992017003,
+      "learning_rate": 9.997382634331271e-06,
+      "loss": 0.7223,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.789348154957315,
+      "learning_rate": 9.997340252290121e-06,
+      "loss": 0.5443,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.8272306825310025,
+      "learning_rate": 9.9972975299522e-06,
+      "loss": 0.6229,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.803005175532133,
+      "learning_rate": 9.997254467320407e-06,
+      "loss": 0.6387,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.122541422198903,
+      "learning_rate": 9.997211064397684e-06,
+      "loss": 0.6494,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.986571555784634,
+      "learning_rate": 9.99716732118698e-06,
+      "loss": 0.5728,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.077846333594792,
+      "learning_rate": 9.997123237691278e-06,
+      "loss": 0.6128,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.149989254225549,
+      "learning_rate": 9.997078813913577e-06,
+      "loss": 0.6121,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 23.778689259682807,
+      "learning_rate": 9.997034049856904e-06,
+      "loss": 0.7438,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.270869381742254,
+      "learning_rate": 9.996988945524305e-06,
+      "loss": 0.5707,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.665436709703485,
+      "learning_rate": 9.996943500918854e-06,
+      "loss": 0.5272,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.93412076881617,
+      "learning_rate": 9.996897716043643e-06,
+      "loss": 0.6326,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.62009804828515,
+      "learning_rate": 9.996851590901794e-06,
+      "loss": 0.7097,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 21.776433221837582,
+      "learning_rate": 9.996805125496441e-06,
+      "loss": 0.4739,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.122435749567897,
+      "learning_rate": 9.996758319830755e-06,
+      "loss": 0.6396,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.495987379441264,
+      "learning_rate": 9.99671117390792e-06,
+      "loss": 0.669,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.698766997167759,
+      "learning_rate": 9.996663687731145e-06,
+      "loss": 0.6147,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.818008943810367,
+      "learning_rate": 9.996615861303666e-06,
+      "loss": 0.6349,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.248520448083664,
+      "learning_rate": 9.996567694628742e-06,
+      "loss": 0.5428,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 101.35172511706202,
+      "learning_rate": 9.996519187709649e-06,
+      "loss": 0.6169,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.900135653342379,
+      "learning_rate": 9.996470340549689e-06,
+      "loss": 0.69,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.685443213797643,
+      "learning_rate": 9.996421153152192e-06,
+      "loss": 0.6541,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.436392883080101,
+      "learning_rate": 9.996371625520505e-06,
+      "loss": 0.6846,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.270404054882102,
+      "learning_rate": 9.996321757658002e-06,
+      "loss": 0.5831,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.268327083720806,
+      "learning_rate": 9.996271549568078e-06,
+      "loss": 0.6085,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.00468971401466,
+      "learning_rate": 9.996221001254154e-06,
+      "loss": 0.6499,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.075969410807537,
+      "learning_rate": 9.996170112719669e-06,
+      "loss": 0.5845,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.580520369334339,
+      "learning_rate": 9.996118883968088e-06,
+      "loss": 0.6192,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.9398664174691955,
+      "learning_rate": 9.996067315002904e-06,
+      "loss": 0.4831,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.50882508596984,
+      "learning_rate": 9.996015405827625e-06,
+      "loss": 0.654,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.39430747599092,
+      "learning_rate": 9.995963156445787e-06,
+      "loss": 0.4961,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.48263443824669,
+      "learning_rate": 9.995910566860946e-06,
+      "loss": 0.606,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.9080771361308715,
+      "learning_rate": 9.995857637076684e-06,
+      "loss": 0.4975,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.50930306399065,
+      "learning_rate": 9.99580436709661e-06,
+      "loss": 0.7414,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.345893788824842,
+      "learning_rate": 9.995750756924342e-06,
+      "loss": 0.7369,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.11371100916381,
+      "learning_rate": 9.99569680656354e-06,
+      "loss": 0.697,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.969892269802246,
+      "learning_rate": 9.995642516017872e-06,
+      "loss": 0.618,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.700095774633456,
+      "learning_rate": 9.995587885291036e-06,
+      "loss": 0.5522,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.748319314742104,
+      "learning_rate": 9.995532914386754e-06,
+      "loss": 0.6077,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.821902264500286,
+      "learning_rate": 9.995477603308768e-06,
+      "loss": 0.5279,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.533381366319548,
+      "learning_rate": 9.995421952060845e-06,
+      "loss": 0.5349,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.830663033052675,
+      "learning_rate": 9.995365960646773e-06,
+      "loss": 0.6222,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.591031269436857,
+      "learning_rate": 9.995309629070367e-06,
+      "loss": 0.6271,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.663607794438116,
+      "learning_rate": 9.995252957335459e-06,
+      "loss": 0.6622,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.230638995643542,
+      "learning_rate": 9.995195945445914e-06,
+      "loss": 0.5175,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.675559830379535,
+      "learning_rate": 9.99513859340561e-06,
+      "loss": 0.6857,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.936348716975601,
+      "learning_rate": 9.995080901218455e-06,
+      "loss": 0.6407,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.455555156222862,
+      "learning_rate": 9.995022868888373e-06,
+      "loss": 0.6763,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.678262892281809,
+      "learning_rate": 9.994964496419322e-06,
+      "loss": 0.5899,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.205284494053673,
+      "learning_rate": 9.994905783815272e-06,
+      "loss": 0.5754,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.455859916105756,
+      "learning_rate": 9.994846731080224e-06,
+      "loss": 0.6098,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.7548875462569855,
+      "learning_rate": 9.994787338218197e-06,
+      "loss": 0.6317,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 14.025150942616104,
+      "learning_rate": 9.994727605233237e-06,
+      "loss": 0.6267,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.13387259847279,
+      "learning_rate": 9.994667532129412e-06,
+      "loss": 0.5627,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.334962357851314,
+      "learning_rate": 9.99460711891081e-06,
+      "loss": 0.641,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.551983236376067,
+      "learning_rate": 9.994546365581548e-06,
+      "loss": 0.5871,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.540640238205132,
+      "learning_rate": 9.994485272145761e-06,
+      "loss": 0.7097,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.905467986236814,
+      "learning_rate": 9.99442383860761e-06,
+      "loss": 0.7604,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.091161431028977,
+      "learning_rate": 9.994362064971279e-06,
+      "loss": 0.6141,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.544148068364372,
+      "learning_rate": 9.994299951240972e-06,
+      "loss": 0.6186,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.319876847105103,
+      "learning_rate": 9.994237497420923e-06,
+      "loss": 0.7262,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.714070498637281,
+      "learning_rate": 9.99417470351538e-06,
+      "loss": 0.6105,
+      "step": 558
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.883475224870057,
+      "learning_rate": 9.994111569528621e-06,
+      "loss": 0.5634,
+      "step": 559
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.091893603867726,
+      "learning_rate": 9.994048095464947e-06,
+      "loss": 0.6174,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.803145601105623,
+      "learning_rate": 9.993984281328677e-06,
+      "loss": 0.5611,
+      "step": 561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.764954191317917,
+      "learning_rate": 9.993920127124159e-06,
+      "loss": 0.6045,
+      "step": 562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.472207648704204,
+      "learning_rate": 9.993855632855762e-06,
+      "loss": 0.5652,
+      "step": 563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 22.428887385881666,
+      "learning_rate": 9.993790798527875e-06,
+      "loss": 0.5271,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.03127005526951,
+      "learning_rate": 9.993725624144915e-06,
+      "loss": 0.6203,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.179064383314232,
+      "learning_rate": 9.993660109711319e-06,
+      "loss": 0.5761,
+      "step": 566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.061340776758832,
+      "learning_rate": 9.99359425523155e-06,
+      "loss": 0.711,
+      "step": 567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.690077215463388,
+      "learning_rate": 9.993528060710089e-06,
+      "loss": 0.6916,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.89844488133037,
+      "learning_rate": 9.993461526151447e-06,
+      "loss": 0.579,
+      "step": 569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.992202414496631,
+      "learning_rate": 9.993394651560154e-06,
+      "loss": 0.7041,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.335257605620757,
+      "learning_rate": 9.993327436940764e-06,
+      "loss": 0.5828,
+      "step": 571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.759940708542796,
+      "learning_rate": 9.99325988229785e-06,
+      "loss": 0.5299,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.833267067326899,
+      "learning_rate": 9.993191987636019e-06,
+      "loss": 0.5819,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.10125885902509,
+      "learning_rate": 9.99312375295989e-06,
+      "loss": 0.5513,
+      "step": 574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.277848280240189,
+      "learning_rate": 9.993055178274109e-06,
+      "loss": 0.6117,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.916171192407227,
+      "learning_rate": 9.992986263583347e-06,
+      "loss": 0.6124,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.318005413658978,
+      "learning_rate": 9.992917008892298e-06,
+      "loss": 0.6228,
+      "step": 577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.367641662037906,
+      "learning_rate": 9.992847414205675e-06,
+      "loss": 0.6885,
+      "step": 578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.632876528481077,
+      "learning_rate": 9.99277747952822e-06,
+      "loss": 0.5519,
+      "step": 579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.574151661803853,
+      "learning_rate": 9.992707204864694e-06,
+      "loss": 0.5908,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.288795133967565,
+      "learning_rate": 9.992636590219881e-06,
+      "loss": 0.7148,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.755003039330308,
+      "learning_rate": 9.992565635598593e-06,
+      "loss": 0.5312,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.989837000218207,
+      "learning_rate": 9.99249434100566e-06,
+      "loss": 0.5815,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.414807741744693,
+      "learning_rate": 9.992422706445935e-06,
+      "loss": 0.6868,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.821902744282863,
+      "learning_rate": 9.992350731924298e-06,
+      "loss": 0.6342,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.788623104817612,
+      "learning_rate": 9.992278417445649e-06,
+      "loss": 0.6197,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.238280296607476,
+      "learning_rate": 9.992205763014913e-06,
+      "loss": 0.6067,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.206766554398455,
+      "learning_rate": 9.992132768637037e-06,
+      "loss": 0.7691,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.2338881629891985,
+      "learning_rate": 9.992059434316992e-06,
+      "loss": 0.5727,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.938791473669545,
+      "learning_rate": 9.991985760059773e-06,
+      "loss": 0.6206,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.402093249877145,
+      "learning_rate": 9.991911745870395e-06,
+      "loss": 0.5762,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.2199943594724125,
+      "learning_rate": 9.991837391753898e-06,
+      "loss": 0.5733,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.703324154383682,
+      "learning_rate": 9.991762697715348e-06,
+      "loss": 0.6413,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.367059234566039,
+      "learning_rate": 9.991687663759826e-06,
+      "loss": 0.6064,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.057241137766459,
+      "learning_rate": 9.991612289892446e-06,
+      "loss": 0.6639,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.593262955917416,
+      "learning_rate": 9.991536576118339e-06,
+      "loss": 0.6553,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.950069094144177,
+      "learning_rate": 9.991460522442663e-06,
+      "loss": 0.6513,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.335584158709535,
+      "learning_rate": 9.991384128870591e-06,
+      "loss": 0.6658,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.781163808366034,
+      "learning_rate": 9.991307395407332e-06,
+      "loss": 0.7118,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.842919744220737,
+      "learning_rate": 9.991230322058107e-06,
+      "loss": 0.755,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.832451989115855,
+      "learning_rate": 9.991152908828167e-06,
+      "loss": 0.6217,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.225654126585404,
+      "learning_rate": 9.991075155722782e-06,
+      "loss": 0.6193,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.238046552560933,
+      "learning_rate": 9.990997062747245e-06,
+      "loss": 0.688,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.749390667978442,
+      "learning_rate": 9.990918629906877e-06,
+      "loss": 0.5609,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.077548902148913,
+      "learning_rate": 9.990839857207017e-06,
+      "loss": 0.6294,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.404411601602331,
+      "learning_rate": 9.99076074465303e-06,
+      "loss": 0.5588,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.132160683682487,
+      "learning_rate": 9.990681292250301e-06,
+      "loss": 0.6405,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.203619841627564,
+      "learning_rate": 9.990601500004243e-06,
+      "loss": 0.5843,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4464117329060797,
+      "learning_rate": 9.99052136792029e-06,
+      "loss": 0.7397,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.517669933190403,
+      "learning_rate": 9.990440896003894e-06,
+      "loss": 0.5096,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.975771603121325,
+      "learning_rate": 9.990360084260541e-06,
+      "loss": 0.4772,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.090423453531876,
+      "learning_rate": 9.99027893269573e-06,
+      "loss": 0.5617,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.8604339155863245,
+      "learning_rate": 9.990197441314989e-06,
+      "loss": 0.5652,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8058851498127744,
+      "learning_rate": 9.990115610123863e-06,
+      "loss": 0.6558,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.252825627456938,
+      "learning_rate": 9.99003343912793e-06,
+      "loss": 0.6908,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.367177471725237,
+      "learning_rate": 9.989950928332783e-06,
+      "loss": 0.5232,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.377773676619592,
+      "learning_rate": 9.98986807774404e-06,
+      "loss": 0.6032,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.077306663856659,
+      "learning_rate": 9.989784887367345e-06,
+      "loss": 0.5744,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.445370529151184,
+      "learning_rate": 9.989701357208359e-06,
+      "loss": 0.6226,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.84356085106096,
+      "learning_rate": 9.989617487272774e-06,
+      "loss": 0.6393,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.948412413370538,
+      "learning_rate": 9.989533277566298e-06,
+      "loss": 0.6352,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.185446683982368,
+      "learning_rate": 9.989448728094669e-06,
+      "loss": 0.6073,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.454669930058769,
+      "learning_rate": 9.98936383886364e-06,
+      "loss": 0.6077,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.42602368333551,
+      "learning_rate": 9.989278609878993e-06,
+      "loss": 0.6335,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.202664392332381,
+      "learning_rate": 9.989193041146535e-06,
+      "loss": 0.5977,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.321857300508276,
+      "learning_rate": 9.98910713267209e-06,
+      "loss": 0.6067,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.708213802837024,
+      "learning_rate": 9.989020884461507e-06,
+      "loss": 0.5397,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.564404988033187,
+      "learning_rate": 9.988934296520663e-06,
+      "loss": 0.588,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0765278807721934,
+      "learning_rate": 9.98884736885545e-06,
+      "loss": 0.7345,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.575621946945146,
+      "learning_rate": 9.988760101471788e-06,
+      "loss": 0.5638,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.54867151142498,
+      "learning_rate": 9.988672494375623e-06,
+      "loss": 0.688,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.259875177938298,
+      "learning_rate": 9.988584547572916e-06,
+      "loss": 0.5889,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.624753269593689,
+      "learning_rate": 9.988496261069661e-06,
+      "loss": 0.5516,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.7229773415359,
+      "learning_rate": 9.988407634871867e-06,
+      "loss": 0.7117,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.809656014341641,
+      "learning_rate": 9.988318668985566e-06,
+      "loss": 0.5613,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 27.46276218593051,
+      "learning_rate": 9.988229363416821e-06,
+      "loss": 0.6446,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.170747922266809,
+      "learning_rate": 9.988139718171713e-06,
+      "loss": 0.6169,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.299351054910534,
+      "learning_rate": 9.988049733256342e-06,
+      "loss": 0.5355,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.871635578101333,
+      "learning_rate": 9.98795940867684e-06,
+      "loss": 0.5083,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.730010687542885,
+      "learning_rate": 9.987868744439357e-06,
+      "loss": 0.5638,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1109377778250753,
+      "learning_rate": 9.987777740550067e-06,
+      "loss": 0.8131,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.554139583600708,
+      "learning_rate": 9.987686397015163e-06,
+      "loss": 0.6374,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.731938150786532,
+      "learning_rate": 9.987594713840871e-06,
+      "loss": 0.658,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.623894089413795,
+      "learning_rate": 9.98750269103343e-06,
+      "loss": 0.5473,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.690448720663107,
+      "learning_rate": 9.987410328599108e-06,
+      "loss": 0.8151,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.529287627658452,
+      "learning_rate": 9.987317626544194e-06,
+      "loss": 0.6852,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.9018034676823685,
+      "learning_rate": 9.987224584875002e-06,
+      "loss": 0.5104,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.69931880633369,
+      "learning_rate": 9.987131203597867e-06,
+      "loss": 0.6177,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.746640339612046,
+      "learning_rate": 9.987037482719146e-06,
+      "loss": 0.5638,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.417143859431394,
+      "learning_rate": 9.986943422245221e-06,
+      "loss": 0.593,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.220608648778223,
+      "learning_rate": 9.986849022182502e-06,
+      "loss": 0.5456,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.62945236957523,
+      "learning_rate": 9.986754282537411e-06,
+      "loss": 0.6175,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.18966131793515,
+      "learning_rate": 9.986659203316401e-06,
+      "loss": 0.7282,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.16201432307554,
+      "learning_rate": 9.986563784525952e-06,
+      "loss": 0.4554,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.85110680372945,
+      "learning_rate": 9.986468026172553e-06,
+      "loss": 0.539,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.342868057740313,
+      "learning_rate": 9.986371928262731e-06,
+      "loss": 0.5553,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.834846862265912,
+      "learning_rate": 9.986275490803027e-06,
+      "loss": 0.5983,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.73557045068984,
+      "learning_rate": 9.98617871380001e-06,
+      "loss": 0.5697,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.1476864702802585,
+      "learning_rate": 9.986081597260267e-06,
+      "loss": 0.5476,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.198715906154269,
+      "learning_rate": 9.985984141190414e-06,
+      "loss": 0.4708,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.135122280291613,
+      "learning_rate": 9.985886345597087e-06,
+      "loss": 0.6619,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.67888232892043,
+      "learning_rate": 9.985788210486943e-06,
+      "loss": 0.6084,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.004484692974607,
+      "learning_rate": 9.985689735866666e-06,
+      "loss": 0.5419,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.43078562674513,
+      "learning_rate": 9.985590921742963e-06,
+      "loss": 0.6214,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.269768693277465,
+      "learning_rate": 9.985491768122563e-06,
+      "loss": 0.6082,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.576354171556185,
+      "learning_rate": 9.985392275012215e-06,
+      "loss": 0.6038,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.561422024014586,
+      "learning_rate": 9.985292442418697e-06,
+      "loss": 0.6731,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.02751012148394,
+      "learning_rate": 9.985192270348805e-06,
+      "loss": 0.602,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9696472960290559,
+      "learning_rate": 9.98509175880936e-06,
+      "loss": 0.7826,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.248905676813904,
+      "learning_rate": 9.98499090780721e-06,
+      "loss": 0.5677,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.310645741188196,
+      "learning_rate": 9.98488971734922e-06,
+      "loss": 0.6379,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.757571983161412,
+      "learning_rate": 9.98478818744228e-06,
+      "loss": 0.5561,
+      "step": 672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.470377612273818,
+      "learning_rate": 9.984686318093304e-06,
+      "loss": 0.5787,
+      "step": 673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.5794047504531665,
+      "learning_rate": 9.98458410930923e-06,
+      "loss": 0.5179,
+      "step": 674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 16.746084084060556,
+      "learning_rate": 9.984481561097017e-06,
+      "loss": 0.7359,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.765848172670895,
+      "learning_rate": 9.984378673463649e-06,
+      "loss": 0.5906,
+      "step": 676
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.702507554083579,
+      "learning_rate": 9.98427544641613e-06,
+      "loss": 0.5035,
+      "step": 677
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.378119597821675,
+      "learning_rate": 9.984171879961491e-06,
+      "loss": 0.5692,
+      "step": 678
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.797407824948369,
+      "learning_rate": 9.984067974106786e-06,
+      "loss": 0.6865,
+      "step": 679
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.664359992236042,
+      "learning_rate": 9.983963728859088e-06,
+      "loss": 0.6102,
+      "step": 680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.210497332770461,
+      "learning_rate": 9.983859144225495e-06,
+      "loss": 0.6743,
+      "step": 681
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.586559018006374,
+      "learning_rate": 9.983754220213132e-06,
+      "loss": 0.6272,
+      "step": 682
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.418216448206815,
+      "learning_rate": 9.983648956829142e-06,
+      "loss": 0.685,
+      "step": 683
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.464759318455064,
+      "learning_rate": 9.983543354080691e-06,
+      "loss": 0.5678,
+      "step": 684
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.055683215476427,
+      "learning_rate": 9.983437411974975e-06,
+      "loss": 0.6179,
+      "step": 685
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.686097651467241,
+      "learning_rate": 9.983331130519203e-06,
+      "loss": 0.4988,
+      "step": 686
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.9775205990780655,
+      "learning_rate": 9.983224509720616e-06,
+      "loss": 0.4273,
+      "step": 687
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.047514472886402,
+      "learning_rate": 9.983117549586471e-06,
+      "loss": 0.6355,
+      "step": 688
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.525647916318052,
+      "learning_rate": 9.983010250124057e-06,
+      "loss": 0.6319,
+      "step": 689
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.735968141202308,
+      "learning_rate": 9.982902611340674e-06,
+      "loss": 0.6582,
+      "step": 690
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 41.18449541042415,
+      "learning_rate": 9.982794633243657e-06,
+      "loss": 0.6554,
+      "step": 691
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.198550611184992,
+      "learning_rate": 9.982686315840355e-06,
+      "loss": 0.5858,
+      "step": 692
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.719824930496458,
+      "learning_rate": 9.982577659138146e-06,
+      "loss": 0.6386,
+      "step": 693
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.955595165205136,
+      "learning_rate": 9.98246866314443e-06,
+      "loss": 0.6943,
+      "step": 694
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 28.520449059937853,
+      "learning_rate": 9.982359327866627e-06,
+      "loss": 0.6501,
+      "step": 695
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.500188579958536,
+      "learning_rate": 9.982249653312182e-06,
+      "loss": 0.5323,
+      "step": 696
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.198485984519784,
+      "learning_rate": 9.982139639488566e-06,
+      "loss": 0.5351,
+      "step": 697
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.25389232142257,
+      "learning_rate": 9.982029286403267e-06,
+      "loss": 0.5311,
+      "step": 698
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9665228939644723,
+      "learning_rate": 9.981918594063804e-06,
+      "loss": 0.78,
+      "step": 699
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.697909202248853,
+      "learning_rate": 9.98180756247771e-06,
+      "loss": 0.5711,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.084489237877124,
+      "learning_rate": 9.981696191652548e-06,
+      "loss": 0.6256,
+      "step": 701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.811610278950454,
+      "learning_rate": 9.981584481595902e-06,
+      "loss": 0.6458,
+      "step": 702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.7701671377983805,
+      "learning_rate": 9.98147243231538e-06,
+      "loss": 0.7164,
+      "step": 703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.200983787457734,
+      "learning_rate": 9.981360043818608e-06,
+      "loss": 0.5772,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 17.78518006030081,
+      "learning_rate": 9.981247316113245e-06,
+      "loss": 0.4713,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.120535492553035,
+      "learning_rate": 9.981134249206962e-06,
+      "loss": 0.5399,
+      "step": 706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.641498037326985,
+      "learning_rate": 9.981020843107462e-06,
+      "loss": 0.6803,
+      "step": 707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.968882161562552,
+      "learning_rate": 9.980907097822465e-06,
+      "loss": 0.5729,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.636176262198662,
+      "learning_rate": 9.980793013359717e-06,
+      "loss": 0.5437,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.518214645617144,
+      "learning_rate": 9.980678589726988e-06,
+      "loss": 0.5947,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.16536037395515,
+      "learning_rate": 9.980563826932069e-06,
+      "loss": 0.5119,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.639531994273837,
+      "learning_rate": 9.980448724982774e-06,
+      "loss": 0.6005,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.720255147308612,
+      "learning_rate": 9.980333283886943e-06,
+      "loss": 0.7013,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.836892787491626,
+      "learning_rate": 9.980217503652434e-06,
+      "loss": 0.5972,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.494404810711529,
+      "learning_rate": 9.980101384287133e-06,
+      "loss": 0.503,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.638573659958426,
+      "learning_rate": 9.979984925798948e-06,
+      "loss": 0.6114,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.027873081802696,
+      "learning_rate": 9.979868128195809e-06,
+      "loss": 0.5686,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.128710974487534,
+      "learning_rate": 9.979750991485665e-06,
+      "loss": 0.6526,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.06580256551704,
+      "learning_rate": 9.979633515676501e-06,
+      "loss": 0.5557,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.314108997163302,
+      "learning_rate": 9.97951570077631e-06,
+      "loss": 0.6367,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.031791741004426,
+      "learning_rate": 9.979397546793115e-06,
+      "loss": 0.5265,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.582717318617625,
+      "learning_rate": 9.979279053734964e-06,
+      "loss": 0.5373,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.41880073968488,
+      "learning_rate": 9.979160221609928e-06,
+      "loss": 0.7054,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.680064681717368,
+      "learning_rate": 9.979041050426093e-06,
+      "loss": 0.6659,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.46753207075893,
+      "learning_rate": 9.978921540191577e-06,
+      "loss": 0.6219,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.721149115123412,
+      "learning_rate": 9.978801690914519e-06,
+      "loss": 0.5883,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.9692714242744955,
+      "learning_rate": 9.978681502603079e-06,
+      "loss": 0.5095,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 20.23895783917436,
+      "learning_rate": 9.978560975265442e-06,
+      "loss": 0.6169,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.868503456370413,
+      "learning_rate": 9.978440108909814e-06,
+      "loss": 0.5978,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.281689747994934,
+      "learning_rate": 9.978318903544429e-06,
+      "loss": 0.698,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.309854768108936,
+      "learning_rate": 9.978197359177537e-06,
+      "loss": 0.5466,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.738669562095467,
+      "learning_rate": 9.978075475817414e-06,
+      "loss": 0.3865,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.536516478397159,
+      "learning_rate": 9.977953253472363e-06,
+      "loss": 0.5288,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 16.380683856204584,
+      "learning_rate": 9.977830692150706e-06,
+      "loss": 0.5674,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.903134732391848,
+      "learning_rate": 9.977707791860787e-06,
+      "loss": 0.5904,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.629225844677529,
+      "learning_rate": 9.977584552610977e-06,
+      "loss": 0.6179,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.636123616894606,
+      "learning_rate": 9.977460974409667e-06,
+      "loss": 0.5362,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 19.00358663946508,
+      "learning_rate": 9.97733705726527e-06,
+      "loss": 0.4896,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.569262450605674,
+      "learning_rate": 9.97721280118623e-06,
+      "loss": 0.6336,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.72743397930331,
+      "learning_rate": 9.977088206181003e-06,
+      "loss": 0.4523,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.594473838160326,
+      "learning_rate": 9.976963272258075e-06,
+      "loss": 0.5313,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.752088444770559,
+      "learning_rate": 9.976837999425954e-06,
+      "loss": 0.6365,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 38.25495813194906,
+      "learning_rate": 9.97671238769317e-06,
+      "loss": 0.5532,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.657041948413244,
+      "learning_rate": 9.976586437068276e-06,
+      "loss": 0.6031,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.694995408927946,
+      "learning_rate": 9.97646014755985e-06,
+      "loss": 0.5143,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.503692376549681,
+      "learning_rate": 9.97633351917649e-06,
+      "loss": 0.4399,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.1564499949508,
+      "learning_rate": 9.976206551926822e-06,
+      "loss": 0.5933,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.519562545486004,
+      "learning_rate": 9.97607924581949e-06,
+      "loss": 0.5907,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.306986564604776,
+      "learning_rate": 9.97595160086316e-06,
+      "loss": 0.4771,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.52822084482565,
+      "learning_rate": 9.975823617066531e-06,
+      "loss": 0.6683,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.387288005805004,
+      "learning_rate": 9.975695294438312e-06,
+      "loss": 0.625,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 15.674574933453343,
+      "learning_rate": 9.975566632987242e-06,
+      "loss": 0.6058,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 11.708964960671533,
+      "learning_rate": 9.975437632722084e-06,
+      "loss": 0.5759,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.710951955554571,
+      "learning_rate": 9.975308293651624e-06,
+      "loss": 0.6337,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.443412058788645,
+      "learning_rate": 9.975178615784666e-06,
+      "loss": 0.5253,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0893937501522006,
+      "learning_rate": 9.975048599130042e-06,
+      "loss": 0.7444,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.481764034869314,
+      "learning_rate": 9.974918243696606e-06,
+      "loss": 0.5127,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.643140088570465,
+      "learning_rate": 9.974787549493232e-06,
+      "loss": 0.6483,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.992821292993867,
+      "learning_rate": 9.974656516528822e-06,
+      "loss": 0.4979,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.307370808774207,
+      "learning_rate": 9.9745251448123e-06,
+      "loss": 0.5985,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.429104994780412,
+      "learning_rate": 9.97439343435261e-06,
+      "loss": 0.6009,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.130155580521042,
+      "learning_rate": 9.97426138515872e-06,
+      "loss": 0.6135,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.559748860973095,
+      "learning_rate": 9.974128997239624e-06,
+      "loss": 0.6061,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.525358606702473,
+      "learning_rate": 9.973996270604335e-06,
+      "loss": 0.5996,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8314152956832876,
+      "learning_rate": 9.973863205261894e-06,
+      "loss": 0.736,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6460933465576617,
+      "learning_rate": 9.97372980122136e-06,
+      "loss": 0.6684,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.508464638063013,
+      "learning_rate": 9.973596058491817e-06,
+      "loss": 0.5451,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.495330924158752,
+      "learning_rate": 9.973461977082375e-06,
+      "loss": 0.4787,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.042000617740993,
+      "learning_rate": 9.97332755700216e-06,
+      "loss": 0.6438,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.781268013795374,
+      "learning_rate": 9.973192798260329e-06,
+      "loss": 0.5298,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.61897001912762,
+      "learning_rate": 9.973057700866055e-06,
+      "loss": 0.539,
+      "step": 771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.4945119015612365,
+      "learning_rate": 9.972922264828542e-06,
+      "loss": 0.6544,
+      "step": 772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.215751577269106,
+      "learning_rate": 9.97278649015701e-06,
+      "loss": 0.6313,
+      "step": 773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.573168682519035,
+      "learning_rate": 9.972650376860705e-06,
+      "loss": 0.5195,
+      "step": 774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.8287612419920825,
+      "learning_rate": 9.972513924948895e-06,
+      "loss": 0.6334,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.422029002581222,
+      "learning_rate": 9.972377134430873e-06,
+      "loss": 0.6082,
+      "step": 776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.591712837026288,
+      "learning_rate": 9.972240005315954e-06,
+      "loss": 0.6142,
+      "step": 777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.243092310449694,
+      "learning_rate": 9.972102537613473e-06,
+      "loss": 0.7699,
+      "step": 778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.904919992987018,
+      "learning_rate": 9.971964731332795e-06,
+      "loss": 0.5581,
+      "step": 779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.327047240503244,
+      "learning_rate": 9.971826586483301e-06,
+      "loss": 0.5433,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.685253127062064,
+      "learning_rate": 9.971688103074401e-06,
+      "loss": 0.5876,
+      "step": 781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 23.967994394414973,
+      "learning_rate": 9.97154928111552e-06,
+      "loss": 0.567,
+      "step": 782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.372599212571878,
+      "learning_rate": 9.97141012061612e-06,
+      "loss": 0.5882,
+      "step": 783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.095130631763505,
+      "learning_rate": 9.971270621585665e-06,
+      "loss": 0.4946,
+      "step": 784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.440071505293244,
+      "learning_rate": 9.971130784033664e-06,
+      "loss": 0.4918,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.036146914494532,
+      "learning_rate": 9.970990607969637e-06,
+      "loss": 0.5531,
+      "step": 786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.044385043909202,
+      "learning_rate": 9.970850093403129e-06,
+      "loss": 0.6459,
+      "step": 787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.712604675195857,
+      "learning_rate": 9.970709240343707e-06,
+      "loss": 0.6034,
+      "step": 788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.138502427737817,
+      "learning_rate": 9.970568048800965e-06,
+      "loss": 0.8257,
+      "step": 789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.299346760517267,
+      "learning_rate": 9.970426518784514e-06,
+      "loss": 0.4618,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.190894925535282,
+      "learning_rate": 9.970284650303993e-06,
+      "loss": 0.512,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.009678581342257,
+      "learning_rate": 9.970142443369064e-06,
+      "loss": 0.7224,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.850389562028894,
+      "learning_rate": 9.969999897989411e-06,
+      "loss": 0.6077,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.376541407397356,
+      "learning_rate": 9.969857014174737e-06,
+      "loss": 0.5257,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.996775255445888,
+      "learning_rate": 9.969713791934778e-06,
+      "loss": 0.5968,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.867412551197331,
+      "learning_rate": 9.969570231279282e-06,
+      "loss": 0.6274,
+      "step": 796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 12.851591646923918,
+      "learning_rate": 9.969426332218024e-06,
+      "loss": 0.578,
+      "step": 797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.15351776597485,
+      "learning_rate": 9.969282094760807e-06,
+      "loss": 0.6067,
+      "step": 798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.629187818981064,
+      "learning_rate": 9.96913751891745e-06,
+      "loss": 0.5998,
+      "step": 799
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.263723865050121,
+      "learning_rate": 9.968992604697797e-06,
+      "loss": 0.6769,
+      "step": 800
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.4089928840741806,
+      "learning_rate": 9.96884735211172e-06,
+      "loss": 0.5469,
+      "step": 801
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.785632984923619,
+      "learning_rate": 9.968701761169108e-06,
+      "loss": 0.5764,
+      "step": 802
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.058059232318288,
+      "learning_rate": 9.968555831879873e-06,
+      "loss": 0.7371,
+      "step": 803
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.064400342737331,
+      "learning_rate": 9.968409564253956e-06,
+      "loss": 0.5183,
+      "step": 804
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 25.723090547023048,
+      "learning_rate": 9.968262958301316e-06,
+      "loss": 0.5182,
+      "step": 805
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.70757539623557,
+      "learning_rate": 9.968116014031935e-06,
+      "loss": 0.5861,
+      "step": 806
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.4478659710943145,
+      "learning_rate": 9.967968731455818e-06,
+      "loss": 0.6447,
+      "step": 807
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.9873299944955285,
+      "learning_rate": 9.967821110582998e-06,
+      "loss": 0.5831,
+      "step": 808
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.093673917902496,
+      "learning_rate": 9.967673151423523e-06,
+      "loss": 0.5847,
+      "step": 809
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.197724791496282,
+      "learning_rate": 9.967524853987474e-06,
+      "loss": 0.5597,
+      "step": 810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3838787074367933,
+      "learning_rate": 9.967376218284944e-06,
+      "loss": 0.8143,
+      "step": 811
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.61919390026761,
+      "learning_rate": 9.967227244326056e-06,
+      "loss": 0.567,
+      "step": 812
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.204227363553488,
+      "learning_rate": 9.967077932120956e-06,
+      "loss": 0.6383,
+      "step": 813
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.506037412292499,
+      "learning_rate": 9.966928281679812e-06,
+      "loss": 0.5619,
+      "step": 814
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7596695259603343,
+      "learning_rate": 9.96677829301281e-06,
+      "loss": 0.7439,
+      "step": 815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.19769033177162,
+      "learning_rate": 9.966627966130168e-06,
+      "loss": 0.6724,
+      "step": 816
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.125551276137998,
+      "learning_rate": 9.966477301042122e-06,
+      "loss": 0.5281,
+      "step": 817
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.147312300736502,
+      "learning_rate": 9.966326297758931e-06,
+      "loss": 0.5774,
+      "step": 818
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.288343854044063,
+      "learning_rate": 9.966174956290877e-06,
+      "loss": 0.5772,
+      "step": 819
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.4379822989467215,
+      "learning_rate": 9.966023276648266e-06,
+      "loss": 0.5121,
+      "step": 820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.0392666509926975,
+      "learning_rate": 9.965871258841427e-06,
+      "loss": 0.5304,
+      "step": 821
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.652994685369217,
+      "learning_rate": 9.96571890288071e-06,
+      "loss": 0.5076,
+      "step": 822
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.817219524933746,
+      "learning_rate": 9.965566208776493e-06,
+      "loss": 0.5858,
+      "step": 823
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.415083411770071,
+      "learning_rate": 9.965413176539174e-06,
+      "loss": 0.5732,
+      "step": 824
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.194316818539082,
+      "learning_rate": 9.96525980617917e-06,
+      "loss": 0.7561,
+      "step": 825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.438725248787346,
+      "learning_rate": 9.965106097706926e-06,
+      "loss": 0.6461,
+      "step": 826
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.318882169195731,
+      "learning_rate": 9.964952051132913e-06,
+      "loss": 0.5384,
+      "step": 827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.861018745262478,
+      "learning_rate": 9.964797666467614e-06,
+      "loss": 0.5789,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.492244526490031,
+      "learning_rate": 9.964642943721547e-06,
+      "loss": 0.7637,
+      "step": 829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.019773820694503,
+      "learning_rate": 9.964487882905247e-06,
+      "loss": 0.6084,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.76344068204411,
+      "learning_rate": 9.964332484029273e-06,
+      "loss": 0.5351,
+      "step": 831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.133357771554578,
+      "learning_rate": 9.964176747104204e-06,
+      "loss": 0.5882,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0384274809284277,
+      "learning_rate": 9.96402067214065e-06,
+      "loss": 0.7124,
+      "step": 833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.758445191382094,
+      "learning_rate": 9.963864259149235e-06,
+      "loss": 0.4956,
+      "step": 834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8596845529369936,
+      "learning_rate": 9.963707508140611e-06,
+      "loss": 0.6024,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.703347308402639,
+      "learning_rate": 9.963550419125454e-06,
+      "loss": 0.4673,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.46608206966379,
+      "learning_rate": 9.96339299211446e-06,
+      "loss": 0.7084,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.43760699597422,
+      "learning_rate": 9.963235227118346e-06,
+      "loss": 0.6155,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.091490757177361,
+      "learning_rate": 9.963077124147859e-06,
+      "loss": 0.6185,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.322799940148098,
+      "learning_rate": 9.962918683213767e-06,
+      "loss": 0.5089,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.746980173997344,
+      "learning_rate": 9.962759904326852e-06,
+      "loss": 0.4782,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.06157671260489,
+      "learning_rate": 9.96260078749793e-06,
+      "loss": 0.6172,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8230972357451765,
+      "learning_rate": 9.962441332737839e-06,
+      "loss": 0.7726,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.17309731737822,
+      "learning_rate": 9.962281540057432e-06,
+      "loss": 0.4661,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.0351117476706,
+      "learning_rate": 9.962121409467595e-06,
+      "loss": 0.5058,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.281661400224745,
+      "learning_rate": 9.96196094097923e-06,
+      "loss": 0.5027,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.351732118428748,
+      "learning_rate": 9.961800134603261e-06,
+      "loss": 0.6633,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.653382709076699,
+      "learning_rate": 9.961638990350644e-06,
+      "loss": 0.5192,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.893581515413597,
+      "learning_rate": 9.961477508232347e-06,
+      "loss": 0.6718,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.509307367963578,
+      "learning_rate": 9.961315688259369e-06,
+      "loss": 0.584,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.92340066377067,
+      "learning_rate": 9.961153530442732e-06,
+      "loss": 0.687,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.328778505073384,
+      "learning_rate": 9.960991034793473e-06,
+      "loss": 0.6137,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 13.549549263580559,
+      "learning_rate": 9.960828201322658e-06,
+      "loss": 0.5462,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.267632226436028,
+      "learning_rate": 9.960665030041379e-06,
+      "loss": 0.4518,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.1409544308909325,
+      "learning_rate": 9.960501520960743e-06,
+      "loss": 0.6651,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.920956358129832,
+      "learning_rate": 9.960337674091887e-06,
+      "loss": 0.5343,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 14.641317901213732,
+      "learning_rate": 9.960173489445965e-06,
+      "loss": 0.5458,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.980207255635076,
+      "learning_rate": 9.960008967034163e-06,
+      "loss": 0.642,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.730141701175137,
+      "learning_rate": 9.95984410686768e-06,
+      "loss": 0.5873,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.934097571827185,
+      "learning_rate": 9.959678908957743e-06,
+      "loss": 0.5799,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.578502305114089,
+      "learning_rate": 9.9595133733156e-06,
+      "loss": 0.6111,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.878053547343118,
+      "learning_rate": 9.959347499952525e-06,
+      "loss": 0.5166,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.02693422241157,
+      "learning_rate": 9.959181288879813e-06,
+      "loss": 0.5076,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.605340890627287,
+      "learning_rate": 9.959014740108782e-06,
+      "loss": 0.497,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.570570357791429,
+      "learning_rate": 9.958847853650774e-06,
+      "loss": 0.613,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.223750262831484,
+      "learning_rate": 9.958680629517152e-06,
+      "loss": 0.6788,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.20552782152081,
+      "learning_rate": 9.958513067719303e-06,
+      "loss": 0.568,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 12.25528501732002,
+      "learning_rate": 9.958345168268639e-06,
+      "loss": 0.6209,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.86911946643418,
+      "learning_rate": 9.95817693117659e-06,
+      "loss": 0.5414,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.788268454633197,
+      "learning_rate": 9.958008356454617e-06,
+      "loss": 0.6991,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.433784185809803,
+      "learning_rate": 9.957839444114197e-06,
+      "loss": 0.609,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.048953939254857,
+      "learning_rate": 9.95767019416683e-06,
+      "loss": 0.5287,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.237142734890177,
+      "learning_rate": 9.957500606624042e-06,
+      "loss": 0.5855,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.991147824756172,
+      "learning_rate": 9.957330681497384e-06,
+      "loss": 0.5781,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.813801817236163,
+      "learning_rate": 9.957160418798425e-06,
+      "loss": 0.5021,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.925698903931939,
+      "learning_rate": 9.956989818538759e-06,
+      "loss": 0.5434,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3723295405825855,
+      "learning_rate": 9.956818880730002e-06,
+      "loss": 0.7923,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.873890641057963,
+      "learning_rate": 9.956647605383798e-06,
+      "loss": 0.615,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.656094299853768,
+      "learning_rate": 9.956475992511807e-06,
+      "loss": 0.4347,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6448281793491284,
+      "learning_rate": 9.956304042125716e-06,
+      "loss": 0.6817,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.922220472604239,
+      "learning_rate": 9.956131754237233e-06,
+      "loss": 0.5953,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.454690082531868,
+      "learning_rate": 9.955959128858091e-06,
+      "loss": 0.4336,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.939275741204424,
+      "learning_rate": 9.955786166000044e-06,
+      "loss": 0.5825,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0819147037577626,
+      "learning_rate": 9.955612865674871e-06,
+      "loss": 0.7782,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.902714155948248,
+      "learning_rate": 9.955439227894375e-06,
+      "loss": 0.5879,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.961142026434629,
+      "learning_rate": 9.955265252670376e-06,
+      "loss": 0.6499,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.595529639764193,
+      "learning_rate": 9.955090940014723e-06,
+      "loss": 0.6727,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.182686138301782,
+      "learning_rate": 9.954916289939287e-06,
+      "loss": 0.4643,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.388324362709051,
+      "learning_rate": 9.954741302455959e-06,
+      "loss": 0.6629,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.660245000977892,
+      "learning_rate": 9.954565977576654e-06,
+      "loss": 0.6961,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.2238876887186665,
+      "learning_rate": 9.954390315313314e-06,
+      "loss": 0.5581,
+      "step": 891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.210112742050761,
+      "learning_rate": 9.954214315677897e-06,
+      "loss": 0.6312,
+      "step": 892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.24206750100229,
+      "learning_rate": 9.954037978682392e-06,
+      "loss": 0.6379,
+      "step": 893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.64304648112321,
+      "learning_rate": 9.953861304338805e-06,
+      "loss": 0.5731,
+      "step": 894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.350431892985222,
+      "learning_rate": 9.953684292659166e-06,
+      "loss": 0.7054,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.504998217162817,
+      "learning_rate": 9.953506943655529e-06,
+      "loss": 0.604,
+      "step": 896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.177053929287642,
+      "learning_rate": 9.95332925733997e-06,
+      "loss": 0.5632,
+      "step": 897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.898830613780346,
+      "learning_rate": 9.95315123372459e-06,
+      "loss": 0.5418,
+      "step": 898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.085828869540821,
+      "learning_rate": 9.952972872821512e-06,
+      "loss": 0.5798,
+      "step": 899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.055096183012736,
+      "learning_rate": 9.95279417464288e-06,
+      "loss": 0.5541,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.817850930684552,
+      "learning_rate": 9.952615139200862e-06,
+      "loss": 0.4682,
+      "step": 901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 22.649569349979355,
+      "learning_rate": 9.952435766507653e-06,
+      "loss": 0.5389,
+      "step": 902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.004713847188265,
+      "learning_rate": 9.952256056575464e-06,
+      "loss": 0.6051,
+      "step": 903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.816947628432787,
+      "learning_rate": 9.952076009416537e-06,
+      "loss": 0.5541,
+      "step": 904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.355269871318229,
+      "learning_rate": 9.951895625043125e-06,
+      "loss": 0.5155,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.828917187675117,
+      "learning_rate": 9.951714903467517e-06,
+      "loss": 0.5931,
+      "step": 906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.360420712659712,
+      "learning_rate": 9.951533844702017e-06,
+      "loss": 0.5783,
+      "step": 907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.79615049899187,
+      "learning_rate": 9.951352448758956e-06,
+      "loss": 0.6458,
+      "step": 908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.379038934009893,
+      "learning_rate": 9.951170715650685e-06,
+      "loss": 0.5682,
+      "step": 909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.8008375401773655,
+      "learning_rate": 9.95098864538958e-06,
+      "loss": 0.5623,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.040901313539463,
+      "learning_rate": 9.95080623798804e-06,
+      "loss": 0.6391,
+      "step": 911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.589758753461778,
+      "learning_rate": 9.950623493458483e-06,
+      "loss": 0.5091,
+      "step": 912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.5460505851490804,
+      "learning_rate": 9.950440411813356e-06,
+      "loss": 0.6023,
+      "step": 913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.801135924195882,
+      "learning_rate": 9.950256993065125e-06,
+      "loss": 0.5696,
+      "step": 914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.554950749302028,
+      "learning_rate": 9.95007323722628e-06,
+      "loss": 0.601,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.693881052095628,
+      "learning_rate": 9.949889144309335e-06,
+      "loss": 0.572,
+      "step": 916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.8372006267959735,
+      "learning_rate": 9.949704714326823e-06,
+      "loss": 0.596,
+      "step": 917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2849203002841576,
+      "learning_rate": 9.949519947291306e-06,
+      "loss": 0.8088,
+      "step": 918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.214478180204056,
+      "learning_rate": 9.949334843215366e-06,
+      "loss": 0.582,
+      "step": 919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.01847532572868,
+      "learning_rate": 9.949149402111604e-06,
+      "loss": 0.5709,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.60961799625377,
+      "learning_rate": 9.948963623992651e-06,
+      "loss": 0.5767,
+      "step": 921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.214497060946947,
+      "learning_rate": 9.948777508871158e-06,
+      "loss": 0.6047,
+      "step": 922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.483719951043064,
+      "learning_rate": 9.948591056759797e-06,
+      "loss": 0.538,
+      "step": 923
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.915902517743382,
+      "learning_rate": 9.948404267671263e-06,
+      "loss": 0.5836,
+      "step": 924
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.869959288958805,
+      "learning_rate": 9.94821714161828e-06,
+      "loss": 0.6284,
+      "step": 925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1837537814010166,
+      "learning_rate": 9.948029678613586e-06,
+      "loss": 0.7785,
+      "step": 926
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.383238733759933,
+      "learning_rate": 9.94784187866995e-06,
+      "loss": 0.5898,
+      "step": 927
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.506981381974643,
+      "learning_rate": 9.947653741800158e-06,
+      "loss": 0.4872,
+      "step": 928
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.064084753352281,
+      "learning_rate": 9.947465268017024e-06,
+      "loss": 0.5348,
+      "step": 929
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.865723770222079,
+      "learning_rate": 9.947276457333378e-06,
+      "loss": 0.539,
+      "step": 930
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.735701082861237,
+      "learning_rate": 9.94708730976208e-06,
+      "loss": 0.5759,
+      "step": 931
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.38588153428739,
+      "learning_rate": 9.94689782531601e-06,
+      "loss": 0.5612,
+      "step": 932
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.663322623712922,
+      "learning_rate": 9.94670800400807e-06,
+      "loss": 0.4689,
+      "step": 933
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.618041159894497,
+      "learning_rate": 9.946517845851187e-06,
+      "loss": 0.6092,
+      "step": 934
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.984030803560398,
+      "learning_rate": 9.94632735085831e-06,
+      "loss": 0.5934,
+      "step": 935
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.595113273965969,
+      "learning_rate": 9.946136519042408e-06,
+      "loss": 0.5892,
+      "step": 936
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 67.44756912905208,
+      "learning_rate": 9.945945350416479e-06,
+      "loss": 0.5662,
+      "step": 937
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.600188571891748,
+      "learning_rate": 9.945753844993541e-06,
+      "loss": 0.5771,
+      "step": 938
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.4024483110610095,
+      "learning_rate": 9.945562002786635e-06,
+      "loss": 0.5601,
+      "step": 939
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.8208250078647,
+      "learning_rate": 9.94536982380882e-06,
+      "loss": 0.5138,
+      "step": 940
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.167227889122572,
+      "learning_rate": 9.945177308073187e-06,
+      "loss": 0.5562,
+      "step": 941
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.162032186533693,
+      "learning_rate": 9.944984455592846e-06,
+      "loss": 0.5471,
+      "step": 942
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.863356916817595,
+      "learning_rate": 9.944791266380924e-06,
+      "loss": 0.5263,
+      "step": 943
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.7594272043079116,
+      "learning_rate": 9.944597740450583e-06,
+      "loss": 0.6288,
+      "step": 944
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.051621676042953,
+      "learning_rate": 9.944403877814996e-06,
+      "loss": 0.4426,
+      "step": 945
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.14171238163337,
+      "learning_rate": 9.944209678487369e-06,
+      "loss": 0.5559,
+      "step": 946
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.937577558151903,
+      "learning_rate": 9.944015142480923e-06,
+      "loss": 0.604,
+      "step": 947
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.284205526367296,
+      "learning_rate": 9.943820269808904e-06,
+      "loss": 0.6417,
+      "step": 948
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.127010754030518,
+      "learning_rate": 9.943625060484582e-06,
+      "loss": 0.5522,
+      "step": 949
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.689732899451311,
+      "learning_rate": 9.943429514521254e-06,
+      "loss": 0.7148,
+      "step": 950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.94794210692499,
+      "learning_rate": 9.943233631932234e-06,
+      "loss": 0.7147,
+      "step": 951
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.5497227742388,
+      "learning_rate": 9.943037412730858e-06,
+      "loss": 0.5527,
+      "step": 952
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.393711639706105,
+      "learning_rate": 9.942840856930492e-06,
+      "loss": 0.5347,
+      "step": 953
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.343843213133239,
+      "learning_rate": 9.942643964544516e-06,
+      "loss": 0.6206,
+      "step": 954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.669212254114528,
+      "learning_rate": 9.94244673558634e-06,
+      "loss": 0.5025,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.51326929885249,
+      "learning_rate": 9.942249170069393e-06,
+      "loss": 0.4682,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.858368324530993,
+      "learning_rate": 9.942051268007132e-06,
+      "loss": 0.6026,
+      "step": 957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.095962843095585,
+      "learning_rate": 9.94185302941303e-06,
+      "loss": 0.5944,
+      "step": 958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 36.823484003675496,
+      "learning_rate": 9.941654454300586e-06,
+      "loss": 0.6166,
+      "step": 959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.851537666921912,
+      "learning_rate": 9.941455542683323e-06,
+      "loss": 0.5607,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.823037021096042,
+      "learning_rate": 9.941256294574786e-06,
+      "loss": 0.5798,
+      "step": 961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.784572317866587,
+      "learning_rate": 9.941056709988544e-06,
+      "loss": 0.5409,
+      "step": 962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.165194720850087,
+      "learning_rate": 9.940856788938186e-06,
+      "loss": 0.5566,
+      "step": 963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.446603158734456,
+      "learning_rate": 9.940656531437327e-06,
+      "loss": 0.6117,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.8379622200798975,
+      "learning_rate": 9.940455937499603e-06,
+      "loss": 0.5367,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.829017300888265,
+      "learning_rate": 9.940255007138672e-06,
+      "loss": 0.5144,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.91199179563306,
+      "learning_rate": 9.940053740368219e-06,
+      "loss": 0.4763,
+      "step": 967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.144099860794626,
+      "learning_rate": 9.93985213720195e-06,
+      "loss": 0.6256,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.124508902084358,
+      "learning_rate": 9.93965019765359e-06,
+      "loss": 0.488,
+      "step": 969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.506509549552956,
+      "learning_rate": 9.939447921736893e-06,
+      "loss": 0.5948,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8011350841240672,
+      "learning_rate": 9.939245309465631e-06,
+      "loss": 0.7534,
+      "step": 971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.13972021453503,
+      "learning_rate": 9.939042360853604e-06,
+      "loss": 0.5561,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.423920102132209,
+      "learning_rate": 9.938839075914629e-06,
+      "loss": 0.4966,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 66.69622410569475,
+      "learning_rate": 9.938635454662548e-06,
+      "loss": 0.599,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.181991580839778,
+      "learning_rate": 9.938431497111232e-06,
+      "loss": 0.5135,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 39.186259626132475,
+      "learning_rate": 9.938227203274563e-06,
+      "loss": 0.578,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.194003510369883,
+      "learning_rate": 9.938022573166457e-06,
+      "loss": 0.5566,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.122704705574467,
+      "learning_rate": 9.937817606800847e-06,
+      "loss": 0.4451,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4289980786181165,
+      "learning_rate": 9.93761230419169e-06,
+      "loss": 0.8147,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.756610515977056,
+      "learning_rate": 9.937406665352966e-06,
+      "loss": 0.5606,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.738188701814876,
+      "learning_rate": 9.93720069029868e-06,
+      "loss": 0.5521,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.122380076182631,
+      "learning_rate": 9.936994379042856e-06,
+      "loss": 0.5134,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.00159917303707,
+      "learning_rate": 9.936787731599543e-06,
+      "loss": 0.454,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.855927032876727,
+      "learning_rate": 9.936580747982814e-06,
+      "loss": 0.5989,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8993881501024035,
+      "learning_rate": 9.936373428206762e-06,
+      "loss": 0.6759,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.597335624412432,
+      "learning_rate": 9.936165772285506e-06,
+      "loss": 0.6631,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.221648666977631,
+      "learning_rate": 9.935957780233188e-06,
+      "loss": 0.5287,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.591120879659366,
+      "learning_rate": 9.935749452063967e-06,
+      "loss": 0.5721,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.82750449864475,
+      "learning_rate": 9.935540787792032e-06,
+      "loss": 0.5537,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.718969545717616,
+      "learning_rate": 9.935331787431592e-06,
+      "loss": 0.5708,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.845575816782693,
+      "learning_rate": 9.935122450996876e-06,
+      "loss": 0.5379,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.724465578298777,
+      "learning_rate": 9.934912778502146e-06,
+      "loss": 0.5541,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.073047213495178,
+      "learning_rate": 9.934702769961672e-06,
+      "loss": 0.5594,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.712857773854929,
+      "learning_rate": 9.934492425389759e-06,
+      "loss": 0.5433,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.699015815541499,
+      "learning_rate": 9.93428174480073e-06,
+      "loss": 0.5894,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 22.25126624401806,
+      "learning_rate": 9.93407072820893e-06,
+      "loss": 0.4482,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1755088853445366,
+      "learning_rate": 9.93385937562873e-06,
+      "loss": 0.7542,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.511419640294802,
+      "learning_rate": 9.93364768707452e-06,
+      "loss": 0.5706,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.895195961570816,
+      "learning_rate": 9.933435662560718e-06,
+      "loss": 0.5802,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.441007508793278,
+      "learning_rate": 9.93322330210176e-06,
+      "loss": 0.4452,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 62.88326905479679,
+      "learning_rate": 9.933010605712106e-06,
+      "loss": 0.4689,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.088392136905184,
+      "learning_rate": 9.932797573406242e-06,
+      "loss": 0.5955,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.3930551951456165,
+      "learning_rate": 9.932584205198673e-06,
+      "loss": 0.6372,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.832629326926635,
+      "learning_rate": 9.932370501103929e-06,
+      "loss": 0.7722,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.389626076841408,
+      "learning_rate": 9.932156461136562e-06,
+      "loss": 0.6148,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.828571532593378,
+      "learning_rate": 9.931942085311146e-06,
+      "loss": 0.6047,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.930691208584589,
+      "learning_rate": 9.931727373642283e-06,
+      "loss": 0.6084,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.563679110981295,
+      "learning_rate": 9.93151232614459e-06,
+      "loss": 0.5284,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.73498524200236,
+      "learning_rate": 9.931296942832712e-06,
+      "loss": 0.5895,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.933392017936969,
+      "learning_rate": 9.931081223721315e-06,
+      "loss": 0.5114,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.60802415790699,
+      "learning_rate": 9.93086516882509e-06,
+      "loss": 0.5676,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.657002119826139,
+      "learning_rate": 9.93064877815875e-06,
+      "loss": 0.5905,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.14888083215417,
+      "learning_rate": 9.930432051737026e-06,
+      "loss": 0.5429,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.940902988378835,
+      "learning_rate": 9.93021498957468e-06,
+      "loss": 0.5532,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.497804779673605,
+      "learning_rate": 9.929997591686493e-06,
+      "loss": 0.4432,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.188137266856303,
+      "learning_rate": 9.929779858087268e-06,
+      "loss": 0.5931,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0685425446899237,
+      "learning_rate": 9.92956178879183e-06,
+      "loss": 0.7692,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.93753200536441,
+      "learning_rate": 9.92934338381503e-06,
+      "loss": 0.6126,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.594923347913501,
+      "learning_rate": 9.929124643171741e-06,
+      "loss": 0.6119,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 21.5666411646536,
+      "learning_rate": 9.928905566876859e-06,
+      "loss": 0.4906,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.084078638974368,
+      "learning_rate": 9.9286861549453e-06,
+      "loss": 0.6376,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.841442112005636,
+      "learning_rate": 9.928466407392006e-06,
+      "loss": 0.4257,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 56.21278273741323,
+      "learning_rate": 9.928246324231939e-06,
+      "loss": 0.4593,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.369498233900541,
+      "learning_rate": 9.928025905480088e-06,
+      "loss": 0.5721,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.430098431973589,
+      "learning_rate": 9.927805151151463e-06,
+      "loss": 0.5779,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.203493244710055,
+      "learning_rate": 9.927584061261095e-06,
+      "loss": 0.554,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.386588049069639,
+      "learning_rate": 9.927362635824039e-06,
+      "loss": 0.6281,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.790171000395867,
+      "learning_rate": 9.927140874855374e-06,
+      "loss": 0.5772,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 17.172405813537637,
+      "learning_rate": 9.9269187783702e-06,
+      "loss": 0.5978,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.812824555115454,
+      "learning_rate": 9.926696346383643e-06,
+      "loss": 0.6367,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.969904837034381,
+      "learning_rate": 9.926473578910846e-06,
+      "loss": 0.4775,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 33.15143911940542,
+      "learning_rate": 9.926250475966981e-06,
+      "loss": 0.6136,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.538756808362451,
+      "learning_rate": 9.92602703756724e-06,
+      "loss": 0.5381,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.889160602533309,
+      "learning_rate": 9.925803263726838e-06,
+      "loss": 0.4986,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.101291539592363,
+      "learning_rate": 9.925579154461013e-06,
+      "loss": 0.5646,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 12.311042220919251,
+      "learning_rate": 9.925354709785025e-06,
+      "loss": 0.5704,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.27426847300758,
+      "learning_rate": 9.92512992971416e-06,
+      "loss": 0.5702,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.541488380042324,
+      "learning_rate": 9.924904814263722e-06,
+      "loss": 0.5152,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.641878520743584,
+      "learning_rate": 9.924679363449042e-06,
+      "loss": 0.5874,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.544314213266222,
+      "learning_rate": 9.92445357728547e-06,
+      "loss": 0.6145,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.49780384507577,
+      "learning_rate": 9.924227455788384e-06,
+      "loss": 0.622,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7966027995948863,
+      "learning_rate": 9.92400099897318e-06,
+      "loss": 0.7094,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.647603429685295,
+      "learning_rate": 9.923774206855282e-06,
+      "loss": 0.5833,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.282712727580346,
+      "learning_rate": 9.923547079450126e-06,
+      "loss": 0.593,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.982142134700442,
+      "learning_rate": 9.923319616773186e-06,
+      "loss": 0.6093,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.685632222420102,
+      "learning_rate": 9.923091818839949e-06,
+      "loss": 0.5,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.387495335832092,
+      "learning_rate": 9.922863685665925e-06,
+      "loss": 0.4966,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.11349447383908,
+      "learning_rate": 9.92263521726665e-06,
+      "loss": 0.6081,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.429566941137038,
+      "learning_rate": 9.922406413657684e-06,
+      "loss": 0.5119,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.634165691873186,
+      "learning_rate": 9.922177274854604e-06,
+      "loss": 0.4871,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.273104068124814,
+      "learning_rate": 9.921947800873014e-06,
+      "loss": 0.5189,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.425103524592953,
+      "learning_rate": 9.921717991728541e-06,
+      "loss": 0.625,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.881521986947523,
+      "learning_rate": 9.921487847436836e-06,
+      "loss": 0.5872,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.959852026848693,
+      "learning_rate": 9.921257368013566e-06,
+      "loss": 0.5312,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.292737493490531,
+      "learning_rate": 9.92102655347443e-06,
+      "loss": 0.6221,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.292117400872386,
+      "learning_rate": 9.920795403835143e-06,
+      "loss": 0.6537,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.884039126532933,
+      "learning_rate": 9.920563919111446e-06,
+      "loss": 0.5718,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.932078613924822,
+      "learning_rate": 9.920332099319104e-06,
+      "loss": 0.5,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.6182147396405355,
+      "learning_rate": 9.920099944473899e-06,
+      "loss": 0.6439,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.629592202613414,
+      "learning_rate": 9.919867454591644e-06,
+      "loss": 0.4945,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.303428041801128,
+      "learning_rate": 9.919634629688166e-06,
+      "loss": 0.5212,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.488087888706214,
+      "learning_rate": 9.919401469779322e-06,
+      "loss": 0.558,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.46116603914721,
+      "learning_rate": 9.91916797488099e-06,
+      "loss": 0.6229,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.516501639819051,
+      "learning_rate": 9.918934145009066e-06,
+      "loss": 0.565,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 19.08825241435375,
+      "learning_rate": 9.918699980179479e-06,
+      "loss": 0.6104,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 29.787350751591184,
+      "learning_rate": 9.918465480408171e-06,
+      "loss": 0.6396,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.393856299367907,
+      "learning_rate": 9.918230645711108e-06,
+      "loss": 0.5427,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.862433471172531,
+      "learning_rate": 9.917995476104286e-06,
+      "loss": 0.6378,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.72840962278625,
+      "learning_rate": 9.917759971603716e-06,
+      "loss": 0.6191,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.9538386402799,
+      "learning_rate": 9.917524132225436e-06,
+      "loss": 0.5264,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.389598292641276,
+      "learning_rate": 9.917287957985504e-06,
+      "loss": 0.6736,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.597174679361349,
+      "learning_rate": 9.917051448900006e-06,
+      "loss": 0.4488,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.755216784636392,
+      "learning_rate": 9.916814604985044e-06,
+      "loss": 0.6063,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.795883073261075,
+      "learning_rate": 9.916577426256745e-06,
+      "loss": 0.5394,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.038942049273042,
+      "learning_rate": 9.916339912731262e-06,
+      "loss": 0.5616,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.481096309319223,
+      "learning_rate": 9.91610206442477e-06,
+      "loss": 0.6055,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.777091449874407,
+      "learning_rate": 9.915863881353463e-06,
+      "loss": 0.6889,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.687598444599011,
+      "learning_rate": 9.915625363533561e-06,
+      "loss": 0.7052,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.4745346181410595,
+      "learning_rate": 9.915386510981306e-06,
+      "loss": 0.5529,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.263617190196953,
+      "learning_rate": 9.915147323712963e-06,
+      "loss": 0.6115,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7536715921241275,
+      "learning_rate": 9.91490780174482e-06,
+      "loss": 0.7584,
+      "step": 1081
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.512141377324871,
+      "learning_rate": 9.914667945093186e-06,
+      "loss": 0.5123,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.126253160087142,
+      "learning_rate": 9.914427753774394e-06,
+      "loss": 0.5902,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9961258801510837,
+      "learning_rate": 9.914187227804802e-06,
+      "loss": 0.7511,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.702765259080442,
+      "learning_rate": 9.913946367200786e-06,
+      "loss": 0.6019,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.228240940761811,
+      "learning_rate": 9.91370517197875e-06,
+      "loss": 0.6132,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.60281829874014,
+      "learning_rate": 9.91346364215512e-06,
+      "loss": 0.5937,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.766562121704931,
+      "learning_rate": 9.913221777746338e-06,
+      "loss": 0.5365,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.902377309702143,
+      "learning_rate": 9.912979578768877e-06,
+      "loss": 0.5337,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.395471889168595,
+      "learning_rate": 9.91273704523923e-06,
+      "loss": 0.5633,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.100296563066587,
+      "learning_rate": 9.912494177173911e-06,
+      "loss": 0.5516,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8735740967803267,
+      "learning_rate": 9.91225097458946e-06,
+      "loss": 0.7724,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.786890782329433,
+      "learning_rate": 9.912007437502436e-06,
+      "loss": 0.5094,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.694748154755322,
+      "learning_rate": 9.911763565929423e-06,
+      "loss": 0.5867,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4216884161540455,
+      "learning_rate": 9.91151935988703e-06,
+      "loss": 0.8053,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.75502786184883,
+      "learning_rate": 9.911274819391882e-06,
+      "loss": 0.549,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.661031021947113,
+      "learning_rate": 9.911029944460637e-06,
+      "loss": 0.5955,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.274687485837838,
+      "learning_rate": 9.910784735109964e-06,
+      "loss": 0.6249,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.502968819654156,
+      "learning_rate": 9.910539191356565e-06,
+      "loss": 0.6061,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.507554011907327,
+      "learning_rate": 9.91029331321716e-06,
+      "loss": 0.4941,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.374185908074267,
+      "learning_rate": 9.910047100708489e-06,
+      "loss": 0.468,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.351778723864339,
+      "learning_rate": 9.90980055384732e-06,
+      "loss": 0.5921,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.11862936237905,
+      "learning_rate": 9.909553672650441e-06,
+      "loss": 0.521,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 61.08246701392925,
+      "learning_rate": 9.909306457134666e-06,
+      "loss": 0.6671,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.925160532832615,
+      "learning_rate": 9.909058907316826e-06,
+      "loss": 0.496,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.817432749406708,
+      "learning_rate": 9.90881102321378e-06,
+      "loss": 0.5597,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.7873681479500645,
+      "learning_rate": 9.90856280484241e-06,
+      "loss": 0.5634,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.147081020740526,
+      "learning_rate": 9.908314252219613e-06,
+      "loss": 0.5059,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.080970336805907,
+      "learning_rate": 9.908065365362319e-06,
+      "loss": 0.5929,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.146081609665039,
+      "learning_rate": 9.907816144287474e-06,
+      "loss": 0.5431,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.388064306024749,
+      "learning_rate": 9.90756658901205e-06,
+      "loss": 0.5504,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.663632420627557,
+      "learning_rate": 9.907316699553039e-06,
+      "loss": 0.4728,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.375644599111833,
+      "learning_rate": 9.907066475927456e-06,
+      "loss": 0.4905,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 37.533060804224085,
+      "learning_rate": 9.906815918152345e-06,
+      "loss": 0.4642,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9469718988732776,
+      "learning_rate": 9.906565026244766e-06,
+      "loss": 0.8595,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.268276583623765,
+      "learning_rate": 9.906313800221801e-06,
+      "loss": 0.5983,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.12749753621969,
+      "learning_rate": 9.90606224010056e-06,
+      "loss": 0.4989,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.542826725082093,
+      "learning_rate": 9.90581034589817e-06,
+      "loss": 0.5019,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.884988528591188,
+      "learning_rate": 9.905558117631791e-06,
+      "loss": 0.6515,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.417593753524638,
+      "learning_rate": 9.90530555531859e-06,
+      "loss": 0.4691,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.323615187274306,
+      "learning_rate": 9.905052658975771e-06,
+      "loss": 0.571,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.65158072290572,
+      "learning_rate": 9.904799428620553e-06,
+      "loss": 0.6345,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.209336200255796,
+      "learning_rate": 9.90454586427018e-06,
+      "loss": 0.6089,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 24.379900043720003,
+      "learning_rate": 9.90429196594192e-06,
+      "loss": 0.587,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.021299071060428,
+      "learning_rate": 9.90403773365306e-06,
+      "loss": 0.4533,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.793936665808148,
+      "learning_rate": 9.903783167420916e-06,
+      "loss": 0.6902,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.802792062811092,
+      "learning_rate": 9.903528267262816e-06,
+      "loss": 0.6193,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.138017084618811,
+      "learning_rate": 9.903273033196124e-06,
+      "loss": 0.5099,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.314996235150481,
+      "learning_rate": 9.903017465238218e-06,
+      "loss": 0.5408,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 16.512057743586645,
+      "learning_rate": 9.9027615634065e-06,
+      "loss": 0.5677,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 15.641607718361028,
+      "learning_rate": 9.902505327718398e-06,
+      "loss": 0.6105,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.837664883062691,
+      "learning_rate": 9.902248758191359e-06,
+      "loss": 0.6353,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.02896416653501,
+      "learning_rate": 9.901991854842857e-06,
+      "loss": 0.5295,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.125517904876455,
+      "learning_rate": 9.901734617690381e-06,
+      "loss": 0.5542,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.149836862691668,
+      "learning_rate": 9.90147704675145e-06,
+      "loss": 0.5946,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.205642481007212,
+      "learning_rate": 9.901219142043605e-06,
+      "loss": 0.582,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.22787633940205,
+      "learning_rate": 9.900960903584407e-06,
+      "loss": 0.5598,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.004932574992003,
+      "learning_rate": 9.90070233139144e-06,
+      "loss": 0.4486,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.405550018284409,
+      "learning_rate": 9.900443425482314e-06,
+      "loss": 0.8827,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.124243543283232,
+      "learning_rate": 9.900184185874655e-06,
+      "loss": 0.6707,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.402245683233225,
+      "learning_rate": 9.89992461258612e-06,
+      "loss": 0.5414,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.220306781863485,
+      "learning_rate": 9.899664705634386e-06,
+      "loss": 0.5379,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.63646822017257,
+      "learning_rate": 9.899404465037148e-06,
+      "loss": 0.5004,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 14.426474486255486,
+      "learning_rate": 9.899143890812125e-06,
+      "loss": 0.4706,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.475583558075543,
+      "learning_rate": 9.89888298297707e-06,
+      "loss": 0.558,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.36635251817758,
+      "learning_rate": 9.898621741549742e-06,
+      "loss": 0.5529,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.64925882892268,
+      "learning_rate": 9.898360166547932e-06,
+      "loss": 0.5101,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.542554545226208,
+      "learning_rate": 9.89809825798945e-06,
+      "loss": 0.6261,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 22.84052212670384,
+      "learning_rate": 9.897836015892138e-06,
+      "loss": 0.5151,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 13.064663399889968,
+      "learning_rate": 9.897573440273844e-06,
+      "loss": 0.6322,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.413801486333614,
+      "learning_rate": 9.897310531152457e-06,
+      "loss": 0.4827,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.908421271459816,
+      "learning_rate": 9.897047288545875e-06,
+      "loss": 0.4695,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.315558629398586,
+      "learning_rate": 9.896783712472024e-06,
+      "loss": 0.6529,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 45.42488165023227,
+      "learning_rate": 9.896519802948854e-06,
+      "loss": 0.5234,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 23.88693504002204,
+      "learning_rate": 9.896255559994334e-06,
+      "loss": 0.6464,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.046036763570308,
+      "learning_rate": 9.89599098362646e-06,
+      "loss": 0.6195,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 17.723568151008674,
+      "learning_rate": 9.895726073863247e-06,
+      "loss": 0.6183,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.528946109415827,
+      "learning_rate": 9.895460830722736e-06,
+      "loss": 0.5624,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.757122626966352,
+      "learning_rate": 9.895195254222986e-06,
+      "loss": 0.6107,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.513027096384175,
+      "learning_rate": 9.894929344382083e-06,
+      "loss": 0.5645,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.722250566504982,
+      "learning_rate": 9.894663101218136e-06,
+      "loss": 0.549,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 12.45034481378329,
+      "learning_rate": 9.894396524749271e-06,
+      "loss": 0.6114,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.00508985491703,
+      "learning_rate": 9.894129614993643e-06,
+      "loss": 0.451,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.8586324525672255,
+      "learning_rate": 9.89386237196943e-06,
+      "loss": 0.4907,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.939840980451002,
+      "learning_rate": 9.893594795694826e-06,
+      "loss": 0.5033,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.679874467748848,
+      "learning_rate": 9.893326886188053e-06,
+      "loss": 0.5056,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.714592685447564,
+      "learning_rate": 9.893058643467354e-06,
+      "loss": 0.5379,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.204109814935201,
+      "learning_rate": 9.892790067550998e-06,
+      "loss": 0.5229,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.127344392708783,
+      "learning_rate": 9.89252115845727e-06,
+      "loss": 0.5437,
+      "step": 1169
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9245829456352053,
+      "learning_rate": 9.892251916204485e-06,
+      "loss": 0.7932,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.472765064092329,
+      "learning_rate": 9.891982340810972e-06,
+      "loss": 0.6551,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.389326412955032,
+      "learning_rate": 9.891712432295096e-06,
+      "loss": 0.5853,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.224376260368423,
+      "learning_rate": 9.891442190675227e-06,
+      "loss": 0.5379,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 11.513806751104555,
+      "learning_rate": 9.891171615969774e-06,
+      "loss": 0.5546,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.733206752148798,
+      "learning_rate": 9.890900708197161e-06,
+      "loss": 0.5763,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.7600943300143,
+      "learning_rate": 9.890629467375834e-06,
+      "loss": 0.5453,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.674936304822756,
+      "learning_rate": 9.890357893524265e-06,
+      "loss": 0.588,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.144998011639295,
+      "learning_rate": 9.890085986660944e-06,
+      "loss": 0.5312,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.531342403974748,
+      "learning_rate": 9.88981374680439e-06,
+      "loss": 0.6357,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.270282146082957,
+      "learning_rate": 9.889541173973138e-06,
+      "loss": 0.5737,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.873362487849457,
+      "learning_rate": 9.889268268185755e-06,
+      "loss": 0.6392,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.70237596739009,
+      "learning_rate": 9.888995029460817e-06,
+      "loss": 0.5859,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.333882898100374,
+      "learning_rate": 9.888721457816934e-06,
+      "loss": 0.562,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.549724888224208,
+      "learning_rate": 9.888447553272738e-06,
+      "loss": 0.4974,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5535808130262836,
+      "learning_rate": 9.888173315846874e-06,
+      "loss": 0.6982,
+      "step": 1185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.091719824692539,
+      "learning_rate": 9.887898745558023e-06,
+      "loss": 0.4316,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.865691843193967,
+      "learning_rate": 9.887623842424878e-06,
+      "loss": 0.61,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.166259539952375,
+      "learning_rate": 9.887348606466158e-06,
+      "loss": 0.7477,
+      "step": 1188
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.594865978620034,
+      "learning_rate": 9.88707303770061e-06,
+      "loss": 0.6265,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 21.63824007298819,
+      "learning_rate": 9.886797136146994e-06,
+      "loss": 0.5147,
+      "step": 1190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.238225158053643,
+      "learning_rate": 9.8865209018241e-06,
+      "loss": 0.5384,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.490276939616244,
+      "learning_rate": 9.886244334750738e-06,
+      "loss": 0.536,
+      "step": 1192
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.867500519295662,
+      "learning_rate": 9.885967434945743e-06,
+      "loss": 0.5622,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.982640936336614,
+      "learning_rate": 9.885690202427968e-06,
+      "loss": 0.4437,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.963745389626448,
+      "learning_rate": 9.885412637216291e-06,
+      "loss": 0.6153,
+      "step": 1195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.751963758727461,
+      "learning_rate": 9.885134739329616e-06,
+      "loss": 0.5524,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.01508879305325,
+      "learning_rate": 9.884856508786862e-06,
+      "loss": 0.5383,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.586361965779996,
+      "learning_rate": 9.88457794560698e-06,
+      "loss": 0.5391,
+      "step": 1198
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8410560781718301,
+      "learning_rate": 9.884299049808936e-06,
+      "loss": 0.7281,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.835025121677635,
+      "learning_rate": 9.884019821411723e-06,
+      "loss": 0.5362,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.165346236230395,
+      "learning_rate": 9.883740260434353e-06,
+      "loss": 0.4263,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.471739852834671,
+      "learning_rate": 9.883460366895867e-06,
+      "loss": 0.5412,
+      "step": 1202
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.398474156824882,
+      "learning_rate": 9.88318014081532e-06,
+      "loss": 0.6164,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.05576446597958,
+      "learning_rate": 9.882899582211798e-06,
+      "loss": 0.5086,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.92168533413764,
+      "learning_rate": 9.882618691104402e-06,
+      "loss": 0.5315,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.458634623693536,
+      "learning_rate": 9.882337467512263e-06,
+      "loss": 0.6662,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.638303018795646,
+      "learning_rate": 9.88205591145453e-06,
+      "loss": 0.5501,
+      "step": 1207
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.059571190265348,
+      "learning_rate": 9.881774022950373e-06,
+      "loss": 0.5668,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.814809766973392,
+      "learning_rate": 9.881491802018991e-06,
+      "loss": 0.4539,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.455714961006747,
+      "learning_rate": 9.881209248679599e-06,
+      "loss": 0.5803,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.2496994382773705,
+      "learning_rate": 9.880926362951441e-06,
+      "loss": 0.5494,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.270433946881091,
+      "learning_rate": 9.880643144853778e-06,
+      "loss": 0.5831,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.552568891025155,
+      "learning_rate": 9.880359594405897e-06,
+      "loss": 0.6068,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.182638977123081,
+      "learning_rate": 9.880075711627104e-06,
+      "loss": 0.6838,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 40.469214514946486,
+      "learning_rate": 9.879791496536735e-06,
+      "loss": 0.5441,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.676899494614954,
+      "learning_rate": 9.879506949154139e-06,
+      "loss": 0.5626,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.692733987423683,
+      "learning_rate": 9.879222069498695e-06,
+      "loss": 0.4922,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.920559371055724,
+      "learning_rate": 9.8789368575898e-06,
+      "loss": 0.5271,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.068324487054468,
+      "learning_rate": 9.878651313446878e-06,
+      "loss": 0.5284,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.640380241100038,
+      "learning_rate": 9.878365437089373e-06,
+      "loss": 0.566,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.917924533238043,
+      "learning_rate": 9.87807922853675e-06,
+      "loss": 0.3511,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.694740553855398,
+      "learning_rate": 9.8777926878085e-06,
+      "loss": 0.4867,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.979767141547916,
+      "learning_rate": 9.877505814924136e-06,
+      "loss": 0.6289,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.482970407196143,
+      "learning_rate": 9.87721860990319e-06,
+      "loss": 0.6045,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8382461733715878,
+      "learning_rate": 9.876931072765221e-06,
+      "loss": 0.7681,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 14.702064638312786,
+      "learning_rate": 9.87664320352981e-06,
+      "loss": 0.4712,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.970343182218778,
+      "learning_rate": 9.87635500221656e-06,
+      "loss": 0.3906,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.024575959995008,
+      "learning_rate": 9.876066468845093e-06,
+      "loss": 0.5317,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.695708777249256,
+      "learning_rate": 9.875777603435059e-06,
+      "loss": 0.6587,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 10.753464621900129,
+      "learning_rate": 9.875488406006129e-06,
+      "loss": 0.5449,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.508349162312325,
+      "learning_rate": 9.875198876577995e-06,
+      "loss": 0.5356,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.810412552395468,
+      "learning_rate": 9.874909015170373e-06,
+      "loss": 0.5901,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.459260457948165,
+      "learning_rate": 9.874618821803001e-06,
+      "loss": 0.6142,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 16.835803701358717,
+      "learning_rate": 9.874328296495641e-06,
+      "loss": 0.6502,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.67828418905167,
+      "learning_rate": 9.874037439268075e-06,
+      "loss": 0.5239,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.97251478655691,
+      "learning_rate": 9.87374625014011e-06,
+      "loss": 0.5738,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.04257670895652,
+      "learning_rate": 9.873454729131572e-06,
+      "loss": 0.4783,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.832430166612986,
+      "learning_rate": 9.87316287626232e-06,
+      "loss": 0.5856,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.966057612001444,
+      "learning_rate": 9.87287069155222e-06,
+      "loss": 0.6112,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.710946101374645,
+      "learning_rate": 9.872578175021168e-06,
+      "loss": 0.5198,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.643371255201494,
+      "learning_rate": 9.872285326689089e-06,
+      "loss": 0.6151,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.501126410696296,
+      "learning_rate": 9.871992146575922e-06,
+      "loss": 0.5601,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 19.94836798630234,
+      "learning_rate": 9.87169863470163e-06,
+      "loss": 0.5905,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.332193830143737,
+      "learning_rate": 9.871404791086202e-06,
+      "loss": 0.5077,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.229607186652667,
+      "learning_rate": 9.871110615749646e-06,
+      "loss": 0.5554,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.861272160124119,
+      "learning_rate": 9.870816108711996e-06,
+      "loss": 0.668,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.909314866041382,
+      "learning_rate": 9.870521269993303e-06,
+      "loss": 0.6622,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.198624234455503,
+      "learning_rate": 9.870226099613646e-06,
+      "loss": 0.6081,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.927420610502496,
+      "learning_rate": 9.869930597593127e-06,
+      "loss": 0.6504,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.877133221691329,
+      "learning_rate": 9.869634763951865e-06,
+      "loss": 0.5679,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.057747290983144,
+      "learning_rate": 9.869338598710007e-06,
+      "loss": 0.552,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.377134016132539,
+      "learning_rate": 9.869042101887721e-06,
+      "loss": 0.4948,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.932705446629709,
+      "learning_rate": 9.868745273505196e-06,
+      "loss": 0.6068,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.704935330636238,
+      "learning_rate": 9.868448113582642e-06,
+      "loss": 0.5097,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.383375658169992,
+      "learning_rate": 9.8681506221403e-06,
+      "loss": 0.5846,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.050112642905055,
+      "learning_rate": 9.867852799198425e-06,
+      "loss": 0.7494,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.348064565644697,
+      "learning_rate": 9.867554644777295e-06,
+      "loss": 0.4591,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.998906081179129,
+      "learning_rate": 9.867256158897219e-06,
+      "loss": 0.6215,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.522895957965173,
+      "learning_rate": 9.866957341578516e-06,
+      "loss": 0.5506,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.652878600905992,
+      "learning_rate": 9.866658192841536e-06,
+      "loss": 0.5359,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.31999861296592,
+      "learning_rate": 9.866358712706654e-06,
+      "loss": 0.6955,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.737905253810084,
+      "learning_rate": 9.866058901194259e-06,
+      "loss": 0.4949,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.410782907500033,
+      "learning_rate": 9.865758758324767e-06,
+      "loss": 0.5175,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.213942313121277,
+      "learning_rate": 9.865458284118619e-06,
+      "loss": 0.5044,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6224117937416471,
+      "learning_rate": 9.865157478596272e-06,
+      "loss": 0.7603,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.479830343156748,
+      "learning_rate": 9.864856341778213e-06,
+      "loss": 0.5352,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.591875639420023,
+      "learning_rate": 9.864554873684947e-06,
+      "loss": 0.5573,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.236912501843659,
+      "learning_rate": 9.864253074337002e-06,
+      "loss": 0.5452,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.412970164975071,
+      "learning_rate": 9.86395094375493e-06,
+      "loss": 0.5584,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 35.262171186170434,
+      "learning_rate": 9.863648481959305e-06,
+      "loss": 0.4194,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.021098079727297,
+      "learning_rate": 9.863345688970723e-06,
+      "loss": 0.5554,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.075815890824359,
+      "learning_rate": 9.863042564809801e-06,
+      "loss": 0.5485,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.781811327731372,
+      "learning_rate": 9.862739109497184e-06,
+      "loss": 0.5871,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.893098622492569,
+      "learning_rate": 9.862435323053533e-06,
+      "loss": 0.6648,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.193415145543045,
+      "learning_rate": 9.862131205499536e-06,
+      "loss": 0.4706,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7126053290394678,
+      "learning_rate": 9.8618267568559e-06,
+      "loss": 0.6811,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.868849209876627,
+      "learning_rate": 9.86152197714336e-06,
+      "loss": 0.483,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.743118996022989,
+      "learning_rate": 9.861216866382668e-06,
+      "loss": 0.6003,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.07620343833528,
+      "learning_rate": 9.860911424594602e-06,
+      "loss": 0.6838,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.767826754506952,
+      "learning_rate": 9.860605651799958e-06,
+      "loss": 0.4692,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.6111062043338755,
+      "learning_rate": 9.860299548019561e-06,
+      "loss": 0.4969,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.612644874936883,
+      "learning_rate": 9.859993113274256e-06,
+      "loss": 0.5829,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.04454463555107,
+      "learning_rate": 9.859686347584906e-06,
+      "loss": 0.4822,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.540636611747615,
+      "learning_rate": 9.859379250972403e-06,
+      "loss": 0.5637,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.373322149994872,
+      "learning_rate": 9.85907182345766e-06,
+      "loss": 0.6149,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.157045987702165,
+      "learning_rate": 9.85876406506161e-06,
+      "loss": 0.5669,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.358965513412099,
+      "learning_rate": 9.858455975805207e-06,
+      "loss": 0.5426,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.990861207601451,
+      "learning_rate": 9.858147555709436e-06,
+      "loss": 0.5759,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.7336554684540335,
+      "learning_rate": 9.857838804795297e-06,
+      "loss": 0.6022,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.419133584918708,
+      "learning_rate": 9.857529723083813e-06,
+      "loss": 0.6075,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.482707708596139,
+      "learning_rate": 9.857220310596032e-06,
+      "loss": 0.5528,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.370641258791128,
+      "learning_rate": 9.856910567353023e-06,
+      "loss": 0.6903,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.046880569352994,
+      "learning_rate": 9.856600493375882e-06,
+      "loss": 0.5473,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.945008898358774,
+      "learning_rate": 9.856290088685719e-06,
+      "loss": 0.6197,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.280116192443807,
+      "learning_rate": 9.855979353303672e-06,
+      "loss": 0.5249,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.345792847960627,
+      "learning_rate": 9.855668287250904e-06,
+      "loss": 0.5609,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 23.558555130368777,
+      "learning_rate": 9.855356890548593e-06,
+      "loss": 0.5587,
+      "step": 1297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.501832879477308,
+      "learning_rate": 9.855045163217944e-06,
+      "loss": 0.4514,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.776201266768339,
+      "learning_rate": 9.854733105280187e-06,
+      "loss": 0.5566,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.612836327620698,
+      "learning_rate": 9.854420716756573e-06,
+      "loss": 0.5273,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.108565806657645,
+      "learning_rate": 9.854107997668369e-06,
+      "loss": 0.6152,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.979573976806221,
+      "learning_rate": 9.853794948036872e-06,
+      "loss": 0.6671,
+      "step": 1302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.926645145788536,
+      "learning_rate": 9.853481567883401e-06,
+      "loss": 0.5057,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.921451079022426,
+      "learning_rate": 9.853167857229295e-06,
+      "loss": 0.5622,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.611374194426139,
+      "learning_rate": 9.852853816095916e-06,
+      "loss": 0.5456,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.7130991701509,
+      "learning_rate": 9.85253944450465e-06,
+      "loss": 0.5997,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.633880790338663,
+      "learning_rate": 9.852224742476902e-06,
+      "loss": 0.6379,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.072183555737692,
+      "learning_rate": 9.851909710034102e-06,
+      "loss": 0.4764,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.464638521454184,
+      "learning_rate": 9.851594347197703e-06,
+      "loss": 0.5951,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.2812480921479015,
+      "learning_rate": 9.851278653989182e-06,
+      "loss": 0.6428,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.299488427302251,
+      "learning_rate": 9.850962630430031e-06,
+      "loss": 0.5041,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.96291677631613,
+      "learning_rate": 9.850646276541777e-06,
+      "loss": 0.5483,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.13785874688469,
+      "learning_rate": 9.850329592345958e-06,
+      "loss": 0.4894,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.162334514831798,
+      "learning_rate": 9.850012577864139e-06,
+      "loss": 0.5149,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.884232961267355,
+      "learning_rate": 9.849695233117907e-06,
+      "loss": 0.5397,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.851861823837599,
+      "learning_rate": 9.849377558128872e-06,
+      "loss": 0.5781,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.914030761597541,
+      "learning_rate": 9.849059552918666e-06,
+      "loss": 0.6185,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.393387060067573,
+      "learning_rate": 9.848741217508947e-06,
+      "loss": 0.5109,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.928601015307863,
+      "learning_rate": 9.848422551921385e-06,
+      "loss": 0.6639,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 13.335740889940912,
+      "learning_rate": 9.848103556177688e-06,
+      "loss": 0.5335,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.606612662546631,
+      "learning_rate": 9.847784230299573e-06,
+      "loss": 0.4964,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.22750197559209,
+      "learning_rate": 9.847464574308787e-06,
+      "loss": 0.4636,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.320269562423957,
+      "learning_rate": 9.847144588227099e-06,
+      "loss": 0.578,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.551539302974614,
+      "learning_rate": 9.846824272076293e-06,
+      "loss": 0.5791,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.84293988825581,
+      "learning_rate": 9.846503625878184e-06,
+      "loss": 0.8032,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.767269525015572,
+      "learning_rate": 9.846182649654606e-06,
+      "loss": 0.5344,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.899994168121567,
+      "learning_rate": 9.845861343427416e-06,
+      "loss": 0.5255,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.11009093733848,
+      "learning_rate": 9.845539707218497e-06,
+      "loss": 0.5944,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.268452030448266,
+      "learning_rate": 9.845217741049747e-06,
+      "loss": 0.569,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.555010391122396,
+      "learning_rate": 9.844895444943093e-06,
+      "loss": 0.6883,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.311738075155416,
+      "learning_rate": 9.84457281892048e-06,
+      "loss": 0.5564,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.266520422260563,
+      "learning_rate": 9.844249863003876e-06,
+      "loss": 0.5085,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.34877333437171,
+      "learning_rate": 9.843926577215278e-06,
+      "loss": 0.4798,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.951088342575049,
+      "learning_rate": 9.843602961576695e-06,
+      "loss": 0.5294,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.909514161510429,
+      "learning_rate": 9.843279016110167e-06,
+      "loss": 0.6593,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.593405887051752,
+      "learning_rate": 9.842954740837753e-06,
+      "loss": 0.5772,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.86187531272817,
+      "learning_rate": 9.842630135781535e-06,
+      "loss": 0.5389,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.5721504876739125,
+      "learning_rate": 9.842305200963613e-06,
+      "loss": 0.6549,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.923072836563848,
+      "learning_rate": 9.84197993640612e-06,
+      "loss": 0.5119,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.257543928164262,
+      "learning_rate": 9.8416543421312e-06,
+      "loss": 0.6092,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.289881661891871,
+      "learning_rate": 9.841328418161027e-06,
+      "loss": 0.5131,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.524628582166992,
+      "learning_rate": 9.841002164517795e-06,
+      "loss": 0.5288,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.047032400299166,
+      "learning_rate": 9.840675581223719e-06,
+      "loss": 0.5944,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.515114201719064,
+      "learning_rate": 9.84034866830104e-06,
+      "loss": 0.4634,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.302935186557862,
+      "learning_rate": 9.840021425772017e-06,
+      "loss": 0.6321,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.527721018159603,
+      "learning_rate": 9.839693853658934e-06,
+      "loss": 0.544,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.04857796103804,
+      "learning_rate": 9.8393659519841e-06,
+      "loss": 0.4567,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.480904224296905,
+      "learning_rate": 9.83903772076984e-06,
+      "loss": 0.5274,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.321314882150382,
+      "learning_rate": 9.83870916003851e-06,
+      "loss": 0.5692,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.647038288667122,
+      "learning_rate": 9.838380269812477e-06,
+      "loss": 0.597,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.426862872562679,
+      "learning_rate": 9.838051050114141e-06,
+      "loss": 0.5096,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.897288654385179,
+      "learning_rate": 9.83772150096592e-06,
+      "loss": 0.5617,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.557738042707938,
+      "learning_rate": 9.837391622390256e-06,
+      "loss": 0.5351,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.096350288590614,
+      "learning_rate": 9.837061414409608e-06,
+      "loss": 0.5209,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0063939072038415,
+      "learning_rate": 9.836730877046466e-06,
+      "loss": 0.531,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.48100498961106,
+      "learning_rate": 9.836400010323337e-06,
+      "loss": 0.6288,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.607679994024629,
+      "learning_rate": 9.836068814262752e-06,
+      "loss": 0.5434,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.911026995856014,
+      "learning_rate": 9.835737288887262e-06,
+      "loss": 0.6504,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.382090826733301,
+      "learning_rate": 9.835405434219442e-06,
+      "loss": 0.5622,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.272014186684673,
+      "learning_rate": 9.835073250281894e-06,
+      "loss": 0.5695,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.1557255019240795,
+      "learning_rate": 9.834740737097234e-06,
+      "loss": 0.543,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.337160147512249,
+      "learning_rate": 9.834407894688108e-06,
+      "loss": 0.5872,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.314342775357393,
+      "learning_rate": 9.834074723077179e-06,
+      "loss": 0.5376,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.02257672942302,
+      "learning_rate": 9.833741222287135e-06,
+      "loss": 0.5571,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.536868198357297,
+      "learning_rate": 9.833407392340686e-06,
+      "loss": 0.5262,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.498845141234568,
+      "learning_rate": 9.833073233260563e-06,
+      "loss": 0.6505,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.134426199702468,
+      "learning_rate": 9.832738745069524e-06,
+      "loss": 0.5537,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.725452151602584,
+      "learning_rate": 9.832403927790343e-06,
+      "loss": 0.554,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.191923214167629,
+      "learning_rate": 9.832068781445824e-06,
+      "loss": 0.4878,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.686506462365005,
+      "learning_rate": 9.83173330605878e-06,
+      "loss": 0.5369,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.988085464911334,
+      "learning_rate": 9.831397501652065e-06,
+      "loss": 0.6913,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.211835424806197,
+      "learning_rate": 9.831061368248541e-06,
+      "loss": 0.5309,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.693968537471919,
+      "learning_rate": 9.830724905871098e-06,
+      "loss": 0.5623,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.52898802008261,
+      "learning_rate": 9.830388114542648e-06,
+      "loss": 0.4891,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.131619670866672,
+      "learning_rate": 9.830050994286124e-06,
+      "loss": 0.5206,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.994994849329319,
+      "learning_rate": 9.829713545124485e-06,
+      "loss": 0.531,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.2302276485291825,
+      "learning_rate": 9.829375767080707e-06,
+      "loss": 0.5228,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.948315180448217,
+      "learning_rate": 9.829037660177792e-06,
+      "loss": 0.6259,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.072455918982279,
+      "learning_rate": 9.828699224438762e-06,
+      "loss": 0.5937,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.704358155849851,
+      "learning_rate": 9.828360459886667e-06,
+      "loss": 0.5723,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9436266270149014,
+      "learning_rate": 9.828021366544574e-06,
+      "loss": 0.4238,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.818724718824392,
+      "learning_rate": 9.82768194443557e-06,
+      "loss": 0.7347,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.782745713213662,
+      "learning_rate": 9.827342193582771e-06,
+      "loss": 0.4903,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.4528068099973614,
+      "learning_rate": 9.827002114009315e-06,
+      "loss": 0.5618,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.085648430425007,
+      "learning_rate": 9.826661705738357e-06,
+      "loss": 0.5892,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.118572859209323,
+      "learning_rate": 9.826320968793075e-06,
+      "loss": 0.5553,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.601778742471703,
+      "learning_rate": 9.825979903196676e-06,
+      "loss": 0.6428,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.55863497642136,
+      "learning_rate": 9.825638508972384e-06,
+      "loss": 0.5525,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.749431014650006,
+      "learning_rate": 9.825296786143445e-06,
+      "loss": 0.5172,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.242137496905102,
+      "learning_rate": 9.824954734733132e-06,
+      "loss": 0.562,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.182757118222098,
+      "learning_rate": 9.824612354764734e-06,
+      "loss": 0.436,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.365251554419271,
+      "learning_rate": 9.824269646261563e-06,
+      "loss": 0.47,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.757781531538353,
+      "learning_rate": 9.823926609246964e-06,
+      "loss": 0.521,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.80722984980495,
+      "learning_rate": 9.823583243744288e-06,
+      "loss": 0.5274,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.080216990445674,
+      "learning_rate": 9.823239549776924e-06,
+      "loss": 0.6697,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.665767420918346,
+      "learning_rate": 9.82289552736827e-06,
+      "loss": 0.5285,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.918232629053387,
+      "learning_rate": 9.822551176541755e-06,
+      "loss": 0.5596,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.28116519229621,
+      "learning_rate": 9.822206497320828e-06,
+      "loss": 0.5857,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.2960184563941635,
+      "learning_rate": 9.82186148972896e-06,
+      "loss": 0.5307,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.0951945647182,
+      "learning_rate": 9.821516153789643e-06,
+      "loss": 0.459,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.419193175419042,
+      "learning_rate": 9.821170489526395e-06,
+      "loss": 0.6279,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.98589580976765,
+      "learning_rate": 9.820824496962753e-06,
+      "loss": 0.5539,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.153141817252099,
+      "learning_rate": 9.820478176122278e-06,
+      "loss": 0.545,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.735912224316904,
+      "learning_rate": 9.820131527028554e-06,
+      "loss": 0.4311,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.981965679258029,
+      "learning_rate": 9.819784549705182e-06,
+      "loss": 0.458,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.160194099841661,
+      "learning_rate": 9.819437244175796e-06,
+      "loss": 0.5716,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.965368091874903,
+      "learning_rate": 9.81908961046404e-06,
+      "loss": 0.5495,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.363784823127846,
+      "learning_rate": 9.81874164859359e-06,
+      "loss": 0.6493,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.333262823128347,
+      "learning_rate": 9.81839335858814e-06,
+      "loss": 0.5146,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.683468922137128,
+      "learning_rate": 9.818044740471407e-06,
+      "loss": 0.5117,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.346019747881089,
+      "learning_rate": 9.817695794267129e-06,
+      "loss": 0.6572,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0292963019127805,
+      "learning_rate": 9.817346519999067e-06,
+      "loss": 0.783,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.947458700186012,
+      "learning_rate": 9.816996917691008e-06,
+      "loss": 0.544,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.980081297090281,
+      "learning_rate": 9.816646987366758e-06,
+      "loss": 0.5484,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.60291779409384,
+      "learning_rate": 9.816296729050145e-06,
+      "loss": 0.5803,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.12684858035981,
+      "learning_rate": 9.815946142765017e-06,
+      "loss": 0.5319,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.088140914678528,
+      "learning_rate": 9.815595228535253e-06,
+      "loss": 0.5927,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.191957224988699,
+      "learning_rate": 9.815243986384746e-06,
+      "loss": 0.5056,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.4518757533072,
+      "learning_rate": 9.814892416337412e-06,
+      "loss": 0.6289,
+      "step": 1419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.076092184639515,
+      "learning_rate": 9.814540518417195e-06,
+      "loss": 0.5975,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.790296345173433,
+      "learning_rate": 9.814188292648054e-06,
+      "loss": 0.5618,
+      "step": 1421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.759914764938543,
+      "learning_rate": 9.813835739053978e-06,
+      "loss": 0.5451,
+      "step": 1422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.262499139984985,
+      "learning_rate": 9.813482857658971e-06,
+      "loss": 0.5268,
+      "step": 1423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.168839555501082,
+      "learning_rate": 9.813129648487066e-06,
+      "loss": 0.4878,
+      "step": 1424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.186149617198051,
+      "learning_rate": 9.81277611156231e-06,
+      "loss": 0.562,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.649300638863543,
+      "learning_rate": 9.812422246908782e-06,
+      "loss": 0.5992,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.886929355318595,
+      "learning_rate": 9.812068054550576e-06,
+      "loss": 0.4694,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.77931824041604,
+      "learning_rate": 9.811713534511813e-06,
+      "loss": 0.6739,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 16.75728615932642,
+      "learning_rate": 9.811358686816634e-06,
+      "loss": 0.5597,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.790333646936081,
+      "learning_rate": 9.811003511489198e-06,
+      "loss": 0.5417,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.7782976782069815,
+      "learning_rate": 9.810648008553698e-06,
+      "loss": 0.5801,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.892334022256463,
+      "learning_rate": 9.810292178034338e-06,
+      "loss": 0.495,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.278267694356492,
+      "learning_rate": 9.809936019955348e-06,
+      "loss": 0.5622,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.490520404763158,
+      "learning_rate": 9.809579534340983e-06,
+      "loss": 0.6188,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.631413668834407,
+      "learning_rate": 9.809222721215516e-06,
+      "loss": 0.55,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.054870158144142,
+      "learning_rate": 9.808865580603246e-06,
+      "loss": 0.7351,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.189299306732304,
+      "learning_rate": 9.808508112528493e-06,
+      "loss": 0.619,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.812589014061041,
+      "learning_rate": 9.808150317015598e-06,
+      "loss": 0.5366,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.890764128490912,
+      "learning_rate": 9.807792194088924e-06,
+      "loss": 0.5425,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.384234300908147,
+      "learning_rate": 9.80743374377286e-06,
+      "loss": 0.6365,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.090489958942003,
+      "learning_rate": 9.807074966091814e-06,
+      "loss": 0.5459,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.900090999075401,
+      "learning_rate": 9.806715861070218e-06,
+      "loss": 0.6259,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.772638239308611,
+      "learning_rate": 9.806356428732522e-06,
+      "loss": 0.6004,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7954650708179436,
+      "learning_rate": 9.805996669103205e-06,
+      "loss": 0.7385,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.638967265634591,
+      "learning_rate": 9.805636582206766e-06,
+      "loss": 0.542,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.83052377996743,
+      "learning_rate": 9.805276168067724e-06,
+      "loss": 0.6712,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.925855616578825,
+      "learning_rate": 9.804915426710619e-06,
+      "loss": 0.6038,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.417490962740051,
+      "learning_rate": 9.80455435816002e-06,
+      "loss": 0.5925,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.495565594345446,
+      "learning_rate": 9.80419296244051e-06,
+      "loss": 0.6497,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.444630765134007,
+      "learning_rate": 9.803831239576704e-06,
+      "loss": 0.6904,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.016917603398411,
+      "learning_rate": 9.803469189593228e-06,
+      "loss": 0.4859,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5531477457543574,
+      "learning_rate": 9.80310681251474e-06,
+      "loss": 0.5581,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.922098748638693,
+      "learning_rate": 9.802744108365913e-06,
+      "loss": 0.5376,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.33873555273009,
+      "learning_rate": 9.80238107717145e-06,
+      "loss": 0.5618,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7433405745906834,
+      "learning_rate": 9.802017718956067e-06,
+      "loss": 0.7308,
+      "step": 1455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.674867810857108,
+      "learning_rate": 9.801654033744509e-06,
+      "loss": 0.5727,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.696083480711868,
+      "learning_rate": 9.801290021561544e-06,
+      "loss": 0.6168,
+      "step": 1457
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 14.226342180272633,
+      "learning_rate": 9.800925682431955e-06,
+      "loss": 0.5734,
+      "step": 1458
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.346123745398971,
+      "learning_rate": 9.800561016380556e-06,
+      "loss": 0.498,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.288854458591826,
+      "learning_rate": 9.800196023432176e-06,
+      "loss": 0.5101,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.918055809610522,
+      "learning_rate": 9.79983070361167e-06,
+      "loss": 0.4769,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.982010230604236,
+      "learning_rate": 9.799465056943917e-06,
+      "loss": 0.4978,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.312033577363671,
+      "learning_rate": 9.799099083453812e-06,
+      "loss": 0.6203,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.997035616050602,
+      "learning_rate": 9.798732783166281e-06,
+      "loss": 0.6,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.01349445067279,
+      "learning_rate": 9.798366156106263e-06,
+      "loss": 0.6094,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.453574008377757,
+      "learning_rate": 9.797999202298728e-06,
+      "loss": 0.5415,
+      "step": 1466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.0745820605414655,
+      "learning_rate": 9.797631921768657e-06,
+      "loss": 0.5272,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.110855340875,
+      "learning_rate": 9.797264314541069e-06,
+      "loss": 0.5098,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.70266472622433,
+      "learning_rate": 9.79689638064099e-06,
+      "loss": 0.5286,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.71925054783178,
+      "learning_rate": 9.796528120093477e-06,
+      "loss": 0.6579,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.704938480411034,
+      "learning_rate": 9.796159532923607e-06,
+      "loss": 0.5948,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.8097242016416475,
+      "learning_rate": 9.795790619156478e-06,
+      "loss": 0.6222,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.7196306582879695,
+      "learning_rate": 9.795421378817214e-06,
+      "loss": 0.5048,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.954766567313106,
+      "learning_rate": 9.795051811930956e-06,
+      "loss": 0.594,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.467488044569055,
+      "learning_rate": 9.79468191852287e-06,
+      "loss": 0.5935,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 13.075159100256746,
+      "learning_rate": 9.794311698618144e-06,
+      "loss": 0.5437,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.209390646893176,
+      "learning_rate": 9.793941152241992e-06,
+      "loss": 0.6093,
+      "step": 1477
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.489192601105646,
+      "learning_rate": 9.793570279419642e-06,
+      "loss": 0.4502,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.731913306637849,
+      "learning_rate": 9.79319908017635e-06,
+      "loss": 0.5275,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.222603986300755,
+      "learning_rate": 9.792827554537394e-06,
+      "loss": 0.5709,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.772245792711384,
+      "learning_rate": 9.792455702528075e-06,
+      "loss": 0.5907,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.563406426079585,
+      "learning_rate": 9.79208352417371e-06,
+      "loss": 0.4891,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.576952811241028,
+      "learning_rate": 9.791711019499644e-06,
+      "loss": 0.615,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.433145205143825,
+      "learning_rate": 9.791338188531244e-06,
+      "loss": 0.5312,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.271889096470585,
+      "learning_rate": 9.7909650312939e-06,
+      "loss": 0.7059,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.53966429885603,
+      "learning_rate": 9.79059154781302e-06,
+      "loss": 0.5432,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.110982084642159,
+      "learning_rate": 9.790217738114037e-06,
+      "loss": 0.5352,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.1075442592600755,
+      "learning_rate": 9.789843602222406e-06,
+      "loss": 0.5354,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.053043421547403,
+      "learning_rate": 9.789469140163604e-06,
+      "loss": 0.4067,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.981768221153617,
+      "learning_rate": 9.78909435196313e-06,
+      "loss": 0.4674,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.3457040398253985,
+      "learning_rate": 9.788719237646503e-06,
+      "loss": 0.6617,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.9477716999275545,
+      "learning_rate": 9.78834379723927e-06,
+      "loss": 0.4599,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.349310682288395,
+      "learning_rate": 9.787968030766997e-06,
+      "loss": 0.5996,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.4717897473990185,
+      "learning_rate": 9.787591938255271e-06,
+      "loss": 0.5099,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.976351562449752,
+      "learning_rate": 9.787215519729702e-06,
+      "loss": 0.682,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 15.177075787357111,
+      "learning_rate": 9.786838775215922e-06,
+      "loss": 0.536,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 12.359116264099091,
+      "learning_rate": 9.786461704739588e-06,
+      "loss": 0.5751,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.439113610106464,
+      "learning_rate": 9.786084308326372e-06,
+      "loss": 0.5613,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.429774925601567,
+      "learning_rate": 9.785706586001977e-06,
+      "loss": 0.4493,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.403242308011926,
+      "learning_rate": 9.785328537792126e-06,
+      "loss": 0.5035,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.347289641371294,
+      "learning_rate": 9.784950163722555e-06,
+      "loss": 0.5773,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.62047436549887,
+      "learning_rate": 9.784571463819037e-06,
+      "loss": 0.5297,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.9792325362031695,
+      "learning_rate": 9.784192438107358e-06,
+      "loss": 0.5502,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.2786315900406935,
+      "learning_rate": 9.783813086613325e-06,
+      "loss": 0.5671,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.251081909324777,
+      "learning_rate": 9.783433409362774e-06,
+      "loss": 0.5479,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 11.263722032405084,
+      "learning_rate": 9.783053406381557e-06,
+      "loss": 0.5958,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.913828832160016,
+      "learning_rate": 9.78267307769555e-06,
+      "loss": 0.5669,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.598788606713529,
+      "learning_rate": 9.782292423330654e-06,
+      "loss": 0.5623,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.676662959127634,
+      "learning_rate": 9.781911443312788e-06,
+      "loss": 0.6038,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.963315349799748,
+      "learning_rate": 9.781530137667895e-06,
+      "loss": 0.6268,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 18.867946287532483,
+      "learning_rate": 9.781148506421944e-06,
+      "loss": 0.5516,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.618551044321482,
+      "learning_rate": 9.780766549600917e-06,
+      "loss": 0.5565,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.941740592283598,
+      "learning_rate": 9.780384267230828e-06,
+      "loss": 0.4293,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.801464219326943,
+      "learning_rate": 9.780001659337704e-06,
+      "loss": 0.4867,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.273150489975363,
+      "learning_rate": 9.779618725947602e-06,
+      "loss": 0.4593,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.973425596701881,
+      "learning_rate": 9.779235467086598e-06,
+      "loss": 0.4577,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.3408401961030405,
+      "learning_rate": 9.77885188278079e-06,
+      "loss": 0.5653,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.762919365210727,
+      "learning_rate": 9.778467973056297e-06,
+      "loss": 0.5991,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.284172144004119,
+      "learning_rate": 9.778083737939265e-06,
+      "loss": 0.4846,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.853615463173143,
+      "learning_rate": 9.777699177455856e-06,
+      "loss": 0.492,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.823364516623068,
+      "learning_rate": 9.777314291632257e-06,
+      "loss": 0.527,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.188143635371158,
+      "learning_rate": 9.776929080494677e-06,
+      "loss": 0.5097,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.184165657153025,
+      "learning_rate": 9.776543544069348e-06,
+      "loss": 0.6805,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.113106966340889,
+      "learning_rate": 9.776157682382523e-06,
+      "loss": 0.5225,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.234532958424037,
+      "learning_rate": 9.775771495460477e-06,
+      "loss": 0.6501,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.1697806951074625,
+      "learning_rate": 9.77538498332951e-06,
+      "loss": 0.5007,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.410729443659244,
+      "learning_rate": 9.774998146015939e-06,
+      "loss": 0.6774,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1191506642318965,
+      "learning_rate": 9.774610983546105e-06,
+      "loss": 0.7388,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 42.7951041327968,
+      "learning_rate": 9.774223495946374e-06,
+      "loss": 0.6296,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.481339079073851,
+      "learning_rate": 9.773835683243133e-06,
+      "loss": 0.5424,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8619175953142686,
+      "learning_rate": 9.77344754546279e-06,
+      "loss": 0.5986,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.004946473865461,
+      "learning_rate": 9.773059082631773e-06,
+      "loss": 0.5967,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.537906354422441,
+      "learning_rate": 9.772670294776537e-06,
+      "loss": 0.5529,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.872563179580495,
+      "learning_rate": 9.772281181923557e-06,
+      "loss": 0.5283,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.807697703514108,
+      "learning_rate": 9.771891744099327e-06,
+      "loss": 0.6221,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.266934233124818,
+      "learning_rate": 9.771501981330369e-06,
+      "loss": 0.5451,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.79542815902704,
+      "learning_rate": 9.771111893643224e-06,
+      "loss": 0.5336,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.993872722191618,
+      "learning_rate": 9.770721481064453e-06,
+      "loss": 0.5575,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0121543153306916,
+      "learning_rate": 9.770330743620644e-06,
+      "loss": 0.831,
+      "step": 1539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.9734262373333245,
+      "learning_rate": 9.769939681338401e-06,
+      "loss": 0.4494,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.90713082867934,
+      "learning_rate": 9.769548294244358e-06,
+      "loss": 0.6162,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.880318948526877,
+      "learning_rate": 9.769156582365163e-06,
+      "loss": 0.4822,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.837035908594699,
+      "learning_rate": 9.768764545727492e-06,
+      "loss": 0.4957,
+      "step": 1543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.889582128136806,
+      "learning_rate": 9.768372184358039e-06,
+      "loss": 0.6073,
+      "step": 1544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.7013891536035795,
+      "learning_rate": 9.767979498283526e-06,
+      "loss": 0.5299,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.548261371660875,
+      "learning_rate": 9.76758648753069e-06,
+      "loss": 0.5439,
+      "step": 1546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.472778447747078,
+      "learning_rate": 9.767193152126292e-06,
+      "loss": 0.664,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1894210624026136,
+      "learning_rate": 9.766799492097118e-06,
+      "loss": 0.7437,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.046212973559433,
+      "learning_rate": 9.766405507469979e-06,
+      "loss": 0.4932,
+      "step": 1549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.396065188664228,
+      "learning_rate": 9.766011198271695e-06,
+      "loss": 0.5981,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 39.68025857616991,
+      "learning_rate": 9.765616564529123e-06,
+      "loss": 0.6189,
+      "step": 1551
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.703462197214206,
+      "learning_rate": 9.765221606269133e-06,
+      "loss": 0.6126,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.241161332251329,
+      "learning_rate": 9.764826323518621e-06,
+      "loss": 0.5408,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.982204097301023,
+      "learning_rate": 9.764430716304504e-06,
+      "loss": 0.516,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.072322093797773,
+      "learning_rate": 9.764034784653722e-06,
+      "loss": 0.5095,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.012524425361757,
+      "learning_rate": 9.763638528593235e-06,
+      "loss": 0.4902,
+      "step": 1556
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.0078478561508035,
+      "learning_rate": 9.763241948150027e-06,
+      "loss": 0.4465,
+      "step": 1557
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.11697839512625,
+      "learning_rate": 9.7628450433511e-06,
+      "loss": 0.5475,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 13.983969390956403,
+      "learning_rate": 9.762447814223485e-06,
+      "loss": 0.6214,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.542412291774303,
+      "learning_rate": 9.762050260794232e-06,
+      "loss": 0.5017,
+      "step": 1560
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.23414240940961,
+      "learning_rate": 9.761652383090412e-06,
+      "loss": 0.5289,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 17.83752918334586,
+      "learning_rate": 9.761254181139117e-06,
+      "loss": 0.5753,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.427992097483075,
+      "learning_rate": 9.760855654967464e-06,
+      "loss": 0.4261,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.449165076270983,
+      "learning_rate": 9.760456804602592e-06,
+      "loss": 0.4947,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.257783327712196,
+      "learning_rate": 9.760057630071657e-06,
+      "loss": 0.5964,
+      "step": 1565
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.1492960424996985,
+      "learning_rate": 9.759658131401847e-06,
+      "loss": 0.4975,
+      "step": 1566
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.869203342311778,
+      "learning_rate": 9.75925830862036e-06,
+      "loss": 0.6203,
+      "step": 1567
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.458329356022955,
+      "learning_rate": 9.758858161754427e-06,
+      "loss": 0.5152,
+      "step": 1568
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.010250131424764,
+      "learning_rate": 9.758457690831294e-06,
+      "loss": 0.5262,
+      "step": 1569
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9782522370243085,
+      "learning_rate": 9.758056895878232e-06,
+      "loss": 0.6553,
+      "step": 1570
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 24.457514928253953,
+      "learning_rate": 9.757655776922533e-06,
+      "loss": 0.5977,
+      "step": 1571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.646879990077473,
+      "learning_rate": 9.75725433399151e-06,
+      "loss": 0.526,
+      "step": 1572
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.352787870362106,
+      "learning_rate": 9.7568525671125e-06,
+      "loss": 0.5007,
+      "step": 1573
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.751118064545964,
+      "learning_rate": 9.756450476312864e-06,
+      "loss": 0.5015,
+      "step": 1574
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.938840371985883,
+      "learning_rate": 9.75604806161998e-06,
+      "loss": 0.6249,
+      "step": 1575
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.115657817352322,
+      "learning_rate": 9.755645323061253e-06,
+      "loss": 0.566,
+      "step": 1576
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.17880161682341,
+      "learning_rate": 9.755242260664107e-06,
+      "loss": 0.4023,
+      "step": 1577
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.39716445810216,
+      "learning_rate": 9.754838874455987e-06,
+      "loss": 0.6725,
+      "step": 1578
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.748301927542967,
+      "learning_rate": 9.754435164464362e-06,
+      "loss": 0.5347,
+      "step": 1579
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.2012815437983555,
+      "learning_rate": 9.754031130716726e-06,
+      "loss": 0.5243,
+      "step": 1580
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.5853992295843335,
+      "learning_rate": 9.75362677324059e-06,
+      "loss": 0.4795,
+      "step": 1581
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.110381690620701,
+      "learning_rate": 9.753222092063486e-06,
+      "loss": 0.7845,
+      "step": 1582
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.940582598967454,
+      "learning_rate": 9.752817087212976e-06,
+      "loss": 0.5132,
+      "step": 1583
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.905799302443108,
+      "learning_rate": 9.752411758716635e-06,
+      "loss": 0.6229,
+      "step": 1584
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.949276629175285,
+      "learning_rate": 9.752006106602065e-06,
+      "loss": 0.5545,
+      "step": 1585
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.755172743056997,
+      "learning_rate": 9.751600130896892e-06,
+      "loss": 0.4659,
+      "step": 1586
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.057624985372731,
+      "learning_rate": 9.751193831628759e-06,
+      "loss": 0.5197,
+      "step": 1587
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.706503895863195,
+      "learning_rate": 9.750787208825334e-06,
+      "loss": 0.4789,
+      "step": 1588
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.186229458277802,
+      "learning_rate": 9.750380262514304e-06,
+      "loss": 0.768,
+      "step": 1589
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.417338606611095,
+      "learning_rate": 9.749972992723381e-06,
+      "loss": 0.5508,
+      "step": 1590
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8743553571220364,
+      "learning_rate": 9.749565399480301e-06,
+      "loss": 0.78,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.707341282625887,
+      "learning_rate": 9.749157482812817e-06,
+      "loss": 0.5865,
+      "step": 1592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.435338128160739,
+      "learning_rate": 9.748749242748705e-06,
+      "loss": 0.526,
+      "step": 1593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9875418946624277,
+      "learning_rate": 9.748340679315766e-06,
+      "loss": 0.731,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.774829024309479,
+      "learning_rate": 9.747931792541822e-06,
+      "loss": 0.5321,
+      "step": 1595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.822878476581114,
+      "learning_rate": 9.747522582454716e-06,
+      "loss": 0.5348,
+      "step": 1596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.888656896305578,
+      "learning_rate": 9.747113049082314e-06,
+      "loss": 0.4571,
+      "step": 1597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.813053459089294,
+      "learning_rate": 9.746703192452502e-06,
+      "loss": 0.534,
+      "step": 1598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.299979107014019,
+      "learning_rate": 9.746293012593189e-06,
+      "loss": 0.5672,
+      "step": 1599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.220764851179996,
+      "learning_rate": 9.745882509532307e-06,
+      "loss": 0.6394,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.15491109911858,
+      "learning_rate": 9.745471683297812e-06,
+      "loss": 0.6696,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.986716621973081,
+      "learning_rate": 9.745060533917675e-06,
+      "loss": 0.5262,
+      "step": 1602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.795238593269832,
+      "learning_rate": 9.744649061419898e-06,
+      "loss": 0.4626,
+      "step": 1603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.552133475551694,
+      "learning_rate": 9.744237265832495e-06,
+      "loss": 0.6876,
+      "step": 1604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.889173445515535,
+      "learning_rate": 9.743825147183513e-06,
+      "loss": 0.549,
+      "step": 1605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.990364720241643,
+      "learning_rate": 9.743412705501013e-06,
+      "loss": 0.6906,
+      "step": 1606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.963352755176244,
+      "learning_rate": 9.742999940813081e-06,
+      "loss": 0.7769,
+      "step": 1607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.206308830558893,
+      "learning_rate": 9.742586853147821e-06,
+      "loss": 0.433,
+      "step": 1608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.101986571077013,
+      "learning_rate": 9.742173442533369e-06,
+      "loss": 0.5274,
+      "step": 1609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.203432000171883,
+      "learning_rate": 9.741759708997871e-06,
+      "loss": 0.5975,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.525453707046135,
+      "learning_rate": 9.741345652569503e-06,
+      "loss": 0.5613,
+      "step": 1611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.623235248867816,
+      "learning_rate": 9.74093127327646e-06,
+      "loss": 0.5773,
+      "step": 1612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.443391186724636,
+      "learning_rate": 9.740516571146959e-06,
+      "loss": 0.5364,
+      "step": 1613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.230766856661008,
+      "learning_rate": 9.740101546209238e-06,
+      "loss": 0.697,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.7954685630825225,
+      "learning_rate": 9.73968619849156e-06,
+      "loss": 0.5007,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.917901644005734,
+      "learning_rate": 9.73927052802221e-06,
+      "loss": 0.4816,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.88071319946045,
+      "learning_rate": 9.738854534829491e-06,
+      "loss": 0.6217,
+      "step": 1617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.604295733563033,
+      "learning_rate": 9.73843821894173e-06,
+      "loss": 0.5167,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.890186343941587,
+      "learning_rate": 9.738021580387278e-06,
+      "loss": 0.577,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.488294257763036,
+      "learning_rate": 9.737604619194506e-06,
+      "loss": 0.531,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.254673772602777,
+      "learning_rate": 9.737187335391804e-06,
+      "loss": 0.4723,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.079038038923764,
+      "learning_rate": 9.736769729007591e-06,
+      "loss": 0.5637,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.421370067308815,
+      "learning_rate": 9.736351800070303e-06,
+      "loss": 0.4949,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.411569779513658,
+      "learning_rate": 9.7359335486084e-06,
+      "loss": 0.5298,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.876676363251411,
+      "learning_rate": 9.73551497465036e-06,
+      "loss": 0.5664,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.182911144680365,
+      "learning_rate": 9.73509607822469e-06,
+      "loss": 0.5687,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.76485322894778,
+      "learning_rate": 9.734676859359912e-06,
+      "loss": 0.5731,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.089213066755627,
+      "learning_rate": 9.734257318084575e-06,
+      "loss": 0.5735,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.248806629716608,
+      "learning_rate": 9.733837454427246e-06,
+      "loss": 0.5391,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.650149421327421,
+      "learning_rate": 9.733417268416517e-06,
+      "loss": 0.6004,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.478045967422278,
+      "learning_rate": 9.732996760081e-06,
+      "loss": 0.5309,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.315480126050631,
+      "learning_rate": 9.732575929449333e-06,
+      "loss": 0.4681,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2582209237105983,
+      "learning_rate": 9.732154776550167e-06,
+      "loss": 0.3978,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.625037413925845,
+      "learning_rate": 9.731733301412186e-06,
+      "loss": 0.6228,
+      "step": 1634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.096166844636771,
+      "learning_rate": 9.731311504064087e-06,
+      "loss": 0.5842,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9582511680155106,
+      "learning_rate": 9.730889384534595e-06,
+      "loss": 0.5191,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 14.579216489636979,
+      "learning_rate": 9.730466942852451e-06,
+      "loss": 0.5869,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.95439658965732,
+      "learning_rate": 9.730044179046424e-06,
+      "loss": 0.5068,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.395076296013264,
+      "learning_rate": 9.729621093145304e-06,
+      "loss": 0.5473,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.42661523772014,
+      "learning_rate": 9.729197685177899e-06,
+      "loss": 0.5512,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.687003276937102,
+      "learning_rate": 9.72877395517304e-06,
+      "loss": 0.4311,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.6656451939795245,
+      "learning_rate": 9.728349903159587e-06,
+      "loss": 0.553,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.777556032148227,
+      "learning_rate": 9.727925529166406e-06,
+      "loss": 0.59,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.922412582877618,
+      "learning_rate": 9.727500833222405e-06,
+      "loss": 0.5961,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.779597127705232,
+      "learning_rate": 9.727075815356499e-06,
+      "loss": 0.6954,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.031838649824728,
+      "learning_rate": 9.726650475597629e-06,
+      "loss": 0.6462,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.517693289377705,
+      "learning_rate": 9.72622481397476e-06,
+      "loss": 0.5945,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.243891888132412,
+      "learning_rate": 9.725798830516881e-06,
+      "loss": 0.5372,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.676266798926628,
+      "learning_rate": 9.725372525252995e-06,
+      "loss": 0.5489,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.651530950263813,
+      "learning_rate": 9.724945898212132e-06,
+      "loss": 0.5728,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 11.533501710491658,
+      "learning_rate": 9.724518949423344e-06,
+      "loss": 0.5088,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.898045921362227,
+      "learning_rate": 9.724091678915707e-06,
+      "loss": 0.5844,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.274863407204796,
+      "learning_rate": 9.723664086718311e-06,
+      "loss": 0.5896,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.898737308949479,
+      "learning_rate": 9.72323617286028e-06,
+      "loss": 0.6165,
+      "step": 1654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.368546571403453,
+      "learning_rate": 9.722807937370748e-06,
+      "loss": 0.6355,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.626641992361353,
+      "learning_rate": 9.722379380278875e-06,
+      "loss": 0.4506,
+      "step": 1656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.083845989231559,
+      "learning_rate": 9.721950501613847e-06,
+      "loss": 0.55,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.802682837489053,
+      "learning_rate": 9.72152130140487e-06,
+      "loss": 0.5199,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9117975148303084,
+      "learning_rate": 9.721091779681166e-06,
+      "loss": 0.7431,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.296258514074307,
+      "learning_rate": 9.720661936471986e-06,
+      "loss": 0.4898,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.589040785160044,
+      "learning_rate": 9.720231771806601e-06,
+      "loss": 0.5766,
+      "step": 1661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.8462396046342935,
+      "learning_rate": 9.719801285714304e-06,
+      "loss": 0.5394,
+      "step": 1662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.930307324627928,
+      "learning_rate": 9.719370478224406e-06,
+      "loss": 0.4929,
+      "step": 1663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.5112017005659295,
+      "learning_rate": 9.718939349366245e-06,
+      "loss": 0.4737,
+      "step": 1664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 18.914703360036626,
+      "learning_rate": 9.71850789916918e-06,
+      "loss": 0.4701,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.212111009245552,
+      "learning_rate": 9.718076127662589e-06,
+      "loss": 0.6068,
+      "step": 1666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.000512889478079,
+      "learning_rate": 9.717644034875874e-06,
+      "loss": 0.6712,
+      "step": 1667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.501317464974374,
+      "learning_rate": 9.71721162083846e-06,
+      "loss": 0.468,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.19767613332669,
+      "learning_rate": 9.71677888557979e-06,
+      "loss": 0.6253,
+      "step": 1669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.029428208691211,
+      "learning_rate": 9.716345829129334e-06,
+      "loss": 0.5202,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.018123415722698,
+      "learning_rate": 9.71591245151658e-06,
+      "loss": 0.776,
+      "step": 1671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.662458088428295,
+      "learning_rate": 9.715478752771038e-06,
+      "loss": 0.508,
+      "step": 1672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.820936439695323,
+      "learning_rate": 9.715044732922245e-06,
+      "loss": 0.5117,
+      "step": 1673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.318715758756439,
+      "learning_rate": 9.714610391999752e-06,
+      "loss": 0.5245,
+      "step": 1674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.588568634305054,
+      "learning_rate": 9.714175730033138e-06,
+      "loss": 0.54,
+      "step": 1675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.583710879072915,
+      "learning_rate": 9.713740747051998e-06,
+      "loss": 0.6742,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.197110290429818,
+      "learning_rate": 9.713305443085955e-06,
+      "loss": 0.4928,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.839914790237485,
+      "learning_rate": 9.712869818164652e-06,
+      "loss": 0.5152,
+      "step": 1678
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.828084057055026,
+      "learning_rate": 9.712433872317752e-06,
+      "loss": 0.4692,
+      "step": 1679
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.718542934223198,
+      "learning_rate": 9.71199760557494e-06,
+      "loss": 0.5393,
+      "step": 1680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.063309370523485,
+      "learning_rate": 9.711561017965927e-06,
+      "loss": 0.5177,
+      "step": 1681
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.025075300219402,
+      "learning_rate": 9.711124109520438e-06,
+      "loss": 0.509,
+      "step": 1682
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.223485153011838,
+      "learning_rate": 9.710686880268228e-06,
+      "loss": 0.4754,
+      "step": 1683
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.446573045743142,
+      "learning_rate": 9.71024933023907e-06,
+      "loss": 0.6093,
+      "step": 1684
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.281111880117413,
+      "learning_rate": 9.70981145946276e-06,
+      "loss": 0.6542,
+      "step": 1685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.596870174894743,
+      "learning_rate": 9.709373267969112e-06,
+      "loss": 0.5604,
+      "step": 1686
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.393160486776664,
+      "learning_rate": 9.708934755787967e-06,
+      "loss": 0.4407,
+      "step": 1687
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.220520543570482,
+      "learning_rate": 9.708495922949187e-06,
+      "loss": 0.6126,
+      "step": 1688
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 16.040241739650167,
+      "learning_rate": 9.708056769482652e-06,
+      "loss": 0.6068,
+      "step": 1689
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.042327942471179,
+      "learning_rate": 9.707617295418266e-06,
+      "loss": 0.5681,
+      "step": 1690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.512583941536613,
+      "learning_rate": 9.70717750078596e-06,
+      "loss": 0.5674,
+      "step": 1691
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.49513347813271,
+      "learning_rate": 9.706737385615677e-06,
+      "loss": 0.4746,
+      "step": 1692
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.35899972023022,
+      "learning_rate": 9.706296949937389e-06,
+      "loss": 0.516,
+      "step": 1693
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 23.29373903251458,
+      "learning_rate": 9.705856193781085e-06,
+      "loss": 0.6025,
+      "step": 1694
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.253348832224467,
+      "learning_rate": 9.705415117176784e-06,
+      "loss": 0.6289,
+      "step": 1695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.005173924469922,
+      "learning_rate": 9.704973720154517e-06,
+      "loss": 0.4678,
+      "step": 1696
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.822019816356409,
+      "learning_rate": 9.704532002744342e-06,
+      "loss": 0.5931,
+      "step": 1697
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.572025303307895,
+      "learning_rate": 9.70408996497634e-06,
+      "loss": 0.5063,
+      "step": 1698
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.139195517882202,
+      "learning_rate": 9.70364760688061e-06,
+      "loss": 0.5469,
+      "step": 1699
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.5772418005770446,
+      "learning_rate": 9.703204928487272e-06,
+      "loss": 0.5759,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.717399416898641,
+      "learning_rate": 9.702761929826479e-06,
+      "loss": 0.589,
+      "step": 1701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.2154403595192305,
+      "learning_rate": 9.702318610928388e-06,
+      "loss": 0.5782,
+      "step": 1702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8143425436469305,
+      "learning_rate": 9.70187497182319e-06,
+      "loss": 0.5273,
+      "step": 1703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 11.674251380229848,
+      "learning_rate": 9.701431012541098e-06,
+      "loss": 0.5022,
+      "step": 1704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.3647779941375635,
+      "learning_rate": 9.70098673311234e-06,
+      "loss": 0.5453,
+      "step": 1705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.571943288221938,
+      "learning_rate": 9.70054213356717e-06,
+      "loss": 0.5816,
+      "step": 1706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.030760116862505,
+      "learning_rate": 9.700097213935864e-06,
+      "loss": 0.513,
+      "step": 1707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.723703305400056,
+      "learning_rate": 9.699651974248721e-06,
+      "loss": 0.5113,
+      "step": 1708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.392934352502651,
+      "learning_rate": 9.699206414536056e-06,
+      "loss": 0.5503,
+      "step": 1709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.768621021627371,
+      "learning_rate": 9.698760534828212e-06,
+      "loss": 0.5037,
+      "step": 1710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.643127735273095,
+      "learning_rate": 9.698314335155552e-06,
+      "loss": 0.5501,
+      "step": 1711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.8600171960399425,
+      "learning_rate": 9.697867815548457e-06,
+      "loss": 0.5613,
+      "step": 1712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.33612270462994,
+      "learning_rate": 9.697420976037337e-06,
+      "loss": 0.4768,
+      "step": 1713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.239773633688117,
+      "learning_rate": 9.696973816652618e-06,
+      "loss": 0.5583,
+      "step": 1714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9648801872653328,
+      "learning_rate": 9.696526337424747e-06,
+      "loss": 0.703,
+      "step": 1715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.096711546864162,
+      "learning_rate": 9.6960785383842e-06,
+      "loss": 0.5311,
+      "step": 1716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.794862364223541,
+      "learning_rate": 9.695630419561468e-06,
+      "loss": 0.5334,
+      "step": 1717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1741579057723572,
+      "learning_rate": 9.695181980987066e-06,
+      "loss": 0.565,
+      "step": 1718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.440498466650316,
+      "learning_rate": 9.69473322269153e-06,
+      "loss": 0.5254,
+      "step": 1719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.968176243871982,
+      "learning_rate": 9.69428414470542e-06,
+      "loss": 0.5701,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.603398963301825,
+      "learning_rate": 9.693834747059315e-06,
+      "loss": 0.4576,
+      "step": 1721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.382071359684356,
+      "learning_rate": 9.693385029783818e-06,
+      "loss": 0.5109,
+      "step": 1722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.475776096222905,
+      "learning_rate": 9.69293499290955e-06,
+      "loss": 0.513,
+      "step": 1723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.116837138919085,
+      "learning_rate": 9.69248463646716e-06,
+      "loss": 0.3944,
+      "step": 1724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 20.77352256358166,
+      "learning_rate": 9.692033960487313e-06,
+      "loss": 0.5678,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.807458164584601,
+      "learning_rate": 9.691582965000701e-06,
+      "loss": 0.5082,
+      "step": 1726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.023400225424119,
+      "learning_rate": 9.69113165003803e-06,
+      "loss": 0.6131,
+      "step": 1727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.779237019614612,
+      "learning_rate": 9.690680015630037e-06,
+      "loss": 0.5582,
+      "step": 1728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.256674037639484,
+      "learning_rate": 9.690228061807473e-06,
+      "loss": 0.546,
+      "step": 1729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.917724316435249,
+      "learning_rate": 9.68977578860112e-06,
+      "loss": 0.5844,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.024579237701929,
+      "learning_rate": 9.689323196041766e-06,
+      "loss": 0.6229,
+      "step": 1731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.223210124488481,
+      "learning_rate": 9.688870284160239e-06,
+      "loss": 0.5529,
+      "step": 1732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.536177025555527,
+      "learning_rate": 9.688417052987376e-06,
+      "loss": 0.5321,
+      "step": 1733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.444684077743848,
+      "learning_rate": 9.687963502554043e-06,
+      "loss": 0.4607,
+      "step": 1734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.5250703522438105,
+      "learning_rate": 9.68750963289112e-06,
+      "loss": 0.6118,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.638356735108921,
+      "learning_rate": 9.68705544402952e-06,
+      "loss": 0.56,
+      "step": 1736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3850080188956393,
+      "learning_rate": 9.686600936000165e-06,
+      "loss": 0.5064,
+      "step": 1737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.605933056712678,
+      "learning_rate": 9.686146108834011e-06,
+      "loss": 0.5614,
+      "step": 1738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9278112905479436,
+      "learning_rate": 9.685690962562024e-06,
+      "loss": 0.4626,
+      "step": 1739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 17.070932256669796,
+      "learning_rate": 9.685235497215202e-06,
+      "loss": 0.5598,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.149788047062106,
+      "learning_rate": 9.684779712824556e-06,
+      "loss": 0.8016,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.445277870759299,
+      "learning_rate": 9.684323609421126e-06,
+      "loss": 0.5783,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.826789837590213,
+      "learning_rate": 9.683867187035971e-06,
+      "loss": 0.5428,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.311483105967433,
+      "learning_rate": 9.683410445700168e-06,
+      "loss": 0.5526,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.24323327245179,
+      "learning_rate": 9.682953385444823e-06,
+      "loss": 0.5252,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.132688826601286,
+      "learning_rate": 9.682496006301056e-06,
+      "loss": 0.5127,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.860126203054165,
+      "learning_rate": 9.682038308300017e-06,
+      "loss": 0.4813,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.44568273801598,
+      "learning_rate": 9.68158029147287e-06,
+      "loss": 0.6347,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.381157746553778,
+      "learning_rate": 9.681121955850803e-06,
+      "loss": 0.6096,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.838196756945976,
+      "learning_rate": 9.680663301465029e-06,
+      "loss": 0.4346,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.007589183385321,
+      "learning_rate": 9.680204328346778e-06,
+      "loss": 0.5757,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3118706151242012,
+      "learning_rate": 9.679745036527308e-06,
+      "loss": 0.7921,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.898965574050543,
+      "learning_rate": 9.679285426037892e-06,
+      "loss": 0.6533,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.219904241440135,
+      "learning_rate": 9.678825496909827e-06,
+      "loss": 0.582,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.504284489026256,
+      "learning_rate": 9.678365249174433e-06,
+      "loss": 0.5343,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.947748175316304,
+      "learning_rate": 9.677904682863051e-06,
+      "loss": 0.5942,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.500021753245767,
+      "learning_rate": 9.677443798007042e-06,
+      "loss": 0.5214,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.681169573566766,
+      "learning_rate": 9.676982594637793e-06,
+      "loss": 0.4857,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.653949819195006,
+      "learning_rate": 9.676521072786707e-06,
+      "loss": 0.5515,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7577104388505853,
+      "learning_rate": 9.676059232485213e-06,
+      "loss": 0.776,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.333030165190266,
+      "learning_rate": 9.675597073764762e-06,
+      "loss": 0.4932,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9081460643281187,
+      "learning_rate": 9.675134596656823e-06,
+      "loss": 0.5327,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6478062147833743,
+      "learning_rate": 9.674671801192888e-06,
+      "loss": 0.5642,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.115948118311903,
+      "learning_rate": 9.674208687404472e-06,
+      "loss": 0.5996,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.411107756548086,
+      "learning_rate": 9.673745255323111e-06,
+      "loss": 0.574,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.977872137306749,
+      "learning_rate": 9.673281504980363e-06,
+      "loss": 0.5815,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.723526822726555,
+      "learning_rate": 9.672817436407808e-06,
+      "loss": 0.4973,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.135156454078351,
+      "learning_rate": 9.672353049637046e-06,
+      "loss": 0.6424,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.37983427898832,
+      "learning_rate": 9.6718883446997e-06,
+      "loss": 0.5424,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.399625586712974,
+      "learning_rate": 9.671423321627414e-06,
+      "loss": 0.5098,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.409042728842514,
+      "learning_rate": 9.670957980451855e-06,
+      "loss": 0.5091,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.9373911658490215,
+      "learning_rate": 9.670492321204709e-06,
+      "loss": 0.4214,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.299650598634061,
+      "learning_rate": 9.670026343917687e-06,
+      "loss": 0.5289,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5837199813692022,
+      "learning_rate": 9.669560048622519e-06,
+      "loss": 0.633,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.7895258636675875,
+      "learning_rate": 9.669093435350957e-06,
+      "loss": 0.5703,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.97557937213238,
+      "learning_rate": 9.668626504134778e-06,
+      "loss": 0.4836,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.258300884946375,
+      "learning_rate": 9.668159255005775e-06,
+      "loss": 0.5385,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.379654015566775,
+      "learning_rate": 9.667691687995767e-06,
+      "loss": 0.6065,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.098838085431207,
+      "learning_rate": 9.667223803136592e-06,
+      "loss": 0.6517,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.20877067184846,
+      "learning_rate": 9.666755600460114e-06,
+      "loss": 0.4394,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.8649504894771,
+      "learning_rate": 9.666287079998213e-06,
+      "loss": 0.5147,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8459294517089986,
+      "learning_rate": 9.665818241782793e-06,
+      "loss": 0.4263,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.105995369122903,
+      "learning_rate": 9.66534908584578e-06,
+      "loss": 0.5269,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.533181941207882,
+      "learning_rate": 9.664879612219123e-06,
+      "loss": 0.4447,
+      "step": 1784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.385400367715767,
+      "learning_rate": 9.66440982093479e-06,
+      "loss": 0.4697,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.826492863723673,
+      "learning_rate": 9.663939712024771e-06,
+      "loss": 0.6432,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.77062268482245,
+      "learning_rate": 9.66346928552108e-06,
+      "loss": 0.6016,
+      "step": 1787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.169779853980782,
+      "learning_rate": 9.662998541455749e-06,
+      "loss": 0.8602,
+      "step": 1788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.178650391986909,
+      "learning_rate": 9.662527479860834e-06,
+      "loss": 0.6468,
+      "step": 1789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.820576180321416,
+      "learning_rate": 9.662056100768413e-06,
+      "loss": 0.4899,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.399860793615325,
+      "learning_rate": 9.661584404210587e-06,
+      "loss": 0.5213,
+      "step": 1791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.343030541340229,
+      "learning_rate": 9.661112390219472e-06,
+      "loss": 0.5497,
+      "step": 1792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.501778211498369,
+      "learning_rate": 9.660640058827212e-06,
+      "loss": 0.5017,
+      "step": 1793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.51534153660858,
+      "learning_rate": 9.66016741006597e-06,
+      "loss": 0.5583,
+      "step": 1794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.189156675502188,
+      "learning_rate": 9.659694443967934e-06,
+      "loss": 0.8327,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7214105925342944,
+      "learning_rate": 9.659221160565308e-06,
+      "loss": 0.8146,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.773963014427417,
+      "learning_rate": 9.658747559890323e-06,
+      "loss": 0.5915,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3786280264580766,
+      "learning_rate": 9.658273641975227e-06,
+      "loss": 0.661,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.394207733106347,
+      "learning_rate": 9.657799406852291e-06,
+      "loss": 0.5308,
+      "step": 1799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.1242807494268074,
+      "learning_rate": 9.657324854553811e-06,
+      "loss": 0.6279,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.991073808059664,
+      "learning_rate": 9.6568499851121e-06,
+      "loss": 0.5757,
+      "step": 1801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.771696341072152,
+      "learning_rate": 9.656374798559495e-06,
+      "loss": 0.6562,
+      "step": 1802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.874388778445121,
+      "learning_rate": 9.655899294928354e-06,
+      "loss": 0.5817,
+      "step": 1803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.008336350437858,
+      "learning_rate": 9.655423474251057e-06,
+      "loss": 0.6464,
+      "step": 1804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.706967930435837,
+      "learning_rate": 9.654947336560005e-06,
+      "loss": 0.6488,
+      "step": 1805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.516376698854688,
+      "learning_rate": 9.654470881887621e-06,
+      "loss": 0.4793,
+      "step": 1806
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.881763703608345,
+      "learning_rate": 9.65399411026635e-06,
+      "loss": 0.4549,
+      "step": 1807
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.089900137349899,
+      "learning_rate": 9.653517021728658e-06,
+      "loss": 0.5585,
+      "step": 1808
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 52.318475573536446,
+      "learning_rate": 9.653039616307031e-06,
+      "loss": 0.5525,
+      "step": 1809
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.2235238068081244,
+      "learning_rate": 9.65256189403398e-06,
+      "loss": 0.4126,
+      "step": 1810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.597766432220077,
+      "learning_rate": 9.652083854942037e-06,
+      "loss": 0.619,
+      "step": 1811
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.759329779699485,
+      "learning_rate": 9.651605499063748e-06,
+      "loss": 0.4974,
+      "step": 1812
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.606508997391844,
+      "learning_rate": 9.651126826431694e-06,
+      "loss": 0.4796,
+      "step": 1813
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.86785708192879,
+      "learning_rate": 9.650647837078468e-06,
+      "loss": 0.5212,
+      "step": 1814
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4695623090625713,
+      "learning_rate": 9.650168531036687e-06,
+      "loss": 0.5395,
+      "step": 1815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.065817029028293,
+      "learning_rate": 9.649688908338989e-06,
+      "loss": 0.6581,
+      "step": 1816
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.270905618288876,
+      "learning_rate": 9.649208969018036e-06,
+      "loss": 0.5712,
+      "step": 1817
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.621795244082398,
+      "learning_rate": 9.648728713106506e-06,
+      "loss": 0.8741,
+      "step": 1818
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.472285081475371,
+      "learning_rate": 9.648248140637107e-06,
+      "loss": 0.4976,
+      "step": 1819
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9386561285236468,
+      "learning_rate": 9.647767251642561e-06,
+      "loss": 0.5617,
+      "step": 1820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.874934920588806,
+      "learning_rate": 9.647286046155615e-06,
+      "loss": 0.58,
+      "step": 1821
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.074780888085148,
+      "learning_rate": 9.646804524209036e-06,
+      "loss": 0.356,
+      "step": 1822
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.082606238435341,
+      "learning_rate": 9.646322685835616e-06,
+      "loss": 0.5094,
+      "step": 1823
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.583358491369057,
+      "learning_rate": 9.645840531068165e-06,
+      "loss": 0.484,
+      "step": 1824
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.317288369274811,
+      "learning_rate": 9.645358059939513e-06,
+      "loss": 0.556,
+      "step": 1825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.047901847058606,
+      "learning_rate": 9.644875272482518e-06,
+      "loss": 0.5456,
+      "step": 1826
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.240446797198185,
+      "learning_rate": 9.644392168730054e-06,
+      "loss": 0.6101,
+      "step": 1827
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.589269631658657,
+      "learning_rate": 9.643908748715019e-06,
+      "loss": 0.6138,
+      "step": 1828
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.474551161923375,
+      "learning_rate": 9.643425012470329e-06,
+      "loss": 0.6008,
+      "step": 1829
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5779731359969034,
+      "learning_rate": 9.642940960028928e-06,
+      "loss": 0.7375,
+      "step": 1830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.211596524466611,
+      "learning_rate": 9.642456591423775e-06,
+      "loss": 0.4836,
+      "step": 1831
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.434641763927256,
+      "learning_rate": 9.641971906687854e-06,
+      "loss": 0.5681,
+      "step": 1832
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.380190145646167,
+      "learning_rate": 9.641486905854172e-06,
+      "loss": 0.6213,
+      "step": 1833
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.945211912058855,
+      "learning_rate": 9.641001588955755e-06,
+      "loss": 0.5109,
+      "step": 1834
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.9763014187968,
+      "learning_rate": 9.640515956025648e-06,
+      "loss": 0.4697,
+      "step": 1835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.123405292673234,
+      "learning_rate": 9.640030007096923e-06,
+      "loss": 0.5332,
+      "step": 1836
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.7035623814323655,
+      "learning_rate": 9.639543742202668e-06,
+      "loss": 0.4864,
+      "step": 1837
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.800407499622365,
+      "learning_rate": 9.639057161376e-06,
+      "loss": 0.6017,
+      "step": 1838
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.450711795058843,
+      "learning_rate": 9.63857026465005e-06,
+      "loss": 0.5106,
+      "step": 1839
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.767092266259318,
+      "learning_rate": 9.638083052057974e-06,
+      "loss": 0.6252,
+      "step": 1840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.583064386724883,
+      "learning_rate": 9.63759552363295e-06,
+      "loss": 0.5644,
+      "step": 1841
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.439602681819176,
+      "learning_rate": 9.637107679408175e-06,
+      "loss": 0.5012,
+      "step": 1842
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.736731107182953,
+      "learning_rate": 9.63661951941687e-06,
+      "loss": 0.4957,
+      "step": 1843
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.302488909442054,
+      "learning_rate": 9.636131043692277e-06,
+      "loss": 0.5712,
+      "step": 1844
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.841090220769926,
+      "learning_rate": 9.635642252267658e-06,
+      "loss": 0.4843,
+      "step": 1845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.781984862436998,
+      "learning_rate": 9.635153145176297e-06,
+      "loss": 0.5408,
+      "step": 1846
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.208970474262599,
+      "learning_rate": 9.634663722451502e-06,
+      "loss": 0.53,
+      "step": 1847
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.351909080296728,
+      "learning_rate": 9.6341739841266e-06,
+      "loss": 0.6072,
+      "step": 1848
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.957901620654066,
+      "learning_rate": 9.633683930234938e-06,
+      "loss": 0.6102,
+      "step": 1849
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.704495039190623,
+      "learning_rate": 9.63319356080989e-06,
+      "loss": 0.5794,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9234252873282105,
+      "learning_rate": 9.632702875884843e-06,
+      "loss": 0.5735,
+      "step": 1851
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.034372897754948,
+      "learning_rate": 9.632211875493216e-06,
+      "loss": 0.6136,
+      "step": 1852
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.714783750828687,
+      "learning_rate": 9.63172055966844e-06,
+      "loss": 0.6469,
+      "step": 1853
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.743518485865195,
+      "learning_rate": 9.631228928443974e-06,
+      "loss": 0.5759,
+      "step": 1854
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1983550114308152,
+      "learning_rate": 9.630736981853294e-06,
+      "loss": 0.6347,
+      "step": 1855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.055984734121403,
+      "learning_rate": 9.6302447199299e-06,
+      "loss": 0.5417,
+      "step": 1856
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.699637872178884,
+      "learning_rate": 9.629752142707314e-06,
+      "loss": 0.5168,
+      "step": 1857
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.67925047004616,
+      "learning_rate": 9.629259250219076e-06,
+      "loss": 0.5778,
+      "step": 1858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.090058031876984,
+      "learning_rate": 9.628766042498752e-06,
+      "loss": 0.6119,
+      "step": 1859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.326599752864527,
+      "learning_rate": 9.628272519579926e-06,
+      "loss": 0.5554,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.220181988546479,
+      "learning_rate": 9.627778681496205e-06,
+      "loss": 0.5308,
+      "step": 1861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.219196093698272,
+      "learning_rate": 9.627284528281218e-06,
+      "loss": 0.4612,
+      "step": 1862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4094070633908005,
+      "learning_rate": 9.626790059968613e-06,
+      "loss": 0.514,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.67699938967154,
+      "learning_rate": 9.62629527659206e-06,
+      "loss": 0.5485,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.775901990704701,
+      "learning_rate": 9.625800178185258e-06,
+      "loss": 0.5457,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 26.362603445089597,
+      "learning_rate": 9.625304764781915e-06,
+      "loss": 0.5288,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.570184873334286,
+      "learning_rate": 9.624809036415767e-06,
+      "loss": 0.5662,
+      "step": 1867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.233753973737347,
+      "learning_rate": 9.624312993120574e-06,
+      "loss": 0.5425,
+      "step": 1868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8955694431757433,
+      "learning_rate": 9.623816634930112e-06,
+      "loss": 0.7989,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.599825115634058,
+      "learning_rate": 9.623319961878179e-06,
+      "loss": 0.5775,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.318154434767641,
+      "learning_rate": 9.6228229739986e-06,
+      "loss": 0.5194,
+      "step": 1871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.905129466026968,
+      "learning_rate": 9.622325671325216e-06,
+      "loss": 0.7781,
+      "step": 1872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.380602616928459,
+      "learning_rate": 9.621828053891892e-06,
+      "loss": 0.5575,
+      "step": 1873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.394098794880531,
+      "learning_rate": 9.621330121732512e-06,
+      "loss": 0.5704,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.881414607355243,
+      "learning_rate": 9.620831874880983e-06,
+      "loss": 0.5144,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.221698839151444,
+      "learning_rate": 9.620333313371235e-06,
+      "loss": 0.5122,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.725053719005712,
+      "learning_rate": 9.619834437237217e-06,
+      "loss": 0.498,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.102462142619845,
+      "learning_rate": 9.6193352465129e-06,
+      "loss": 0.6102,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.265321491700388,
+      "learning_rate": 9.618835741232277e-06,
+      "loss": 0.4979,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.840580586048226,
+      "learning_rate": 9.618335921429362e-06,
+      "loss": 0.674,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.792913070314878,
+      "learning_rate": 9.617835787138189e-06,
+      "loss": 0.5306,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.855384928449102,
+      "learning_rate": 9.61733533839282e-06,
+      "loss": 0.5594,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.306411711576674,
+      "learning_rate": 9.616834575227326e-06,
+      "loss": 0.6144,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.775036542441297,
+      "learning_rate": 9.616333497675812e-06,
+      "loss": 0.5208,
+      "step": 1884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6909212688674131,
+      "learning_rate": 9.615832105772399e-06,
+      "loss": 0.7146,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7272458139540499,
+      "learning_rate": 9.615330399551226e-06,
+      "loss": 0.8315,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.28437352136229,
+      "learning_rate": 9.614828379046461e-06,
+      "loss": 0.5171,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.099633919464887,
+      "learning_rate": 9.614326044292288e-06,
+      "loss": 0.5738,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.676782762192034,
+      "learning_rate": 9.613823395322911e-06,
+      "loss": 0.6112,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.078925253055688,
+      "learning_rate": 9.613320432172564e-06,
+      "loss": 0.4859,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.404267923427067,
+      "learning_rate": 9.61281715487549e-06,
+      "loss": 0.5655,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.2440018595684945,
+      "learning_rate": 9.612313563465964e-06,
+      "loss": 0.6087,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.210888067399672,
+      "learning_rate": 9.61180965797828e-06,
+      "loss": 0.5012,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.943549082227026,
+      "learning_rate": 9.611305438446745e-06,
+      "loss": 0.4754,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.333153960624742,
+      "learning_rate": 9.610800904905702e-06,
+      "loss": 0.5321,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.793877992452497,
+      "learning_rate": 9.610296057389502e-06,
+      "loss": 0.5538,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.111448028941146,
+      "learning_rate": 9.609790895932525e-06,
+      "loss": 0.6373,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.072404719442253,
+      "learning_rate": 9.60928542056917e-06,
+      "loss": 0.5142,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.205839404324228,
+      "learning_rate": 9.608779631333857e-06,
+      "loss": 0.5701,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.620691348540193,
+      "learning_rate": 9.608273528261029e-06,
+      "loss": 0.564,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.81014655155425,
+      "learning_rate": 9.60776711138515e-06,
+      "loss": 0.441,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.124372792027468,
+      "learning_rate": 9.607260380740703e-06,
+      "loss": 0.5516,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.2010916999205765,
+      "learning_rate": 9.606753336362195e-06,
+      "loss": 0.548,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.514006701013328,
+      "learning_rate": 9.606245978284153e-06,
+      "loss": 0.4948,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.360990400479164,
+      "learning_rate": 9.605738306541126e-06,
+      "loss": 0.451,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6631699447811483,
+      "learning_rate": 9.605230321167686e-06,
+      "loss": 0.5068,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.463025445935594,
+      "learning_rate": 9.604722022198421e-06,
+      "loss": 0.4893,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.583999365509948,
+      "learning_rate": 9.604213409667947e-06,
+      "loss": 0.5594,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6368913988280216,
+      "learning_rate": 9.603704483610897e-06,
+      "loss": 0.5493,
+      "step": 1909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.158492567683161,
+      "learning_rate": 9.603195244061926e-06,
+      "loss": 0.5552,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.885840983939337,
+      "learning_rate": 9.602685691055714e-06,
+      "loss": 0.6057,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.375952597122533,
+      "learning_rate": 9.602175824626953e-06,
+      "loss": 0.5795,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.082157274625304,
+      "learning_rate": 9.60166564481037e-06,
+      "loss": 0.5434,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9549663851960664,
+      "learning_rate": 9.601155151640703e-06,
+      "loss": 0.51,
+      "step": 1914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.486909359175219,
+      "learning_rate": 9.600644345152712e-06,
+      "loss": 0.6672,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.966975155508751,
+      "learning_rate": 9.600133225381183e-06,
+      "loss": 0.5028,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.347953758674539,
+      "learning_rate": 9.59962179236092e-06,
+      "loss": 0.6629,
+      "step": 1917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.896160551564654,
+      "learning_rate": 9.599110046126751e-06,
+      "loss": 0.6087,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.633659828861575,
+      "learning_rate": 9.598597986713522e-06,
+      "loss": 0.4504,
+      "step": 1919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.982100702486805,
+      "learning_rate": 9.598085614156103e-06,
+      "loss": 0.4868,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.222925064055202,
+      "learning_rate": 9.597572928489385e-06,
+      "loss": 0.4981,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.432663060083922,
+      "learning_rate": 9.597059929748279e-06,
+      "loss": 0.5089,
+      "step": 1922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.772233057959861,
+      "learning_rate": 9.596546617967714e-06,
+      "loss": 0.6411,
+      "step": 1923
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.587630203903501,
+      "learning_rate": 9.59603299318265e-06,
+      "loss": 0.5506,
+      "step": 1924
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.984106518644022,
+      "learning_rate": 9.59551905542806e-06,
+      "loss": 0.5963,
+      "step": 1925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.94999419740949,
+      "learning_rate": 9.59500480473894e-06,
+      "loss": 0.5579,
+      "step": 1926
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.984444708946869,
+      "learning_rate": 9.594490241150311e-06,
+      "loss": 0.5315,
+      "step": 1927
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.992002628394882,
+      "learning_rate": 9.593975364697211e-06,
+      "loss": 0.5587,
+      "step": 1928
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.977563158005267,
+      "learning_rate": 9.593460175414702e-06,
+      "loss": 0.4734,
+      "step": 1929
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.763239348176118,
+      "learning_rate": 9.592944673337863e-06,
+      "loss": 0.4723,
+      "step": 1930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.237173570864725,
+      "learning_rate": 9.592428858501799e-06,
+      "loss": 0.4604,
+      "step": 1931
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.9117767904369956,
+      "learning_rate": 9.591912730941635e-06,
+      "loss": 0.5708,
+      "step": 1932
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.304138663950965,
+      "learning_rate": 9.591396290692519e-06,
+      "loss": 0.5646,
+      "step": 1933
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.999287998239806,
+      "learning_rate": 9.590879537789613e-06,
+      "loss": 0.5437,
+      "step": 1934
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.0441520278004335,
+      "learning_rate": 9.59036247226811e-06,
+      "loss": 0.4551,
+      "step": 1935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.184637374789556,
+      "learning_rate": 9.58984509416322e-06,
+      "loss": 0.6069,
+      "step": 1936
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.313389560172534,
+      "learning_rate": 9.589327403510172e-06,
+      "loss": 0.6354,
+      "step": 1937
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.266517703374887,
+      "learning_rate": 9.58880940034422e-06,
+      "loss": 0.5461,
+      "step": 1938
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.531100538135785,
+      "learning_rate": 9.588291084700636e-06,
+      "loss": 0.6123,
+      "step": 1939
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.900115640313341,
+      "learning_rate": 9.587772456614716e-06,
+      "loss": 0.4862,
+      "step": 1940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.210175293329938,
+      "learning_rate": 9.587253516121778e-06,
+      "loss": 0.461,
+      "step": 1941
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8945764707363355,
+      "learning_rate": 9.586734263257156e-06,
+      "loss": 0.8189,
+      "step": 1942
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.44194593303873,
+      "learning_rate": 9.586214698056213e-06,
+      "loss": 0.5609,
+      "step": 1943
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.529681261318594,
+      "learning_rate": 9.585694820554325e-06,
+      "loss": 0.6411,
+      "step": 1944
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.906248471317555,
+      "learning_rate": 9.585174630786896e-06,
+      "loss": 0.7033,
+      "step": 1945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.848011110577012,
+      "learning_rate": 9.584654128789346e-06,
+      "loss": 0.5074,
+      "step": 1946
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.92865285698009,
+      "learning_rate": 9.584133314597124e-06,
+      "loss": 0.4859,
+      "step": 1947
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 18.109107440646852,
+      "learning_rate": 9.583612188245688e-06,
+      "loss": 0.6591,
+      "step": 1948
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.0216168504772125,
+      "learning_rate": 9.58309074977053e-06,
+      "loss": 0.5176,
+      "step": 1949
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6357244996010163,
+      "learning_rate": 9.582568999207157e-06,
+      "loss": 0.7614,
+      "step": 1950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.98429214941143,
+      "learning_rate": 9.582046936591096e-06,
+      "loss": 0.5477,
+      "step": 1951
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.420938654179466,
+      "learning_rate": 9.581524561957897e-06,
+      "loss": 0.5136,
+      "step": 1952
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.982122024940341,
+      "learning_rate": 9.581001875343133e-06,
+      "loss": 0.5066,
+      "step": 1953
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.16234567116975,
+      "learning_rate": 9.580478876782397e-06,
+      "loss": 0.4694,
+      "step": 1954
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.995611495709285,
+      "learning_rate": 9.5799555663113e-06,
+      "loss": 0.494,
+      "step": 1955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.845268934929785,
+      "learning_rate": 9.57943194396548e-06,
+      "loss": 0.5487,
+      "step": 1956
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.816133027856764,
+      "learning_rate": 9.578908009780592e-06,
+      "loss": 0.6399,
+      "step": 1957
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.090759300009877,
+      "learning_rate": 9.578383763792314e-06,
+      "loss": 0.5219,
+      "step": 1958
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3013311504361242,
+      "learning_rate": 9.577859206036346e-06,
+      "loss": 0.5116,
+      "step": 1959
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.6947418797779585,
+      "learning_rate": 9.577334336548405e-06,
+      "loss": 0.6252,
+      "step": 1960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.822675123889666,
+      "learning_rate": 9.576809155364236e-06,
+      "loss": 0.5163,
+      "step": 1961
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.676844186316745,
+      "learning_rate": 9.576283662519598e-06,
+      "loss": 0.4832,
+      "step": 1962
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.716296105310007,
+      "learning_rate": 9.575757858050279e-06,
+      "loss": 0.6176,
+      "step": 1963
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.368003917631037,
+      "learning_rate": 9.575231741992079e-06,
+      "loss": 0.5728,
+      "step": 1964
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.019730840559109,
+      "learning_rate": 9.574705314380827e-06,
+      "loss": 0.5539,
+      "step": 1965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.332381552029035,
+      "learning_rate": 9.574178575252369e-06,
+      "loss": 0.6371,
+      "step": 1966
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.209306439153875,
+      "learning_rate": 9.573651524642577e-06,
+      "loss": 0.6104,
+      "step": 1967
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.51173108721652,
+      "learning_rate": 9.573124162587338e-06,
+      "loss": 0.4865,
+      "step": 1968
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7304467301077104,
+      "learning_rate": 9.572596489122561e-06,
+      "loss": 0.5296,
+      "step": 1969
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.9006540095445805,
+      "learning_rate": 9.572068504284182e-06,
+      "loss": 0.6062,
+      "step": 1970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8229589029769637,
+      "learning_rate": 9.571540208108152e-06,
+      "loss": 0.5719,
+      "step": 1971
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.236616386587533,
+      "learning_rate": 9.571011600630448e-06,
+      "loss": 0.5911,
+      "step": 1972
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.729360277757883,
+      "learning_rate": 9.570482681887066e-06,
+      "loss": 0.6204,
+      "step": 1973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.611597006094529,
+      "learning_rate": 9.56995345191402e-06,
+      "loss": 0.5306,
+      "step": 1974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.761633801917894,
+      "learning_rate": 9.569423910747348e-06,
+      "loss": 0.6117,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.157031507436332,
+      "learning_rate": 9.568894058423114e-06,
+      "loss": 0.4609,
+      "step": 1976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.4306912022572496,
+      "learning_rate": 9.568363894977394e-06,
+      "loss": 0.5593,
+      "step": 1977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6950981568704084,
+      "learning_rate": 9.567833420446293e-06,
+      "loss": 0.4259,
+      "step": 1978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.050290771244189,
+      "learning_rate": 9.567302634865932e-06,
+      "loss": 0.5795,
+      "step": 1979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.074486649805231,
+      "learning_rate": 9.566771538272455e-06,
+      "loss": 0.5755,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.1110905916784555,
+      "learning_rate": 9.56624013070203e-06,
+      "loss": 0.5293,
+      "step": 1981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.421709212852072,
+      "learning_rate": 9.56570841219084e-06,
+      "loss": 0.6301,
+      "step": 1982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.2277339731370205,
+      "learning_rate": 9.565176382775094e-06,
+      "loss": 0.5589,
+      "step": 1983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.900131283753099,
+      "learning_rate": 9.56464404249102e-06,
+      "loss": 0.5541,
+      "step": 1984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.824642071306635,
+      "learning_rate": 9.56411139137487e-06,
+      "loss": 0.4947,
+      "step": 1985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.528972573870201,
+      "learning_rate": 9.563578429462917e-06,
+      "loss": 0.565,
+      "step": 1986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.854595998050608,
+      "learning_rate": 9.563045156791448e-06,
+      "loss": 0.5543,
+      "step": 1987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.914527165711704,
+      "learning_rate": 9.562511573396777e-06,
+      "loss": 0.5253,
+      "step": 1988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.453061635736494,
+      "learning_rate": 9.561977679315245e-06,
+      "loss": 0.4956,
+      "step": 1989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.467883683745534,
+      "learning_rate": 9.561443474583201e-06,
+      "loss": 0.4505,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.235296158572335,
+      "learning_rate": 9.560908959237027e-06,
+      "loss": 0.5604,
+      "step": 1991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.299746568412442,
+      "learning_rate": 9.560374133313118e-06,
+      "loss": 0.4311,
+      "step": 1992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.534653188821914,
+      "learning_rate": 9.559838996847892e-06,
+      "loss": 0.7752,
+      "step": 1993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.4892708368294585,
+      "learning_rate": 9.559303549877792e-06,
+      "loss": 0.5702,
+      "step": 1994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3500121903352915,
+      "learning_rate": 9.55876779243928e-06,
+      "loss": 0.4959,
+      "step": 1995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.723415857513551,
+      "learning_rate": 9.558231724568838e-06,
+      "loss": 0.8023,
+      "step": 1996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.0263780124820014,
+      "learning_rate": 9.557695346302969e-06,
+      "loss": 0.4943,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7439025755288098,
+      "learning_rate": 9.557158657678199e-06,
+      "loss": 0.7608,
+      "step": 1998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.352563392057546,
+      "learning_rate": 9.556621658731072e-06,
+      "loss": 0.5827,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.99217997923918,
+      "learning_rate": 9.556084349498158e-06,
+      "loss": 0.5107,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.908565082023536,
+      "learning_rate": 9.555546730016045e-06,
+      "loss": 0.5938,
+      "step": 2001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.796166912037611,
+      "learning_rate": 9.555008800321341e-06,
+      "loss": 0.4899,
+      "step": 2002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.07775909080919,
+      "learning_rate": 9.554470560450679e-06,
+      "loss": 0.4673,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.245665961772408,
+      "learning_rate": 9.553932010440707e-06,
+      "loss": 0.5111,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.602442104803588,
+      "learning_rate": 9.553393150328102e-06,
+      "loss": 0.4923,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.776065358603938,
+      "learning_rate": 9.552853980149555e-06,
+      "loss": 0.4298,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.633354799931569,
+      "learning_rate": 9.552314499941783e-06,
+      "loss": 0.4608,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5807671135415906,
+      "learning_rate": 9.551774709741522e-06,
+      "loss": 0.4769,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.99527353983725,
+      "learning_rate": 9.551234609585527e-06,
+      "loss": 0.4696,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6896995042470557,
+      "learning_rate": 9.550694199510578e-06,
+      "loss": 0.5382,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.412568800706428,
+      "learning_rate": 9.550153479553475e-06,
+      "loss": 0.5527,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6706483509202474,
+      "learning_rate": 9.54961244975104e-06,
+      "loss": 0.545,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.306497850177073,
+      "learning_rate": 9.549071110140111e-06,
+      "loss": 0.4741,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.831038630635109,
+      "learning_rate": 9.548529460757554e-06,
+      "loss": 0.6164,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.937099894769051,
+      "learning_rate": 9.547987501640252e-06,
+      "loss": 0.5045,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.206058863187371,
+      "learning_rate": 9.54744523282511e-06,
+      "loss": 0.6427,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.405509118386091,
+      "learning_rate": 9.546902654349054e-06,
+      "loss": 0.5075,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.893718490829987,
+      "learning_rate": 9.54635976624903e-06,
+      "loss": 0.5654,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.268747210781893,
+      "learning_rate": 9.545816568562009e-06,
+      "loss": 0.5557,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.448810753097637,
+      "learning_rate": 9.545273061324978e-06,
+      "loss": 0.5358,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.189216848969711,
+      "learning_rate": 9.544729244574949e-06,
+      "loss": 0.6001,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.543916240259153,
+      "learning_rate": 9.544185118348953e-06,
+      "loss": 0.5483,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.943125506488578,
+      "learning_rate": 9.54364068268404e-06,
+      "loss": 0.5565,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.951506065453001,
+      "learning_rate": 9.543095937617286e-06,
+      "loss": 0.5254,
+      "step": 2024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6659584399858276,
+      "learning_rate": 9.542550883185788e-06,
+      "loss": 0.5011,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.34597228882672,
+      "learning_rate": 9.54200551942666e-06,
+      "loss": 0.6539,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.593507846819904,
+      "learning_rate": 9.541459846377037e-06,
+      "loss": 0.5773,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.743201106667412,
+      "learning_rate": 9.540913864074077e-06,
+      "loss": 0.7414,
+      "step": 2028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.909370694670903,
+      "learning_rate": 9.540367572554962e-06,
+      "loss": 0.7139,
+      "step": 2029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7822921410043917,
+      "learning_rate": 9.53982097185689e-06,
+      "loss": 0.4305,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.507381167184284,
+      "learning_rate": 9.539274062017083e-06,
+      "loss": 0.519,
+      "step": 2031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.460774778817747,
+      "learning_rate": 9.538726843072784e-06,
+      "loss": 0.5551,
+      "step": 2032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.684631990306953,
+      "learning_rate": 9.53817931506125e-06,
+      "loss": 0.5338,
+      "step": 2033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.605926687533489,
+      "learning_rate": 9.537631478019774e-06,
+      "loss": 0.6328,
+      "step": 2034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.764942531664048,
+      "learning_rate": 9.537083331985657e-06,
+      "loss": 0.4982,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.217389362404793,
+      "learning_rate": 9.536534876996223e-06,
+      "loss": 0.6333,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.291364302527904,
+      "learning_rate": 9.535986113088826e-06,
+      "loss": 0.6228,
+      "step": 2037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1266942278069747,
+      "learning_rate": 9.535437040300828e-06,
+      "loss": 0.7663,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.486625218930649,
+      "learning_rate": 9.53488765866962e-06,
+      "loss": 0.5722,
+      "step": 2039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.467031062875899,
+      "learning_rate": 9.534337968232614e-06,
+      "loss": 0.5771,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.304510930621847,
+      "learning_rate": 9.533787969027241e-06,
+      "loss": 0.4722,
+      "step": 2041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.593879693313447,
+      "learning_rate": 9.533237661090953e-06,
+      "loss": 0.5623,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8012733807328734,
+      "learning_rate": 9.532687044461226e-06,
+      "loss": 0.478,
+      "step": 2043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.590547845836493,
+      "learning_rate": 9.53213611917555e-06,
+      "loss": 0.5827,
+      "step": 2044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.45281339768516,
+      "learning_rate": 9.531584885271443e-06,
+      "loss": 0.6591,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.347663523194424,
+      "learning_rate": 9.531033342786442e-06,
+      "loss": 0.6034,
+      "step": 2046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.419877042645565,
+      "learning_rate": 9.530481491758104e-06,
+      "loss": 0.4111,
+      "step": 2047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.568276442996724,
+      "learning_rate": 9.529929332224007e-06,
+      "loss": 0.4969,
+      "step": 2048
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.172934617193751,
+      "learning_rate": 9.529376864221753e-06,
+      "loss": 0.5406,
+      "step": 2049
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.85354394920581,
+      "learning_rate": 9.52882408778896e-06,
+      "loss": 0.5575,
+      "step": 2050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.095140222592422,
+      "learning_rate": 9.528271002963269e-06,
+      "loss": 0.5577,
+      "step": 2051
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.936212634159569,
+      "learning_rate": 9.527717609782348e-06,
+      "loss": 0.6073,
+      "step": 2052
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.342477212108584,
+      "learning_rate": 9.527163908283875e-06,
+      "loss": 0.5637,
+      "step": 2053
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.553135390256062,
+      "learning_rate": 9.526609898505555e-06,
+      "loss": 0.5814,
+      "step": 2054
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8599635149874987,
+      "learning_rate": 9.526055580485116e-06,
+      "loss": 0.7688,
+      "step": 2055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.018270794548384,
+      "learning_rate": 9.525500954260305e-06,
+      "loss": 0.5644,
+      "step": 2056
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.161743980377216,
+      "learning_rate": 9.524946019868887e-06,
+      "loss": 0.6061,
+      "step": 2057
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.2911084700280995,
+      "learning_rate": 9.524390777348654e-06,
+      "loss": 0.5694,
+      "step": 2058
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.03490695681306,
+      "learning_rate": 9.52383522673741e-06,
+      "loss": 0.5383,
+      "step": 2059
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.179506767278171,
+      "learning_rate": 9.523279368072992e-06,
+      "loss": 0.5357,
+      "step": 2060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.988649704158613,
+      "learning_rate": 9.522723201393246e-06,
+      "loss": 0.4078,
+      "step": 2061
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.331623080557472,
+      "learning_rate": 9.52216672673605e-06,
+      "loss": 0.5539,
+      "step": 2062
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.766342028805837,
+      "learning_rate": 9.521609944139291e-06,
+      "loss": 0.5385,
+      "step": 2063
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.27744463295059,
+      "learning_rate": 9.52105285364089e-06,
+      "loss": 0.6029,
+      "step": 2064
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.16197217254959,
+      "learning_rate": 9.520495455278777e-06,
+      "loss": 0.57,
+      "step": 2065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.638619468950895,
+      "learning_rate": 9.519937749090913e-06,
+      "loss": 0.5675,
+      "step": 2066
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.35059037303091,
+      "learning_rate": 9.519379735115272e-06,
+      "loss": 0.5188,
+      "step": 2067
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.650086410572488,
+      "learning_rate": 9.518821413389853e-06,
+      "loss": 0.4813,
+      "step": 2068
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.759206845635247,
+      "learning_rate": 9.518262783952675e-06,
+      "loss": 0.497,
+      "step": 2069
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.043061489045207,
+      "learning_rate": 9.517703846841779e-06,
+      "loss": 0.6079,
+      "step": 2070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.65703616430118,
+      "learning_rate": 9.517144602095228e-06,
+      "loss": 0.5702,
+      "step": 2071
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.780652271951478,
+      "learning_rate": 9.516585049751099e-06,
+      "loss": 0.6399,
+      "step": 2072
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.307158739074851,
+      "learning_rate": 9.5160251898475e-06,
+      "loss": 0.5655,
+      "step": 2073
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.793422926654128,
+      "learning_rate": 9.515465022422551e-06,
+      "loss": 0.6042,
+      "step": 2074
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.288602660389193,
+      "learning_rate": 9.5149045475144e-06,
+      "loss": 0.5532,
+      "step": 2075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.201815386311646,
+      "learning_rate": 9.514343765161211e-06,
+      "loss": 0.5534,
+      "step": 2076
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.258677122432349,
+      "learning_rate": 9.513782675401173e-06,
+      "loss": 0.5072,
+      "step": 2077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.492543670180325,
+      "learning_rate": 9.513221278272491e-06,
+      "loss": 0.4878,
+      "step": 2078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.168962355531413,
+      "learning_rate": 9.512659573813396e-06,
+      "loss": 0.6598,
+      "step": 2079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.460103586933416,
+      "learning_rate": 9.512097562062138e-06,
+      "loss": 0.4734,
+      "step": 2080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.745675714478258,
+      "learning_rate": 9.511535243056983e-06,
+      "loss": 0.4914,
+      "step": 2081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.63944381734001,
+      "learning_rate": 9.510972616836227e-06,
+      "loss": 0.5431,
+      "step": 2082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.554584995343043,
+      "learning_rate": 9.51040968343818e-06,
+      "loss": 0.543,
+      "step": 2083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.239602819648189,
+      "learning_rate": 9.509846442901176e-06,
+      "loss": 0.4832,
+      "step": 2084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.476400968713387,
+      "learning_rate": 9.50928289526357e-06,
+      "loss": 0.6007,
+      "step": 2085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.018604479734722,
+      "learning_rate": 9.508719040563738e-06,
+      "loss": 0.5408,
+      "step": 2086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.855565415957361,
+      "learning_rate": 9.508154878840072e-06,
+      "loss": 0.4801,
+      "step": 2087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.76934477765262,
+      "learning_rate": 9.507590410130993e-06,
+      "loss": 0.6558,
+      "step": 2088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.270653865561689,
+      "learning_rate": 9.507025634474936e-06,
+      "loss": 0.5176,
+      "step": 2089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.978268136710612,
+      "learning_rate": 9.506460551910361e-06,
+      "loss": 0.4436,
+      "step": 2090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.357523716919445,
+      "learning_rate": 9.505895162475749e-06,
+      "loss": 0.5465,
+      "step": 2091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.210430391384749,
+      "learning_rate": 9.505329466209597e-06,
+      "loss": 0.4845,
+      "step": 2092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.1054795081913555,
+      "learning_rate": 9.50476346315043e-06,
+      "loss": 0.5962,
+      "step": 2093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 18.45738831663633,
+      "learning_rate": 9.504197153336788e-06,
+      "loss": 0.4972,
+      "step": 2094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.39456230856888,
+      "learning_rate": 9.503630536807236e-06,
+      "loss": 0.5207,
+      "step": 2095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.549545513542016,
+      "learning_rate": 9.503063613600356e-06,
+      "loss": 0.5793,
+      "step": 2096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.194715683389301,
+      "learning_rate": 9.502496383754753e-06,
+      "loss": 0.4776,
+      "step": 2097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.97423190164509,
+      "learning_rate": 9.501928847309056e-06,
+      "loss": 0.5752,
+      "step": 2098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.696912368060387,
+      "learning_rate": 9.50136100430191e-06,
+      "loss": 0.4692,
+      "step": 2099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 22.09457153451169,
+      "learning_rate": 9.50079285477198e-06,
+      "loss": 0.5328,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.693091004267522,
+      "learning_rate": 9.500224398757958e-06,
+      "loss": 0.4648,
+      "step": 2101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.47016151281045,
+      "learning_rate": 9.49965563629855e-06,
+      "loss": 0.5958,
+      "step": 2102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.623665045086284,
+      "learning_rate": 9.499086567432491e-06,
+      "loss": 0.5485,
+      "step": 2103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.5518137241899215,
+      "learning_rate": 9.498517192198527e-06,
+      "loss": 0.5614,
+      "step": 2104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.53024627611184,
+      "learning_rate": 9.497947510635435e-06,
+      "loss": 0.5209,
+      "step": 2105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.295388588896483,
+      "learning_rate": 9.497377522782003e-06,
+      "loss": 0.4902,
+      "step": 2106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.232325773709331,
+      "learning_rate": 9.496807228677047e-06,
+      "loss": 0.6621,
+      "step": 2107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.061723093310357,
+      "learning_rate": 9.4962366283594e-06,
+      "loss": 0.5377,
+      "step": 2108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.640835619947837,
+      "learning_rate": 9.495665721867919e-06,
+      "loss": 0.4845,
+      "step": 2109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.570669707948493,
+      "learning_rate": 9.495094509241482e-06,
+      "loss": 0.5192,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9035389429170009,
+      "learning_rate": 9.494522990518982e-06,
+      "loss": 0.7319,
+      "step": 2111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.85837705495963,
+      "learning_rate": 9.493951165739338e-06,
+      "loss": 0.5596,
+      "step": 2112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.290438164421028,
+      "learning_rate": 9.49337903494149e-06,
+      "loss": 0.3965,
+      "step": 2113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.739333104227114,
+      "learning_rate": 9.492806598164398e-06,
+      "loss": 0.5165,
+      "step": 2114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.934076701472655,
+      "learning_rate": 9.492233855447041e-06,
+      "loss": 0.4429,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.43956682856479,
+      "learning_rate": 9.49166080682842e-06,
+      "loss": 0.4227,
+      "step": 2116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.136339457944208,
+      "learning_rate": 9.491087452347558e-06,
+      "loss": 0.5087,
+      "step": 2117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.200843119700275,
+      "learning_rate": 9.490513792043498e-06,
+      "loss": 0.6005,
+      "step": 2118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.586546373514695,
+      "learning_rate": 9.489939825955304e-06,
+      "loss": 0.5116,
+      "step": 2119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.14423644304595,
+      "learning_rate": 9.489365554122058e-06,
+      "loss": 0.4516,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.023976294062274,
+      "learning_rate": 9.488790976582868e-06,
+      "loss": 0.4933,
+      "step": 2121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.886280980457648,
+      "learning_rate": 9.48821609337686e-06,
+      "loss": 0.4735,
+      "step": 2122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.613233143284594,
+      "learning_rate": 9.48764090454318e-06,
+      "loss": 0.5273,
+      "step": 2123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.844415821645146,
+      "learning_rate": 9.487065410120997e-06,
+      "loss": 0.5455,
+      "step": 2124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.53945828938682,
+      "learning_rate": 9.486489610149499e-06,
+      "loss": 0.5402,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.849710120543149,
+      "learning_rate": 9.485913504667894e-06,
+      "loss": 0.5813,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.61840799139907,
+      "learning_rate": 9.485337093715414e-06,
+      "loss": 0.6903,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.7904515745930265,
+      "learning_rate": 9.484760377331311e-06,
+      "loss": 0.519,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.8009327992878745,
+      "learning_rate": 9.484183355554854e-06,
+      "loss": 0.5169,
+      "step": 2129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.409548004876113,
+      "learning_rate": 9.483606028425338e-06,
+      "loss": 0.4911,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 13.774310379168654,
+      "learning_rate": 9.483028395982076e-06,
+      "loss": 0.5909,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.929476109749393,
+      "learning_rate": 9.4824504582644e-06,
+      "loss": 0.5149,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.346976176286347,
+      "learning_rate": 9.481872215311669e-06,
+      "loss": 0.596,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.368439970889571,
+      "learning_rate": 9.481293667163254e-06,
+      "loss": 0.5532,
+      "step": 2134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.056118512287302,
+      "learning_rate": 9.480714813858555e-06,
+      "loss": 0.5654,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.480425741139197,
+      "learning_rate": 9.48013565543699e-06,
+      "loss": 0.6494,
+      "step": 2136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.146834183637632,
+      "learning_rate": 9.479556191937996e-06,
+      "loss": 0.5334,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.264798038058563,
+      "learning_rate": 9.478976423401028e-06,
+      "loss": 0.4655,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.365703141797061,
+      "learning_rate": 9.478396349865571e-06,
+      "loss": 0.4985,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.373408450939855,
+      "learning_rate": 9.477815971371125e-06,
+      "loss": 0.4914,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.738195095164051,
+      "learning_rate": 9.477235287957208e-06,
+      "loss": 0.7482,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.062924159268675,
+      "learning_rate": 9.476654299663365e-06,
+      "loss": 0.56,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.162302522520806,
+      "learning_rate": 9.476073006529155e-06,
+      "loss": 0.4827,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.31041961840823,
+      "learning_rate": 9.475491408594167e-06,
+      "loss": 0.4541,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.227406880111323,
+      "learning_rate": 9.474909505898003e-06,
+      "loss": 0.5096,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.706924690654652,
+      "learning_rate": 9.474327298480284e-06,
+      "loss": 0.4344,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.69670115545592,
+      "learning_rate": 9.473744786380661e-06,
+      "loss": 0.4881,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.263588382152765,
+      "learning_rate": 9.473161969638797e-06,
+      "loss": 0.5311,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.11174033127515,
+      "learning_rate": 9.472578848294382e-06,
+      "loss": 0.4883,
+      "step": 2149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.680009569476396,
+      "learning_rate": 9.471995422387124e-06,
+      "loss": 0.5616,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.812984248388266,
+      "learning_rate": 9.471411691956748e-06,
+      "loss": 0.4226,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.309738378260744,
+      "learning_rate": 9.470827657043006e-06,
+      "loss": 0.5635,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.914155167742361,
+      "learning_rate": 9.47024331768567e-06,
+      "loss": 0.6374,
+      "step": 2153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.384376011137604,
+      "learning_rate": 9.469658673924528e-06,
+      "loss": 0.5295,
+      "step": 2154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.1116321598456524,
+      "learning_rate": 9.469073725799393e-06,
+      "loss": 0.496,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.535259559972968,
+      "learning_rate": 9.468488473350097e-06,
+      "loss": 0.5951,
+      "step": 2156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.994196240944015,
+      "learning_rate": 9.467902916616492e-06,
+      "loss": 0.5803,
+      "step": 2157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.661617143387513,
+      "learning_rate": 9.467317055638455e-06,
+      "loss": 0.5152,
+      "step": 2158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.085070180716848,
+      "learning_rate": 9.466730890455877e-06,
+      "loss": 0.5071,
+      "step": 2159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.929112233797199,
+      "learning_rate": 9.466144421108677e-06,
+      "loss": 0.5337,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.665823626307388,
+      "learning_rate": 9.465557647636787e-06,
+      "loss": 0.5686,
+      "step": 2161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.43718820198857,
+      "learning_rate": 9.464970570080165e-06,
+      "loss": 0.5669,
+      "step": 2162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.257910941304369,
+      "learning_rate": 9.464383188478789e-06,
+      "loss": 0.4395,
+      "step": 2163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.288330376836399,
+      "learning_rate": 9.46379550287266e-06,
+      "loss": 0.4251,
+      "step": 2164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.57812410441956,
+      "learning_rate": 9.46320751330179e-06,
+      "loss": 0.5086,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.459485701830878,
+      "learning_rate": 9.462619219806224e-06,
+      "loss": 0.4821,
+      "step": 2166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.0392733662445,
+      "learning_rate": 9.462030622426019e-06,
+      "loss": 0.5656,
+      "step": 2167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.106529130191995,
+      "learning_rate": 9.461441721201259e-06,
+      "loss": 0.4929,
+      "step": 2168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.681408220392834,
+      "learning_rate": 9.460852516172042e-06,
+      "loss": 0.5513,
+      "step": 2169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.290384889765048,
+      "learning_rate": 9.460263007378494e-06,
+      "loss": 0.5964,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.7332957065661585,
+      "learning_rate": 9.459673194860754e-06,
+      "loss": 0.4954,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.382715420642043,
+      "learning_rate": 9.45908307865899e-06,
+      "loss": 0.4404,
+      "step": 2172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.66933426926404,
+      "learning_rate": 9.458492658813382e-06,
+      "loss": 0.6315,
+      "step": 2173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.0790887430101765,
+      "learning_rate": 9.457901935364137e-06,
+      "loss": 0.6344,
+      "step": 2174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.199261589200539,
+      "learning_rate": 9.457310908351483e-06,
+      "loss": 0.5328,
+      "step": 2175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.423821553109532,
+      "learning_rate": 9.456719577815662e-06,
+      "loss": 0.5638,
+      "step": 2176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 38.23694732180162,
+      "learning_rate": 9.456127943796944e-06,
+      "loss": 0.4019,
+      "step": 2177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.158513389559331,
+      "learning_rate": 9.455536006335615e-06,
+      "loss": 0.7513,
+      "step": 2178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.412498504837751,
+      "learning_rate": 9.454943765471985e-06,
+      "loss": 0.5608,
+      "step": 2179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.683949683829999,
+      "learning_rate": 9.45435122124638e-06,
+      "loss": 0.5928,
+      "step": 2180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.289764300377733,
+      "learning_rate": 9.453758373699152e-06,
+      "loss": 0.5727,
+      "step": 2181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.175004681692688,
+      "learning_rate": 9.453165222870671e-06,
+      "loss": 0.5917,
+      "step": 2182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.251718958037843,
+      "learning_rate": 9.452571768801329e-06,
+      "loss": 0.5417,
+      "step": 2183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.28735894680072,
+      "learning_rate": 9.451978011531535e-06,
+      "loss": 0.555,
+      "step": 2184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.69066817097477,
+      "learning_rate": 9.451383951101722e-06,
+      "loss": 0.5281,
+      "step": 2185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.25737785490691,
+      "learning_rate": 9.450789587552347e-06,
+      "loss": 0.5161,
+      "step": 2186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.291879908776177,
+      "learning_rate": 9.450194920923877e-06,
+      "loss": 0.5404,
+      "step": 2187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.018766171427838,
+      "learning_rate": 9.44959995125681e-06,
+      "loss": 0.4882,
+      "step": 2188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.104067042330813,
+      "learning_rate": 9.44900467859166e-06,
+      "loss": 0.5392,
+      "step": 2189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.042870760179191,
+      "learning_rate": 9.448409102968961e-06,
+      "loss": 0.5077,
+      "step": 2190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.36046505428304,
+      "learning_rate": 9.447813224429272e-06,
+      "loss": 0.5312,
+      "step": 2191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.723209326304114,
+      "learning_rate": 9.447217043013168e-06,
+      "loss": 0.5278,
+      "step": 2192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.613376757942838,
+      "learning_rate": 9.446620558761249e-06,
+      "loss": 0.6287,
+      "step": 2193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.701030207168896,
+      "learning_rate": 9.446023771714127e-06,
+      "loss": 0.5505,
+      "step": 2194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.800966634155746,
+      "learning_rate": 9.445426681912445e-06,
+      "loss": 0.5151,
+      "step": 2195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.05096510192036,
+      "learning_rate": 9.444829289396862e-06,
+      "loss": 0.5746,
+      "step": 2196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.94943011626233,
+      "learning_rate": 9.444231594208056e-06,
+      "loss": 0.5573,
+      "step": 2197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 27.75914698420612,
+      "learning_rate": 9.44363359638673e-06,
+      "loss": 0.3763,
+      "step": 2198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.248489959563991,
+      "learning_rate": 9.443035295973602e-06,
+      "loss": 0.4894,
+      "step": 2199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 36.84380197089049,
+      "learning_rate": 9.442436693009414e-06,
+      "loss": 0.4787,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.516540693541943,
+      "learning_rate": 9.441837787534932e-06,
+      "loss": 0.5867,
+      "step": 2201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.045231760934765,
+      "learning_rate": 9.441238579590935e-06,
+      "loss": 0.5145,
+      "step": 2202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.91657062077295,
+      "learning_rate": 9.440639069218227e-06,
+      "loss": 0.4618,
+      "step": 2203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.809606007767243,
+      "learning_rate": 9.440039256457632e-06,
+      "loss": 0.5559,
+      "step": 2204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.4781139414706095,
+      "learning_rate": 9.439439141349995e-06,
+      "loss": 0.5197,
+      "step": 2205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.287686574774902,
+      "learning_rate": 9.438838723936184e-06,
+      "loss": 0.5147,
+      "step": 2206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.970287740082146,
+      "learning_rate": 9.438238004257079e-06,
+      "loss": 0.513,
+      "step": 2207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7253863379016512,
+      "learning_rate": 9.43763698235359e-06,
+      "loss": 0.7648,
+      "step": 2208
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.25808169631474,
+      "learning_rate": 9.437035658266644e-06,
+      "loss": 0.482,
+      "step": 2209
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.558241053709693,
+      "learning_rate": 9.436434032037188e-06,
+      "loss": 0.5035,
+      "step": 2210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.265341547587227,
+      "learning_rate": 9.435832103706189e-06,
+      "loss": 0.5759,
+      "step": 2211
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.159894324650029,
+      "learning_rate": 9.435229873314638e-06,
+      "loss": 0.5226,
+      "step": 2212
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.648587848089845,
+      "learning_rate": 9.434627340903541e-06,
+      "loss": 0.5643,
+      "step": 2213
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.890212764058937,
+      "learning_rate": 9.434024506513931e-06,
+      "loss": 0.4147,
+      "step": 2214
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.821131776971145,
+      "learning_rate": 9.433421370186857e-06,
+      "loss": 0.5612,
+      "step": 2215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.418354037470477,
+      "learning_rate": 9.432817931963388e-06,
+      "loss": 0.496,
+      "step": 2216
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.325581155343542,
+      "learning_rate": 9.432214191884619e-06,
+      "loss": 0.4652,
+      "step": 2217
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.452012238626056,
+      "learning_rate": 9.43161014999166e-06,
+      "loss": 0.5312,
+      "step": 2218
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.424411176803011,
+      "learning_rate": 9.431005806325644e-06,
+      "loss": 0.4848,
+      "step": 2219
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.066626450035868,
+      "learning_rate": 9.430401160927725e-06,
+      "loss": 0.5817,
+      "step": 2220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.244342050226193,
+      "learning_rate": 9.429796213839075e-06,
+      "loss": 0.5147,
+      "step": 2221
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.826148882973916,
+      "learning_rate": 9.429190965100887e-06,
+      "loss": 0.5625,
+      "step": 2222
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.789713136899307,
+      "learning_rate": 9.42858541475438e-06,
+      "loss": 0.8172,
+      "step": 2223
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.675197122281034,
+      "learning_rate": 9.427979562840787e-06,
+      "loss": 0.4872,
+      "step": 2224
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 22.168071157272344,
+      "learning_rate": 9.427373409401364e-06,
+      "loss": 0.5002,
+      "step": 2225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 22.849214570106675,
+      "learning_rate": 9.426766954477387e-06,
+      "loss": 0.571,
+      "step": 2226
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.140935244062847,
+      "learning_rate": 9.426160198110153e-06,
+      "loss": 0.5821,
+      "step": 2227
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.577775924402136,
+      "learning_rate": 9.425553140340979e-06,
+      "loss": 0.6185,
+      "step": 2228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 24.562463858080452,
+      "learning_rate": 9.424945781211204e-06,
+      "loss": 0.6654,
+      "step": 2229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.971623049186782,
+      "learning_rate": 9.424338120762186e-06,
+      "loss": 0.4588,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.431094364153378,
+      "learning_rate": 9.423730159035304e-06,
+      "loss": 0.4917,
+      "step": 2231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.226852725109403,
+      "learning_rate": 9.423121896071959e-06,
+      "loss": 0.5352,
+      "step": 2232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.13701534895968,
+      "learning_rate": 9.422513331913568e-06,
+      "loss": 0.4814,
+      "step": 2233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.889749035886165,
+      "learning_rate": 9.421904466601573e-06,
+      "loss": 0.4725,
+      "step": 2234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.977191280730132,
+      "learning_rate": 9.421295300177435e-06,
+      "loss": 0.5329,
+      "step": 2235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.35269563171114,
+      "learning_rate": 9.420685832682635e-06,
+      "loss": 0.544,
+      "step": 2236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 60.72401295285078,
+      "learning_rate": 9.420076064158676e-06,
+      "loss": 0.4862,
+      "step": 2237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.179790636229752,
+      "learning_rate": 9.419465994647081e-06,
+      "loss": 0.5215,
+      "step": 2238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.092836484756052,
+      "learning_rate": 9.418855624189393e-06,
+      "loss": 0.5707,
+      "step": 2239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.355832671686557,
+      "learning_rate": 9.418244952827174e-06,
+      "loss": 0.594,
+      "step": 2240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.873080823702406,
+      "learning_rate": 9.417633980602008e-06,
+      "loss": 0.4787,
+      "step": 2241
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.765943119614061,
+      "learning_rate": 9.417022707555501e-06,
+      "loss": 0.4682,
+      "step": 2242
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.21208248106156,
+      "learning_rate": 9.416411133729277e-06,
+      "loss": 0.5317,
+      "step": 2243
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.170382789510501,
+      "learning_rate": 9.415799259164981e-06,
+      "loss": 0.5001,
+      "step": 2244
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 19.00624739706448,
+      "learning_rate": 9.415187083904282e-06,
+      "loss": 0.4564,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.057416824830243,
+      "learning_rate": 9.414574607988862e-06,
+      "loss": 0.397,
+      "step": 2246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.21835641367099,
+      "learning_rate": 9.413961831460431e-06,
+      "loss": 0.4729,
+      "step": 2247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.098989183283907,
+      "learning_rate": 9.413348754360716e-06,
+      "loss": 0.5046,
+      "step": 2248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.111804812715874,
+      "learning_rate": 9.412735376731464e-06,
+      "loss": 0.568,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.02041538810531,
+      "learning_rate": 9.412121698614443e-06,
+      "loss": 0.5862,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.948978337466123,
+      "learning_rate": 9.411507720051442e-06,
+      "loss": 0.5148,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.045564334421218,
+      "learning_rate": 9.41089344108427e-06,
+      "loss": 0.5522,
+      "step": 2252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 43.652481736093705,
+      "learning_rate": 9.41027886175476e-06,
+      "loss": 0.4459,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 16.221726256725585,
+      "learning_rate": 9.409663982104756e-06,
+      "loss": 0.5425,
+      "step": 2254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 47.548054753218636,
+      "learning_rate": 9.409048802176134e-06,
+      "loss": 0.4921,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.851227162179512,
+      "learning_rate": 9.408433322010783e-06,
+      "loss": 0.5776,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 39.22053828785807,
+      "learning_rate": 9.407817541650614e-06,
+      "loss": 0.534,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.708365076941863,
+      "learning_rate": 9.407201461137559e-06,
+      "loss": 0.4735,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.109390158223597,
+      "learning_rate": 9.406585080513571e-06,
+      "loss": 0.545,
+      "step": 2259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.843323247099348,
+      "learning_rate": 9.405968399820622e-06,
+      "loss": 0.4589,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.061242868964864,
+      "learning_rate": 9.405351419100709e-06,
+      "loss": 0.5136,
+      "step": 2261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.833543035573967,
+      "learning_rate": 9.404734138395839e-06,
+      "loss": 0.5325,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 31.80471581648115,
+      "learning_rate": 9.404116557748049e-06,
+      "loss": 0.4713,
+      "step": 2263
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.984021577480863,
+      "learning_rate": 9.403498677199396e-06,
+      "loss": 0.6136,
+      "step": 2264
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.82730116565881,
+      "learning_rate": 9.402880496791954e-06,
+      "loss": 0.5376,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.81634634411593,
+      "learning_rate": 9.402262016567813e-06,
+      "loss": 0.6084,
+      "step": 2266
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.26742108929409,
+      "learning_rate": 9.401643236569096e-06,
+      "loss": 0.5353,
+      "step": 2267
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.380481785628877,
+      "learning_rate": 9.401024156837936e-06,
+      "loss": 0.5239,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.711060852261038,
+      "learning_rate": 9.40040477741649e-06,
+      "loss": 0.5452,
+      "step": 2269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.759371629518975,
+      "learning_rate": 9.399785098346936e-06,
+      "loss": 0.5486,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.646751753213888,
+      "learning_rate": 9.399165119671468e-06,
+      "loss": 0.4627,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.278770251302826,
+      "learning_rate": 9.398544841432308e-06,
+      "loss": 0.5368,
+      "step": 2272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 13.24232730002813,
+      "learning_rate": 9.397924263671691e-06,
+      "loss": 0.5564,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.876777204028357,
+      "learning_rate": 9.397303386431878e-06,
+      "loss": 0.546,
+      "step": 2274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.166143925110894,
+      "learning_rate": 9.396682209755146e-06,
+      "loss": 0.491,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.879585815198128,
+      "learning_rate": 9.396060733683793e-06,
+      "loss": 0.5252,
+      "step": 2276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.632477646580025,
+      "learning_rate": 9.395438958260146e-06,
+      "loss": 0.541,
+      "step": 2277
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9548584103143343,
+      "learning_rate": 9.394816883526536e-06,
+      "loss": 0.7753,
+      "step": 2278
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.708278070592595,
+      "learning_rate": 9.39419450952533e-06,
+      "loss": 0.4623,
+      "step": 2279
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 34.161667880978385,
+      "learning_rate": 9.393571836298905e-06,
+      "loss": 0.4443,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 29.10386971689025,
+      "learning_rate": 9.392948863889666e-06,
+      "loss": 0.5517,
+      "step": 2281
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 15.821279386273579,
+      "learning_rate": 9.392325592340032e-06,
+      "loss": 0.6074,
+      "step": 2282
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0398858833883367,
+      "learning_rate": 9.391702021692445e-06,
+      "loss": 0.7424,
+      "step": 2283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 85.31904616835222,
+      "learning_rate": 9.391078151989368e-06,
+      "loss": 0.4554,
+      "step": 2284
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 21.441467670962243,
+      "learning_rate": 9.390453983273285e-06,
+      "loss": 0.635,
+      "step": 2285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.309262234132588,
+      "learning_rate": 9.389829515586696e-06,
+      "loss": 0.4278,
+      "step": 2286
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.686765941423536,
+      "learning_rate": 9.389204748972127e-06,
+      "loss": 0.5126,
+      "step": 2287
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 17.240450889030242,
+      "learning_rate": 9.388579683472121e-06,
+      "loss": 0.5491,
+      "step": 2288
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 133.52939692626083,
+      "learning_rate": 9.387954319129244e-06,
+      "loss": 0.5743,
+      "step": 2289
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.869561307329194,
+      "learning_rate": 9.38732865598608e-06,
+      "loss": 0.4859,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.340544092498176,
+      "learning_rate": 9.38670269408523e-06,
+      "loss": 0.5265,
+      "step": 2291
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.559571569121182,
+      "learning_rate": 9.386076433469322e-06,
+      "loss": 0.4589,
+      "step": 2292
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.091224189155943,
+      "learning_rate": 9.385449874181003e-06,
+      "loss": 0.4484,
+      "step": 2293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.421112659905548,
+      "learning_rate": 9.384823016262938e-06,
+      "loss": 0.8656,
+      "step": 2294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.930640356633312,
+      "learning_rate": 9.384195859757812e-06,
+      "loss": 0.6038,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.315522371330909,
+      "learning_rate": 9.383568404708333e-06,
+      "loss": 0.518,
+      "step": 2296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 20.108020930129392,
+      "learning_rate": 9.382940651157228e-06,
+      "loss": 0.4693,
+      "step": 2297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.479330940591886,
+      "learning_rate": 9.382312599147244e-06,
+      "loss": 0.5009,
+      "step": 2298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.062673878287933,
+      "learning_rate": 9.381684248721145e-06,
+      "loss": 0.5585,
+      "step": 2299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.30916624835139,
+      "learning_rate": 9.381055599921725e-06,
+      "loss": 0.5765,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.346162749087836,
+      "learning_rate": 9.38042665279179e-06,
+      "loss": 0.5533,
+      "step": 2301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.264732470067964,
+      "learning_rate": 9.379797407374165e-06,
+      "loss": 0.4581,
+      "step": 2302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.874542797867765,
+      "learning_rate": 9.379167863711703e-06,
+      "loss": 0.4674,
+      "step": 2303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.47165244956606,
+      "learning_rate": 9.378538021847272e-06,
+      "loss": 0.6096,
+      "step": 2304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.272524808683883,
+      "learning_rate": 9.377907881823761e-06,
+      "loss": 0.5509,
+      "step": 2305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.955838062738934,
+      "learning_rate": 9.37727744368408e-06,
+      "loss": 0.4803,
+      "step": 2306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.944238756882251,
+      "learning_rate": 9.376646707471159e-06,
+      "loss": 0.558,
+      "step": 2307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.982680605972252,
+      "learning_rate": 9.37601567322795e-06,
+      "loss": 0.4967,
+      "step": 2308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.494779029096494,
+      "learning_rate": 9.37538434099742e-06,
+      "loss": 0.5726,
+      "step": 2309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.66652546544279,
+      "learning_rate": 9.37475271082256e-06,
+      "loss": 0.5324,
+      "step": 2310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.67985185368034,
+      "learning_rate": 9.374120782746388e-06,
+      "loss": 0.5045,
+      "step": 2311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.98148454496502,
+      "learning_rate": 9.373488556811927e-06,
+      "loss": 0.4627,
+      "step": 2312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.572694200282372,
+      "learning_rate": 9.372856033062235e-06,
+      "loss": 0.525,
+      "step": 2313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.546685786120024,
+      "learning_rate": 9.37222321154038e-06,
+      "loss": 0.5559,
+      "step": 2314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.889183162949253,
+      "learning_rate": 9.371590092289455e-06,
+      "loss": 0.5297,
+      "step": 2315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.031602547077577,
+      "learning_rate": 9.370956675352576e-06,
+      "loss": 0.5777,
+      "step": 2316
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.540157237795812,
+      "learning_rate": 9.370322960772871e-06,
+      "loss": 0.4579,
+      "step": 2317
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.233341362520561,
+      "learning_rate": 9.369688948593497e-06,
+      "loss": 0.4696,
+      "step": 2318
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.850806876690324,
+      "learning_rate": 9.369054638857627e-06,
+      "loss": 0.4447,
+      "step": 2319
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.032106752622177,
+      "learning_rate": 9.368420031608452e-06,
+      "loss": 0.5676,
+      "step": 2320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.167325011609509,
+      "learning_rate": 9.367785126889187e-06,
+      "loss": 0.4917,
+      "step": 2321
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.530767015144027,
+      "learning_rate": 9.367149924743068e-06,
+      "loss": 0.4078,
+      "step": 2322
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.312256032926584,
+      "learning_rate": 9.366514425213349e-06,
+      "loss": 0.5242,
+      "step": 2323
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.916526629899498,
+      "learning_rate": 9.365878628343302e-06,
+      "loss": 0.4167,
+      "step": 2324
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.487923079521345,
+      "learning_rate": 9.365242534176225e-06,
+      "loss": 0.5435,
+      "step": 2325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.839361004640072,
+      "learning_rate": 9.364606142755433e-06,
+      "loss": 0.5928,
+      "step": 2326
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.841129103957419,
+      "learning_rate": 9.36396945412426e-06,
+      "loss": 0.453,
+      "step": 2327
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.787718082227851,
+      "learning_rate": 9.363332468326061e-06,
+      "loss": 0.4568,
+      "step": 2328
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.503375649190755,
+      "learning_rate": 9.362695185404215e-06,
+      "loss": 0.5762,
+      "step": 2329
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.64117975661933,
+      "learning_rate": 9.362057605402115e-06,
+      "loss": 0.5282,
+      "step": 2330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.075827923023441,
+      "learning_rate": 9.36141972836318e-06,
+      "loss": 0.4321,
+      "step": 2331
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.911537271964783,
+      "learning_rate": 9.360781554330844e-06,
+      "loss": 0.6591,
+      "step": 2332
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.548611623925508,
+      "learning_rate": 9.360143083348567e-06,
+      "loss": 0.4993,
+      "step": 2333
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.588831288551137,
+      "learning_rate": 9.359504315459822e-06,
+      "loss": 0.5781,
+      "step": 2334
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.606459707039292,
+      "learning_rate": 9.358865250708109e-06,
+      "loss": 0.5648,
+      "step": 2335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.104393040354369,
+      "learning_rate": 9.358225889136946e-06,
+      "loss": 0.4552,
+      "step": 2336
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.190747973515009,
+      "learning_rate": 9.35758623078987e-06,
+      "loss": 0.5503,
+      "step": 2337
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.592855026050138,
+      "learning_rate": 9.356946275710437e-06,
+      "loss": 0.5997,
+      "step": 2338
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 26.99190064149443,
+      "learning_rate": 9.356306023942228e-06,
+      "loss": 0.4143,
+      "step": 2339
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.522386199379512,
+      "learning_rate": 9.355665475528839e-06,
+      "loss": 0.5079,
+      "step": 2340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.3465449805730225,
+      "learning_rate": 9.355024630513889e-06,
+      "loss": 0.4527,
+      "step": 2341
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.11068684320371,
+      "learning_rate": 9.354383488941019e-06,
+      "loss": 0.5467,
+      "step": 2342
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.600024314127963,
+      "learning_rate": 9.353742050853884e-06,
+      "loss": 0.5529,
+      "step": 2343
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.755254770739032,
+      "learning_rate": 9.353100316296167e-06,
+      "loss": 0.4993,
+      "step": 2344
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.947007393983167,
+      "learning_rate": 9.352458285311565e-06,
+      "loss": 0.5639,
+      "step": 2345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 19.35708631725734,
+      "learning_rate": 9.351815957943797e-06,
+      "loss": 0.5142,
+      "step": 2346
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.813469418075959,
+      "learning_rate": 9.351173334236605e-06,
+      "loss": 0.5627,
+      "step": 2347
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.171921802783174,
+      "learning_rate": 9.350530414233746e-06,
+      "loss": 0.5326,
+      "step": 2348
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.193263144311574,
+      "learning_rate": 9.349887197979004e-06,
+      "loss": 0.6476,
+      "step": 2349
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.908428189036473,
+      "learning_rate": 9.349243685516174e-06,
+      "loss": 0.5055,
+      "step": 2350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.830605118405026,
+      "learning_rate": 9.348599876889081e-06,
+      "loss": 0.8096,
+      "step": 2351
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.140200743070189,
+      "learning_rate": 9.347955772141563e-06,
+      "loss": 0.4666,
+      "step": 2352
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.020015966754402,
+      "learning_rate": 9.34731137131748e-06,
+      "loss": 0.4987,
+      "step": 2353
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.597559438692725,
+      "learning_rate": 9.346666674460716e-06,
+      "loss": 0.5077,
+      "step": 2354
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.149065567871522,
+      "learning_rate": 9.346021681615169e-06,
+      "loss": 0.5382,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.085801363991614,
+      "learning_rate": 9.345376392824761e-06,
+      "loss": 0.5162,
+      "step": 2356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.792955325273377,
+      "learning_rate": 9.344730808133433e-06,
+      "loss": 0.5437,
+      "step": 2357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.567598938087936,
+      "learning_rate": 9.344084927585149e-06,
+      "loss": 0.5875,
+      "step": 2358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.71096497843664,
+      "learning_rate": 9.343438751223888e-06,
+      "loss": 0.4153,
+      "step": 2359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 51.78664838790106,
+      "learning_rate": 9.34279227909365e-06,
+      "loss": 0.5184,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.894311367552748,
+      "learning_rate": 9.342145511238461e-06,
+      "loss": 0.4844,
+      "step": 2361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9280109615818493,
+      "learning_rate": 9.341498447702363e-06,
+      "loss": 0.8138,
+      "step": 2362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.91572950035025,
+      "learning_rate": 9.340851088529415e-06,
+      "loss": 0.5088,
+      "step": 2363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.1949204325167635,
+      "learning_rate": 9.3402034337637e-06,
+      "loss": 0.4296,
+      "step": 2364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.575329152429918,
+      "learning_rate": 9.339555483449321e-06,
+      "loss": 0.5402,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.668073919414452,
+      "learning_rate": 9.338907237630401e-06,
+      "loss": 0.5979,
+      "step": 2366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.519526056359781,
+      "learning_rate": 9.338258696351083e-06,
+      "loss": 0.5291,
+      "step": 2367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.717596159887797,
+      "learning_rate": 9.33760985965553e-06,
+      "loss": 0.4465,
+      "step": 2368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.056820246707437,
+      "learning_rate": 9.336960727587923e-06,
+      "loss": 0.5121,
+      "step": 2369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.781966667642694,
+      "learning_rate": 9.336311300192467e-06,
+      "loss": 0.6447,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.758891732529526,
+      "learning_rate": 9.335661577513383e-06,
+      "loss": 0.5871,
+      "step": 2371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.153157842836043,
+      "learning_rate": 9.335011559594915e-06,
+      "loss": 0.4796,
+      "step": 2372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.472586439628785,
+      "learning_rate": 9.33436124648133e-06,
+      "loss": 0.5337,
+      "step": 2373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.709194599219141,
+      "learning_rate": 9.333710638216906e-06,
+      "loss": 0.4851,
+      "step": 2374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0285908929040746,
+      "learning_rate": 9.333059734845949e-06,
+      "loss": 0.7983,
+      "step": 2375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.862581714530034,
+      "learning_rate": 9.332408536412783e-06,
+      "loss": 0.5228,
+      "step": 2376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.495929663923735,
+      "learning_rate": 9.33175704296175e-06,
+      "loss": 0.4912,
+      "step": 2377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.218231939993082,
+      "learning_rate": 9.331105254537216e-06,
+      "loss": 0.5161,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.569802085301074,
+      "learning_rate": 9.330453171183564e-06,
+      "loss": 0.5997,
+      "step": 2379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.421631012332951,
+      "learning_rate": 9.329800792945198e-06,
+      "loss": 0.5116,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.091755698988333,
+      "learning_rate": 9.329148119866541e-06,
+      "loss": 0.7114,
+      "step": 2381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.905959777320177,
+      "learning_rate": 9.32849515199204e-06,
+      "loss": 0.4839,
+      "step": 2382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.089690725613659,
+      "learning_rate": 9.327841889366156e-06,
+      "loss": 0.5296,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.225962852244641,
+      "learning_rate": 9.327188332033376e-06,
+      "loss": 0.5952,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.760375531864953,
+      "learning_rate": 9.326534480038202e-06,
+      "loss": 0.4785,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.77487561953259,
+      "learning_rate": 9.325880333425159e-06,
+      "loss": 0.4916,
+      "step": 2386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.948420716560626,
+      "learning_rate": 9.325225892238793e-06,
+      "loss": 0.5097,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.402902976531848,
+      "learning_rate": 9.324571156523667e-06,
+      "loss": 0.675,
+      "step": 2388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.916872834177485,
+      "learning_rate": 9.323916126324365e-06,
+      "loss": 0.5743,
+      "step": 2389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.635471726873575,
+      "learning_rate": 9.323260801685497e-06,
+      "loss": 0.5712,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.153251352983008,
+      "learning_rate": 9.322605182651679e-06,
+      "loss": 0.6369,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.532452620651443,
+      "learning_rate": 9.321949269267563e-06,
+      "loss": 0.497,
+      "step": 2392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.183107505015973,
+      "learning_rate": 9.32129306157781e-06,
+      "loss": 0.5554,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.441843712917574,
+      "learning_rate": 9.320636559627105e-06,
+      "loss": 0.4662,
+      "step": 2394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.584412122699714,
+      "learning_rate": 9.319979763460155e-06,
+      "loss": 0.4373,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 18.171820214815057,
+      "learning_rate": 9.319322673121683e-06,
+      "loss": 0.5317,
+      "step": 2396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.440053043090249,
+      "learning_rate": 9.318665288656435e-06,
+      "loss": 0.514,
+      "step": 2397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.62671030902793,
+      "learning_rate": 9.318007610109177e-06,
+      "loss": 0.6336,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.667415527310038,
+      "learning_rate": 9.317349637524691e-06,
+      "loss": 0.5599,
+      "step": 2399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.51606268162133,
+      "learning_rate": 9.316691370947783e-06,
+      "loss": 0.51,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.14046026500281,
+      "learning_rate": 9.316032810423281e-06,
+      "loss": 0.5086,
+      "step": 2401
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.504704806671771,
+      "learning_rate": 9.315373955996026e-06,
+      "loss": 0.5511,
+      "step": 2402
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.55634225980576,
+      "learning_rate": 9.314714807710885e-06,
+      "loss": 0.5369,
+      "step": 2403
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.869131647502941,
+      "learning_rate": 9.314055365612744e-06,
+      "loss": 0.5392,
+      "step": 2404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.173349151591526,
+      "learning_rate": 9.313395629746507e-06,
+      "loss": 0.6019,
+      "step": 2405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 26.722660097809168,
+      "learning_rate": 9.312735600157098e-06,
+      "loss": 0.6151,
+      "step": 2406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 25.71574216060442,
+      "learning_rate": 9.312075276889463e-06,
+      "loss": 0.6171,
+      "step": 2407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.523452530920682,
+      "learning_rate": 9.311414659988569e-06,
+      "loss": 0.5545,
+      "step": 2408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.784210763785886,
+      "learning_rate": 9.310753749499398e-06,
+      "loss": 0.5595,
+      "step": 2409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.007886101368136,
+      "learning_rate": 9.310092545466957e-06,
+      "loss": 0.4204,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 13.198332695856191,
+      "learning_rate": 9.30943104793627e-06,
+      "loss": 0.448,
+      "step": 2411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6245101740785093,
+      "learning_rate": 9.308769256952383e-06,
+      "loss": 0.7048,
+      "step": 2412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.794588111840986,
+      "learning_rate": 9.30810717256036e-06,
+      "loss": 0.5397,
+      "step": 2413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 26.220010098313967,
+      "learning_rate": 9.307444794805287e-06,
+      "loss": 0.5558,
+      "step": 2414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.200313212500605,
+      "learning_rate": 9.30678212373227e-06,
+      "loss": 0.5801,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.850426192963258,
+      "learning_rate": 9.306119159386431e-06,
+      "loss": 0.4965,
+      "step": 2416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.028084040245703,
+      "learning_rate": 9.305455901812917e-06,
+      "loss": 0.4787,
+      "step": 2417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 12.473507121044989,
+      "learning_rate": 9.304792351056892e-06,
+      "loss": 0.5096,
+      "step": 2418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.0214175190237,
+      "learning_rate": 9.304128507163541e-06,
+      "loss": 0.4999,
+      "step": 2419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.525805062522057,
+      "learning_rate": 9.303464370178072e-06,
+      "loss": 0.4725,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.569127724689583,
+      "learning_rate": 9.302799940145704e-06,
+      "loss": 0.4947,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 32.28844609291413,
+      "learning_rate": 9.302135217111686e-06,
+      "loss": 0.4523,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.69840156149032,
+      "learning_rate": 9.30147020112128e-06,
+      "loss": 0.4878,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.189655231533726,
+      "learning_rate": 9.300804892219775e-06,
+      "loss": 0.446,
+      "step": 2424
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.30167774464952,
+      "learning_rate": 9.30013929045247e-06,
+      "loss": 0.5032,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 19.597652228578706,
+      "learning_rate": 9.299473395864693e-06,
+      "loss": 0.4216,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.452629630977732,
+      "learning_rate": 9.298807208501789e-06,
+      "loss": 0.5197,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 20.44426467180994,
+      "learning_rate": 9.29814072840912e-06,
+      "loss": 0.5568,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.189740582600365,
+      "learning_rate": 9.29747395563207e-06,
+      "loss": 0.5123,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.570526516140736,
+      "learning_rate": 9.296806890216047e-06,
+      "loss": 0.5345,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.324280300100078,
+      "learning_rate": 9.296139532206472e-06,
+      "loss": 0.4999,
+      "step": 2431
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.477180041465752,
+      "learning_rate": 9.29547188164879e-06,
+      "loss": 0.507,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.517001374073121,
+      "learning_rate": 9.294803938588465e-06,
+      "loss": 0.5918,
+      "step": 2433
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.188661979563058,
+      "learning_rate": 9.29413570307098e-06,
+      "loss": 0.6147,
+      "step": 2434
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 43.07311414233131,
+      "learning_rate": 9.293467175141843e-06,
+      "loss": 0.5019,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.776816454169472,
+      "learning_rate": 9.292798354846572e-06,
+      "loss": 0.5381,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.6484888568462885,
+      "learning_rate": 9.292129242230714e-06,
+      "loss": 0.6838,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.807140299450204,
+      "learning_rate": 9.291459837339833e-06,
+      "loss": 0.6563,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.925462203556657,
+      "learning_rate": 9.290790140219511e-06,
+      "loss": 0.5623,
+      "step": 2439
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.546758400955122,
+      "learning_rate": 9.290120150915353e-06,
+      "loss": 0.5194,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.784183530768108,
+      "learning_rate": 9.28944986947298e-06,
+      "loss": 0.5002,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.57234408298958,
+      "learning_rate": 9.288779295938035e-06,
+      "loss": 0.4849,
+      "step": 2442
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.49834798751913,
+      "learning_rate": 9.288108430356187e-06,
+      "loss": 0.4188,
+      "step": 2443
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.767598475740847,
+      "learning_rate": 9.287437272773111e-06,
+      "loss": 0.5047,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.255161194218507,
+      "learning_rate": 9.286765823234513e-06,
+      "loss": 0.5996,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.198739287675478,
+      "learning_rate": 9.28609408178612e-06,
+      "loss": 0.4375,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.448057463891829,
+      "learning_rate": 9.285422048473667e-06,
+      "loss": 0.4166,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.30994355429091,
+      "learning_rate": 9.284749723342923e-06,
+      "loss": 0.5336,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.9855953141820635,
+      "learning_rate": 9.284077106439666e-06,
+      "loss": 0.4575,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.397335948266328,
+      "learning_rate": 9.2834041978097e-06,
+      "loss": 0.6253,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.000371405335432,
+      "learning_rate": 9.282730997498847e-06,
+      "loss": 0.5261,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.726473919023052,
+      "learning_rate": 9.282057505552949e-06,
+      "loss": 0.5747,
+      "step": 2452
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.949545817493783,
+      "learning_rate": 9.28138372201787e-06,
+      "loss": 0.5404,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.73155747377183,
+      "learning_rate": 9.280709646939486e-06,
+      "loss": 0.4835,
+      "step": 2454
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.444274898234705,
+      "learning_rate": 9.280035280363704e-06,
+      "loss": 0.5811,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.665743149452462,
+      "learning_rate": 9.279360622336443e-06,
+      "loss": 0.5999,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.945447721029234,
+      "learning_rate": 9.278685672903645e-06,
+      "loss": 0.4698,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.121389498048725,
+      "learning_rate": 9.27801043211127e-06,
+      "loss": 0.5593,
+      "step": 2458
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.5808933993508125,
+      "learning_rate": 9.2773349000053e-06,
+      "loss": 0.483,
+      "step": 2459
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.962917281391372,
+      "learning_rate": 9.276659076631735e-06,
+      "loss": 0.549,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.822893669458935,
+      "learning_rate": 9.275982962036597e-06,
+      "loss": 0.5248,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.223769517400132,
+      "learning_rate": 9.275306556265924e-06,
+      "loss": 0.6039,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.32504922550695,
+      "learning_rate": 9.27462985936578e-06,
+      "loss": 0.4871,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.720489975184853,
+      "learning_rate": 9.273952871382241e-06,
+      "loss": 0.5329,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.482527723811488,
+      "learning_rate": 9.273275592361409e-06,
+      "loss": 0.5907,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.078218839483586,
+      "learning_rate": 9.272598022349401e-06,
+      "loss": 0.4178,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.023480824477423,
+      "learning_rate": 9.27192016139236e-06,
+      "loss": 0.5299,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.093166793762716,
+      "learning_rate": 9.271242009536446e-06,
+      "loss": 0.7247,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.801831781740145,
+      "learning_rate": 9.270563566827834e-06,
+      "loss": 0.525,
+      "step": 2469
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.416150563414154,
+      "learning_rate": 9.269884833312727e-06,
+      "loss": 0.4695,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8187143879785572,
+      "learning_rate": 9.269205809037341e-06,
+      "loss": 0.7693,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.970392139337754,
+      "learning_rate": 9.268526494047915e-06,
+      "loss": 0.6106,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.87727537705961,
+      "learning_rate": 9.267846888390709e-06,
+      "loss": 0.5622,
+      "step": 2473
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.9389625659066425,
+      "learning_rate": 9.267166992111999e-06,
+      "loss": 0.5577,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.657036593068147,
+      "learning_rate": 9.266486805258085e-06,
+      "loss": 0.448,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.741941338521555,
+      "learning_rate": 9.265806327875283e-06,
+      "loss": 0.5614,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.796742311271932,
+      "learning_rate": 9.265125560009931e-06,
+      "loss": 0.3837,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.8488264561308725,
+      "learning_rate": 9.264444501708387e-06,
+      "loss": 0.4743,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.196050517325064,
+      "learning_rate": 9.263763153017027e-06,
+      "loss": 0.5692,
+      "step": 2479
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.591948189495586,
+      "learning_rate": 9.263081513982248e-06,
+      "loss": 0.5877,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.560698499177985,
+      "learning_rate": 9.262399584650467e-06,
+      "loss": 0.5185,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.052309492995523,
+      "learning_rate": 9.261717365068122e-06,
+      "loss": 0.5485,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 46.93110519108371,
+      "learning_rate": 9.261034855281665e-06,
+      "loss": 0.5838,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.710703172303583,
+      "learning_rate": 9.260352055337576e-06,
+      "loss": 0.4912,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.03631974991413,
+      "learning_rate": 9.25966896528235e-06,
+      "loss": 0.5277,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.104918748467975,
+      "learning_rate": 9.2589855851625e-06,
+      "loss": 0.6222,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9243048971529242,
+      "learning_rate": 9.258301915024562e-06,
+      "loss": 0.7484,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.222136407621248,
+      "learning_rate": 9.257617954915091e-06,
+      "loss": 0.6115,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.757582215831938,
+      "learning_rate": 9.256933704880662e-06,
+      "loss": 0.6426,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.12920650684392,
+      "learning_rate": 9.256249164967873e-06,
+      "loss": 0.5708,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.652958899064322,
+      "learning_rate": 9.255564335223331e-06,
+      "loss": 0.5465,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.623714519394234,
+      "learning_rate": 9.254879215693675e-06,
+      "loss": 0.4032,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 22.1191026597587,
+      "learning_rate": 9.254193806425557e-06,
+      "loss": 0.4683,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.713996117803841,
+      "learning_rate": 9.253508107465649e-06,
+      "loss": 0.5062,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.492303245525071,
+      "learning_rate": 9.25282211886065e-06,
+      "loss": 0.4343,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 30.298879748299004,
+      "learning_rate": 9.252135840657264e-06,
+      "loss": 0.5667,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.432676933030404,
+      "learning_rate": 9.251449272902228e-06,
+      "loss": 0.4809,
+      "step": 2497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 28.26416947822469,
+      "learning_rate": 9.250762415642297e-06,
+      "loss": 0.5254,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.33568859444821,
+      "learning_rate": 9.25007526892424e-06,
+      "loss": 0.4291,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.909476565648416,
+      "learning_rate": 9.249387832794846e-06,
+      "loss": 0.4621,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.198270712266282,
+      "learning_rate": 9.24870010730093e-06,
+      "loss": 0.4676,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.425200335869107,
+      "learning_rate": 9.248012092489323e-06,
+      "loss": 0.4467,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.19987058004298,
+      "learning_rate": 9.247323788406875e-06,
+      "loss": 0.3878,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.212242801383888,
+      "learning_rate": 9.246635195100456e-06,
+      "loss": 0.8276,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.925433909959022,
+      "learning_rate": 9.245946312616955e-06,
+      "loss": 0.5354,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.424861537913615,
+      "learning_rate": 9.245257141003286e-06,
+      "loss": 0.487,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.584149192249618,
+      "learning_rate": 9.244567680306373e-06,
+      "loss": 0.4522,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 62.91421109454137,
+      "learning_rate": 9.243877930573171e-06,
+      "loss": 0.4803,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.27125538295301,
+      "learning_rate": 9.243187891850644e-06,
+      "loss": 0.3923,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 35.211647896280795,
+      "learning_rate": 9.242497564185783e-06,
+      "loss": 0.5536,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 14.580530434339302,
+      "learning_rate": 9.241806947625597e-06,
+      "loss": 0.4654,
+      "step": 2511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.098420149614952,
+      "learning_rate": 9.241116042217112e-06,
+      "loss": 0.4818,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 39.4798364656663,
+      "learning_rate": 9.240424848007377e-06,
+      "loss": 0.6168,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 27.424302975420247,
+      "learning_rate": 9.239733365043457e-06,
+      "loss": 0.5216,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 33.38063410119438,
+      "learning_rate": 9.239041593372442e-06,
+      "loss": 0.5337,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.404511840955932,
+      "learning_rate": 9.238349533041439e-06,
+      "loss": 0.4084,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.684002451259648,
+      "learning_rate": 9.23765718409757e-06,
+      "loss": 0.5143,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.415242179955897,
+      "learning_rate": 9.236964546587984e-06,
+      "loss": 0.5341,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.806998021797452,
+      "learning_rate": 9.236271620559846e-06,
+      "loss": 0.6447,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4324762624944185,
+      "learning_rate": 9.235578406060341e-06,
+      "loss": 0.837,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.565782958642714,
+      "learning_rate": 9.234884903136674e-06,
+      "loss": 0.574,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.774044577179643,
+      "learning_rate": 9.23419111183607e-06,
+      "loss": 0.5678,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.762462158565581,
+      "learning_rate": 9.233497032205772e-06,
+      "loss": 0.513,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.794159859542799,
+      "learning_rate": 9.232802664293044e-06,
+      "loss": 0.4896,
+      "step": 2524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.485404807296382,
+      "learning_rate": 9.232108008145172e-06,
+      "loss": 0.5039,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 90.32505248963406,
+      "learning_rate": 9.231413063809453e-06,
+      "loss": 0.4611,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.289030718991766,
+      "learning_rate": 9.230717831333216e-06,
+      "loss": 0.5405,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.165671914562447,
+      "learning_rate": 9.2300223107638e-06,
+      "loss": 0.4546,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 15.121020705938383,
+      "learning_rate": 9.229326502148568e-06,
+      "loss": 0.5096,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.835921601202983,
+      "learning_rate": 9.2286304055349e-06,
+      "loss": 0.4441,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.580285691286598,
+      "learning_rate": 9.227934020970198e-06,
+      "loss": 0.5563,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 32.48615750367395,
+      "learning_rate": 9.227237348501882e-06,
+      "loss": 0.3973,
+      "step": 2532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 17.40107543459545,
+      "learning_rate": 9.226540388177397e-06,
+      "loss": 0.5311,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.52022996618547,
+      "learning_rate": 9.225843140044196e-06,
+      "loss": 0.5827,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.39529459870494,
+      "learning_rate": 9.22514560414976e-06,
+      "loss": 0.5156,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.989286212428821,
+      "learning_rate": 9.224447780541594e-06,
+      "loss": 0.5168,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.439815048318808,
+      "learning_rate": 9.223749669267208e-06,
+      "loss": 0.4889,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 22.630330471775615,
+      "learning_rate": 9.223051270374147e-06,
+      "loss": 0.4263,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 25.179541835736448,
+      "learning_rate": 9.222352583909965e-06,
+      "loss": 0.6594,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.47621975635947,
+      "learning_rate": 9.221653609922243e-06,
+      "loss": 0.4813,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.332370020825211,
+      "learning_rate": 9.220954348458576e-06,
+      "loss": 0.559,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.375319915578155,
+      "learning_rate": 9.22025479956658e-06,
+      "loss": 0.8182,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.056674740702954,
+      "learning_rate": 9.219554963293892e-06,
+      "loss": 0.592,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.429206834741212,
+      "learning_rate": 9.218854839688167e-06,
+      "loss": 0.5436,
+      "step": 2544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 20.44853206466905,
+      "learning_rate": 9.218154428797082e-06,
+      "loss": 0.5043,
+      "step": 2545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.91549191655946,
+      "learning_rate": 9.21745373066833e-06,
+      "loss": 0.4897,
+      "step": 2546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.417201521908805,
+      "learning_rate": 9.216752745349623e-06,
+      "loss": 0.6356,
+      "step": 2547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.873005685659253,
+      "learning_rate": 9.216051472888703e-06,
+      "loss": 0.5248,
+      "step": 2548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.146367653772502,
+      "learning_rate": 9.215349913333318e-06,
+      "loss": 0.3668,
+      "step": 2549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.600295432833862,
+      "learning_rate": 9.21464806673124e-06,
+      "loss": 0.5756,
+      "step": 2550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.657426145864715,
+      "learning_rate": 9.213945933130267e-06,
+      "loss": 0.5762,
+      "step": 2551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.629493988686306,
+      "learning_rate": 9.213243512578205e-06,
+      "loss": 0.5793,
+      "step": 2552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.729813669760075,
+      "learning_rate": 9.212540805122887e-06,
+      "loss": 0.5318,
+      "step": 2553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.902671855330587,
+      "learning_rate": 9.211837810812167e-06,
+      "loss": 0.4145,
+      "step": 2554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.677669255188045,
+      "learning_rate": 9.211134529693916e-06,
+      "loss": 0.4995,
+      "step": 2555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.269643516887017,
+      "learning_rate": 9.210430961816021e-06,
+      "loss": 0.441,
+      "step": 2556
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.894536804746,
+      "learning_rate": 9.209727107226395e-06,
+      "loss": 0.4907,
+      "step": 2557
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.336126978178077,
+      "learning_rate": 9.209022965972966e-06,
+      "loss": 0.5906,
+      "step": 2558
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.108651582159833,
+      "learning_rate": 9.208318538103682e-06,
+      "loss": 0.6205,
+      "step": 2559
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 30.485036427846637,
+      "learning_rate": 9.207613823666515e-06,
+      "loss": 0.5276,
+      "step": 2560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.077889118071428,
+      "learning_rate": 9.206908822709448e-06,
+      "loss": 0.4931,
+      "step": 2561
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.16899943335678,
+      "learning_rate": 9.20620353528049e-06,
+      "loss": 0.3961,
+      "step": 2562
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.4917613016215805,
+      "learning_rate": 9.205497961427671e-06,
+      "loss": 0.5282,
+      "step": 2563
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.783857823861926,
+      "learning_rate": 9.204792101199036e-06,
+      "loss": 0.5466,
+      "step": 2564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.99083500039673,
+      "learning_rate": 9.204085954642648e-06,
+      "loss": 0.7573,
+      "step": 2565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.133505309768289,
+      "learning_rate": 9.203379521806597e-06,
+      "loss": 0.4989,
+      "step": 2566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.974075518155099,
+      "learning_rate": 9.202672802738983e-06,
+      "loss": 0.5398,
+      "step": 2567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.9834007132912825,
+      "learning_rate": 9.201965797487936e-06,
+      "loss": 0.5111,
+      "step": 2568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.552259079500377,
+      "learning_rate": 9.201258506101594e-06,
+      "loss": 0.6328,
+      "step": 2569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.880508562060243,
+      "learning_rate": 9.200550928628125e-06,
+      "loss": 0.4546,
+      "step": 2570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.77204773838034,
+      "learning_rate": 9.19984306511571e-06,
+      "loss": 0.4323,
+      "step": 2571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.777188724447253,
+      "learning_rate": 9.19913491561255e-06,
+      "loss": 0.5118,
+      "step": 2572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.531603890827958,
+      "learning_rate": 9.198426480166872e-06,
+      "loss": 0.3921,
+      "step": 2573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.589970461068415,
+      "learning_rate": 9.197717758826912e-06,
+      "loss": 0.6527,
+      "step": 2574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.794497373388573,
+      "learning_rate": 9.197008751640932e-06,
+      "loss": 0.583,
+      "step": 2575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.348356145936648,
+      "learning_rate": 9.196299458657214e-06,
+      "loss": 0.5283,
+      "step": 2576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 22.259248784026234,
+      "learning_rate": 9.195589879924057e-06,
+      "loss": 0.4851,
+      "step": 2577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.863248117331584,
+      "learning_rate": 9.194880015489777e-06,
+      "loss": 0.4054,
+      "step": 2578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.8738164990025306,
+      "learning_rate": 9.194169865402718e-06,
+      "loss": 0.4803,
+      "step": 2579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.476797825768649,
+      "learning_rate": 9.193459429711236e-06,
+      "loss": 0.6095,
+      "step": 2580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.785755875856864,
+      "learning_rate": 9.192748708463706e-06,
+      "loss": 0.5441,
+      "step": 2581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.37561045281538,
+      "learning_rate": 9.192037701708527e-06,
+      "loss": 0.5721,
+      "step": 2582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 21.192321750512498,
+      "learning_rate": 9.191326409494117e-06,
+      "loss": 0.5907,
+      "step": 2583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.251914675512925,
+      "learning_rate": 9.190614831868909e-06,
+      "loss": 0.5632,
+      "step": 2584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.551472414252631,
+      "learning_rate": 9.189902968881361e-06,
+      "loss": 0.6058,
+      "step": 2585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.924256135763597,
+      "learning_rate": 9.189190820579945e-06,
+      "loss": 0.4775,
+      "step": 2586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.992756026332936,
+      "learning_rate": 9.188478387013158e-06,
+      "loss": 0.5073,
+      "step": 2587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.755848760740513,
+      "learning_rate": 9.187765668229511e-06,
+      "loss": 0.5445,
+      "step": 2588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.795934082978122,
+      "learning_rate": 9.187052664277538e-06,
+      "loss": 0.5182,
+      "step": 2589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.370498918210606,
+      "learning_rate": 9.186339375205792e-06,
+      "loss": 0.5483,
+      "step": 2590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 54.739919479181914,
+      "learning_rate": 9.185625801062844e-06,
+      "loss": 0.5733,
+      "step": 2591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.224226119420543,
+      "learning_rate": 9.184911941897284e-06,
+      "loss": 0.5832,
+      "step": 2592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9291885152989812,
+      "learning_rate": 9.184197797757726e-06,
+      "loss": 0.7579,
+      "step": 2593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.158713547915966,
+      "learning_rate": 9.183483368692798e-06,
+      "loss": 0.5346,
+      "step": 2594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 28.36482847742256,
+      "learning_rate": 9.18276865475115e-06,
+      "loss": 0.614,
+      "step": 2595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 30.040135825995684,
+      "learning_rate": 9.182053655981448e-06,
+      "loss": 0.5405,
+      "step": 2596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.741222409335834,
+      "learning_rate": 9.181338372432384e-06,
+      "loss": 0.7818,
+      "step": 2597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.316948756631708,
+      "learning_rate": 9.180622804152665e-06,
+      "loss": 0.4693,
+      "step": 2598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.757057910597084,
+      "learning_rate": 9.17990695119102e-06,
+      "loss": 0.4854,
+      "step": 2599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3361897908881875,
+      "learning_rate": 9.17919081359619e-06,
+      "loss": 0.8122,
+      "step": 2600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.332700524507388,
+      "learning_rate": 9.178474391416945e-06,
+      "loss": 0.5953,
+      "step": 2601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.629208250788219,
+      "learning_rate": 9.177757684702067e-06,
+      "loss": 0.4675,
+      "step": 2602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.545090867783461,
+      "learning_rate": 9.177040693500363e-06,
+      "loss": 0.5637,
+      "step": 2603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 28.683282257352296,
+      "learning_rate": 9.176323417860657e-06,
+      "loss": 0.4617,
+      "step": 2604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.369200233673714,
+      "learning_rate": 9.175605857831792e-06,
+      "loss": 0.4633,
+      "step": 2605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.529693331631004,
+      "learning_rate": 9.174888013462631e-06,
+      "loss": 0.4757,
+      "step": 2606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.462569302392129,
+      "learning_rate": 9.174169884802053e-06,
+      "loss": 0.5424,
+      "step": 2607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.921053903142892,
+      "learning_rate": 9.173451471898964e-06,
+      "loss": 0.5522,
+      "step": 2608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.283160838146824,
+      "learning_rate": 9.17273277480228e-06,
+      "loss": 0.4727,
+      "step": 2609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 33.8057795531322,
+      "learning_rate": 9.172013793560945e-06,
+      "loss": 0.5187,
+      "step": 2610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 55.942159589408746,
+      "learning_rate": 9.171294528223918e-06,
+      "loss": 0.5217,
+      "step": 2611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.691268718933056,
+      "learning_rate": 9.170574978840177e-06,
+      "loss": 0.5367,
+      "step": 2612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.886203100024414,
+      "learning_rate": 9.169855145458719e-06,
+      "loss": 0.5165,
+      "step": 2613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 32.11882770547248,
+      "learning_rate": 9.169135028128562e-06,
+      "loss": 0.5901,
+      "step": 2614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.425565495314975,
+      "learning_rate": 9.168414626898742e-06,
+      "loss": 0.4466,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.555984613356024,
+      "learning_rate": 9.167693941818319e-06,
+      "loss": 0.6382,
+      "step": 2616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.506740480477227,
+      "learning_rate": 9.166972972936366e-06,
+      "loss": 0.5585,
+      "step": 2617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 145.84078595823232,
+      "learning_rate": 9.166251720301975e-06,
+      "loss": 0.517,
+      "step": 2618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.401884429388244,
+      "learning_rate": 9.165530183964264e-06,
+      "loss": 0.4205,
+      "step": 2619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.381156871507773,
+      "learning_rate": 9.164808363972366e-06,
+      "loss": 0.607,
+      "step": 2620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.961008340209103,
+      "learning_rate": 9.164086260375434e-06,
+      "loss": 0.6249,
+      "step": 2621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.389230743669525,
+      "learning_rate": 9.163363873222637e-06,
+      "loss": 0.5441,
+      "step": 2622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.237010032548141,
+      "learning_rate": 9.162641202563167e-06,
+      "loss": 0.4583,
+      "step": 2623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.265014796921574,
+      "learning_rate": 9.16191824844624e-06,
+      "loss": 0.6341,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.190677904537326,
+      "learning_rate": 9.161195010921078e-06,
+      "loss": 0.3715,
+      "step": 2625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.053086064197794,
+      "learning_rate": 9.160471490036938e-06,
+      "loss": 0.5014,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.767712269581398,
+      "learning_rate": 9.15974768584308e-06,
+      "loss": 0.5729,
+      "step": 2627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.80920458173317,
+      "learning_rate": 9.1590235983888e-06,
+      "loss": 0.513,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.888724045773712,
+      "learning_rate": 9.158299227723404e-06,
+      "loss": 0.5009,
+      "step": 2629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.885365569208778,
+      "learning_rate": 9.157574573896214e-06,
+      "loss": 0.5174,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.956146434316692,
+      "learning_rate": 9.156849636956577e-06,
+      "loss": 0.5004,
+      "step": 2631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.95252167472639,
+      "learning_rate": 9.15612441695386e-06,
+      "loss": 0.4796,
+      "step": 2632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.507886779878856,
+      "learning_rate": 9.155398913937447e-06,
+      "loss": 0.4917,
+      "step": 2633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.833616891996787,
+      "learning_rate": 9.15467312795674e-06,
+      "loss": 0.6351,
+      "step": 2634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.16676818949695,
+      "learning_rate": 9.153947059061164e-06,
+      "loss": 0.5741,
+      "step": 2635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.239587265390325,
+      "learning_rate": 9.153220707300159e-06,
+      "loss": 0.5786,
+      "step": 2636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.667357714495669,
+      "learning_rate": 9.152494072723185e-06,
+      "loss": 0.5181,
+      "step": 2637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.91888217015337,
+      "learning_rate": 9.151767155379726e-06,
+      "loss": 0.5634,
+      "step": 2638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.9204639603483,
+      "learning_rate": 9.151039955319283e-06,
+      "loss": 0.5116,
+      "step": 2639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.88772845347364,
+      "learning_rate": 9.15031247259137e-06,
+      "loss": 0.43,
+      "step": 2640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.08685373340777,
+      "learning_rate": 9.149584707245529e-06,
+      "loss": 0.5801,
+      "step": 2641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.213894781418435,
+      "learning_rate": 9.148856659331316e-06,
+      "loss": 0.4946,
+      "step": 2642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 17.508389251196004,
+      "learning_rate": 9.148128328898309e-06,
+      "loss": 0.5453,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 18.671213353461162,
+      "learning_rate": 9.147399715996104e-06,
+      "loss": 0.5326,
+      "step": 2644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.569214914728917,
+      "learning_rate": 9.146670820674317e-06,
+      "loss": 0.5294,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.061035982864139,
+      "learning_rate": 9.145941642982579e-06,
+      "loss": 0.5537,
+      "step": 2646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 14.896795160462839,
+      "learning_rate": 9.145212182970546e-06,
+      "loss": 0.5069,
+      "step": 2647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.20568054148256,
+      "learning_rate": 9.144482440687893e-06,
+      "loss": 0.4416,
+      "step": 2648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.431978398575177,
+      "learning_rate": 9.14375241618431e-06,
+      "loss": 0.5661,
+      "step": 2649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.586332709242388,
+      "learning_rate": 9.143022109509509e-06,
+      "loss": 0.5664,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.889385729988712,
+      "learning_rate": 9.14229152071322e-06,
+      "loss": 0.5289,
+      "step": 2651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.78519753647162,
+      "learning_rate": 9.141560649845193e-06,
+      "loss": 0.5387,
+      "step": 2652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8148492250441912,
+      "learning_rate": 9.140829496955199e-06,
+      "loss": 0.7852,
+      "step": 2653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.169892683052464,
+      "learning_rate": 9.140098062093023e-06,
+      "loss": 0.585,
+      "step": 2654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.074948289169805,
+      "learning_rate": 9.139366345308476e-06,
+      "loss": 0.4314,
+      "step": 2655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.517278976978637,
+      "learning_rate": 9.138634346651382e-06,
+      "loss": 0.4916,
+      "step": 2656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.834739225664602,
+      "learning_rate": 9.137902066171586e-06,
+      "loss": 0.4737,
+      "step": 2657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 23.541458096972303,
+      "learning_rate": 9.137169503918958e-06,
+      "loss": 0.62,
+      "step": 2658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.711961359879195,
+      "learning_rate": 9.136436659943377e-06,
+      "loss": 0.5638,
+      "step": 2659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.761390887943612,
+      "learning_rate": 9.13570353429475e-06,
+      "loss": 0.5029,
+      "step": 2660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.742894492901046,
+      "learning_rate": 9.134970127022998e-06,
+      "loss": 0.56,
+      "step": 2661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.283812325446239,
+      "learning_rate": 9.134236438178062e-06,
+      "loss": 0.5206,
+      "step": 2662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4739282423546576,
+      "learning_rate": 9.133502467809905e-06,
+      "loss": 0.79,
+      "step": 2663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.787635192214204,
+      "learning_rate": 9.132768215968503e-06,
+      "loss": 0.5492,
+      "step": 2664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.24531042330498,
+      "learning_rate": 9.132033682703863e-06,
+      "loss": 0.4757,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2496617732216606,
+      "learning_rate": 9.131298868065996e-06,
+      "loss": 0.7401,
+      "step": 2666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.197578303617057,
+      "learning_rate": 9.130563772104944e-06,
+      "loss": 0.4295,
+      "step": 2667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.663797569757676,
+      "learning_rate": 9.12982839487076e-06,
+      "loss": 0.4781,
+      "step": 2668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.490031141753914,
+      "learning_rate": 9.129092736413523e-06,
+      "loss": 0.5966,
+      "step": 2669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.071013244944368,
+      "learning_rate": 9.12835679678333e-06,
+      "loss": 0.5965,
+      "step": 2670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.522197896086466,
+      "learning_rate": 9.127620576030287e-06,
+      "loss": 0.486,
+      "step": 2671
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.272421020590846,
+      "learning_rate": 9.126884074204537e-06,
+      "loss": 0.4963,
+      "step": 2672
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.467738986001912,
+      "learning_rate": 9.126147291356226e-06,
+      "loss": 0.503,
+      "step": 2673
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.1141896284105,
+      "learning_rate": 9.125410227535531e-06,
+      "loss": 0.5695,
+      "step": 2674
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.648507177482564,
+      "learning_rate": 9.124672882792638e-06,
+      "loss": 0.4857,
+      "step": 2675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.1179818061284,
+      "learning_rate": 9.123935257177758e-06,
+      "loss": 0.5686,
+      "step": 2676
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.870964955301107,
+      "learning_rate": 9.123197350741122e-06,
+      "loss": 0.4595,
+      "step": 2677
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.149358771226192,
+      "learning_rate": 9.122459163532977e-06,
+      "loss": 0.5488,
+      "step": 2678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.411454985268062,
+      "learning_rate": 9.121720695603591e-06,
+      "loss": 0.6145,
+      "step": 2679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.740516105962255,
+      "learning_rate": 9.120981947003248e-06,
+      "loss": 0.542,
+      "step": 2680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.443762066885018,
+      "learning_rate": 9.120242917782256e-06,
+      "loss": 0.5937,
+      "step": 2681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.712688444197136,
+      "learning_rate": 9.119503607990939e-06,
+      "loss": 0.4324,
+      "step": 2682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.672688104686376,
+      "learning_rate": 9.118764017679639e-06,
+      "loss": 0.4638,
+      "step": 2683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.005030888702159,
+      "learning_rate": 9.11802414689872e-06,
+      "loss": 0.4597,
+      "step": 2684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.916779748770953,
+      "learning_rate": 9.117283995698566e-06,
+      "loss": 0.5506,
+      "step": 2685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.053146016808885,
+      "learning_rate": 9.116543564129578e-06,
+      "loss": 0.5479,
+      "step": 2686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0612958712181038,
+      "learning_rate": 9.115802852242172e-06,
+      "loss": 0.7628,
+      "step": 2687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.362661532988498,
+      "learning_rate": 9.115061860086789e-06,
+      "loss": 0.6077,
+      "step": 2688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.559208295212759,
+      "learning_rate": 9.11432058771389e-06,
+      "loss": 0.5043,
+      "step": 2689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.677755433437897,
+      "learning_rate": 9.113579035173949e-06,
+      "loss": 0.5876,
+      "step": 2690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.628133299695853,
+      "learning_rate": 9.112837202517465e-06,
+      "loss": 0.5163,
+      "step": 2691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.38217846063007,
+      "learning_rate": 9.112095089794949e-06,
+      "loss": 0.5792,
+      "step": 2692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.665923954730058,
+      "learning_rate": 9.11135269705694e-06,
+      "loss": 0.514,
+      "step": 2693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.222066326154389,
+      "learning_rate": 9.110610024353992e-06,
+      "loss": 0.5672,
+      "step": 2694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 14.93654726590634,
+      "learning_rate": 9.109867071736676e-06,
+      "loss": 0.5939,
+      "step": 2695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 20.131438149119894,
+      "learning_rate": 9.109123839255584e-06,
+      "loss": 0.4105,
+      "step": 2696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.509407994010484,
+      "learning_rate": 9.108380326961326e-06,
+      "loss": 0.5602,
+      "step": 2697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.580340250908982,
+      "learning_rate": 9.107636534904534e-06,
+      "loss": 0.537,
+      "step": 2698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.296750809206602,
+      "learning_rate": 9.106892463135854e-06,
+      "loss": 0.5315,
+      "step": 2699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.719252433209351,
+      "learning_rate": 9.106148111705957e-06,
+      "loss": 0.4292,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.934341042637498,
+      "learning_rate": 9.105403480665528e-06,
+      "loss": 0.3975,
+      "step": 2701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.4946023554731065,
+      "learning_rate": 9.104658570065274e-06,
+      "loss": 0.5177,
+      "step": 2702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.196427226197173,
+      "learning_rate": 9.10391337995592e-06,
+      "loss": 0.3826,
+      "step": 2703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.610373111415772,
+      "learning_rate": 9.10316791038821e-06,
+      "loss": 0.5242,
+      "step": 2704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.250637692795149,
+      "learning_rate": 9.102422161412909e-06,
+      "loss": 0.5557,
+      "step": 2705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.31831212714815,
+      "learning_rate": 9.101676133080796e-06,
+      "loss": 0.5787,
+      "step": 2706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.599183690832428,
+      "learning_rate": 9.100929825442674e-06,
+      "loss": 0.5328,
+      "step": 2707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.037063097978371,
+      "learning_rate": 9.100183238549362e-06,
+      "loss": 0.4955,
+      "step": 2708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.904115617700754,
+      "learning_rate": 9.099436372451701e-06,
+      "loss": 0.4174,
+      "step": 2709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.082020058371917,
+      "learning_rate": 9.098689227200548e-06,
+      "loss": 0.5449,
+      "step": 2710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.927301017084358,
+      "learning_rate": 9.097941802846781e-06,
+      "loss": 0.5799,
+      "step": 2711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.016188007414298,
+      "learning_rate": 9.097194099441297e-06,
+      "loss": 0.5859,
+      "step": 2712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 18.40305954345557,
+      "learning_rate": 9.096446117035009e-06,
+      "loss": 0.528,
+      "step": 2713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 33.49697953234482,
+      "learning_rate": 9.095697855678855e-06,
+      "loss": 0.4962,
+      "step": 2714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.020478555808113,
+      "learning_rate": 9.094949315423784e-06,
+      "loss": 0.5323,
+      "step": 2715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.975604626365348,
+      "learning_rate": 9.094200496320772e-06,
+      "loss": 0.4053,
+      "step": 2716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7521421207622392,
+      "learning_rate": 9.093451398420808e-06,
+      "loss": 0.7873,
+      "step": 2717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 31.39528547274426,
+      "learning_rate": 9.0927020217749e-06,
+      "loss": 0.4823,
+      "step": 2718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.096670818845132,
+      "learning_rate": 9.091952366434082e-06,
+      "loss": 0.5824,
+      "step": 2719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.937075811491798,
+      "learning_rate": 9.0912024324494e-06,
+      "loss": 0.5681,
+      "step": 2720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.14964932693497,
+      "learning_rate": 9.090452219871922e-06,
+      "loss": 0.4736,
+      "step": 2721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.22527602912618,
+      "learning_rate": 9.089701728752733e-06,
+      "loss": 0.4404,
+      "step": 2722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.215694116954673,
+      "learning_rate": 9.088950959142939e-06,
+      "loss": 0.6079,
+      "step": 2723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.85379376077754,
+      "learning_rate": 9.088199911093664e-06,
+      "loss": 0.6108,
+      "step": 2724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.857331019149804,
+      "learning_rate": 9.08744858465605e-06,
+      "loss": 0.617,
+      "step": 2725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.365749288159382,
+      "learning_rate": 9.086696979881262e-06,
+      "loss": 0.5876,
+      "step": 2726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.114395454960682,
+      "learning_rate": 9.085945096820477e-06,
+      "loss": 0.422,
+      "step": 2727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 47.31001569583989,
+      "learning_rate": 9.085192935524897e-06,
+      "loss": 0.5245,
+      "step": 2728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.10741609261972,
+      "learning_rate": 9.084440496045742e-06,
+      "loss": 0.5743,
+      "step": 2729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.408160499593413,
+      "learning_rate": 9.083687778434246e-06,
+      "loss": 0.5959,
+      "step": 2730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.725381274801224,
+      "learning_rate": 9.08293478274167e-06,
+      "loss": 0.5912,
+      "step": 2731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 28.392330681057956,
+      "learning_rate": 9.082181509019288e-06,
+      "loss": 0.6532,
+      "step": 2732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.637465355361893,
+      "learning_rate": 9.081427957318392e-06,
+      "loss": 0.4454,
+      "step": 2733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.443401009155364,
+      "learning_rate": 9.0806741276903e-06,
+      "loss": 0.4663,
+      "step": 2734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.281479843772267,
+      "learning_rate": 9.07992002018634e-06,
+      "loss": 0.5372,
+      "step": 2735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.046426316490415,
+      "learning_rate": 9.079165634857869e-06,
+      "loss": 0.6067,
+      "step": 2736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.656508886533214,
+      "learning_rate": 9.078410971756253e-06,
+      "loss": 0.4176,
+      "step": 2737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.33475728374104,
+      "learning_rate": 9.077656030932882e-06,
+      "loss": 0.4959,
+      "step": 2738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.366637927911914,
+      "learning_rate": 9.076900812439164e-06,
+      "loss": 0.4865,
+      "step": 2739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.964342624454791,
+      "learning_rate": 9.076145316326529e-06,
+      "loss": 0.5578,
+      "step": 2740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.005255147834033,
+      "learning_rate": 9.075389542646417e-06,
+      "loss": 0.5247,
+      "step": 2741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.040498937791626,
+      "learning_rate": 9.074633491450298e-06,
+      "loss": 0.6778,
+      "step": 2742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.617683946785215,
+      "learning_rate": 9.073877162789655e-06,
+      "loss": 0.5069,
+      "step": 2743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.926380695659714,
+      "learning_rate": 9.073120556715989e-06,
+      "loss": 0.53,
+      "step": 2744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 37.01384270646107,
+      "learning_rate": 9.072363673280821e-06,
+      "loss": 0.4953,
+      "step": 2745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.1874186895675,
+      "learning_rate": 9.071606512535694e-06,
+      "loss": 0.5173,
+      "step": 2746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.184306028690244,
+      "learning_rate": 9.070849074532166e-06,
+      "loss": 0.5982,
+      "step": 2747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.838547962326583,
+      "learning_rate": 9.070091359321814e-06,
+      "loss": 0.4313,
+      "step": 2748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.617157598656815,
+      "learning_rate": 9.069333366956237e-06,
+      "loss": 0.4897,
+      "step": 2749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.372446266170055,
+      "learning_rate": 9.06857509748705e-06,
+      "loss": 0.493,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9215496451340814,
+      "learning_rate": 9.067816550965886e-06,
+      "loss": 0.795,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.695534313937628,
+      "learning_rate": 9.067057727444403e-06,
+      "loss": 0.5407,
+      "step": 2752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.933728043704859,
+      "learning_rate": 9.06629862697427e-06,
+      "loss": 0.5553,
+      "step": 2753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 52.851117368786156,
+      "learning_rate": 9.06553924960718e-06,
+      "loss": 0.531,
+      "step": 2754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.271804567172774,
+      "learning_rate": 9.064779595394843e-06,
+      "loss": 0.4923,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.421667108200077,
+      "learning_rate": 9.064019664388988e-06,
+      "loss": 0.5323,
+      "step": 2756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.706494410843068,
+      "learning_rate": 9.063259456641362e-06,
+      "loss": 0.4193,
+      "step": 2757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.766153391971345,
+      "learning_rate": 9.062498972203732e-06,
+      "loss": 0.4623,
+      "step": 2758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 26.016586169133017,
+      "learning_rate": 9.061738211127884e-06,
+      "loss": 0.6741,
+      "step": 2759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.837869550970042,
+      "learning_rate": 9.060977173465624e-06,
+      "loss": 0.5942,
+      "step": 2760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.037339847110202,
+      "learning_rate": 9.060215859268775e-06,
+      "loss": 0.5351,
+      "step": 2761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.788182150900585,
+      "learning_rate": 9.059454268589175e-06,
+      "loss": 0.5802,
+      "step": 2762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.704688925448552,
+      "learning_rate": 9.05869240147869e-06,
+      "loss": 0.6187,
+      "step": 2763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.664428034278409,
+      "learning_rate": 9.057930257989196e-06,
+      "loss": 0.4784,
+      "step": 2764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.279891498344112,
+      "learning_rate": 9.057167838172596e-06,
+      "loss": 0.4348,
+      "step": 2765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 72.14826090533124,
+      "learning_rate": 9.056405142080802e-06,
+      "loss": 0.5481,
+      "step": 2766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.328794568593729,
+      "learning_rate": 9.055642169765755e-06,
+      "loss": 0.5387,
+      "step": 2767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.465779488465414,
+      "learning_rate": 9.054878921279407e-06,
+      "loss": 0.8288,
+      "step": 2768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 23.76754878803984,
+      "learning_rate": 9.054115396673733e-06,
+      "loss": 0.4749,
+      "step": 2769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.49291434757482,
+      "learning_rate": 9.053351596000727e-06,
+      "loss": 0.5667,
+      "step": 2770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.950374473604903,
+      "learning_rate": 9.052587519312398e-06,
+      "loss": 0.5418,
+      "step": 2771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.542379203968213,
+      "learning_rate": 9.051823166660778e-06,
+      "loss": 0.5251,
+      "step": 2772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 263.5018321933999,
+      "learning_rate": 9.051058538097914e-06,
+      "loss": 0.5352,
+      "step": 2773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7320069978128503,
+      "learning_rate": 9.050293633675874e-06,
+      "loss": 0.7585,
+      "step": 2774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 19.000208714897926,
+      "learning_rate": 9.049528453446749e-06,
+      "loss": 0.5288,
+      "step": 2775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.197806400243332,
+      "learning_rate": 9.04876299746264e-06,
+      "loss": 0.5847,
+      "step": 2776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.398696321613588,
+      "learning_rate": 9.047997265775671e-06,
+      "loss": 0.5942,
+      "step": 2777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.799406256297857,
+      "learning_rate": 9.047231258437987e-06,
+      "loss": 0.4589,
+      "step": 2778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.4328264300236,
+      "learning_rate": 9.04646497550175e-06,
+      "loss": 0.5637,
+      "step": 2779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.959873039192779,
+      "learning_rate": 9.045698417019138e-06,
+      "loss": 0.6661,
+      "step": 2780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 34.64621807878258,
+      "learning_rate": 9.044931583042353e-06,
+      "loss": 0.6013,
+      "step": 2781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.0953207231415,
+      "learning_rate": 9.044164473623611e-06,
+      "loss": 0.6048,
+      "step": 2782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.118677644067436,
+      "learning_rate": 9.043397088815149e-06,
+      "loss": 0.5626,
+      "step": 2783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.353159678667806,
+      "learning_rate": 9.042629428669222e-06,
+      "loss": 0.4991,
+      "step": 2784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 28.29980701652979,
+      "learning_rate": 9.04186149323811e-06,
+      "loss": 0.5249,
+      "step": 2785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.356789532893897,
+      "learning_rate": 9.041093282574096e-06,
+      "loss": 0.5346,
+      "step": 2786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.077401377439228,
+      "learning_rate": 9.0403247967295e-06,
+      "loss": 0.4401,
+      "step": 2787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.903409987263079,
+      "learning_rate": 9.039556035756647e-06,
+      "loss": 0.5686,
+      "step": 2788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.462397990918333,
+      "learning_rate": 9.03878699970789e-06,
+      "loss": 0.5107,
+      "step": 2789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.166850353538697,
+      "learning_rate": 9.038017688635596e-06,
+      "loss": 0.5575,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.45968025126134,
+      "learning_rate": 9.037248102592151e-06,
+      "loss": 0.5869,
+      "step": 2791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.0689987125572,
+      "learning_rate": 9.03647824162996e-06,
+      "loss": 0.5566,
+      "step": 2792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.477922857316594,
+      "learning_rate": 9.035708105801448e-06,
+      "loss": 0.5281,
+      "step": 2793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.262092286374091,
+      "learning_rate": 9.034937695159059e-06,
+      "loss": 0.4405,
+      "step": 2794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.3926021067397985,
+      "learning_rate": 9.034167009755252e-06,
+      "loss": 0.4898,
+      "step": 2795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.916759715739614,
+      "learning_rate": 9.03339604964251e-06,
+      "loss": 0.5131,
+      "step": 2796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.816515929945446,
+      "learning_rate": 9.03262481487333e-06,
+      "loss": 0.4521,
+      "step": 2797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.759763717173987,
+      "learning_rate": 9.031853305500229e-06,
+      "loss": 0.5493,
+      "step": 2798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.505359107293858,
+      "learning_rate": 9.031081521575744e-06,
+      "loss": 0.5549,
+      "step": 2799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.224780833608214,
+      "learning_rate": 9.030309463152434e-06,
+      "loss": 0.518,
+      "step": 2800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.446368942423616,
+      "learning_rate": 9.029537130282866e-06,
+      "loss": 0.4854,
+      "step": 2801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.604756430485724,
+      "learning_rate": 9.028764523019637e-06,
+      "loss": 0.468,
+      "step": 2802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.436343204214612,
+      "learning_rate": 9.027991641415358e-06,
+      "loss": 0.5239,
+      "step": 2803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 29.027473386467257,
+      "learning_rate": 9.027218485522656e-06,
+      "loss": 0.5161,
+      "step": 2804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.2440137249630143,
+      "learning_rate": 9.026445055394183e-06,
+      "loss": 0.7465,
+      "step": 2805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.021263024944314,
+      "learning_rate": 9.025671351082603e-06,
+      "loss": 0.5629,
+      "step": 2806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.633413197976209,
+      "learning_rate": 9.024897372640604e-06,
+      "loss": 0.5478,
+      "step": 2807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.164248740264703,
+      "learning_rate": 9.024123120120892e-06,
+      "loss": 0.5168,
+      "step": 2808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6125458312244008,
+      "learning_rate": 9.023348593576185e-06,
+      "loss": 0.6813,
+      "step": 2809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.848187013915542,
+      "learning_rate": 9.022573793059229e-06,
+      "loss": 0.517,
+      "step": 2810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.829890662900217,
+      "learning_rate": 9.021798718622783e-06,
+      "loss": 0.5574,
+      "step": 2811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.213597942743643,
+      "learning_rate": 9.021023370319627e-06,
+      "loss": 0.5664,
+      "step": 2812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.352273584453511,
+      "learning_rate": 9.020247748202558e-06,
+      "loss": 0.4392,
+      "step": 2813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.746817921010699,
+      "learning_rate": 9.01947185232439e-06,
+      "loss": 0.6286,
+      "step": 2814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.25490876296854,
+      "learning_rate": 9.018695682737965e-06,
+      "loss": 0.5377,
+      "step": 2815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.36451470066395,
+      "learning_rate": 9.01791923949613e-06,
+      "loss": 0.4806,
+      "step": 2816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.30532602365199,
+      "learning_rate": 9.01714252265176e-06,
+      "loss": 0.4887,
+      "step": 2817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.74820001105749,
+      "learning_rate": 9.016365532257747e-06,
+      "loss": 0.5026,
+      "step": 2818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.693864627067207,
+      "learning_rate": 9.015588268367e-06,
+      "loss": 0.5755,
+      "step": 2819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.68924846857956,
+      "learning_rate": 9.014810731032445e-06,
+      "loss": 0.6505,
+      "step": 2820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.465813008335646,
+      "learning_rate": 9.014032920307031e-06,
+      "loss": 0.5247,
+      "step": 2821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.988601753379998,
+      "learning_rate": 9.013254836243723e-06,
+      "loss": 0.5301,
+      "step": 2822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.172326062182924,
+      "learning_rate": 9.012476478895506e-06,
+      "loss": 0.5561,
+      "step": 2823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 21.95821906253171,
+      "learning_rate": 9.01169784831538e-06,
+      "loss": 0.501,
+      "step": 2824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.464168013461083,
+      "learning_rate": 9.01091894455637e-06,
+      "loss": 0.4949,
+      "step": 2825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 17.783408756352785,
+      "learning_rate": 9.010139767671514e-06,
+      "loss": 0.596,
+      "step": 2826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 215.879102464359,
+      "learning_rate": 9.009360317713869e-06,
+      "loss": 0.5391,
+      "step": 2827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.899524306785344,
+      "learning_rate": 9.008580594736514e-06,
+      "loss": 0.5357,
+      "step": 2828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.547985747373032,
+      "learning_rate": 9.007800598792543e-06,
+      "loss": 0.4862,
+      "step": 2829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.431997442452401,
+      "learning_rate": 9.007020329935073e-06,
+      "loss": 0.6164,
+      "step": 2830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.464133501099823,
+      "learning_rate": 9.006239788217234e-06,
+      "loss": 0.5482,
+      "step": 2831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.00404739433719,
+      "learning_rate": 9.005458973692182e-06,
+      "loss": 0.5483,
+      "step": 2832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.298898821490261,
+      "learning_rate": 9.004677886413082e-06,
+      "loss": 0.4787,
+      "step": 2833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.007449860962884,
+      "learning_rate": 9.003896526433124e-06,
+      "loss": 0.4122,
+      "step": 2834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.009007871325455,
+      "learning_rate": 9.003114893805516e-06,
+      "loss": 0.5013,
+      "step": 2835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.869573547807148,
+      "learning_rate": 9.002332988583484e-06,
+      "loss": 0.5236,
+      "step": 2836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 21.311126809168503,
+      "learning_rate": 9.00155081082027e-06,
+      "loss": 0.564,
+      "step": 2837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 45.47280366094639,
+      "learning_rate": 9.000768360569142e-06,
+      "loss": 0.5578,
+      "step": 2838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 60.03788566105684,
+      "learning_rate": 8.999985637883373e-06,
+      "loss": 0.4824,
+      "step": 2839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.147081390118608,
+      "learning_rate": 8.99920264281627e-06,
+      "loss": 0.4313,
+      "step": 2840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.08661085023728,
+      "learning_rate": 8.99841937542115e-06,
+      "loss": 0.5969,
+      "step": 2841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.516019319447087,
+      "learning_rate": 8.99763583575135e-06,
+      "loss": 0.6087,
+      "step": 2842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.808639147964319,
+      "learning_rate": 8.996852023860222e-06,
+      "loss": 0.5544,
+      "step": 2843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.276847245749828,
+      "learning_rate": 8.996067939801146e-06,
+      "loss": 0.5082,
+      "step": 2844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.300631929691994,
+      "learning_rate": 8.99528358362751e-06,
+      "loss": 0.4687,
+      "step": 2845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.4050319429747775,
+      "learning_rate": 8.994498955392728e-06,
+      "loss": 0.4969,
+      "step": 2846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.884465833231607,
+      "learning_rate": 8.993714055150229e-06,
+      "loss": 0.5586,
+      "step": 2847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.393290806621557,
+      "learning_rate": 8.99292888295346e-06,
+      "loss": 0.5076,
+      "step": 2848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 21.28003292269745,
+      "learning_rate": 8.992143438855889e-06,
+      "loss": 0.4478,
+      "step": 2849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.10880250073296,
+      "learning_rate": 8.991357722911e-06,
+      "loss": 0.7262,
+      "step": 2850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.008104386158156,
+      "learning_rate": 8.990571735172298e-06,
+      "loss": 0.4837,
+      "step": 2851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.58505283177673,
+      "learning_rate": 8.989785475693306e-06,
+      "loss": 0.4811,
+      "step": 2852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.390346357094746,
+      "learning_rate": 8.988998944527562e-06,
+      "loss": 0.5875,
+      "step": 2853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.897142683509593,
+      "learning_rate": 8.988212141728628e-06,
+      "loss": 0.6061,
+      "step": 2854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.142819519650137,
+      "learning_rate": 8.987425067350081e-06,
+      "loss": 0.4147,
+      "step": 2855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8300719937280143,
+      "learning_rate": 8.986637721445518e-06,
+      "loss": 0.8582,
+      "step": 2856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.21655999419346,
+      "learning_rate": 8.985850104068551e-06,
+      "loss": 0.6735,
+      "step": 2857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8869012669405487,
+      "learning_rate": 8.985062215272818e-06,
+      "loss": 0.7473,
+      "step": 2858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.550097766365163,
+      "learning_rate": 8.984274055111965e-06,
+      "loss": 0.5028,
+      "step": 2859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8475788294711413,
+      "learning_rate": 8.983485623639668e-06,
+      "loss": 0.7216,
+      "step": 2860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.593909968604044,
+      "learning_rate": 8.98269692090961e-06,
+      "loss": 0.542,
+      "step": 2861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.486865750952003,
+      "learning_rate": 8.981907946975502e-06,
+      "loss": 0.521,
+      "step": 2862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.748992364962016,
+      "learning_rate": 8.981118701891069e-06,
+      "loss": 0.5383,
+      "step": 2863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.312810004931288,
+      "learning_rate": 8.980329185710053e-06,
+      "loss": 0.5423,
+      "step": 2864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.05148475392791,
+      "learning_rate": 8.97953939848622e-06,
+      "loss": 0.6042,
+      "step": 2865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.71128962351382,
+      "learning_rate": 8.97874934027335e-06,
+      "loss": 0.5229,
+      "step": 2866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.701156373084743,
+      "learning_rate": 8.977959011125241e-06,
+      "loss": 0.5103,
+      "step": 2867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.5189766336205155,
+      "learning_rate": 8.977168411095711e-06,
+      "loss": 0.5354,
+      "step": 2868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8406541379541812,
+      "learning_rate": 8.9763775402386e-06,
+      "loss": 0.7627,
+      "step": 2869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.486992641212783,
+      "learning_rate": 8.975586398607756e-06,
+      "loss": 0.4613,
+      "step": 2870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.520587363916132,
+      "learning_rate": 8.974794986257059e-06,
+      "loss": 0.4714,
+      "step": 2871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.5440306050624155,
+      "learning_rate": 8.974003303240395e-06,
+      "loss": 0.4714,
+      "step": 2872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.767804065572135,
+      "learning_rate": 8.97321134961168e-06,
+      "loss": 0.637,
+      "step": 2873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7827015744921348,
+      "learning_rate": 8.972419125424837e-06,
+      "loss": 0.8083,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.061796234132379,
+      "learning_rate": 8.971626630733816e-06,
+      "loss": 0.5873,
+      "step": 2875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.908107198339962,
+      "learning_rate": 8.970833865592583e-06,
+      "loss": 0.5997,
+      "step": 2876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8708629292691,
+      "learning_rate": 8.97004083005512e-06,
+      "loss": 0.7856,
+      "step": 2877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.87236426895913,
+      "learning_rate": 8.96924752417543e-06,
+      "loss": 0.5649,
+      "step": 2878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.23232790208286,
+      "learning_rate": 8.968453948007534e-06,
+      "loss": 0.6392,
+      "step": 2879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.308927331197944,
+      "learning_rate": 8.967660101605469e-06,
+      "loss": 0.5356,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.927713866205094,
+      "learning_rate": 8.966865985023294e-06,
+      "loss": 0.4861,
+      "step": 2881
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.337157088921224,
+      "learning_rate": 8.966071598315086e-06,
+      "loss": 0.4617,
+      "step": 2882
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.512630474180724,
+      "learning_rate": 8.965276941534937e-06,
+      "loss": 0.3244,
+      "step": 2883
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.208585167207364,
+      "learning_rate": 8.964482014736962e-06,
+      "loss": 0.6029,
+      "step": 2884
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.043077009887672,
+      "learning_rate": 8.963686817975287e-06,
+      "loss": 0.5038,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.559644019306596,
+      "learning_rate": 8.962891351304068e-06,
+      "loss": 0.4397,
+      "step": 2886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.1326256828622245,
+      "learning_rate": 8.96209561477747e-06,
+      "loss": 0.6156,
+      "step": 2887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 20.755734007873603,
+      "learning_rate": 8.961299608449679e-06,
+      "loss": 0.5481,
+      "step": 2888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.984755839289959,
+      "learning_rate": 8.960503332374898e-06,
+      "loss": 0.49,
+      "step": 2889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.91684403714639,
+      "learning_rate": 8.959706786607351e-06,
+      "loss": 0.4768,
+      "step": 2890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.221056195671551,
+      "learning_rate": 8.95890997120128e-06,
+      "loss": 0.594,
+      "step": 2891
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.648904703799865,
+      "learning_rate": 8.958112886210946e-06,
+      "loss": 0.5519,
+      "step": 2892
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.6481445897161,
+      "learning_rate": 8.957315531690622e-06,
+      "loss": 0.4964,
+      "step": 2893
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.268646673074581,
+      "learning_rate": 8.956517907694609e-06,
+      "loss": 0.4726,
+      "step": 2894
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.507227656091766,
+      "learning_rate": 8.95572001427722e-06,
+      "loss": 0.5743,
+      "step": 2895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.806766589265957,
+      "learning_rate": 8.954921851492788e-06,
+      "loss": 0.3888,
+      "step": 2896
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1013851116469406,
+      "learning_rate": 8.954123419395664e-06,
+      "loss": 0.8047,
+      "step": 2897
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.456812572632221,
+      "learning_rate": 8.95332471804022e-06,
+      "loss": 0.524,
+      "step": 2898
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.487202016983794,
+      "learning_rate": 8.952525747480841e-06,
+      "loss": 0.5615,
+      "step": 2899
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.65410750739627,
+      "learning_rate": 8.951726507771932e-06,
+      "loss": 0.5886,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.49827455353579,
+      "learning_rate": 8.950926998967924e-06,
+      "loss": 0.4942,
+      "step": 2901
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.675160793344542,
+      "learning_rate": 8.950127221123253e-06,
+      "loss": 0.5462,
+      "step": 2902
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.056304872126134,
+      "learning_rate": 8.949327174292385e-06,
+      "loss": 0.4716,
+      "step": 2903
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.852718210826236,
+      "learning_rate": 8.9485268585298e-06,
+      "loss": 0.4475,
+      "step": 2904
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.553650333159188,
+      "learning_rate": 8.947726273889991e-06,
+      "loss": 0.4133,
+      "step": 2905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.748867756686195,
+      "learning_rate": 8.946925420427479e-06,
+      "loss": 0.5689,
+      "step": 2906
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.707510668451992,
+      "learning_rate": 8.946124298196798e-06,
+      "loss": 0.5504,
+      "step": 2907
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.816165070689593,
+      "learning_rate": 8.945322907252498e-06,
+      "loss": 0.6133,
+      "step": 2908
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.265261913362786,
+      "learning_rate": 8.944521247649154e-06,
+      "loss": 0.4681,
+      "step": 2909
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.007159858392999,
+      "learning_rate": 8.943719319441353e-06,
+      "loss": 0.4674,
+      "step": 2910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.905486489813297,
+      "learning_rate": 8.942917122683703e-06,
+      "loss": 0.4749,
+      "step": 2911
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.52940008802827,
+      "learning_rate": 8.94211465743083e-06,
+      "loss": 0.4276,
+      "step": 2912
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.779138368798066,
+      "learning_rate": 8.94131192373738e-06,
+      "loss": 0.4824,
+      "step": 2913
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.870011080087158,
+      "learning_rate": 8.940508921658012e-06,
+      "loss": 0.5747,
+      "step": 2914
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.7856058920989515,
+      "learning_rate": 8.93970565124741e-06,
+      "loss": 0.553,
+      "step": 2915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.32186867819715,
+      "learning_rate": 8.938902112560273e-06,
+      "loss": 0.5149,
+      "step": 2916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.58495892364279,
+      "learning_rate": 8.938098305651318e-06,
+      "loss": 0.4755,
+      "step": 2917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.463917949249232,
+      "learning_rate": 8.93729423057528e-06,
+      "loss": 0.4802,
+      "step": 2918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.058673076323188,
+      "learning_rate": 8.936489887386915e-06,
+      "loss": 0.6121,
+      "step": 2919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.2486226742092965,
+      "learning_rate": 8.935685276140991e-06,
+      "loss": 0.5129,
+      "step": 2920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.726858170359161,
+      "learning_rate": 8.934880396892304e-06,
+      "loss": 0.5233,
+      "step": 2921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.035096631798945,
+      "learning_rate": 8.934075249695658e-06,
+      "loss": 0.421,
+      "step": 2922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.528312466013703,
+      "learning_rate": 8.933269834605883e-06,
+      "loss": 0.4173,
+      "step": 2923
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.645213977452396,
+      "learning_rate": 8.932464151677822e-06,
+      "loss": 0.4611,
+      "step": 2924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.155220309717565,
+      "learning_rate": 8.93165820096634e-06,
+      "loss": 0.5569,
+      "step": 2925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.079967570248401,
+      "learning_rate": 8.930851982526318e-06,
+      "loss": 0.472,
+      "step": 2926
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.455891679696161,
+      "learning_rate": 8.930045496412657e-06,
+      "loss": 0.4795,
+      "step": 2927
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.329621573645845,
+      "learning_rate": 8.929238742680273e-06,
+      "loss": 0.4917,
+      "step": 2928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.200181157084174,
+      "learning_rate": 8.928431721384102e-06,
+      "loss": 0.4735,
+      "step": 2929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.937663911883791,
+      "learning_rate": 8.927624432579102e-06,
+      "loss": 0.3283,
+      "step": 2930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.280969032579474,
+      "learning_rate": 8.926816876320246e-06,
+      "loss": 0.4051,
+      "step": 2931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.724176002144755,
+      "learning_rate": 8.92600905266252e-06,
+      "loss": 0.5837,
+      "step": 2932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.765614709273704,
+      "learning_rate": 8.925200961660937e-06,
+      "loss": 0.4545,
+      "step": 2933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 25.659534533315483,
+      "learning_rate": 8.924392603370524e-06,
+      "loss": 0.5552,
+      "step": 2934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.42107363586265,
+      "learning_rate": 8.923583977846326e-06,
+      "loss": 0.5032,
+      "step": 2935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.113813346333188,
+      "learning_rate": 8.922775085143408e-06,
+      "loss": 0.5463,
+      "step": 2936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.597434913556013,
+      "learning_rate": 8.921965925316852e-06,
+      "loss": 0.5231,
+      "step": 2937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.648755211233201,
+      "learning_rate": 8.921156498421755e-06,
+      "loss": 0.5924,
+      "step": 2938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.83992588190359,
+      "learning_rate": 8.920346804513238e-06,
+      "loss": 0.4324,
+      "step": 2939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.14676197229462,
+      "learning_rate": 8.919536843646437e-06,
+      "loss": 0.5835,
+      "step": 2940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.482283752197011,
+      "learning_rate": 8.918726615876507e-06,
+      "loss": 0.6215,
+      "step": 2941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.135620569125924,
+      "learning_rate": 8.917916121258621e-06,
+      "loss": 0.4347,
+      "step": 2942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.758010814134142,
+      "learning_rate": 8.91710535984797e-06,
+      "loss": 0.6543,
+      "step": 2943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.815101012223806,
+      "learning_rate": 8.916294331699766e-06,
+      "loss": 0.4591,
+      "step": 2944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.501777912090571,
+      "learning_rate": 8.91548303686923e-06,
+      "loss": 0.5951,
+      "step": 2945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.25543693329816,
+      "learning_rate": 8.914671475411614e-06,
+      "loss": 0.4236,
+      "step": 2946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.634394506342137,
+      "learning_rate": 8.913859647382176e-06,
+      "loss": 0.4632,
+      "step": 2947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.522951477000963,
+      "learning_rate": 8.913047552836205e-06,
+      "loss": 0.4299,
+      "step": 2948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.57549883311086,
+      "learning_rate": 8.912235191828995e-06,
+      "loss": 0.6029,
+      "step": 2949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.010494597931514,
+      "learning_rate": 8.911422564415865e-06,
+      "loss": 0.4905,
+      "step": 2950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.138675256263033,
+      "learning_rate": 8.910609670652153e-06,
+      "loss": 0.5307,
+      "step": 2951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.116363824164534,
+      "learning_rate": 8.909796510593215e-06,
+      "loss": 0.4331,
+      "step": 2952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.0384870510644975,
+      "learning_rate": 8.90898308429442e-06,
+      "loss": 0.5159,
+      "step": 2953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.429304883525253,
+      "learning_rate": 8.90816939181116e-06,
+      "loss": 0.6177,
+      "step": 2954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.225802635224781,
+      "learning_rate": 8.907355433198845e-06,
+      "loss": 0.514,
+      "step": 2955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.979877207255527,
+      "learning_rate": 8.906541208512902e-06,
+      "loss": 0.515,
+      "step": 2956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.574393777551384,
+      "learning_rate": 8.905726717808774e-06,
+      "loss": 0.5443,
+      "step": 2957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.96343538035591,
+      "learning_rate": 8.904911961141926e-06,
+      "loss": 0.481,
+      "step": 2958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.046703106728671,
+      "learning_rate": 8.90409693856784e-06,
+      "loss": 0.6155,
+      "step": 2959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.57743816737572,
+      "learning_rate": 8.903281650142012e-06,
+      "loss": 0.6278,
+      "step": 2960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.5816498639519585,
+      "learning_rate": 8.902466095919964e-06,
+      "loss": 0.4129,
+      "step": 2961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.3819901615315,
+      "learning_rate": 8.901650275957227e-06,
+      "loss": 0.4782,
+      "step": 2962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.161303377741943,
+      "learning_rate": 8.90083419030936e-06,
+      "loss": 0.6386,
+      "step": 2963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.5039579462732,
+      "learning_rate": 8.90001783903193e-06,
+      "loss": 0.5857,
+      "step": 2964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.769968850413322,
+      "learning_rate": 8.899201222180529e-06,
+      "loss": 0.4512,
+      "step": 2965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.351948397765586,
+      "learning_rate": 8.898384339810766e-06,
+      "loss": 0.433,
+      "step": 2966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.006921180982117,
+      "learning_rate": 8.897567191978265e-06,
+      "loss": 0.5502,
+      "step": 2967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.068337949785725,
+      "learning_rate": 8.896749778738671e-06,
+      "loss": 0.4793,
+      "step": 2968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.965824575636162,
+      "learning_rate": 8.895932100147648e-06,
+      "loss": 0.4419,
+      "step": 2969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.043620121381536,
+      "learning_rate": 8.895114156260874e-06,
+      "loss": 0.5734,
+      "step": 2970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.2719769385953335,
+      "learning_rate": 8.894295947134046e-06,
+      "loss": 0.4683,
+      "step": 2971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.99611407607436,
+      "learning_rate": 8.893477472822884e-06,
+      "loss": 0.4603,
+      "step": 2972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.79409742511492,
+      "learning_rate": 8.892658733383122e-06,
+      "loss": 0.4934,
+      "step": 2973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.999292770428283,
+      "learning_rate": 8.891839728870511e-06,
+      "loss": 0.4336,
+      "step": 2974
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.526466962205202,
+      "learning_rate": 8.89102045934082e-06,
+      "loss": 0.6588,
+      "step": 2975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.402811841018941,
+      "learning_rate": 8.890200924849842e-06,
+      "loss": 0.4589,
+      "step": 2976
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.804659436892478,
+      "learning_rate": 8.88938112545338e-06,
+      "loss": 0.4676,
+      "step": 2977
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.808522721054803,
+      "learning_rate": 8.888561061207261e-06,
+      "loss": 0.4942,
+      "step": 2978
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.646499315937168,
+      "learning_rate": 8.887740732167327e-06,
+      "loss": 0.5678,
+      "step": 2979
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.275216783419541,
+      "learning_rate": 8.886920138389438e-06,
+      "loss": 0.5227,
+      "step": 2980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.64382194646607,
+      "learning_rate": 8.886099279929475e-06,
+      "loss": 0.5648,
+      "step": 2981
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.688249413010286,
+      "learning_rate": 8.885278156843332e-06,
+      "loss": 0.4501,
+      "step": 2982
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.510204204467477,
+      "learning_rate": 8.884456769186924e-06,
+      "loss": 0.5665,
+      "step": 2983
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.163736157972321,
+      "learning_rate": 8.883635117016185e-06,
+      "loss": 0.478,
+      "step": 2984
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.868581771321491,
+      "learning_rate": 8.882813200387068e-06,
+      "loss": 0.5786,
+      "step": 2985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.553553392061765,
+      "learning_rate": 8.881991019355537e-06,
+      "loss": 0.5707,
+      "step": 2986
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.869348711677056,
+      "learning_rate": 8.881168573977584e-06,
+      "loss": 0.4482,
+      "step": 2987
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.541956065917342,
+      "learning_rate": 8.880345864309211e-06,
+      "loss": 0.6538,
+      "step": 2988
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.285528237080124,
+      "learning_rate": 8.87952289040644e-06,
+      "loss": 0.5813,
+      "step": 2989
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.993893583958915,
+      "learning_rate": 8.878699652325315e-06,
+      "loss": 0.4686,
+      "step": 2990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.49434492815772,
+      "learning_rate": 8.877876150121893e-06,
+      "loss": 0.5686,
+      "step": 2991
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.896543341069095,
+      "learning_rate": 8.877052383852252e-06,
+      "loss": 0.5346,
+      "step": 2992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.062462597384151,
+      "learning_rate": 8.876228353572486e-06,
+      "loss": 0.6245,
+      "step": 2993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9124793424449007,
+      "learning_rate": 8.875404059338704e-06,
+      "loss": 0.7993,
+      "step": 2994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.012019135435576,
+      "learning_rate": 8.874579501207044e-06,
+      "loss": 0.4418,
+      "step": 2995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.230924266409916,
+      "learning_rate": 8.873754679233647e-06,
+      "loss": 0.5475,
+      "step": 2996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.865521218961268,
+      "learning_rate": 8.872929593474689e-06,
+      "loss": 0.3735,
+      "step": 2997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.380648181929019,
+      "learning_rate": 8.872104243986348e-06,
+      "loss": 0.5902,
+      "step": 2998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.955831302575026,
+      "learning_rate": 8.871278630824827e-06,
+      "loss": 0.4683,
+      "step": 2999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2400635319983278,
+      "learning_rate": 8.87045275404635e-06,
+      "loss": 0.738,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.00234281248117,
+      "learning_rate": 8.869626613707151e-06,
+      "loss": 0.446,
+      "step": 3001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.930253645508625,
+      "learning_rate": 8.868800209863492e-06,
+      "loss": 0.5074,
+      "step": 3002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5589118470187824,
+      "learning_rate": 8.867973542571642e-06,
+      "loss": 0.8103,
+      "step": 3003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.766564090723355,
+      "learning_rate": 8.867146611887896e-06,
+      "loss": 0.6242,
+      "step": 3004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.94633876002113,
+      "learning_rate": 8.866319417868565e-06,
+      "loss": 0.5123,
+      "step": 3005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.797869359520703,
+      "learning_rate": 8.865491960569976e-06,
+      "loss": 0.5191,
+      "step": 3006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 22.913637506965664,
+      "learning_rate": 8.864664240048476e-06,
+      "loss": 0.4432,
+      "step": 3007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.170953658332873,
+      "learning_rate": 8.86383625636043e-06,
+      "loss": 0.6969,
+      "step": 3008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.136392401641178,
+      "learning_rate": 8.863008009562218e-06,
+      "loss": 0.4946,
+      "step": 3009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.270872966381769,
+      "learning_rate": 8.862179499710239e-06,
+      "loss": 0.8153,
+      "step": 3010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.704955523664846,
+      "learning_rate": 8.861350726860913e-06,
+      "loss": 0.4649,
+      "step": 3011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.225792238024328,
+      "learning_rate": 8.860521691070677e-06,
+      "loss": 0.3519,
+      "step": 3012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.329750247273381,
+      "learning_rate": 8.85969239239598e-06,
+      "loss": 0.5114,
+      "step": 3013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7591084301245858,
+      "learning_rate": 8.8588628308933e-06,
+      "loss": 0.7631,
+      "step": 3014
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.809282371751308,
+      "learning_rate": 8.85803300661912e-06,
+      "loss": 0.5085,
+      "step": 3015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.977010683084114,
+      "learning_rate": 8.857202919629952e-06,
+      "loss": 0.4815,
+      "step": 3016
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.82788080623641,
+      "learning_rate": 8.856372569982318e-06,
+      "loss": 0.4395,
+      "step": 3017
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.7527909004139195,
+      "learning_rate": 8.855541957732765e-06,
+      "loss": 0.481,
+      "step": 3018
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.095661283147822,
+      "learning_rate": 8.854711082937851e-06,
+      "loss": 0.4274,
+      "step": 3019
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.35935750382468,
+      "learning_rate": 8.853879945654156e-06,
+      "loss": 0.6179,
+      "step": 3020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.577037702879169,
+      "learning_rate": 8.853048545938277e-06,
+      "loss": 0.5213,
+      "step": 3021
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.614185629850594,
+      "learning_rate": 8.852216883846825e-06,
+      "loss": 0.558,
+      "step": 3022
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.601299973443291,
+      "learning_rate": 8.85138495943644e-06,
+      "loss": 0.4823,
+      "step": 3023
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.659876074574062,
+      "learning_rate": 8.850552772763765e-06,
+      "loss": 0.5315,
+      "step": 3024
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.986285152286158,
+      "learning_rate": 8.84972032388547e-06,
+      "loss": 0.5221,
+      "step": 3025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.183286840994839,
+      "learning_rate": 8.848887612858246e-06,
+      "loss": 0.5079,
+      "step": 3026
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.788115759132308,
+      "learning_rate": 8.848054639738791e-06,
+      "loss": 0.5227,
+      "step": 3027
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.04853835347385,
+      "learning_rate": 8.84722140458383e-06,
+      "loss": 0.6269,
+      "step": 3028
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.841249399791439,
+      "learning_rate": 8.846387907450101e-06,
+      "loss": 0.4702,
+      "step": 3029
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2544268481097145,
+      "learning_rate": 8.845554148394363e-06,
+      "loss": 0.7942,
+      "step": 3030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.777109936704585,
+      "learning_rate": 8.84472012747339e-06,
+      "loss": 0.4833,
+      "step": 3031
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.038213269512797,
+      "learning_rate": 8.843885844743975e-06,
+      "loss": 0.4824,
+      "step": 3032
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.865815109072118,
+      "learning_rate": 8.84305130026293e-06,
+      "loss": 0.5428,
+      "step": 3033
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.4920733341724794,
+      "learning_rate": 8.842216494087083e-06,
+      "loss": 0.4733,
+      "step": 3034
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.7209558993461895,
+      "learning_rate": 8.84138142627328e-06,
+      "loss": 0.4818,
+      "step": 3035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.333114987265285,
+      "learning_rate": 8.840546096878389e-06,
+      "loss": 0.4935,
+      "step": 3036
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.105148310813636,
+      "learning_rate": 8.83971050595929e-06,
+      "loss": 0.5934,
+      "step": 3037
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.0287662613371165,
+      "learning_rate": 8.83887465357288e-06,
+      "loss": 0.5465,
+      "step": 3038
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.224265356489045,
+      "learning_rate": 8.838038539776079e-06,
+      "loss": 0.4399,
+      "step": 3039
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.02322150426316,
+      "learning_rate": 8.837202164625825e-06,
+      "loss": 0.3814,
+      "step": 3040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.471416551479726,
+      "learning_rate": 8.83636552817907e-06,
+      "loss": 0.4514,
+      "step": 3041
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.665846705952398,
+      "learning_rate": 8.835528630492785e-06,
+      "loss": 0.5621,
+      "step": 3042
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.564460942089369,
+      "learning_rate": 8.834691471623958e-06,
+      "loss": 0.5866,
+      "step": 3043
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.855721795851184,
+      "learning_rate": 8.833854051629595e-06,
+      "loss": 0.524,
+      "step": 3044
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.270380394245796,
+      "learning_rate": 8.833016370566725e-06,
+      "loss": 0.5612,
+      "step": 3045
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.496483191167532,
+      "learning_rate": 8.832178428492385e-06,
+      "loss": 0.5516,
+      "step": 3046
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.792292624459204,
+      "learning_rate": 8.83134022546364e-06,
+      "loss": 0.5574,
+      "step": 3047
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.323740555750685,
+      "learning_rate": 8.830501761537567e-06,
+      "loss": 0.4544,
+      "step": 3048
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.696596547892314,
+      "learning_rate": 8.829663036771257e-06,
+      "loss": 0.5678,
+      "step": 3049
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.9683979055847995,
+      "learning_rate": 8.828824051221829e-06,
+      "loss": 0.4942,
+      "step": 3050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.316599452565282,
+      "learning_rate": 8.827984804946411e-06,
+      "loss": 0.5783,
+      "step": 3051
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.933549020506499,
+      "learning_rate": 8.827145298002153e-06,
+      "loss": 0.5292,
+      "step": 3052
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.6403861533869915,
+      "learning_rate": 8.826305530446224e-06,
+      "loss": 0.5063,
+      "step": 3053
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.818385528389851,
+      "learning_rate": 8.825465502335805e-06,
+      "loss": 0.6228,
+      "step": 3054
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.711785767537801,
+      "learning_rate": 8.824625213728099e-06,
+      "loss": 0.5541,
+      "step": 3055
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.467433368483134,
+      "learning_rate": 8.823784664680326e-06,
+      "loss": 0.5855,
+      "step": 3056
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0546551520423924,
+      "learning_rate": 8.822943855249723e-06,
+      "loss": 0.7696,
+      "step": 3057
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.909057511479784,
+      "learning_rate": 8.82210278549355e-06,
+      "loss": 0.4411,
+      "step": 3058
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.90703413166901,
+      "learning_rate": 8.821261455469075e-06,
+      "loss": 0.5207,
+      "step": 3059
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.751359248919362,
+      "learning_rate": 8.82041986523359e-06,
+      "loss": 0.4337,
+      "step": 3060
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.045861133951616,
+      "learning_rate": 8.819578014844404e-06,
+      "loss": 0.5415,
+      "step": 3061
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.550037269976995,
+      "learning_rate": 8.818735904358841e-06,
+      "loss": 0.5266,
+      "step": 3062
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.854178756516456,
+      "learning_rate": 8.817893533834251e-06,
+      "loss": 0.4652,
+      "step": 3063
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.094520432683973,
+      "learning_rate": 8.81705090332799e-06,
+      "loss": 0.5331,
+      "step": 3064
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.615562259125296,
+      "learning_rate": 8.816208012897441e-06,
+      "loss": 0.6545,
+      "step": 3065
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 39.84843241333077,
+      "learning_rate": 8.815364862599998e-06,
+      "loss": 0.5403,
+      "step": 3066
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.397474536672723,
+      "learning_rate": 8.814521452493077e-06,
+      "loss": 0.7747,
+      "step": 3067
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.148614084883523,
+      "learning_rate": 8.813677782634112e-06,
+      "loss": 0.498,
+      "step": 3068
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.228329420867545,
+      "learning_rate": 8.81283385308055e-06,
+      "loss": 0.4897,
+      "step": 3069
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.141205709860436,
+      "learning_rate": 8.811989663889863e-06,
+      "loss": 0.5228,
+      "step": 3070
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.121184866534787,
+      "learning_rate": 8.811145215119534e-06,
+      "loss": 0.5241,
+      "step": 3071
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.06037056778049,
+      "learning_rate": 8.810300506827068e-06,
+      "loss": 0.4884,
+      "step": 3072
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.195948478121801,
+      "learning_rate": 8.809455539069982e-06,
+      "loss": 0.4126,
+      "step": 3073
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.711417078325077,
+      "learning_rate": 8.808610311905818e-06,
+      "loss": 0.4278,
+      "step": 3074
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.917283548321658,
+      "learning_rate": 8.807764825392133e-06,
+      "loss": 0.5255,
+      "step": 3075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.191807063788515,
+      "learning_rate": 8.806919079586498e-06,
+      "loss": 0.5374,
+      "step": 3076
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.76820702419045,
+      "learning_rate": 8.806073074546505e-06,
+      "loss": 0.5153,
+      "step": 3077
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.7818816836834275,
+      "learning_rate": 8.805226810329766e-06,
+      "loss": 0.5149,
+      "step": 3078
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.964661334763512,
+      "learning_rate": 8.804380286993904e-06,
+      "loss": 0.5034,
+      "step": 3079
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.894307476663724,
+      "learning_rate": 8.803533504596566e-06,
+      "loss": 0.6253,
+      "step": 3080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.010030106131815,
+      "learning_rate": 8.802686463195413e-06,
+      "loss": 0.5663,
+      "step": 3081
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.932346081647962,
+      "learning_rate": 8.801839162848126e-06,
+      "loss": 0.6142,
+      "step": 3082
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.687512642274916,
+      "learning_rate": 8.800991603612402e-06,
+      "loss": 0.6213,
+      "step": 3083
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.017871675750355,
+      "learning_rate": 8.800143785545953e-06,
+      "loss": 0.4505,
+      "step": 3084
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.392583846257396,
+      "learning_rate": 8.799295708706518e-06,
+      "loss": 0.588,
+      "step": 3085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.608528347731047,
+      "learning_rate": 8.798447373151841e-06,
+      "loss": 0.5321,
+      "step": 3086
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.815511383474546,
+      "learning_rate": 8.797598778939696e-06,
+      "loss": 0.5463,
+      "step": 3087
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.089984175243977,
+      "learning_rate": 8.796749926127861e-06,
+      "loss": 0.5607,
+      "step": 3088
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.155567407849853,
+      "learning_rate": 8.795900814774145e-06,
+      "loss": 0.4721,
+      "step": 3089
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.965927170857018,
+      "learning_rate": 8.795051444936367e-06,
+      "loss": 0.3918,
+      "step": 3090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.703520135964868,
+      "learning_rate": 8.794201816672365e-06,
+      "loss": 0.5803,
+      "step": 3091
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.163542518087752,
+      "learning_rate": 8.793351930039996e-06,
+      "loss": 0.4917,
+      "step": 3092
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.492348660544584,
+      "learning_rate": 8.792501785097131e-06,
+      "loss": 0.5055,
+      "step": 3093
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.32036274486464,
+      "learning_rate": 8.791651381901665e-06,
+      "loss": 0.4655,
+      "step": 3094
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.869623570768005,
+      "learning_rate": 8.790800720511503e-06,
+      "loss": 0.3866,
+      "step": 3095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.035732670662835,
+      "learning_rate": 8.789949800984575e-06,
+      "loss": 0.5996,
+      "step": 3096
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.860889545495081,
+      "learning_rate": 8.78909862337882e-06,
+      "loss": 0.8553,
+      "step": 3097
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.257446768697058,
+      "learning_rate": 8.788247187752202e-06,
+      "loss": 0.5059,
+      "step": 3098
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.3114125295980426,
+      "learning_rate": 8.787395494162702e-06,
+      "loss": 0.523,
+      "step": 3099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.367871673628775,
+      "learning_rate": 8.786543542668313e-06,
+      "loss": 0.5407,
+      "step": 3100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.046263601248345,
+      "learning_rate": 8.785691333327052e-06,
+      "loss": 0.6541,
+      "step": 3101
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.908015352643416,
+      "learning_rate": 8.784838866196949e-06,
+      "loss": 0.5668,
+      "step": 3102
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.82239587227461,
+      "learning_rate": 8.783986141336054e-06,
+      "loss": 0.4457,
+      "step": 3103
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.244352695517485,
+      "learning_rate": 8.783133158802432e-06,
+      "loss": 0.5879,
+      "step": 3104
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.994017624431569,
+      "learning_rate": 8.78227991865417e-06,
+      "loss": 0.4379,
+      "step": 3105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.280895710490437,
+      "learning_rate": 8.781426420949369e-06,
+      "loss": 0.5068,
+      "step": 3106
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.842357156994781,
+      "learning_rate": 8.780572665746148e-06,
+      "loss": 0.5265,
+      "step": 3107
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.925975436556227,
+      "learning_rate": 8.779718653102644e-06,
+      "loss": 0.4867,
+      "step": 3108
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.634385667608447,
+      "learning_rate": 8.778864383077011e-06,
+      "loss": 0.5677,
+      "step": 3109
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.975375550389698,
+      "learning_rate": 8.778009855727423e-06,
+      "loss": 0.5445,
+      "step": 3110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.418564685077493,
+      "learning_rate": 8.777155071112066e-06,
+      "loss": 0.4896,
+      "step": 3111
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.820456065989358,
+      "learning_rate": 8.77630002928915e-06,
+      "loss": 0.4279,
+      "step": 3112
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.969974553998659,
+      "learning_rate": 8.7754447303169e-06,
+      "loss": 0.5438,
+      "step": 3113
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.59517510527474,
+      "learning_rate": 8.774589174253556e-06,
+      "loss": 0.48,
+      "step": 3114
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.020181660452273,
+      "learning_rate": 8.77373336115738e-06,
+      "loss": 0.6478,
+      "step": 3115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.868879729667111,
+      "learning_rate": 8.772877291086646e-06,
+      "loss": 0.4602,
+      "step": 3116
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.877072069465568,
+      "learning_rate": 8.77202096409965e-06,
+      "loss": 0.6129,
+      "step": 3117
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.485605507605042,
+      "learning_rate": 8.771164380254704e-06,
+      "loss": 0.5349,
+      "step": 3118
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.612175006472354,
+      "learning_rate": 8.770307539610137e-06,
+      "loss": 0.4411,
+      "step": 3119
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.451805428303379,
+      "learning_rate": 8.769450442224299e-06,
+      "loss": 0.5614,
+      "step": 3120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.512383704257717,
+      "learning_rate": 8.76859308815555e-06,
+      "loss": 0.5593,
+      "step": 3121
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.463197868630576,
+      "learning_rate": 8.767735477462276e-06,
+      "loss": 0.464,
+      "step": 3122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.663519143267324,
+      "learning_rate": 8.766877610202875e-06,
+      "loss": 0.6009,
+      "step": 3123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.821700269383237,
+      "learning_rate": 8.766019486435762e-06,
+      "loss": 0.4707,
+      "step": 3124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 25.306864700199128,
+      "learning_rate": 8.765161106219375e-06,
+      "loss": 0.4921,
+      "step": 3125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.435549438595912,
+      "learning_rate": 8.764302469612165e-06,
+      "loss": 0.4402,
+      "step": 3126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.401940916742324,
+      "learning_rate": 8.7634435766726e-06,
+      "loss": 0.5104,
+      "step": 3127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.345828152067807,
+      "learning_rate": 8.762584427459164e-06,
+      "loss": 0.5148,
+      "step": 3128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.300344353102462,
+      "learning_rate": 8.761725022030368e-06,
+      "loss": 0.4627,
+      "step": 3129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.595832489672834,
+      "learning_rate": 8.760865360444729e-06,
+      "loss": 0.4791,
+      "step": 3130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.882507943588694,
+      "learning_rate": 8.760005442760789e-06,
+      "loss": 0.6127,
+      "step": 3131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.950719700648115,
+      "learning_rate": 8.759145269037102e-06,
+      "loss": 0.5408,
+      "step": 3132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.553761727753334,
+      "learning_rate": 8.758284839332242e-06,
+      "loss": 0.5322,
+      "step": 3133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.869421219344318,
+      "learning_rate": 8.757424153704803e-06,
+      "loss": 0.6449,
+      "step": 3134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.618804405651809,
+      "learning_rate": 8.756563212213392e-06,
+      "loss": 0.5098,
+      "step": 3135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.663574044391224,
+      "learning_rate": 8.755702014916637e-06,
+      "loss": 0.5612,
+      "step": 3136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.026896026533697,
+      "learning_rate": 8.754840561873177e-06,
+      "loss": 0.4334,
+      "step": 3137
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.791666348936777,
+      "learning_rate": 8.753978853141679e-06,
+      "loss": 0.4855,
+      "step": 3138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.645713403460909,
+      "learning_rate": 8.753116888780819e-06,
+      "loss": 0.4335,
+      "step": 3139
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.38537242697612,
+      "learning_rate": 8.752254668849294e-06,
+      "loss": 0.446,
+      "step": 3140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.102583585078727,
+      "learning_rate": 8.751392193405815e-06,
+      "loss": 0.6543,
+      "step": 3141
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.962088422746222,
+      "learning_rate": 8.750529462509115e-06,
+      "loss": 0.4905,
+      "step": 3142
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.772772743496453,
+      "learning_rate": 8.749666476217941e-06,
+      "loss": 0.5927,
+      "step": 3143
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.082503875629026,
+      "learning_rate": 8.748803234591058e-06,
+      "loss": 0.5283,
+      "step": 3144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.118913690831027,
+      "learning_rate": 8.747939737687251e-06,
+      "loss": 0.605,
+      "step": 3145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.422754520592527,
+      "learning_rate": 8.74707598556532e-06,
+      "loss": 0.5142,
+      "step": 3146
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.862191030545244,
+      "learning_rate": 8.74621197828408e-06,
+      "loss": 0.5339,
+      "step": 3147
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.575547197641861,
+      "learning_rate": 8.74534771590237e-06,
+      "loss": 0.4307,
+      "step": 3148
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.168372633797603,
+      "learning_rate": 8.74448319847904e-06,
+      "loss": 0.6179,
+      "step": 3149
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.990426907121055,
+      "learning_rate": 8.74361842607296e-06,
+      "loss": 0.4784,
+      "step": 3150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.0579501011174095,
+      "learning_rate": 8.742753398743018e-06,
+      "loss": 0.5859,
+      "step": 3151
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.864819348438543,
+      "learning_rate": 8.741888116548116e-06,
+      "loss": 0.6041,
+      "step": 3152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.811340885302906,
+      "learning_rate": 8.74102257954718e-06,
+      "loss": 0.5503,
+      "step": 3153
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.46851506464409,
+      "learning_rate": 8.740156787799145e-06,
+      "loss": 0.4585,
+      "step": 3154
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.776290708901364,
+      "learning_rate": 8.739290741362972e-06,
+      "loss": 0.5084,
+      "step": 3155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.262899779601605,
+      "learning_rate": 8.738424440297631e-06,
+      "loss": 0.4141,
+      "step": 3156
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.139471994906992,
+      "learning_rate": 8.737557884662116e-06,
+      "loss": 0.4661,
+      "step": 3157
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 17.905533217426942,
+      "learning_rate": 8.736691074515432e-06,
+      "loss": 0.5898,
+      "step": 3158
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.825522563618698,
+      "learning_rate": 8.735824009916612e-06,
+      "loss": 0.4746,
+      "step": 3159
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.519471760220561,
+      "learning_rate": 8.73495669092469e-06,
+      "loss": 0.4887,
+      "step": 3160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.405540326520006,
+      "learning_rate": 8.734089117598734e-06,
+      "loss": 0.4788,
+      "step": 3161
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.915151910980554,
+      "learning_rate": 8.733221289997818e-06,
+      "loss": 0.5117,
+      "step": 3162
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 13.49569686462085,
+      "learning_rate": 8.73235320818104e-06,
+      "loss": 0.4875,
+      "step": 3163
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 12.614222168247139,
+      "learning_rate": 8.731484872207509e-06,
+      "loss": 0.5106,
+      "step": 3164
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.205637723917499,
+      "learning_rate": 8.730616282136359e-06,
+      "loss": 0.6394,
+      "step": 3165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.175197837077288,
+      "learning_rate": 8.729747438026735e-06,
+      "loss": 0.572,
+      "step": 3166
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.659601598517307,
+      "learning_rate": 8.728878339937798e-06,
+      "loss": 0.3847,
+      "step": 3167
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.520533446276183,
+      "learning_rate": 8.728008987928735e-06,
+      "loss": 0.5013,
+      "step": 3168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.130811136854678,
+      "learning_rate": 8.727139382058744e-06,
+      "loss": 0.6197,
+      "step": 3169
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.566263359819261,
+      "learning_rate": 8.726269522387042e-06,
+      "loss": 0.5063,
+      "step": 3170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.030064097418066,
+      "learning_rate": 8.725399408972857e-06,
+      "loss": 0.4237,
+      "step": 3171
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.792622373894809,
+      "learning_rate": 8.724529041875446e-06,
+      "loss": 0.5278,
+      "step": 3172
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.538674127729665,
+      "learning_rate": 8.723658421154076e-06,
+      "loss": 0.5705,
+      "step": 3173
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 17.52993316605919,
+      "learning_rate": 8.72278754686803e-06,
+      "loss": 0.4568,
+      "step": 3174
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.320756550895997,
+      "learning_rate": 8.721916419076612e-06,
+      "loss": 0.4543,
+      "step": 3175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6041401591378286,
+      "learning_rate": 8.721045037839143e-06,
+      "loss": 0.7474,
+      "step": 3176
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.5560798701544805,
+      "learning_rate": 8.720173403214958e-06,
+      "loss": 0.5052,
+      "step": 3177
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3095415035512445,
+      "learning_rate": 8.719301515263414e-06,
+      "loss": 0.833,
+      "step": 3178
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.44910436110739,
+      "learning_rate": 8.718429374043882e-06,
+      "loss": 0.5411,
+      "step": 3179
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.854274753556016,
+      "learning_rate": 8.71755697961575e-06,
+      "loss": 0.6299,
+      "step": 3180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.688728101468122,
+      "learning_rate": 8.716684332038425e-06,
+      "loss": 0.5721,
+      "step": 3181
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.503799136069436,
+      "learning_rate": 8.715811431371331e-06,
+      "loss": 0.5782,
+      "step": 3182
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.683123317478004,
+      "learning_rate": 8.714938277673906e-06,
+      "loss": 0.5055,
+      "step": 3183
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.002405756852948,
+      "learning_rate": 8.714064871005611e-06,
+      "loss": 0.4834,
+      "step": 3184
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.735982912342488,
+      "learning_rate": 8.71319121142592e-06,
+      "loss": 0.4421,
+      "step": 3185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.485071869137226,
+      "learning_rate": 8.712317298994324e-06,
+      "loss": 0.5867,
+      "step": 3186
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.2431901167269395,
+      "learning_rate": 8.711443133770335e-06,
+      "loss": 0.4408,
+      "step": 3187
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.098544337695139,
+      "learning_rate": 8.710568715813476e-06,
+      "loss": 0.5214,
+      "step": 3188
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.609826706659392,
+      "learning_rate": 8.709694045183296e-06,
+      "loss": 0.4494,
+      "step": 3189
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.008324989428806,
+      "learning_rate": 8.708819121939353e-06,
+      "loss": 0.4919,
+      "step": 3190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.731190046596526,
+      "learning_rate": 8.707943946141228e-06,
+      "loss": 0.5028,
+      "step": 3191
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.085821132526478,
+      "learning_rate": 8.707068517848513e-06,
+      "loss": 0.5714,
+      "step": 3192
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.336131103775909,
+      "learning_rate": 8.706192837120822e-06,
+      "loss": 0.4644,
+      "step": 3193
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.931131144285032,
+      "learning_rate": 8.705316904017787e-06,
+      "loss": 0.664,
+      "step": 3194
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.135523763276625,
+      "learning_rate": 8.704440718599053e-06,
+      "loss": 0.5362,
+      "step": 3195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.127666283519735,
+      "learning_rate": 8.703564280924287e-06,
+      "loss": 0.5024,
+      "step": 3196
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.462862231581521,
+      "learning_rate": 8.702687591053165e-06,
+      "loss": 0.4878,
+      "step": 3197
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.21762146227931,
+      "learning_rate": 8.701810649045392e-06,
+      "loss": 0.5176,
+      "step": 3198
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.32899996387958,
+      "learning_rate": 8.70093345496068e-06,
+      "loss": 0.5772,
+      "step": 3199
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.896244702980314,
+      "learning_rate": 8.700056008858763e-06,
+      "loss": 0.5838,
+      "step": 3200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.39133491455738,
+      "learning_rate": 8.699178310799392e-06,
+      "loss": 0.4654,
+      "step": 3201
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.7191801039892,
+      "learning_rate": 8.698300360842333e-06,
+      "loss": 0.529,
+      "step": 3202
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6109046350657725,
+      "learning_rate": 8.697422159047372e-06,
+      "loss": 0.8562,
+      "step": 3203
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.70850596275779,
+      "learning_rate": 8.696543705474312e-06,
+      "loss": 0.5599,
+      "step": 3204
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.128208574455213,
+      "learning_rate": 8.695665000182967e-06,
+      "loss": 0.4652,
+      "step": 3205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.647999619820688,
+      "learning_rate": 8.694786043233177e-06,
+      "loss": 0.5486,
+      "step": 3206
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.00803246499216,
+      "learning_rate": 8.693906834684794e-06,
+      "loss": 0.5887,
+      "step": 3207
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.414719094445699,
+      "learning_rate": 8.693027374597687e-06,
+      "loss": 0.4807,
+      "step": 3208
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8528763179731733,
+      "learning_rate": 8.692147663031744e-06,
+      "loss": 0.7676,
+      "step": 3209
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.082326417731869,
+      "learning_rate": 8.69126770004687e-06,
+      "loss": 0.5338,
+      "step": 3210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.13298073805937,
+      "learning_rate": 8.690387485702987e-06,
+      "loss": 0.5305,
+      "step": 3211
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.024137684817608,
+      "learning_rate": 8.689507020060033e-06,
+      "loss": 0.4995,
+      "step": 3212
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.4201415666567,
+      "learning_rate": 8.688626303177965e-06,
+      "loss": 0.4642,
+      "step": 3213
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.865431356236683,
+      "learning_rate": 8.687745335116752e-06,
+      "loss": 0.4099,
+      "step": 3214
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 20.676683289149494,
+      "learning_rate": 8.68686411593639e-06,
+      "loss": 0.5361,
+      "step": 3215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 15.917836838953832,
+      "learning_rate": 8.685982645696882e-06,
+      "loss": 0.4616,
+      "step": 3216
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.728047295908786,
+      "learning_rate": 8.685100924458251e-06,
+      "loss": 0.5118,
+      "step": 3217
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.237704491347566,
+      "learning_rate": 8.684218952280545e-06,
+      "loss": 0.489,
+      "step": 3218
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.523633827077221,
+      "learning_rate": 8.683336729223814e-06,
+      "loss": 0.4404,
+      "step": 3219
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.234199158030292,
+      "learning_rate": 8.68245425534814e-06,
+      "loss": 0.4404,
+      "step": 3220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.3248271424742875,
+      "learning_rate": 8.681571530713613e-06,
+      "loss": 0.4844,
+      "step": 3221
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 18.345140232998062,
+      "learning_rate": 8.68068855538034e-06,
+      "loss": 0.5511,
+      "step": 3222
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.544745621697084,
+      "learning_rate": 8.679805329408453e-06,
+      "loss": 0.4082,
+      "step": 3223
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.91215207137784,
+      "learning_rate": 8.678921852858091e-06,
+      "loss": 0.5369,
+      "step": 3224
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.069709437821796,
+      "learning_rate": 8.678038125789418e-06,
+      "loss": 0.5034,
+      "step": 3225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.402302674434615,
+      "learning_rate": 8.677154148262612e-06,
+      "loss": 0.4694,
+      "step": 3226
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.1729659059041015,
+      "learning_rate": 8.676269920337866e-06,
+      "loss": 0.5222,
+      "step": 3227
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.786665176378358,
+      "learning_rate": 8.675385442075393e-06,
+      "loss": 0.4666,
+      "step": 3228
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.603045365973014,
+      "learning_rate": 8.674500713535421e-06,
+      "loss": 0.4831,
+      "step": 3229
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.790409032677996,
+      "learning_rate": 8.673615734778199e-06,
+      "loss": 0.4547,
+      "step": 3230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.669933751412064,
+      "learning_rate": 8.672730505863986e-06,
+      "loss": 0.4594,
+      "step": 3231
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.285012447465219,
+      "learning_rate": 8.671845026853066e-06,
+      "loss": 0.5793,
+      "step": 3232
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.557586446808952,
+      "learning_rate": 8.670959297805734e-06,
+      "loss": 0.4504,
+      "step": 3233
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.103100503236476,
+      "learning_rate": 8.670073318782306e-06,
+      "loss": 0.5729,
+      "step": 3234
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.752914120995557,
+      "learning_rate": 8.66918708984311e-06,
+      "loss": 0.6348,
+      "step": 3235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.987756915287799,
+      "learning_rate": 8.668300611048497e-06,
+      "loss": 0.5196,
+      "step": 3236
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.212210545395039,
+      "learning_rate": 8.667413882458833e-06,
+      "loss": 0.5463,
+      "step": 3237
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.164571218737885,
+      "learning_rate": 8.666526904134498e-06,
+      "loss": 0.5751,
+      "step": 3238
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.58434742767651,
+      "learning_rate": 8.66563967613589e-06,
+      "loss": 0.4337,
+      "step": 3239
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.36869453810587,
+      "learning_rate": 8.664752198523433e-06,
+      "loss": 0.4583,
+      "step": 3240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.2392996315688585,
+      "learning_rate": 8.663864471357548e-06,
+      "loss": 0.5191,
+      "step": 3241
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.739225933587631,
+      "learning_rate": 8.662976494698696e-06,
+      "loss": 0.5824,
+      "step": 3242
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.441565030189706,
+      "learning_rate": 8.662088268607341e-06,
+      "loss": 0.5413,
+      "step": 3243
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.267952532390524,
+      "learning_rate": 8.661199793143966e-06,
+      "loss": 0.4899,
+      "step": 3244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.246757836206019,
+      "learning_rate": 8.660311068369072e-06,
+      "loss": 0.4786,
+      "step": 3245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.035742767844782,
+      "learning_rate": 8.659422094343179e-06,
+      "loss": 0.4522,
+      "step": 3246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.230386267728627,
+      "learning_rate": 8.65853287112682e-06,
+      "loss": 0.5158,
+      "step": 3247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.683596294317697,
+      "learning_rate": 8.65764339878055e-06,
+      "loss": 0.452,
+      "step": 3248
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.669129870061039,
+      "learning_rate": 8.656753677364936e-06,
+      "loss": 0.5274,
+      "step": 3249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.518332278469669,
+      "learning_rate": 8.655863706940566e-06,
+      "loss": 0.5124,
+      "step": 3250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.71005778061207,
+      "learning_rate": 8.65497348756804e-06,
+      "loss": 0.4587,
+      "step": 3251
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.360207583615762,
+      "learning_rate": 8.654083019307981e-06,
+      "loss": 0.4581,
+      "step": 3252
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.555632422198725,
+      "learning_rate": 8.653192302221024e-06,
+      "loss": 0.4704,
+      "step": 3253
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.088781939384636,
+      "learning_rate": 8.652301336367823e-06,
+      "loss": 0.5063,
+      "step": 3254
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.338340068393466,
+      "learning_rate": 8.651410121809052e-06,
+      "loss": 0.5199,
+      "step": 3255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.8557380575195,
+      "learning_rate": 8.650518658605393e-06,
+      "loss": 0.5585,
+      "step": 3256
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.403938662261037,
+      "learning_rate": 8.649626946817556e-06,
+      "loss": 0.4349,
+      "step": 3257
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.329004195630482,
+      "learning_rate": 8.64873498650626e-06,
+      "loss": 0.5091,
+      "step": 3258
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.483761390577433,
+      "learning_rate": 8.647842777732246e-06,
+      "loss": 0.4861,
+      "step": 3259
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.565568678965891,
+      "learning_rate": 8.646950320556265e-06,
+      "loss": 0.4435,
+      "step": 3260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.107848547312912,
+      "learning_rate": 8.646057615039094e-06,
+      "loss": 0.4896,
+      "step": 3261
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.962236059455415,
+      "learning_rate": 8.645164661241518e-06,
+      "loss": 0.5888,
+      "step": 3262
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9539175406519067,
+      "learning_rate": 8.644271459224347e-06,
+      "loss": 0.756,
+      "step": 3263
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.841623612122568,
+      "learning_rate": 8.643378009048401e-06,
+      "loss": 0.6146,
+      "step": 3264
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.923859518149449,
+      "learning_rate": 8.642484310774525e-06,
+      "loss": 0.4757,
+      "step": 3265
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.29373530492515,
+      "learning_rate": 8.64159036446357e-06,
+      "loss": 0.5352,
+      "step": 3266
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9134594376962242,
+      "learning_rate": 8.640696170176413e-06,
+      "loss": 0.8411,
+      "step": 3267
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 36.53430596330527,
+      "learning_rate": 8.639801727973945e-06,
+      "loss": 0.5667,
+      "step": 3268
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.119949268818998,
+      "learning_rate": 8.638907037917071e-06,
+      "loss": 0.6057,
+      "step": 3269
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.47571304143162,
+      "learning_rate": 8.638012100066718e-06,
+      "loss": 0.5722,
+      "step": 3270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.65227055177224,
+      "learning_rate": 8.637116914483825e-06,
+      "loss": 0.5697,
+      "step": 3271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.524746336105169,
+      "learning_rate": 8.636221481229353e-06,
+      "loss": 0.4098,
+      "step": 3272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9866157987286879,
+      "learning_rate": 8.635325800364273e-06,
+      "loss": 0.8054,
+      "step": 3273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.909871129517537,
+      "learning_rate": 8.634429871949583e-06,
+      "loss": 0.5717,
+      "step": 3274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.698100380387049,
+      "learning_rate": 8.633533696046285e-06,
+      "loss": 0.6568,
+      "step": 3275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.54322504607567,
+      "learning_rate": 8.632637272715409e-06,
+      "loss": 0.5278,
+      "step": 3276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 19.25753476169495,
+      "learning_rate": 8.631740602017996e-06,
+      "loss": 0.5557,
+      "step": 3277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.05179007269998,
+      "learning_rate": 8.630843684015107e-06,
+      "loss": 0.5754,
+      "step": 3278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.56152572194328,
+      "learning_rate": 8.629946518767815e-06,
+      "loss": 0.6411,
+      "step": 3279
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.027249790189915,
+      "learning_rate": 8.629049106337216e-06,
+      "loss": 0.5431,
+      "step": 3280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.258835679196402,
+      "learning_rate": 8.628151446784419e-06,
+      "loss": 0.574,
+      "step": 3281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.604877250516282,
+      "learning_rate": 8.627253540170548e-06,
+      "loss": 0.5572,
+      "step": 3282
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.282753403637403,
+      "learning_rate": 8.626355386556751e-06,
+      "loss": 0.5029,
+      "step": 3283
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.858700920774058,
+      "learning_rate": 8.625456986004185e-06,
+      "loss": 0.5516,
+      "step": 3284
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.71200602799304,
+      "learning_rate": 8.624558338574026e-06,
+      "loss": 0.4045,
+      "step": 3285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 28.180781259156568,
+      "learning_rate": 8.623659444327473e-06,
+      "loss": 0.4644,
+      "step": 3286
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.305072576062851,
+      "learning_rate": 8.622760303325733e-06,
+      "loss": 0.5162,
+      "step": 3287
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.321244404208016,
+      "learning_rate": 8.621860915630034e-06,
+      "loss": 0.4254,
+      "step": 3288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.04255741947051,
+      "learning_rate": 8.62096128130162e-06,
+      "loss": 0.4998,
+      "step": 3289
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.345682246095693,
+      "learning_rate": 8.620061400401753e-06,
+      "loss": 0.4218,
+      "step": 3290
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.118954538935983,
+      "learning_rate": 8.619161272991711e-06,
+      "loss": 0.5872,
+      "step": 3291
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.17130561248148,
+      "learning_rate": 8.618260899132788e-06,
+      "loss": 0.4174,
+      "step": 3292
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.914449024304987,
+      "learning_rate": 8.617360278886295e-06,
+      "loss": 0.6585,
+      "step": 3293
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.816652875742417,
+      "learning_rate": 8.616459412313562e-06,
+      "loss": 0.3734,
+      "step": 3294
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.328202331516604,
+      "learning_rate": 8.615558299475933e-06,
+      "loss": 0.4992,
+      "step": 3295
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.962039955869098,
+      "learning_rate": 8.61465694043477e-06,
+      "loss": 0.5062,
+      "step": 3296
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.839150099721058,
+      "learning_rate": 8.61375533525145e-06,
+      "loss": 0.4109,
+      "step": 3297
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.030937765654298,
+      "learning_rate": 8.612853483987372e-06,
+      "loss": 0.5217,
+      "step": 3298
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.7060463317058305,
+      "learning_rate": 8.611951386703943e-06,
+      "loss": 0.5125,
+      "step": 3299
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.3809391428454205,
+      "learning_rate": 8.611049043462596e-06,
+      "loss": 0.5463,
+      "step": 3300
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.163637930489662,
+      "learning_rate": 8.610146454324776e-06,
+      "loss": 0.4095,
+      "step": 3301
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.211144601211126,
+      "learning_rate": 8.609243619351944e-06,
+      "loss": 0.5294,
+      "step": 3302
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.337748291964507,
+      "learning_rate": 8.608340538605579e-06,
+      "loss": 0.4345,
+      "step": 3303
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.886467221203555,
+      "learning_rate": 8.607437212147179e-06,
+      "loss": 0.4796,
+      "step": 3304
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.188077210040449,
+      "learning_rate": 8.606533640038253e-06,
+      "loss": 0.5662,
+      "step": 3305
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.885508205898436,
+      "learning_rate": 8.605629822340333e-06,
+      "loss": 0.4693,
+      "step": 3306
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.280466415525275,
+      "learning_rate": 8.604725759114965e-06,
+      "loss": 0.5318,
+      "step": 3307
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.366755305403572,
+      "learning_rate": 8.60382145042371e-06,
+      "loss": 0.5417,
+      "step": 3308
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.54315823165151,
+      "learning_rate": 8.602916896328148e-06,
+      "loss": 0.4235,
+      "step": 3309
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.265078917705435,
+      "learning_rate": 8.602012096889877e-06,
+      "loss": 0.5652,
+      "step": 3310
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.14210917540955,
+      "learning_rate": 8.601107052170508e-06,
+      "loss": 0.5463,
+      "step": 3311
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.771250248662253,
+      "learning_rate": 8.60020176223167e-06,
+      "loss": 0.4585,
+      "step": 3312
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.917433317570234,
+      "learning_rate": 8.59929622713501e-06,
+      "loss": 0.567,
+      "step": 3313
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.6002152229453115,
+      "learning_rate": 8.598390446942193e-06,
+      "loss": 0.5074,
+      "step": 3314
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.104513461758285,
+      "learning_rate": 8.597484421714894e-06,
+      "loss": 0.5799,
+      "step": 3315
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.38276706773683,
+      "learning_rate": 8.596578151514816e-06,
+      "loss": 0.5382,
+      "step": 3316
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.143900139934584,
+      "learning_rate": 8.595671636403663e-06,
+      "loss": 0.4178,
+      "step": 3317
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.725498936352634,
+      "learning_rate": 8.594764876443174e-06,
+      "loss": 0.4855,
+      "step": 3318
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.281439130843477,
+      "learning_rate": 8.593857871695089e-06,
+      "loss": 0.5362,
+      "step": 3319
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.238259820075442,
+      "learning_rate": 8.592950622221173e-06,
+      "loss": 0.4991,
+      "step": 3320
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.274526407555415,
+      "learning_rate": 8.592043128083206e-06,
+      "loss": 0.5582,
+      "step": 3321
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 25.0156804688774,
+      "learning_rate": 8.591135389342984e-06,
+      "loss": 0.5566,
+      "step": 3322
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.310809990759658,
+      "learning_rate": 8.59022740606232e-06,
+      "loss": 0.5052,
+      "step": 3323
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.545792911129993,
+      "learning_rate": 8.589319178303042e-06,
+      "loss": 0.535,
+      "step": 3324
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.151038663671084,
+      "learning_rate": 8.588410706127e-06,
+      "loss": 0.4844,
+      "step": 3325
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.902910615545473,
+      "learning_rate": 8.587501989596054e-06,
+      "loss": 0.8366,
+      "step": 3326
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.2275017094716905,
+      "learning_rate": 8.586593028772086e-06,
+      "loss": 0.4576,
+      "step": 3327
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.856525435546007,
+      "learning_rate": 8.585683823716988e-06,
+      "loss": 0.4465,
+      "step": 3328
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.443650195412311,
+      "learning_rate": 8.584774374492678e-06,
+      "loss": 0.6176,
+      "step": 3329
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.704010167611658,
+      "learning_rate": 8.583864681161083e-06,
+      "loss": 0.5496,
+      "step": 3330
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.2301392732706615,
+      "learning_rate": 8.582954743784149e-06,
+      "loss": 0.4578,
+      "step": 3331
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.951124058563851,
+      "learning_rate": 8.582044562423836e-06,
+      "loss": 0.5143,
+      "step": 3332
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.272521643551233,
+      "learning_rate": 8.58113413714213e-06,
+      "loss": 0.5889,
+      "step": 3333
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.811374436374216,
+      "learning_rate": 8.58022346800102e-06,
+      "loss": 0.4787,
+      "step": 3334
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.217157162981239,
+      "learning_rate": 8.579312555062524e-06,
+      "loss": 0.4285,
+      "step": 3335
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.2966820775804,
+      "learning_rate": 8.578401398388667e-06,
+      "loss": 0.5067,
+      "step": 3336
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.809059661838157,
+      "learning_rate": 8.577489998041499e-06,
+      "loss": 0.5488,
+      "step": 3337
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 16.754773111856345,
+      "learning_rate": 8.576578354083077e-06,
+      "loss": 0.4993,
+      "step": 3338
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.78568682242596,
+      "learning_rate": 8.575666466575485e-06,
+      "loss": 0.5132,
+      "step": 3339
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.8544073869581466,
+      "learning_rate": 8.574754335580815e-06,
+      "loss": 0.4836,
+      "step": 3340
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.5020640873167315,
+      "learning_rate": 8.57384196116118e-06,
+      "loss": 0.5599,
+      "step": 3341
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.363385821530349,
+      "learning_rate": 8.57292934337871e-06,
+      "loss": 0.4718,
+      "step": 3342
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.8745746323155,
+      "learning_rate": 8.572016482295548e-06,
+      "loss": 0.5697,
+      "step": 3343
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.774749873417641,
+      "learning_rate": 8.57110337797386e-06,
+      "loss": 0.4512,
+      "step": 3344
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.876456070781328,
+      "learning_rate": 8.570190030475818e-06,
+      "loss": 0.5687,
+      "step": 3345
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.99239980809926,
+      "learning_rate": 8.569276439863624e-06,
+      "loss": 0.5407,
+      "step": 3346
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.06304163791042,
+      "learning_rate": 8.568362606199483e-06,
+      "loss": 0.5298,
+      "step": 3347
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.134179149222783,
+      "learning_rate": 8.567448529545625e-06,
+      "loss": 0.5362,
+      "step": 3348
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.146465908090399,
+      "learning_rate": 8.566534209964299e-06,
+      "loss": 0.4841,
+      "step": 3349
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.662581066595625,
+      "learning_rate": 8.565619647517759e-06,
+      "loss": 0.4685,
+      "step": 3350
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.520108209995202,
+      "learning_rate": 8.564704842268287e-06,
+      "loss": 0.4757,
+      "step": 3351
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.414858575946362,
+      "learning_rate": 8.56378979427818e-06,
+      "loss": 0.5125,
+      "step": 3352
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.993980683252084,
+      "learning_rate": 8.56287450360974e-06,
+      "loss": 0.6125,
+      "step": 3353
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 14.183850614275041,
+      "learning_rate": 8.561958970325303e-06,
+      "loss": 0.508,
+      "step": 3354
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.507234821531577,
+      "learning_rate": 8.561043194487207e-06,
+      "loss": 0.4264,
+      "step": 3355
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.87676949738523,
+      "learning_rate": 8.560127176157815e-06,
+      "loss": 0.4519,
+      "step": 3356
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.262312351662039,
+      "learning_rate": 8.559210915399506e-06,
+      "loss": 0.5018,
+      "step": 3357
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.852921856813953,
+      "learning_rate": 8.558294412274665e-06,
+      "loss": 0.5194,
+      "step": 3358
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.992233481644572,
+      "learning_rate": 8.557377666845711e-06,
+      "loss": 0.5756,
+      "step": 3359
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.808659851128842,
+      "learning_rate": 8.556460679175067e-06,
+      "loss": 0.6228,
+      "step": 3360
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.055455606956162,
+      "learning_rate": 8.555543449325175e-06,
+      "loss": 0.4453,
+      "step": 3361
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.097739127466468,
+      "learning_rate": 8.554625977358495e-06,
+      "loss": 0.4606,
+      "step": 3362
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.849808278243612,
+      "learning_rate": 8.553708263337503e-06,
+      "loss": 0.523,
+      "step": 3363
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.034472201230639,
+      "learning_rate": 8.55279030732469e-06,
+      "loss": 0.7748,
+      "step": 3364
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.723161169559033,
+      "learning_rate": 8.551872109382566e-06,
+      "loss": 0.4775,
+      "step": 3365
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.658875430849247,
+      "learning_rate": 8.550953669573654e-06,
+      "loss": 0.3737,
+      "step": 3366
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.902935236336449,
+      "learning_rate": 8.550034987960498e-06,
+      "loss": 0.3853,
+      "step": 3367
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.081712744005605,
+      "learning_rate": 8.549116064605657e-06,
+      "loss": 0.5713,
+      "step": 3368
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.68540030145988,
+      "learning_rate": 8.548196899571704e-06,
+      "loss": 0.5683,
+      "step": 3369
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.864035564938673,
+      "learning_rate": 8.547277492921231e-06,
+      "loss": 0.5655,
+      "step": 3370
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8234076428802395,
+      "learning_rate": 8.546357844716843e-06,
+      "loss": 0.8152,
+      "step": 3371
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.581231187071657,
+      "learning_rate": 8.545437955021168e-06,
+      "loss": 0.5243,
+      "step": 3372
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.995954559761051,
+      "learning_rate": 8.544517823896843e-06,
+      "loss": 0.6078,
+      "step": 3373
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.375290959164477,
+      "learning_rate": 8.543597451406525e-06,
+      "loss": 0.4739,
+      "step": 3374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.62830315186741,
+      "learning_rate": 8.54267683761289e-06,
+      "loss": 0.5399,
+      "step": 3375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.580998287542458,
+      "learning_rate": 8.541755982578626e-06,
+      "loss": 0.4508,
+      "step": 3376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.198733938261928,
+      "learning_rate": 8.54083488636644e-06,
+      "loss": 0.5378,
+      "step": 3377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.187205797111046,
+      "learning_rate": 8.539913549039052e-06,
+      "loss": 0.5955,
+      "step": 3378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.950467061850621,
+      "learning_rate": 8.538991970659204e-06,
+      "loss": 0.4689,
+      "step": 3379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6426871209251583,
+      "learning_rate": 8.53807015128965e-06,
+      "loss": 0.7255,
+      "step": 3380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.695103726928884,
+      "learning_rate": 8.537148090993161e-06,
+      "loss": 0.6265,
+      "step": 3381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.861826181610671,
+      "learning_rate": 8.53622578983253e-06,
+      "loss": 0.4838,
+      "step": 3382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.905023941159508,
+      "learning_rate": 8.535303247870554e-06,
+      "loss": 0.5071,
+      "step": 3383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.138518332999341,
+      "learning_rate": 8.53438046517006e-06,
+      "loss": 0.5478,
+      "step": 3384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.877694600264366,
+      "learning_rate": 8.533457441793884e-06,
+      "loss": 0.4318,
+      "step": 3385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.757174545389768,
+      "learning_rate": 8.532534177804877e-06,
+      "loss": 0.4927,
+      "step": 3386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.71354702072817,
+      "learning_rate": 8.531610673265914e-06,
+      "loss": 0.5421,
+      "step": 3387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.024315723396342,
+      "learning_rate": 8.530686928239878e-06,
+      "loss": 0.51,
+      "step": 3388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.986080103162126,
+      "learning_rate": 8.529762942789672e-06,
+      "loss": 0.5679,
+      "step": 3389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.629416103571385,
+      "learning_rate": 8.528838716978215e-06,
+      "loss": 0.5567,
+      "step": 3390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.298522172684395,
+      "learning_rate": 8.527914250868448e-06,
+      "loss": 0.4843,
+      "step": 3391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.830628431495397,
+      "learning_rate": 8.526989544523316e-06,
+      "loss": 0.5087,
+      "step": 3392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.435712777673372,
+      "learning_rate": 8.526064598005792e-06,
+      "loss": 0.4578,
+      "step": 3393
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.093541292902064,
+      "learning_rate": 8.525139411378858e-06,
+      "loss": 0.5151,
+      "step": 3394
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.212894391414852,
+      "learning_rate": 8.524213984705518e-06,
+      "loss": 0.8505,
+      "step": 3395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.847218355560523,
+      "learning_rate": 8.523288318048787e-06,
+      "loss": 0.5315,
+      "step": 3396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.006288115196405,
+      "learning_rate": 8.522362411471698e-06,
+      "loss": 0.4659,
+      "step": 3397
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.767327557094911,
+      "learning_rate": 8.521436265037307e-06,
+      "loss": 0.5198,
+      "step": 3398
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8739000970247721,
+      "learning_rate": 8.520509878808672e-06,
+      "loss": 0.8052,
+      "step": 3399
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.557321070355123,
+      "learning_rate": 8.519583252848882e-06,
+      "loss": 0.49,
+      "step": 3400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.776280370255304,
+      "learning_rate": 8.518656387221032e-06,
+      "loss": 0.5424,
+      "step": 3401
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.675868379002005,
+      "learning_rate": 8.517729281988241e-06,
+      "loss": 0.3469,
+      "step": 3402
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.335967973790378,
+      "learning_rate": 8.51680193721364e-06,
+      "loss": 0.5564,
+      "step": 3403
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1306941560885226,
+      "learning_rate": 8.515874352960377e-06,
+      "loss": 0.7085,
+      "step": 3404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.03242443230205,
+      "learning_rate": 8.514946529291615e-06,
+      "loss": 0.6313,
+      "step": 3405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.748316341294603,
+      "learning_rate": 8.514018466270536e-06,
+      "loss": 0.5358,
+      "step": 3406
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.662242570258182,
+      "learning_rate": 8.513090163960336e-06,
+      "loss": 0.5566,
+      "step": 3407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.208998765992376,
+      "learning_rate": 8.51216162242423e-06,
+      "loss": 0.4458,
+      "step": 3408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.4113931427505895,
+      "learning_rate": 8.511232841725444e-06,
+      "loss": 0.5059,
+      "step": 3409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.822743565154,
+      "learning_rate": 8.510303821927226e-06,
+      "loss": 0.5553,
+      "step": 3410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.603602284187942,
+      "learning_rate": 8.50937456309284e-06,
+      "loss": 0.4925,
+      "step": 3411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.588960193968673,
+      "learning_rate": 8.508445065285564e-06,
+      "loss": 0.624,
+      "step": 3412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.647323831776027,
+      "learning_rate": 8.50751532856869e-06,
+      "loss": 0.5003,
+      "step": 3413
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.399063839386002,
+      "learning_rate": 8.506585353005529e-06,
+      "loss": 0.5548,
+      "step": 3414
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.969498482934343,
+      "learning_rate": 8.505655138659413e-06,
+      "loss": 0.5699,
+      "step": 3415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.8092839272518155,
+      "learning_rate": 8.50472468559368e-06,
+      "loss": 0.5079,
+      "step": 3416
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.594344624362778,
+      "learning_rate": 8.503793993871691e-06,
+      "loss": 0.3896,
+      "step": 3417
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.075156969149434,
+      "learning_rate": 8.502863063556826e-06,
+      "loss": 0.4529,
+      "step": 3418
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.535636403603174,
+      "learning_rate": 8.501931894712471e-06,
+      "loss": 0.5114,
+      "step": 3419
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.719552197879754,
+      "learning_rate": 8.50100048740204e-06,
+      "loss": 0.7365,
+      "step": 3420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.033053912036917,
+      "learning_rate": 8.500068841688953e-06,
+      "loss": 0.6043,
+      "step": 3421
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.667203776901603,
+      "learning_rate": 8.499136957636656e-06,
+      "loss": 0.6064,
+      "step": 3422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.1640461914694935,
+      "learning_rate": 8.4982048353086e-06,
+      "loss": 0.4679,
+      "step": 3423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.295690133737806,
+      "learning_rate": 8.497272474768266e-06,
+      "loss": 0.55,
+      "step": 3424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.275893047177973,
+      "learning_rate": 8.496339876079137e-06,
+      "loss": 0.4599,
+      "step": 3425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.466317758937274,
+      "learning_rate": 8.495407039304722e-06,
+      "loss": 0.5442,
+      "step": 3426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.471685018749936,
+      "learning_rate": 8.494473964508542e-06,
+      "loss": 0.5146,
+      "step": 3427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.183601255700262,
+      "learning_rate": 8.493540651754135e-06,
+      "loss": 0.533,
+      "step": 3428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.255345928671082,
+      "learning_rate": 8.492607101105058e-06,
+      "loss": 0.4401,
+      "step": 3429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.739083081783443,
+      "learning_rate": 8.49167331262488e-06,
+      "loss": 0.5574,
+      "step": 3430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.1059210654472835,
+      "learning_rate": 8.490739286377186e-06,
+      "loss": 0.5374,
+      "step": 3431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.67407504525516,
+      "learning_rate": 8.48980502242558e-06,
+      "loss": 0.5577,
+      "step": 3432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.538545323019135,
+      "learning_rate": 8.488870520833684e-06,
+      "loss": 0.5102,
+      "step": 3433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.243860728846915,
+      "learning_rate": 8.48793578166513e-06,
+      "loss": 0.5001,
+      "step": 3434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.813572456369876,
+      "learning_rate": 8.487000804983572e-06,
+      "loss": 0.4314,
+      "step": 3435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.583789226831637,
+      "learning_rate": 8.486065590852677e-06,
+      "loss": 0.488,
+      "step": 3436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.794112899309258,
+      "learning_rate": 8.485130139336129e-06,
+      "loss": 0.409,
+      "step": 3437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.218724520221874,
+      "learning_rate": 8.484194450497627e-06,
+      "loss": 0.5115,
+      "step": 3438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.507622336061647,
+      "learning_rate": 8.483258524400888e-06,
+      "loss": 0.5916,
+      "step": 3439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.143240570102486,
+      "learning_rate": 8.482322361109644e-06,
+      "loss": 0.4572,
+      "step": 3440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.9444296163984705,
+      "learning_rate": 8.481385960687644e-06,
+      "loss": 0.5266,
+      "step": 3441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.162335626756961,
+      "learning_rate": 8.480449323198657e-06,
+      "loss": 0.5519,
+      "step": 3442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.431876486106576,
+      "learning_rate": 8.479512448706454e-06,
+      "loss": 0.6186,
+      "step": 3443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 24.52132179061023,
+      "learning_rate": 8.478575337274842e-06,
+      "loss": 0.5003,
+      "step": 3444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 65.91387900930023,
+      "learning_rate": 8.477637988967629e-06,
+      "loss": 0.4999,
+      "step": 3445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.990245415234607,
+      "learning_rate": 8.476700403848643e-06,
+      "loss": 0.5047,
+      "step": 3446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.727078750267724,
+      "learning_rate": 8.475762581981734e-06,
+      "loss": 0.4866,
+      "step": 3447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.833694291337748,
+      "learning_rate": 8.47482452343076e-06,
+      "loss": 0.4333,
+      "step": 3448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.117460770958319,
+      "learning_rate": 8.4738862282596e-06,
+      "loss": 0.5997,
+      "step": 3449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.568268693134318,
+      "learning_rate": 8.472947696532148e-06,
+      "loss": 0.7039,
+      "step": 3450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.7275434846419255,
+      "learning_rate": 8.472008928312313e-06,
+      "loss": 0.5076,
+      "step": 3451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.590385101017826,
+      "learning_rate": 8.471069923664019e-06,
+      "loss": 0.4533,
+      "step": 3452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.008439686565441,
+      "learning_rate": 8.470130682651214e-06,
+      "loss": 0.5284,
+      "step": 3453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.054676077285254,
+      "learning_rate": 8.46919120533785e-06,
+      "loss": 0.5252,
+      "step": 3454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.70444296690228,
+      "learning_rate": 8.468251491787904e-06,
+      "loss": 0.4487,
+      "step": 3455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 14.343992685427772,
+      "learning_rate": 8.467311542065368e-06,
+      "loss": 0.4748,
+      "step": 3456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.416183732411804,
+      "learning_rate": 8.466371356234246e-06,
+      "loss": 0.614,
+      "step": 3457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.607252396489128,
+      "learning_rate": 8.46543093435856e-06,
+      "loss": 0.5174,
+      "step": 3458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.0638023700353525,
+      "learning_rate": 8.464490276502351e-06,
+      "loss": 0.4916,
+      "step": 3459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.521808244279732,
+      "learning_rate": 8.46354938272967e-06,
+      "loss": 0.5882,
+      "step": 3460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.750380235102867,
+      "learning_rate": 8.462608253104591e-06,
+      "loss": 0.4621,
+      "step": 3461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.956567010433936,
+      "learning_rate": 8.4616668876912e-06,
+      "loss": 0.5092,
+      "step": 3462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.611291491935021,
+      "learning_rate": 8.460725286553601e-06,
+      "loss": 0.4943,
+      "step": 3463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.976291730481552,
+      "learning_rate": 8.45978344975591e-06,
+      "loss": 0.4938,
+      "step": 3464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.066231852735436,
+      "learning_rate": 8.458841377362263e-06,
+      "loss": 0.5927,
+      "step": 3465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.21394607613989,
+      "learning_rate": 8.457899069436813e-06,
+      "loss": 0.4751,
+      "step": 3466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.48198248724378,
+      "learning_rate": 8.456956526043725e-06,
+      "loss": 0.5097,
+      "step": 3467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.082174252158246,
+      "learning_rate": 8.456013747247181e-06,
+      "loss": 0.4802,
+      "step": 3468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.204288467570324,
+      "learning_rate": 8.45507073311138e-06,
+      "loss": 0.4853,
+      "step": 3469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.724482638031853,
+      "learning_rate": 8.454127483700543e-06,
+      "loss": 0.5901,
+      "step": 3470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.06992555463435,
+      "learning_rate": 8.453183999078893e-06,
+      "loss": 0.5894,
+      "step": 3471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.449356379872386,
+      "learning_rate": 8.452240279310681e-06,
+      "loss": 0.5393,
+      "step": 3472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8816589849020564,
+      "learning_rate": 8.451296324460172e-06,
+      "loss": 0.7593,
+      "step": 3473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.793153435414808,
+      "learning_rate": 8.450352134591641e-06,
+      "loss": 0.6156,
+      "step": 3474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.249830481252067,
+      "learning_rate": 8.449407709769386e-06,
+      "loss": 0.4826,
+      "step": 3475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.916567038041206,
+      "learning_rate": 8.448463050057719e-06,
+      "loss": 0.7024,
+      "step": 3476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.8933914260029425,
+      "learning_rate": 8.447518155520962e-06,
+      "loss": 0.4926,
+      "step": 3477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.3630793249706565,
+      "learning_rate": 8.446573026223463e-06,
+      "loss": 0.545,
+      "step": 3478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 19.82545783908855,
+      "learning_rate": 8.445627662229582e-06,
+      "loss": 0.5005,
+      "step": 3479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.2076099841311185,
+      "learning_rate": 8.44468206360369e-06,
+      "loss": 0.419,
+      "step": 3480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.271084780795182,
+      "learning_rate": 8.44373623041018e-06,
+      "loss": 0.49,
+      "step": 3481
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.628787354457532,
+      "learning_rate": 8.44279016271346e-06,
+      "loss": 0.3757,
+      "step": 3482
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 23.924439220015007,
+      "learning_rate": 8.441843860577951e-06,
+      "loss": 0.5426,
+      "step": 3483
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.553608962821373,
+      "learning_rate": 8.440897324068093e-06,
+      "loss": 0.4748,
+      "step": 3484
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.8778985807327535,
+      "learning_rate": 8.439950553248342e-06,
+      "loss": 0.4095,
+      "step": 3485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.871073756797628,
+      "learning_rate": 8.439003548183167e-06,
+      "loss": 0.5146,
+      "step": 3486
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.733430546394343,
+      "learning_rate": 8.438056308937057e-06,
+      "loss": 0.5501,
+      "step": 3487
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.6455619825998635,
+      "learning_rate": 8.437108835574512e-06,
+      "loss": 0.4946,
+      "step": 3488
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.013102626669617,
+      "learning_rate": 8.436161128160055e-06,
+      "loss": 0.5064,
+      "step": 3489
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8855749807787652,
+      "learning_rate": 8.435213186758214e-06,
+      "loss": 0.7696,
+      "step": 3490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.089308335543846,
+      "learning_rate": 8.434265011433546e-06,
+      "loss": 0.4729,
+      "step": 3491
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.124114887571558,
+      "learning_rate": 8.433316602250613e-06,
+      "loss": 0.5385,
+      "step": 3492
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.896315550539924,
+      "learning_rate": 8.432367959274002e-06,
+      "loss": 0.4912,
+      "step": 3493
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.536202742470563,
+      "learning_rate": 8.43141908256831e-06,
+      "loss": 0.4696,
+      "step": 3494
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.302767248831405,
+      "learning_rate": 8.430469972198149e-06,
+      "loss": 0.4663,
+      "step": 3495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.803239427199292,
+      "learning_rate": 8.42952062822815e-06,
+      "loss": 0.5857,
+      "step": 3496
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.929103722305215,
+      "learning_rate": 8.42857105072296e-06,
+      "loss": 0.5504,
+      "step": 3497
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.241233670587391,
+      "learning_rate": 8.42762123974724e-06,
+      "loss": 0.4944,
+      "step": 3498
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.264855588923915,
+      "learning_rate": 8.426671195365672e-06,
+      "loss": 0.4783,
+      "step": 3499
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.450571934994326,
+      "learning_rate": 8.425720917642943e-06,
+      "loss": 0.4744,
+      "step": 3500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.9005430953902165,
+      "learning_rate": 8.424770406643768e-06,
+      "loss": 0.5654,
+      "step": 3501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.07428312802422,
+      "learning_rate": 8.423819662432868e-06,
+      "loss": 0.5644,
+      "step": 3502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.991023376485304,
+      "learning_rate": 8.42286868507499e-06,
+      "loss": 0.4945,
+      "step": 3503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 21.792357978916193,
+      "learning_rate": 8.421917474634888e-06,
+      "loss": 0.5011,
+      "step": 3504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.523069263784069,
+      "learning_rate": 8.420966031177336e-06,
+      "loss": 0.5331,
+      "step": 3505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.706392435016193,
+      "learning_rate": 8.420014354767123e-06,
+      "loss": 0.4341,
+      "step": 3506
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.080471178876911,
+      "learning_rate": 8.419062445469053e-06,
+      "loss": 0.6146,
+      "step": 3507
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.996540881449119,
+      "learning_rate": 8.418110303347949e-06,
+      "loss": 0.4451,
+      "step": 3508
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.630145187221245,
+      "learning_rate": 8.417157928468645e-06,
+      "loss": 0.4447,
+      "step": 3509
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 16.24617256662968,
+      "learning_rate": 8.416205320895995e-06,
+      "loss": 0.5563,
+      "step": 3510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.860812657601553,
+      "learning_rate": 8.41525248069487e-06,
+      "loss": 0.5283,
+      "step": 3511
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.090719818672214,
+      "learning_rate": 8.414299407930146e-06,
+      "loss": 0.5408,
+      "step": 3512
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.17306483548666,
+      "learning_rate": 8.413346102666733e-06,
+      "loss": 0.6017,
+      "step": 3513
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.582688930480258,
+      "learning_rate": 8.41239256496954e-06,
+      "loss": 0.5069,
+      "step": 3514
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.275760495081853,
+      "learning_rate": 8.411438794903503e-06,
+      "loss": 0.6237,
+      "step": 3515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.573503856928824,
+      "learning_rate": 8.410484792533566e-06,
+      "loss": 0.5495,
+      "step": 3516
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.228477814067196,
+      "learning_rate": 8.409530557924695e-06,
+      "loss": 0.5534,
+      "step": 3517
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.399827502558166,
+      "learning_rate": 8.408576091141869e-06,
+      "loss": 0.5579,
+      "step": 3518
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.673246292312127,
+      "learning_rate": 8.40762139225008e-06,
+      "loss": 0.5107,
+      "step": 3519
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.226970823882237,
+      "learning_rate": 8.406666461314342e-06,
+      "loss": 0.5607,
+      "step": 3520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.866622587259412,
+      "learning_rate": 8.40571129839968e-06,
+      "loss": 0.4796,
+      "step": 3521
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.777457876207507,
+      "learning_rate": 8.404755903571136e-06,
+      "loss": 0.4956,
+      "step": 3522
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.200615677303817,
+      "learning_rate": 8.40380027689377e-06,
+      "loss": 0.4272,
+      "step": 3523
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.066856644813287,
+      "learning_rate": 8.402844418432654e-06,
+      "loss": 0.4567,
+      "step": 3524
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.9227936427638905,
+      "learning_rate": 8.40188832825288e-06,
+      "loss": 0.4814,
+      "step": 3525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.277426666758466,
+      "learning_rate": 8.400932006419551e-06,
+      "loss": 0.5042,
+      "step": 3526
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.236015537267004,
+      "learning_rate": 8.39997545299779e-06,
+      "loss": 0.6389,
+      "step": 3527
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.329065770680831,
+      "learning_rate": 8.399018668052734e-06,
+      "loss": 0.6067,
+      "step": 3528
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.121236376350083,
+      "learning_rate": 8.398061651649536e-06,
+      "loss": 0.5209,
+      "step": 3529
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.847292331659146,
+      "learning_rate": 8.397104403853362e-06,
+      "loss": 0.3926,
+      "step": 3530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.869857917011103,
+      "learning_rate": 8.396146924729399e-06,
+      "loss": 0.5146,
+      "step": 3531
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 11.315706521638845,
+      "learning_rate": 8.395189214342846e-06,
+      "loss": 0.4259,
+      "step": 3532
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.715051458830305,
+      "learning_rate": 8.394231272758919e-06,
+      "loss": 0.5543,
+      "step": 3533
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.827667228013227,
+      "learning_rate": 8.39327310004285e-06,
+      "loss": 0.4639,
+      "step": 3534
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.75509284708867,
+      "learning_rate": 8.392314696259888e-06,
+      "loss": 0.465,
+      "step": 3535
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 17.03195161881997,
+      "learning_rate": 8.39135606147529e-06,
+      "loss": 0.4895,
+      "step": 3536
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8211142553187025,
+      "learning_rate": 8.390397195754342e-06,
+      "loss": 0.7847,
+      "step": 3537
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.757396254039364,
+      "learning_rate": 8.389438099162335e-06,
+      "loss": 0.6215,
+      "step": 3538
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.654427257355364,
+      "learning_rate": 8.38847877176458e-06,
+      "loss": 0.5184,
+      "step": 3539
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.303917675975541,
+      "learning_rate": 8.387519213626403e-06,
+      "loss": 0.5317,
+      "step": 3540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.543514160960338,
+      "learning_rate": 8.386559424813144e-06,
+      "loss": 0.4766,
+      "step": 3541
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.912240789935337,
+      "learning_rate": 8.385599405390162e-06,
+      "loss": 0.4743,
+      "step": 3542
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.041544683721531,
+      "learning_rate": 8.38463915542283e-06,
+      "loss": 0.5787,
+      "step": 3543
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.2094394719107155,
+      "learning_rate": 8.383678674976536e-06,
+      "loss": 0.5519,
+      "step": 3544
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.176929011716309,
+      "learning_rate": 8.382717964116687e-06,
+      "loss": 0.5056,
+      "step": 3545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.166361735077347,
+      "learning_rate": 8.3817570229087e-06,
+      "loss": 0.4964,
+      "step": 3546
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.99008702317138,
+      "learning_rate": 8.38079585141801e-06,
+      "loss": 0.6176,
+      "step": 3547
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.056615008975451,
+      "learning_rate": 8.379834449710073e-06,
+      "loss": 0.574,
+      "step": 3548
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.438738532925512,
+      "learning_rate": 8.378872817850354e-06,
+      "loss": 0.5693,
+      "step": 3549
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.726781846972347,
+      "learning_rate": 8.377910955904335e-06,
+      "loss": 0.4501,
+      "step": 3550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.176418937454682,
+      "learning_rate": 8.376948863937515e-06,
+      "loss": 0.3994,
+      "step": 3551
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.427712803116572,
+      "learning_rate": 8.375986542015408e-06,
+      "loss": 0.4276,
+      "step": 3552
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.695294143160045,
+      "learning_rate": 8.375023990203545e-06,
+      "loss": 0.5175,
+      "step": 3553
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.560931270797628,
+      "learning_rate": 8.37406120856747e-06,
+      "loss": 0.5829,
+      "step": 3554
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.586120069999142,
+      "learning_rate": 8.373098197172745e-06,
+      "loss": 0.4703,
+      "step": 3555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 15.356573867413829,
+      "learning_rate": 8.372134956084948e-06,
+      "loss": 0.5383,
+      "step": 3556
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.220606895948785,
+      "learning_rate": 8.371171485369668e-06,
+      "loss": 0.5012,
+      "step": 3557
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.081257518571126,
+      "learning_rate": 8.370207785092519e-06,
+      "loss": 0.5177,
+      "step": 3558
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.3005766792709315,
+      "learning_rate": 8.369243855319119e-06,
+      "loss": 0.4605,
+      "step": 3559
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.974367494461294,
+      "learning_rate": 8.368279696115108e-06,
+      "loss": 0.6104,
+      "step": 3560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.756847166285601,
+      "learning_rate": 8.367315307546145e-06,
+      "loss": 0.5125,
+      "step": 3561
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.813017530426597,
+      "learning_rate": 8.366350689677895e-06,
+      "loss": 0.4552,
+      "step": 3562
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.800610287003505,
+      "learning_rate": 8.36538584257605e-06,
+      "loss": 0.5456,
+      "step": 3563
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.722107783508615,
+      "learning_rate": 8.364420766306307e-06,
+      "loss": 0.5219,
+      "step": 3564
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 25.112756081045728,
+      "learning_rate": 8.363455460934387e-06,
+      "loss": 0.5558,
+      "step": 3565
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.245976698107647,
+      "learning_rate": 8.36248992652602e-06,
+      "loss": 0.5519,
+      "step": 3566
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.938009675359292,
+      "learning_rate": 8.361524163146957e-06,
+      "loss": 0.5856,
+      "step": 3567
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.537584817679035,
+      "learning_rate": 8.360558170862962e-06,
+      "loss": 0.4955,
+      "step": 3568
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.438333780952223,
+      "learning_rate": 8.359591949739813e-06,
+      "loss": 0.6121,
+      "step": 3569
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.0870164411655905,
+      "learning_rate": 8.358625499843307e-06,
+      "loss": 0.5188,
+      "step": 3570
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.46676895278791,
+      "learning_rate": 8.357658821239254e-06,
+      "loss": 0.5054,
+      "step": 3571
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.971973859882658,
+      "learning_rate": 8.356691913993481e-06,
+      "loss": 0.4381,
+      "step": 3572
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.750856654326321,
+      "learning_rate": 8.355724778171831e-06,
+      "loss": 0.4274,
+      "step": 3573
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.204561107824764,
+      "learning_rate": 8.35475741384016e-06,
+      "loss": 0.5644,
+      "step": 3574
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.349777980228698,
+      "learning_rate": 8.353789821064341e-06,
+      "loss": 0.629,
+      "step": 3575
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.955491075098183,
+      "learning_rate": 8.352821999910268e-06,
+      "loss": 0.6415,
+      "step": 3576
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.321524871653928,
+      "learning_rate": 8.351853950443837e-06,
+      "loss": 0.3719,
+      "step": 3577
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7794721926488166,
+      "learning_rate": 8.350885672730972e-06,
+      "loss": 0.7945,
+      "step": 3578
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.928590577717362,
+      "learning_rate": 8.34991716683761e-06,
+      "loss": 0.5412,
+      "step": 3579
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.527732474146418,
+      "learning_rate": 8.348948432829703e-06,
+      "loss": 0.3816,
+      "step": 3580
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6616162808953934,
+      "learning_rate": 8.34797947077321e-06,
+      "loss": 0.7363,
+      "step": 3581
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 39.29624649770848,
+      "learning_rate": 8.34701028073412e-06,
+      "loss": 0.5652,
+      "step": 3582
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.720510871091731,
+      "learning_rate": 8.346040862778431e-06,
+      "loss": 0.558,
+      "step": 3583
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.5253146927101,
+      "learning_rate": 8.345071216972151e-06,
+      "loss": 0.5475,
+      "step": 3584
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.784810928188096,
+      "learning_rate": 8.344101343381312e-06,
+      "loss": 0.5304,
+      "step": 3585
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.586611700512844,
+      "learning_rate": 8.343131242071957e-06,
+      "loss": 0.5103,
+      "step": 3586
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 24.49167715931628,
+      "learning_rate": 8.342160913110145e-06,
+      "loss": 0.5093,
+      "step": 3587
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.999457034399358,
+      "learning_rate": 8.341190356561953e-06,
+      "loss": 0.5354,
+      "step": 3588
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.845348979953848,
+      "learning_rate": 8.340219572493471e-06,
+      "loss": 0.5397,
+      "step": 3589
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.772111571911976,
+      "learning_rate": 8.339248560970804e-06,
+      "loss": 0.4634,
+      "step": 3590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.08559535043936,
+      "learning_rate": 8.338277322060075e-06,
+      "loss": 0.5304,
+      "step": 3591
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 22.78479097815828,
+      "learning_rate": 8.33730585582742e-06,
+      "loss": 0.5612,
+      "step": 3592
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.309226803629712,
+      "learning_rate": 8.336334162338992e-06,
+      "loss": 0.4608,
+      "step": 3593
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.205079602076425,
+      "learning_rate": 8.335362241660959e-06,
+      "loss": 0.5163,
+      "step": 3594
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.644959898786618,
+      "learning_rate": 8.334390093859504e-06,
+      "loss": 0.3568,
+      "step": 3595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.919113119911759,
+      "learning_rate": 8.333417719000828e-06,
+      "loss": 0.5861,
+      "step": 3596
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.92955664957474,
+      "learning_rate": 8.332445117151143e-06,
+      "loss": 0.5042,
+      "step": 3597
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.484324959863743,
+      "learning_rate": 8.33147228837668e-06,
+      "loss": 0.4711,
+      "step": 3598
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.767815500890393,
+      "learning_rate": 8.330499232743683e-06,
+      "loss": 0.515,
+      "step": 3599
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.117889259688594,
+      "learning_rate": 8.329525950318416e-06,
+      "loss": 0.5205,
+      "step": 3600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.198351025796937,
+      "learning_rate": 8.32855244116715e-06,
+      "loss": 0.4586,
+      "step": 3601
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.70197777700214,
+      "learning_rate": 8.327578705356183e-06,
+      "loss": 0.4916,
+      "step": 3602
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.203246039204899,
+      "learning_rate": 8.326604742951816e-06,
+      "loss": 0.5224,
+      "step": 3603
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 33.719993732353174,
+      "learning_rate": 8.325630554020377e-06,
+      "loss": 0.4738,
+      "step": 3604
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.609737721658647,
+      "learning_rate": 8.3246561386282e-06,
+      "loss": 0.4993,
+      "step": 3605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.6620548187703115,
+      "learning_rate": 8.32368149684164e-06,
+      "loss": 0.4578,
+      "step": 3606
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.27817365512824,
+      "learning_rate": 8.322706628727067e-06,
+      "loss": 0.4977,
+      "step": 3607
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.334886814092892,
+      "learning_rate": 8.321731534350864e-06,
+      "loss": 0.5605,
+      "step": 3608
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.958657534655327,
+      "learning_rate": 8.320756213779426e-06,
+      "loss": 0.4636,
+      "step": 3609
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.222541277714898,
+      "learning_rate": 8.319780667079176e-06,
+      "loss": 0.5887,
+      "step": 3610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.8557235460919195,
+      "learning_rate": 8.31880489431654e-06,
+      "loss": 0.3967,
+      "step": 3611
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.5092879418614205,
+      "learning_rate": 8.317828895557966e-06,
+      "loss": 0.4628,
+      "step": 3612
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.143497723438257,
+      "learning_rate": 8.316852670869913e-06,
+      "loss": 0.8211,
+      "step": 3613
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.030421725038982,
+      "learning_rate": 8.315876220318859e-06,
+      "loss": 0.4699,
+      "step": 3614
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.2778252919187505,
+      "learning_rate": 8.314899543971294e-06,
+      "loss": 0.4987,
+      "step": 3615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.829543747400511,
+      "learning_rate": 8.313922641893728e-06,
+      "loss": 0.4442,
+      "step": 3616
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.344555782712774,
+      "learning_rate": 8.312945514152681e-06,
+      "loss": 0.4199,
+      "step": 3617
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.301078072181946,
+      "learning_rate": 8.311968160814694e-06,
+      "loss": 0.5205,
+      "step": 3618
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.316013135095086,
+      "learning_rate": 8.310990581946317e-06,
+      "loss": 0.558,
+      "step": 3619
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.220595052798193,
+      "learning_rate": 8.310012777614123e-06,
+      "loss": 0.4811,
+      "step": 3620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.8964108095307735,
+      "learning_rate": 8.309034747884694e-06,
+      "loss": 0.4897,
+      "step": 3621
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.950473106873088,
+      "learning_rate": 8.308056492824628e-06,
+      "loss": 0.4896,
+      "step": 3622
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.03963685842277,
+      "learning_rate": 8.30707801250054e-06,
+      "loss": 0.4949,
+      "step": 3623
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.331046416203326,
+      "learning_rate": 8.306099306979064e-06,
+      "loss": 0.5534,
+      "step": 3624
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.699465441159886,
+      "learning_rate": 8.305120376326843e-06,
+      "loss": 0.5486,
+      "step": 3625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.530725955755005,
+      "learning_rate": 8.304141220610535e-06,
+      "loss": 0.4624,
+      "step": 3626
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.143631657655193,
+      "learning_rate": 8.303161839896822e-06,
+      "loss": 0.546,
+      "step": 3627
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.525197358140269,
+      "learning_rate": 8.302182234252392e-06,
+      "loss": 0.4984,
+      "step": 3628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.459911597548828,
+      "learning_rate": 8.301202403743951e-06,
+      "loss": 0.4847,
+      "step": 3629
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7846589537343196,
+      "learning_rate": 8.300222348438223e-06,
+      "loss": 0.7349,
+      "step": 3630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.4870094270592,
+      "learning_rate": 8.299242068401946e-06,
+      "loss": 0.6399,
+      "step": 3631
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.8180672543435925,
+      "learning_rate": 8.298261563701871e-06,
+      "loss": 0.489,
+      "step": 3632
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.267938615907452,
+      "learning_rate": 8.297280834404765e-06,
+      "loss": 0.5614,
+      "step": 3633
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.509919933910249,
+      "learning_rate": 8.296299880577415e-06,
+      "loss": 0.4835,
+      "step": 3634
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.796991507441733,
+      "learning_rate": 8.295318702286617e-06,
+      "loss": 0.4154,
+      "step": 3635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.806530673785464,
+      "learning_rate": 8.294337299599185e-06,
+      "loss": 0.4889,
+      "step": 3636
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.943413575233301,
+      "learning_rate": 8.293355672581949e-06,
+      "loss": 0.5635,
+      "step": 3637
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.220859750330224,
+      "learning_rate": 8.292373821301753e-06,
+      "loss": 0.516,
+      "step": 3638
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.470215490406219,
+      "learning_rate": 8.291391745825456e-06,
+      "loss": 0.6026,
+      "step": 3639
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.089446664586613,
+      "learning_rate": 8.290409446219936e-06,
+      "loss": 0.5484,
+      "step": 3640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9110834166459743,
+      "learning_rate": 8.289426922552079e-06,
+      "loss": 0.7353,
+      "step": 3641
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.5647689104503595,
+      "learning_rate": 8.288444174888794e-06,
+      "loss": 0.5448,
+      "step": 3642
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.476606997528038,
+      "learning_rate": 8.287461203297001e-06,
+      "loss": 0.3921,
+      "step": 3643
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.2672080722922106,
+      "learning_rate": 8.286478007843633e-06,
+      "loss": 0.5906,
+      "step": 3644
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.683863030814304,
+      "learning_rate": 8.285494588595645e-06,
+      "loss": 0.5809,
+      "step": 3645
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.03490325410489,
+      "learning_rate": 8.284510945620004e-06,
+      "loss": 0.5937,
+      "step": 3646
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.974727309615757,
+      "learning_rate": 8.283527078983686e-06,
+      "loss": 0.5193,
+      "step": 3647
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.523806413449996,
+      "learning_rate": 8.282542988753695e-06,
+      "loss": 0.5662,
+      "step": 3648
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.977156760810557,
+      "learning_rate": 8.28155867499704e-06,
+      "loss": 0.3974,
+      "step": 3649
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.4883077433551195,
+      "learning_rate": 8.280574137780746e-06,
+      "loss": 0.4018,
+      "step": 3650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.903565633170126,
+      "learning_rate": 8.279589377171862e-06,
+      "loss": 0.5841,
+      "step": 3651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.012320801817599,
+      "learning_rate": 8.27860439323744e-06,
+      "loss": 0.4723,
+      "step": 3652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.996629744792759,
+      "learning_rate": 8.277619186044551e-06,
+      "loss": 0.5087,
+      "step": 3653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 18.743577493263643,
+      "learning_rate": 8.276633755660293e-06,
+      "loss": 0.5029,
+      "step": 3654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.857388683037292,
+      "learning_rate": 8.275648102151762e-06,
+      "loss": 0.517,
+      "step": 3655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.034850671436515,
+      "learning_rate": 8.274662225586078e-06,
+      "loss": 0.4381,
+      "step": 3656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.589801059340985,
+      "learning_rate": 8.273676126030376e-06,
+      "loss": 0.5112,
+      "step": 3657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.543321234114773,
+      "learning_rate": 8.272689803551803e-06,
+      "loss": 0.508,
+      "step": 3658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.926561764771141,
+      "learning_rate": 8.271703258217525e-06,
+      "loss": 0.5115,
+      "step": 3659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.67608452634555,
+      "learning_rate": 8.270716490094721e-06,
+      "loss": 0.5322,
+      "step": 3660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.050842603276816,
+      "learning_rate": 8.269729499250586e-06,
+      "loss": 0.5892,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.282270689403786,
+      "learning_rate": 8.268742285752329e-06,
+      "loss": 0.4864,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.7219637626193744,
+      "learning_rate": 8.267754849667174e-06,
+      "loss": 0.5394,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.913599462158424,
+      "learning_rate": 8.266767191062362e-06,
+      "loss": 0.5888,
+      "step": 3664
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.454559201136258,
+      "learning_rate": 8.265779310005149e-06,
+      "loss": 0.5395,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.27658594743028,
+      "learning_rate": 8.264791206562804e-06,
+      "loss": 0.4863,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.916942868405446,
+      "learning_rate": 8.263802880802612e-06,
+      "loss": 0.6535,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.306653867254724,
+      "learning_rate": 8.262814332791876e-06,
+      "loss": 0.4296,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.588468775992327,
+      "learning_rate": 8.261825562597909e-06,
+      "loss": 0.5835,
+      "step": 3669
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.752669897681185,
+      "learning_rate": 8.260836570288045e-06,
+      "loss": 0.5051,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.558337687811942,
+      "learning_rate": 8.259847355929626e-06,
+      "loss": 0.4501,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.573337316531747,
+      "learning_rate": 8.258857919590017e-06,
+      "loss": 0.401,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.264974908409792,
+      "learning_rate": 8.257868261336592e-06,
+      "loss": 0.5678,
+      "step": 3673
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.103953584942702,
+      "learning_rate": 8.256878381236743e-06,
+      "loss": 0.6479,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.683725856947736,
+      "learning_rate": 8.255888279357875e-06,
+      "loss": 0.5571,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.069194724069089,
+      "learning_rate": 8.254897955767413e-06,
+      "loss": 0.5379,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.591143305935272,
+      "learning_rate": 8.253907410532792e-06,
+      "loss": 0.4515,
+      "step": 3677
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.943497592652056,
+      "learning_rate": 8.252916643721463e-06,
+      "loss": 0.478,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.585705219351978,
+      "learning_rate": 8.251925655400892e-06,
+      "loss": 0.5453,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.318723022221155,
+      "learning_rate": 8.250934445638562e-06,
+      "loss": 0.5492,
+      "step": 3680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.700277964886863,
+      "learning_rate": 8.249943014501971e-06,
+      "loss": 0.3857,
+      "step": 3681
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.395962899292703,
+      "learning_rate": 8.24895136205863e-06,
+      "loss": 0.5381,
+      "step": 3682
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.319561978593938,
+      "learning_rate": 8.247959488376066e-06,
+      "loss": 0.5704,
+      "step": 3683
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.157136702035869,
+      "learning_rate": 8.246967393521821e-06,
+      "loss": 0.4798,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8597482386264,
+      "learning_rate": 8.245975077563454e-06,
+      "loss": 0.566,
+      "step": 3685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.092997183221051,
+      "learning_rate": 8.244982540568537e-06,
+      "loss": 0.5637,
+      "step": 3686
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.870977037993715,
+      "learning_rate": 8.243989782604656e-06,
+      "loss": 0.6358,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.581913267195646,
+      "learning_rate": 8.242996803739412e-06,
+      "loss": 0.499,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 17.244191151257855,
+      "learning_rate": 8.242003604040426e-06,
+      "loss": 0.5095,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.25686815492249,
+      "learning_rate": 8.241010183575328e-06,
+      "loss": 0.4625,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9408372125420383,
+      "learning_rate": 8.240016542411769e-06,
+      "loss": 0.5342,
+      "step": 3691
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.199083406764285,
+      "learning_rate": 8.239022680617406e-06,
+      "loss": 0.5268,
+      "step": 3692
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.953462530655232,
+      "learning_rate": 8.238028598259923e-06,
+      "loss": 0.4283,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6361534577162744,
+      "learning_rate": 8.237034295407005e-06,
+      "loss": 0.7694,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.63805576902322,
+      "learning_rate": 8.236039772126367e-06,
+      "loss": 0.4281,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.740660203264187,
+      "learning_rate": 8.235045028485728e-06,
+      "loss": 0.4496,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.528623823824217,
+      "learning_rate": 8.234050064552826e-06,
+      "loss": 0.5022,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.052846138318852,
+      "learning_rate": 8.233054880395412e-06,
+      "loss": 0.4707,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.688367255764407,
+      "learning_rate": 8.232059476081258e-06,
+      "loss": 0.5374,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.743450806582542,
+      "learning_rate": 8.231063851678142e-06,
+      "loss": 0.4854,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.530748505791039,
+      "learning_rate": 8.230068007253867e-06,
+      "loss": 0.5821,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.550434115688681,
+      "learning_rate": 8.22907194287624e-06,
+      "loss": 0.6488,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.234885841006097,
+      "learning_rate": 8.228075658613089e-06,
+      "loss": 0.5714,
+      "step": 3703
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.929698204864925,
+      "learning_rate": 8.227079154532261e-06,
+      "loss": 0.5166,
+      "step": 3704
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.514907283268624,
+      "learning_rate": 8.22608243070161e-06,
+      "loss": 0.4646,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.949926521724741,
+      "learning_rate": 8.22508548718901e-06,
+      "loss": 0.4084,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.067700181692851,
+      "learning_rate": 8.22408832406235e-06,
+      "loss": 0.6889,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2665904708986324,
+      "learning_rate": 8.223090941389526e-06,
+      "loss": 0.7447,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.308236609602225,
+      "learning_rate": 8.222093339238464e-06,
+      "loss": 0.4883,
+      "step": 3709
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.677118588595357,
+      "learning_rate": 8.221095517677089e-06,
+      "loss": 0.549,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.721854354712263,
+      "learning_rate": 8.220097476773351e-06,
+      "loss": 0.5726,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.513853317412811,
+      "learning_rate": 8.219099216595213e-06,
+      "loss": 0.5456,
+      "step": 3712
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.927791517465652,
+      "learning_rate": 8.218100737210652e-06,
+      "loss": 0.5929,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.617548923774926,
+      "learning_rate": 8.217102038687659e-06,
+      "loss": 0.4977,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.562856077062456,
+      "learning_rate": 8.216103121094242e-06,
+      "loss": 0.4837,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.087837314757011,
+      "learning_rate": 8.21510398449842e-06,
+      "loss": 0.5309,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.11824655374052,
+      "learning_rate": 8.214104628968236e-06,
+      "loss": 0.3932,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.508284329650937,
+      "learning_rate": 8.213105054571735e-06,
+      "loss": 0.4379,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.8350398743222485,
+      "learning_rate": 8.212105261376987e-06,
+      "loss": 0.552,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.888549301625491,
+      "learning_rate": 8.211105249452074e-06,
+      "loss": 0.4874,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.564075814380509,
+      "learning_rate": 8.21010501886509e-06,
+      "loss": 0.584,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 19.5834790493329,
+      "learning_rate": 8.209104569684148e-06,
+      "loss": 0.504,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.876678821470523,
+      "learning_rate": 8.208103901977373e-06,
+      "loss": 0.4339,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.189985417132193,
+      "learning_rate": 8.207103015812909e-06,
+      "loss": 0.4459,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6451115193278971,
+      "learning_rate": 8.206101911258907e-06,
+      "loss": 0.7565,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.38824164258767,
+      "learning_rate": 8.205100588383543e-06,
+      "loss": 0.5254,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.434325025074155,
+      "learning_rate": 8.204099047254998e-06,
+      "loss": 0.5435,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.839686424676175,
+      "learning_rate": 8.203097287941476e-06,
+      "loss": 0.4476,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.197413215860973,
+      "learning_rate": 8.20209531051119e-06,
+      "loss": 0.4996,
+      "step": 3729
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.218936168020379,
+      "learning_rate": 8.201093115032373e-06,
+      "loss": 0.4654,
+      "step": 3730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.145089587340447,
+      "learning_rate": 8.200090701573267e-06,
+      "loss": 0.4554,
+      "step": 3731
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.237965172306517,
+      "learning_rate": 8.199088070202133e-06,
+      "loss": 0.5088,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.011197332067799,
+      "learning_rate": 8.198085220987246e-06,
+      "loss": 0.5111,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.9618746802402445,
+      "learning_rate": 8.197082153996896e-06,
+      "loss": 0.5971,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.777503119414467,
+      "learning_rate": 8.196078869299386e-06,
+      "loss": 0.6249,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.16781537388829,
+      "learning_rate": 8.195075366963039e-06,
+      "loss": 0.4482,
+      "step": 3736
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.728651278688712,
+      "learning_rate": 8.194071647056184e-06,
+      "loss": 0.4743,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.033499663990609,
+      "learning_rate": 8.193067709647174e-06,
+      "loss": 0.6486,
+      "step": 3738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.003854902684285,
+      "learning_rate": 8.192063554804371e-06,
+      "loss": 0.5674,
+      "step": 3739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.30510409820181,
+      "learning_rate": 8.191059182596151e-06,
+      "loss": 0.4871,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.195696004071126,
+      "learning_rate": 8.190054593090912e-06,
+      "loss": 0.538,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.929869377271293,
+      "learning_rate": 8.18904978635706e-06,
+      "loss": 0.5483,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.017624101604865,
+      "learning_rate": 8.188044762463017e-06,
+      "loss": 0.6053,
+      "step": 3743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.1167433008549,
+      "learning_rate": 8.187039521477221e-06,
+      "loss": 0.5069,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 44.5843536865377,
+      "learning_rate": 8.186034063468125e-06,
+      "loss": 0.6692,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.343891094273244,
+      "learning_rate": 8.185028388504198e-06,
+      "loss": 0.517,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 22.369119880302506,
+      "learning_rate": 8.184022496653918e-06,
+      "loss": 0.3881,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.95683013667758,
+      "learning_rate": 8.183016387985784e-06,
+      "loss": 0.654,
+      "step": 3748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.266346729921759,
+      "learning_rate": 8.182010062568311e-06,
+      "loss": 0.5158,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.482313077608056,
+      "learning_rate": 8.18100352047002e-06,
+      "loss": 0.5579,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.373733538315702,
+      "learning_rate": 8.179996761759452e-06,
+      "loss": 0.3596,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.790305687598856,
+      "learning_rate": 8.178989786505166e-06,
+      "loss": 0.4135,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.88488019082936,
+      "learning_rate": 8.17798259477573e-06,
+      "loss": 0.48,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.38908882197007,
+      "learning_rate": 8.176975186639734e-06,
+      "loss": 0.5324,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.257980540566506,
+      "learning_rate": 8.175967562165771e-06,
+      "loss": 0.5503,
+      "step": 3755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.366618128219251,
+      "learning_rate": 8.174959721422462e-06,
+      "loss": 0.5907,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.211885359280256,
+      "learning_rate": 8.173951664478432e-06,
+      "loss": 0.5149,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.696772448771698,
+      "learning_rate": 8.17294339140233e-06,
+      "loss": 0.5369,
+      "step": 3758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.618675619017282,
+      "learning_rate": 8.17193490226281e-06,
+      "loss": 0.4449,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.749674145177838,
+      "learning_rate": 8.170926197128547e-06,
+      "loss": 0.527,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.216030584198778,
+      "learning_rate": 8.169917276068232e-06,
+      "loss": 0.4638,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.78194434685047,
+      "learning_rate": 8.168908139150565e-06,
+      "loss": 0.5167,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.050808844816927,
+      "learning_rate": 8.167898786444265e-06,
+      "loss": 0.6558,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.810010849245106,
+      "learning_rate": 8.166889218018064e-06,
+      "loss": 0.4411,
+      "step": 3764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.7897682916218605,
+      "learning_rate": 8.16587943394071e-06,
+      "loss": 0.4945,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.0658905511881,
+      "learning_rate": 8.164869434280964e-06,
+      "loss": 0.7186,
+      "step": 3766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.505331143661067,
+      "learning_rate": 8.163859219107604e-06,
+      "loss": 0.4699,
+      "step": 3767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.181773166476152,
+      "learning_rate": 8.162848788489419e-06,
+      "loss": 0.5669,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.774761197300455,
+      "learning_rate": 8.161838142495215e-06,
+      "loss": 0.5074,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.81251753387367,
+      "learning_rate": 8.160827281193816e-06,
+      "loss": 0.4646,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.187316909594648,
+      "learning_rate": 8.159816204654053e-06,
+      "loss": 0.4354,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.962074234375996,
+      "learning_rate": 8.158804912944777e-06,
+      "loss": 0.4777,
+      "step": 3772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 28.97844927489227,
+      "learning_rate": 8.157793406134854e-06,
+      "loss": 0.5298,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.923661659758274,
+      "learning_rate": 8.15678168429316e-06,
+      "loss": 0.399,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.407653918590597,
+      "learning_rate": 8.155769747488593e-06,
+      "loss": 0.3884,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.78691374754676,
+      "learning_rate": 8.154757595790059e-06,
+      "loss": 0.3973,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 21.322371820652275,
+      "learning_rate": 8.153745229266482e-06,
+      "loss": 0.6206,
+      "step": 3777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.365593742465589,
+      "learning_rate": 8.152732647986798e-06,
+      "loss": 0.5695,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.28226586801012,
+      "learning_rate": 8.151719852019961e-06,
+      "loss": 0.4489,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.557340973663642,
+      "learning_rate": 8.150706841434936e-06,
+      "loss": 0.6,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.214670929850804,
+      "learning_rate": 8.149693616300707e-06,
+      "loss": 0.5239,
+      "step": 3781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.503703167739467,
+      "learning_rate": 8.148680176686267e-06,
+      "loss": 0.5446,
+      "step": 3782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.032993672209306,
+      "learning_rate": 8.147666522660632e-06,
+      "loss": 0.5638,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.050958023061375,
+      "learning_rate": 8.14665265429282e-06,
+      "loss": 0.503,
+      "step": 3784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.964228051135589,
+      "learning_rate": 8.14563857165188e-06,
+      "loss": 0.5259,
+      "step": 3785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.252252480854292,
+      "learning_rate": 8.144624274806859e-06,
+      "loss": 0.4182,
+      "step": 3786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.732185575287397,
+      "learning_rate": 8.143609763826828e-06,
+      "loss": 0.479,
+      "step": 3787
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.92628035507929,
+      "learning_rate": 8.142595038780874e-06,
+      "loss": 0.7883,
+      "step": 3788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.961440514185693,
+      "learning_rate": 8.14158009973809e-06,
+      "loss": 0.6228,
+      "step": 3789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.08902211678695,
+      "learning_rate": 8.140564946767593e-06,
+      "loss": 0.4548,
+      "step": 3790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 21.39874616839451,
+      "learning_rate": 8.139549579938507e-06,
+      "loss": 0.4194,
+      "step": 3791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.01237549297968,
+      "learning_rate": 8.138533999319978e-06,
+      "loss": 0.5037,
+      "step": 3792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.737830031036061,
+      "learning_rate": 8.137518204981159e-06,
+      "loss": 0.4628,
+      "step": 3793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.210818129645409,
+      "learning_rate": 8.136502196991223e-06,
+      "loss": 0.4785,
+      "step": 3794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.755384080447318,
+      "learning_rate": 8.135485975419357e-06,
+      "loss": 0.4739,
+      "step": 3795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.621743180070064,
+      "learning_rate": 8.134469540334758e-06,
+      "loss": 0.5511,
+      "step": 3796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1262688031306793,
+      "learning_rate": 8.133452891806644e-06,
+      "loss": 0.8262,
+      "step": 3797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.4664363861407805,
+      "learning_rate": 8.132436029904241e-06,
+      "loss": 0.5207,
+      "step": 3798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.653353613768965,
+      "learning_rate": 8.131418954696792e-06,
+      "loss": 0.5482,
+      "step": 3799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.6009707026194,
+      "learning_rate": 8.130401666253562e-06,
+      "loss": 0.5157,
+      "step": 3800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.52466149505616,
+      "learning_rate": 8.129384164643816e-06,
+      "loss": 0.5012,
+      "step": 3801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.903737598417261,
+      "learning_rate": 8.128366449936848e-06,
+      "loss": 0.5321,
+      "step": 3802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0706753500808994,
+      "learning_rate": 8.127348522201954e-06,
+      "loss": 0.8349,
+      "step": 3803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.4775322551325605,
+      "learning_rate": 8.126330381508453e-06,
+      "loss": 0.45,
+      "step": 3804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.350568818416195,
+      "learning_rate": 8.125312027925678e-06,
+      "loss": 0.4768,
+      "step": 3805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.716843823948012,
+      "learning_rate": 8.124293461522971e-06,
+      "loss": 0.4839,
+      "step": 3806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.5623230180564205,
+      "learning_rate": 8.123274682369694e-06,
+      "loss": 0.497,
+      "step": 3807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.666659796873235,
+      "learning_rate": 8.12225569053522e-06,
+      "loss": 0.5224,
+      "step": 3808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.427987613539861,
+      "learning_rate": 8.12123648608894e-06,
+      "loss": 0.487,
+      "step": 3809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.660724210431715,
+      "learning_rate": 8.120217069100254e-06,
+      "loss": 0.417,
+      "step": 3810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.852860254877519,
+      "learning_rate": 8.11919743963858e-06,
+      "loss": 0.5963,
+      "step": 3811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.591643451549886,
+      "learning_rate": 8.118177597773355e-06,
+      "loss": 0.6105,
+      "step": 3812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.6393750624748895,
+      "learning_rate": 8.11715754357402e-06,
+      "loss": 0.5506,
+      "step": 3813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.936520662297092,
+      "learning_rate": 8.116137277110039e-06,
+      "loss": 0.5007,
+      "step": 3814
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.453251340501355,
+      "learning_rate": 8.115116798450888e-06,
+      "loss": 0.498,
+      "step": 3815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.081112254903822,
+      "learning_rate": 8.114096107666056e-06,
+      "loss": 0.6097,
+      "step": 3816
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.1453090945788205,
+      "learning_rate": 8.113075204825049e-06,
+      "loss": 0.4111,
+      "step": 3817
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.290948251496725,
+      "learning_rate": 8.112054089997385e-06,
+      "loss": 0.5425,
+      "step": 3818
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.77973071276735,
+      "learning_rate": 8.111032763252595e-06,
+      "loss": 0.5337,
+      "step": 3819
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.216340385678647,
+      "learning_rate": 8.110011224660231e-06,
+      "loss": 0.5075,
+      "step": 3820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.1383883004825055,
+      "learning_rate": 8.108989474289854e-06,
+      "loss": 0.5604,
+      "step": 3821
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.140965956537602,
+      "learning_rate": 8.107967512211037e-06,
+      "loss": 0.4832,
+      "step": 3822
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.113408368954123,
+      "learning_rate": 8.106945338493377e-06,
+      "loss": 0.4623,
+      "step": 3823
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.4403808598535885,
+      "learning_rate": 8.105922953206476e-06,
+      "loss": 0.4867,
+      "step": 3824
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.075488100498784,
+      "learning_rate": 8.104900356419954e-06,
+      "loss": 0.4744,
+      "step": 3825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.509651209759349,
+      "learning_rate": 8.103877548203446e-06,
+      "loss": 0.4849,
+      "step": 3826
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.308812405225606,
+      "learning_rate": 8.102854528626602e-06,
+      "loss": 0.3992,
+      "step": 3827
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.737888815666421,
+      "learning_rate": 8.101831297759082e-06,
+      "loss": 0.523,
+      "step": 3828
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.71247254908851,
+      "learning_rate": 8.100807855670565e-06,
+      "loss": 0.4558,
+      "step": 3829
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.533384205664273,
+      "learning_rate": 8.099784202430745e-06,
+      "loss": 0.4012,
+      "step": 3830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.385093576600733,
+      "learning_rate": 8.098760338109325e-06,
+      "loss": 0.4341,
+      "step": 3831
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.704688116471297,
+      "learning_rate": 8.097736262776028e-06,
+      "loss": 0.6202,
+      "step": 3832
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.205001131725457,
+      "learning_rate": 8.096711976500586e-06,
+      "loss": 0.4106,
+      "step": 3833
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.51302807921795,
+      "learning_rate": 8.095687479352752e-06,
+      "loss": 0.5312,
+      "step": 3834
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9401042577352692,
+      "learning_rate": 8.094662771402288e-06,
+      "loss": 0.7978,
+      "step": 3835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.147097808479476,
+      "learning_rate": 8.093637852718972e-06,
+      "loss": 0.5292,
+      "step": 3836
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.678414972376531,
+      "learning_rate": 8.092612723372597e-06,
+      "loss": 0.4571,
+      "step": 3837
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.792122699858384,
+      "learning_rate": 8.091587383432969e-06,
+      "loss": 0.5016,
+      "step": 3838
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.19001209802955,
+      "learning_rate": 8.09056183296991e-06,
+      "loss": 0.5142,
+      "step": 3839
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.005285485028755,
+      "learning_rate": 8.089536072053255e-06,
+      "loss": 0.4089,
+      "step": 3840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.404062591823338,
+      "learning_rate": 8.088510100752853e-06,
+      "loss": 0.4711,
+      "step": 3841
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.214666465138622,
+      "learning_rate": 8.087483919138569e-06,
+      "loss": 0.5736,
+      "step": 3842
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.946173311329015,
+      "learning_rate": 8.086457527280282e-06,
+      "loss": 0.6279,
+      "step": 3843
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.317695092619702,
+      "learning_rate": 8.085430925247882e-06,
+      "loss": 0.6706,
+      "step": 3844
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.652995785690427,
+      "learning_rate": 8.08440411311128e-06,
+      "loss": 0.5481,
+      "step": 3845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 32.49879301859775,
+      "learning_rate": 8.083377090940395e-06,
+      "loss": 0.5684,
+      "step": 3846
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.29478747364572,
+      "learning_rate": 8.082349858805163e-06,
+      "loss": 0.4322,
+      "step": 3847
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.4459721339768,
+      "learning_rate": 8.081322416775535e-06,
+      "loss": 0.4914,
+      "step": 3848
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.9830995113657774,
+      "learning_rate": 8.080294764921473e-06,
+      "loss": 0.4523,
+      "step": 3849
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.189155167199814,
+      "learning_rate": 8.079266903312958e-06,
+      "loss": 0.5374,
+      "step": 3850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.453298589374432,
+      "learning_rate": 8.078238832019982e-06,
+      "loss": 0.5685,
+      "step": 3851
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 46.38813785541571,
+      "learning_rate": 8.077210551112551e-06,
+      "loss": 0.5214,
+      "step": 3852
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.002636190258081,
+      "learning_rate": 8.076182060660686e-06,
+      "loss": 0.4744,
+      "step": 3853
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.1971002675502875,
+      "learning_rate": 8.075153360734424e-06,
+      "loss": 0.4331,
+      "step": 3854
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.956266469653313,
+      "learning_rate": 8.074124451403817e-06,
+      "loss": 0.4733,
+      "step": 3855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.470501910525475,
+      "learning_rate": 8.073095332738925e-06,
+      "loss": 0.4712,
+      "step": 3856
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.80384875795671,
+      "learning_rate": 8.07206600480983e-06,
+      "loss": 0.4725,
+      "step": 3857
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.885941758541443,
+      "learning_rate": 8.071036467686622e-06,
+      "loss": 0.6276,
+      "step": 3858
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.986328408382248,
+      "learning_rate": 8.07000672143941e-06,
+      "loss": 0.5302,
+      "step": 3859
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.684672289408606,
+      "learning_rate": 8.068976766138314e-06,
+      "loss": 0.5208,
+      "step": 3860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.789887523209176,
+      "learning_rate": 8.067946601853469e-06,
+      "loss": 0.547,
+      "step": 3861
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.78654494461196,
+      "learning_rate": 8.066916228655026e-06,
+      "loss": 0.4983,
+      "step": 3862
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.521772459776451,
+      "learning_rate": 8.065885646613147e-06,
+      "loss": 0.4547,
+      "step": 3863
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.9791730854135245,
+      "learning_rate": 8.06485485579801e-06,
+      "loss": 0.5907,
+      "step": 3864
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.074436492283673,
+      "learning_rate": 8.063823856279811e-06,
+      "loss": 0.4603,
+      "step": 3865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.61996364351894,
+      "learning_rate": 8.062792648128752e-06,
+      "loss": 0.5211,
+      "step": 3866
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.134912078520959,
+      "learning_rate": 8.061761231415057e-06,
+      "loss": 0.4453,
+      "step": 3867
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.743104409663411,
+      "learning_rate": 8.06072960620896e-06,
+      "loss": 0.4488,
+      "step": 3868
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.874757725230186,
+      "learning_rate": 8.05969777258071e-06,
+      "loss": 0.4196,
+      "step": 3869
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.008424313677871,
+      "learning_rate": 8.058665730600568e-06,
+      "loss": 0.6025,
+      "step": 3870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.129222267355131,
+      "learning_rate": 8.057633480338814e-06,
+      "loss": 0.4922,
+      "step": 3871
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.804914650023312,
+      "learning_rate": 8.05660102186574e-06,
+      "loss": 0.3867,
+      "step": 3872
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.596853540009678,
+      "learning_rate": 8.055568355251651e-06,
+      "loss": 0.5911,
+      "step": 3873
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.816596069408019,
+      "learning_rate": 8.054535480566865e-06,
+      "loss": 0.4217,
+      "step": 3874
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.449870916947366,
+      "learning_rate": 8.05350239788172e-06,
+      "loss": 0.449,
+      "step": 3875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.52391617174595,
+      "learning_rate": 8.052469107266563e-06,
+      "loss": 0.3967,
+      "step": 3876
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.55536494434901,
+      "learning_rate": 8.051435608791756e-06,
+      "loss": 0.5584,
+      "step": 3877
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.781182972869908,
+      "learning_rate": 8.050401902527676e-06,
+      "loss": 0.5598,
+      "step": 3878
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.577569815061196,
+      "learning_rate": 8.049367988544713e-06,
+      "loss": 0.4541,
+      "step": 3879
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.38582868487762,
+      "learning_rate": 8.048333866913274e-06,
+      "loss": 0.5308,
+      "step": 3880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 17.370940138550367,
+      "learning_rate": 8.047299537703775e-06,
+      "loss": 0.5007,
+      "step": 3881
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.964425405641547,
+      "learning_rate": 8.046265000986652e-06,
+      "loss": 0.5616,
+      "step": 3882
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.27192544617393,
+      "learning_rate": 8.04523025683235e-06,
+      "loss": 0.4724,
+      "step": 3883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.493789933355774,
+      "learning_rate": 8.044195305311333e-06,
+      "loss": 0.4978,
+      "step": 3884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.26266538424539,
+      "learning_rate": 8.043160146494076e-06,
+      "loss": 0.5901,
+      "step": 3885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.2151626654502605,
+      "learning_rate": 8.042124780451066e-06,
+      "loss": 0.4782,
+      "step": 3886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0646097646656547,
+      "learning_rate": 8.041089207252811e-06,
+      "loss": 0.8396,
+      "step": 3887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.410889958676457,
+      "learning_rate": 8.040053426969828e-06,
+      "loss": 0.5872,
+      "step": 3888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.547080994910914,
+      "learning_rate": 8.039017439672647e-06,
+      "loss": 0.4745,
+      "step": 3889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.430632869905207,
+      "learning_rate": 8.037981245431814e-06,
+      "loss": 0.4999,
+      "step": 3890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 36.02502095071812,
+      "learning_rate": 8.036944844317893e-06,
+      "loss": 0.4477,
+      "step": 3891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.586675108165876,
+      "learning_rate": 8.035908236401453e-06,
+      "loss": 0.5451,
+      "step": 3892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.857979214443661,
+      "learning_rate": 8.034871421753087e-06,
+      "loss": 0.5479,
+      "step": 3893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.871935907470284,
+      "learning_rate": 8.033834400443395e-06,
+      "loss": 0.5264,
+      "step": 3894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.499308234983453,
+      "learning_rate": 8.032797172542994e-06,
+      "loss": 0.449,
+      "step": 3895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.109686373362138,
+      "learning_rate": 8.031759738122516e-06,
+      "loss": 0.5217,
+      "step": 3896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.155792707835912,
+      "learning_rate": 8.030722097252604e-06,
+      "loss": 0.4488,
+      "step": 3897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.6546467337753805,
+      "learning_rate": 8.029684250003917e-06,
+      "loss": 0.5735,
+      "step": 3898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.584925511883239,
+      "learning_rate": 8.02864619644713e-06,
+      "loss": 0.6404,
+      "step": 3899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.083250692290925,
+      "learning_rate": 8.027607936652926e-06,
+      "loss": 0.5425,
+      "step": 3900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.526489117063721,
+      "learning_rate": 8.02656947069201e-06,
+      "loss": 0.7235,
+      "step": 3901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.854619095526233,
+      "learning_rate": 8.025530798635092e-06,
+      "loss": 0.4648,
+      "step": 3902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.73222358758858,
+      "learning_rate": 8.024491920552906e-06,
+      "loss": 0.4121,
+      "step": 3903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.673481497754393,
+      "learning_rate": 8.023452836516192e-06,
+      "loss": 0.4616,
+      "step": 3904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.629500184935063,
+      "learning_rate": 8.02241354659571e-06,
+      "loss": 0.5059,
+      "step": 3905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.88757773016436,
+      "learning_rate": 8.021374050862226e-06,
+      "loss": 0.5425,
+      "step": 3906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 14.193504011137357,
+      "learning_rate": 8.02033434938653e-06,
+      "loss": 0.4008,
+      "step": 3907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.366111508909193,
+      "learning_rate": 8.019294442239419e-06,
+      "loss": 0.5351,
+      "step": 3908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.660144164999226,
+      "learning_rate": 8.018254329491703e-06,
+      "loss": 0.4854,
+      "step": 3909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.344262989396654,
+      "learning_rate": 8.017214011214217e-06,
+      "loss": 0.5139,
+      "step": 3910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.674337727880567,
+      "learning_rate": 8.016173487477794e-06,
+      "loss": 0.5163,
+      "step": 3911
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.457075501537396,
+      "learning_rate": 8.015132758353292e-06,
+      "loss": 0.4031,
+      "step": 3912
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.570828478647574,
+      "learning_rate": 8.014091823911583e-06,
+      "loss": 0.5136,
+      "step": 3913
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.161432645387363,
+      "learning_rate": 8.013050684223546e-06,
+      "loss": 0.6258,
+      "step": 3914
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.303413837313135,
+      "learning_rate": 8.01200933936008e-06,
+      "loss": 0.4531,
+      "step": 3915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.886590813352778,
+      "learning_rate": 8.010967789392094e-06,
+      "loss": 0.8119,
+      "step": 3916
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.2673834748110995,
+      "learning_rate": 8.009926034390516e-06,
+      "loss": 0.4605,
+      "step": 3917
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.579817310399272,
+      "learning_rate": 8.008884074426285e-06,
+      "loss": 0.6035,
+      "step": 3918
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.29018348254262,
+      "learning_rate": 8.007841909570349e-06,
+      "loss": 0.5874,
+      "step": 3919
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 19.09301519888663,
+      "learning_rate": 8.00679953989368e-06,
+      "loss": 0.534,
+      "step": 3920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.928882800321041,
+      "learning_rate": 8.005756965467256e-06,
+      "loss": 0.5108,
+      "step": 3921
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.057051077685055,
+      "learning_rate": 8.004714186362071e-06,
+      "loss": 0.4465,
+      "step": 3922
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.515306311484553,
+      "learning_rate": 8.003671202649138e-06,
+      "loss": 0.5883,
+      "step": 3923
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.513495036836654,
+      "learning_rate": 8.002628014399474e-06,
+      "loss": 0.4883,
+      "step": 3924
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.95131560054958,
+      "learning_rate": 8.00158462168412e-06,
+      "loss": 0.4948,
+      "step": 3925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.216790726741248,
+      "learning_rate": 8.000541024574123e-06,
+      "loss": 0.7012,
+      "step": 3926
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.57738194675473,
+      "learning_rate": 7.99949722314055e-06,
+      "loss": 0.5038,
+      "step": 3927
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.145550710067029,
+      "learning_rate": 7.998453217454479e-06,
+      "loss": 0.5443,
+      "step": 3928
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.3246262573384,
+      "learning_rate": 7.997409007587e-06,
+      "loss": 0.5985,
+      "step": 3929
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.039074116668168,
+      "learning_rate": 7.996364593609223e-06,
+      "loss": 0.6284,
+      "step": 3930
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3673311477845225,
+      "learning_rate": 7.995319975592262e-06,
+      "loss": 0.8855,
+      "step": 3931
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.377983856103112,
+      "learning_rate": 7.994275153607256e-06,
+      "loss": 0.5157,
+      "step": 3932
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.989678133639987,
+      "learning_rate": 7.993230127725352e-06,
+      "loss": 0.5585,
+      "step": 3933
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.358722950971842,
+      "learning_rate": 7.99218489801771e-06,
+      "loss": 0.5029,
+      "step": 3934
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.870315116299098,
+      "learning_rate": 7.991139464555507e-06,
+      "loss": 0.5681,
+      "step": 3935
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.156288274113374,
+      "learning_rate": 7.990093827409929e-06,
+      "loss": 0.4498,
+      "step": 3936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.514431483722879,
+      "learning_rate": 7.989047986652184e-06,
+      "loss": 0.494,
+      "step": 3937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.004107946363983,
+      "learning_rate": 7.988001942353486e-06,
+      "loss": 0.4732,
+      "step": 3938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.094338542909673,
+      "learning_rate": 7.98695569458507e-06,
+      "loss": 0.6224,
+      "step": 3939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.127385835696428,
+      "learning_rate": 7.985909243418176e-06,
+      "loss": 0.4189,
+      "step": 3940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.359809002713606,
+      "learning_rate": 7.984862588924063e-06,
+      "loss": 0.591,
+      "step": 3941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.123440590322994,
+      "learning_rate": 7.983815731174008e-06,
+      "loss": 0.5647,
+      "step": 3942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.743936900414463,
+      "learning_rate": 7.982768670239292e-06,
+      "loss": 0.5176,
+      "step": 3943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.486755225657864,
+      "learning_rate": 7.981721406191219e-06,
+      "loss": 0.5391,
+      "step": 3944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.985461998876378,
+      "learning_rate": 7.9806739391011e-06,
+      "loss": 0.4114,
+      "step": 3945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.024364900592188,
+      "learning_rate": 7.979626269040266e-06,
+      "loss": 0.3882,
+      "step": 3946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8216151554491524,
+      "learning_rate": 7.978578396080058e-06,
+      "loss": 0.8309,
+      "step": 3947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.803114455660785,
+      "learning_rate": 7.977530320291831e-06,
+      "loss": 0.5752,
+      "step": 3948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.760751979270808,
+      "learning_rate": 7.976482041746954e-06,
+      "loss": 0.4648,
+      "step": 3949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 23.172106662409735,
+      "learning_rate": 7.97543356051681e-06,
+      "loss": 0.5147,
+      "step": 3950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.616606882441578,
+      "learning_rate": 7.974384876672796e-06,
+      "loss": 0.6073,
+      "step": 3951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.828959894398432,
+      "learning_rate": 7.973335990286323e-06,
+      "loss": 0.5582,
+      "step": 3952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.98676247797655,
+      "learning_rate": 7.972286901428817e-06,
+      "loss": 0.5652,
+      "step": 3953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.124220638296681,
+      "learning_rate": 7.971237610171714e-06,
+      "loss": 0.4919,
+      "step": 3954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.144399690556098,
+      "learning_rate": 7.970188116586466e-06,
+      "loss": 0.4906,
+      "step": 3955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.5237645099770365,
+      "learning_rate": 7.96913842074454e-06,
+      "loss": 0.4892,
+      "step": 3956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.15310260051625,
+      "learning_rate": 7.968088522717416e-06,
+      "loss": 0.6691,
+      "step": 3957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.367862550110012,
+      "learning_rate": 7.967038422576588e-06,
+      "loss": 0.394,
+      "step": 3958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.661982304184752,
+      "learning_rate": 7.96598812039356e-06,
+      "loss": 0.5207,
+      "step": 3959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.647905428368102,
+      "learning_rate": 7.964937616239856e-06,
+      "loss": 0.5351,
+      "step": 3960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 17.29137315985265,
+      "learning_rate": 7.963886910187009e-06,
+      "loss": 0.5244,
+      "step": 3961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.244985470314943,
+      "learning_rate": 7.962836002306569e-06,
+      "loss": 0.6009,
+      "step": 3962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.301216971571447,
+      "learning_rate": 7.961784892670097e-06,
+      "loss": 0.4949,
+      "step": 3963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.804527868539367,
+      "learning_rate": 7.96073358134917e-06,
+      "loss": 0.5229,
+      "step": 3964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.652370652889273,
+      "learning_rate": 7.959682068415375e-06,
+      "loss": 0.5091,
+      "step": 3965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.451577659933083,
+      "learning_rate": 7.958630353940317e-06,
+      "loss": 0.4742,
+      "step": 3966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.528482316705981,
+      "learning_rate": 7.957578437995614e-06,
+      "loss": 0.4061,
+      "step": 3967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.7362894535017,
+      "learning_rate": 7.956526320652898e-06,
+      "loss": 0.4564,
+      "step": 3968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.0861243685822926,
+      "learning_rate": 7.955474001983811e-06,
+      "loss": 0.525,
+      "step": 3969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.676504093826561,
+      "learning_rate": 7.95442148206001e-06,
+      "loss": 0.6043,
+      "step": 3970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.382924222070004,
+      "learning_rate": 7.953368760953171e-06,
+      "loss": 0.5735,
+      "step": 3971
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.711745310943468,
+      "learning_rate": 7.952315838734978e-06,
+      "loss": 0.55,
+      "step": 3972
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.888880502179827,
+      "learning_rate": 7.951262715477129e-06,
+      "loss": 0.5977,
+      "step": 3973
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.5534305789950515,
+      "learning_rate": 7.950209391251337e-06,
+      "loss": 0.4148,
+      "step": 3974
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.857711878005167,
+      "learning_rate": 7.94915586612933e-06,
+      "loss": 0.5569,
+      "step": 3975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.875294870536916,
+      "learning_rate": 7.94810214018285e-06,
+      "loss": 0.5925,
+      "step": 3976
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.970570359776238,
+      "learning_rate": 7.947048213483646e-06,
+      "loss": 0.4333,
+      "step": 3977
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.808448522781344,
+      "learning_rate": 7.945994086103493e-06,
+      "loss": 0.4383,
+      "step": 3978
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.42042059896875,
+      "learning_rate": 7.944939758114165e-06,
+      "loss": 0.3681,
+      "step": 3979
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.293698003483842,
+      "learning_rate": 7.943885229587461e-06,
+      "loss": 0.5105,
+      "step": 3980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.461543661568397,
+      "learning_rate": 7.94283050059519e-06,
+      "loss": 0.4578,
+      "step": 3981
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 22.52063950650416,
+      "learning_rate": 7.941775571209174e-06,
+      "loss": 0.5,
+      "step": 3982
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.88524579206497,
+      "learning_rate": 7.940720441501246e-06,
+      "loss": 0.4676,
+      "step": 3983
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.731800786741002,
+      "learning_rate": 7.93966511154326e-06,
+      "loss": 0.4938,
+      "step": 3984
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.183219170614759,
+      "learning_rate": 7.938609581407078e-06,
+      "loss": 0.4025,
+      "step": 3985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.4099593941144075,
+      "learning_rate": 7.937553851164574e-06,
+      "loss": 0.5397,
+      "step": 3986
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.201500438299795,
+      "learning_rate": 7.936497920887643e-06,
+      "loss": 0.579,
+      "step": 3987
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.032947115262496,
+      "learning_rate": 7.935441790648189e-06,
+      "loss": 0.5809,
+      "step": 3988
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.5636997745326955,
+      "learning_rate": 7.934385460518124e-06,
+      "loss": 0.5588,
+      "step": 3989
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.951485557249397,
+      "learning_rate": 7.933328930569385e-06,
+      "loss": 0.5579,
+      "step": 3990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6611784509853988,
+      "learning_rate": 7.932272200873916e-06,
+      "loss": 0.7792,
+      "step": 3991
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.322316905486785,
+      "learning_rate": 7.931215271503676e-06,
+      "loss": 0.5223,
+      "step": 3992
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.54028691107987,
+      "learning_rate": 7.930158142530635e-06,
+      "loss": 0.4867,
+      "step": 3993
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.337896543309396,
+      "learning_rate": 7.92910081402678e-06,
+      "loss": 0.465,
+      "step": 3994
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.277721834378312,
+      "learning_rate": 7.92804328606411e-06,
+      "loss": 0.4629,
+      "step": 3995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.527317076969615,
+      "learning_rate": 7.92698555871464e-06,
+      "loss": 0.4781,
+      "step": 3996
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.862950673782374,
+      "learning_rate": 7.925927632050392e-06,
+      "loss": 0.5218,
+      "step": 3997
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.483794464704518,
+      "learning_rate": 7.92486950614341e-06,
+      "loss": 0.613,
+      "step": 3998
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.479423966726486,
+      "learning_rate": 7.923811181065751e-06,
+      "loss": 0.5792,
+      "step": 3999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.955369990504279,
+      "learning_rate": 7.922752656889474e-06,
+      "loss": 0.5544,
+      "step": 4000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.255767029654107,
+      "learning_rate": 7.921693933686666e-06,
+      "loss": 0.5182,
+      "step": 4001
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.075625718978498,
+      "learning_rate": 7.920635011529417e-06,
+      "loss": 0.428,
+      "step": 4002
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.013643288883886,
+      "learning_rate": 7.919575890489838e-06,
+      "loss": 0.5214,
+      "step": 4003
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.750525596312109,
+      "learning_rate": 7.91851657064005e-06,
+      "loss": 0.4933,
+      "step": 4004
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.816204773289573,
+      "learning_rate": 7.917457052052186e-06,
+      "loss": 0.4169,
+      "step": 4005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 16.55120692338697,
+      "learning_rate": 7.916397334798398e-06,
+      "loss": 0.5884,
+      "step": 4006
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.291359614473228,
+      "learning_rate": 7.915337418950847e-06,
+      "loss": 0.5672,
+      "step": 4007
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 50.030462805857645,
+      "learning_rate": 7.914277304581706e-06,
+      "loss": 0.4255,
+      "step": 4008
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.239718802197119,
+      "learning_rate": 7.913216991763168e-06,
+      "loss": 0.435,
+      "step": 4009
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.993208880303607,
+      "learning_rate": 7.912156480567432e-06,
+      "loss": 0.5269,
+      "step": 4010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.843723592088379,
+      "learning_rate": 7.911095771066715e-06,
+      "loss": 0.5373,
+      "step": 4011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.642640896913838,
+      "learning_rate": 7.91003486333325e-06,
+      "loss": 0.528,
+      "step": 4012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.265632917075285,
+      "learning_rate": 7.908973757439275e-06,
+      "loss": 0.4987,
+      "step": 4013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.232361448832585,
+      "learning_rate": 7.907912453457049e-06,
+      "loss": 0.4957,
+      "step": 4014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.56478163162116,
+      "learning_rate": 7.906850951458842e-06,
+      "loss": 0.4461,
+      "step": 4015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.230839133384235,
+      "learning_rate": 7.905789251516938e-06,
+      "loss": 0.5249,
+      "step": 4016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.650211583097345,
+      "learning_rate": 7.904727353703633e-06,
+      "loss": 0.4598,
+      "step": 4017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.056598121267331,
+      "learning_rate": 7.903665258091242e-06,
+      "loss": 0.5091,
+      "step": 4018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.321010838243698,
+      "learning_rate": 7.902602964752082e-06,
+      "loss": 0.4004,
+      "step": 4019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.612424483195902,
+      "learning_rate": 7.901540473758494e-06,
+      "loss": 0.463,
+      "step": 4020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.284924905942143,
+      "learning_rate": 7.900477785182829e-06,
+      "loss": 0.4931,
+      "step": 4021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.818462912570032,
+      "learning_rate": 7.899414899097453e-06,
+      "loss": 0.5853,
+      "step": 4022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 55.85670993735534,
+      "learning_rate": 7.898351815574741e-06,
+      "loss": 0.5492,
+      "step": 4023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5815363715684896,
+      "learning_rate": 7.897288534687085e-06,
+      "loss": 0.7336,
+      "step": 4024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.833652550062547,
+      "learning_rate": 7.89622505650689e-06,
+      "loss": 0.5431,
+      "step": 4025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.6960175835737505,
+      "learning_rate": 7.895161381106573e-06,
+      "loss": 0.5912,
+      "step": 4026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.53141831708279,
+      "learning_rate": 7.89409750855857e-06,
+      "loss": 0.4802,
+      "step": 4027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.039679446420092,
+      "learning_rate": 7.893033438935318e-06,
+      "loss": 0.518,
+      "step": 4028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.922372724445731,
+      "learning_rate": 7.891969172309283e-06,
+      "loss": 0.5844,
+      "step": 4029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.390442817721276,
+      "learning_rate": 7.890904708752935e-06,
+      "loss": 0.5788,
+      "step": 4030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.734402952746699,
+      "learning_rate": 7.889840048338757e-06,
+      "loss": 0.4801,
+      "step": 4031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.869221054491615,
+      "learning_rate": 7.88877519113925e-06,
+      "loss": 0.5601,
+      "step": 4032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.347217211483554,
+      "learning_rate": 7.887710137226925e-06,
+      "loss": 0.5324,
+      "step": 4033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.816274708194143,
+      "learning_rate": 7.886644886674306e-06,
+      "loss": 0.5438,
+      "step": 4034
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.559652271118928,
+      "learning_rate": 7.885579439553935e-06,
+      "loss": 0.761,
+      "step": 4035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.654591973942951,
+      "learning_rate": 7.884513795938362e-06,
+      "loss": 0.5018,
+      "step": 4036
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.813458021160164,
+      "learning_rate": 7.883447955900152e-06,
+      "loss": 0.5338,
+      "step": 4037
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.034251086589538,
+      "learning_rate": 7.882381919511889e-06,
+      "loss": 0.4879,
+      "step": 4038
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 22.807917390242228,
+      "learning_rate": 7.881315686846159e-06,
+      "loss": 0.4763,
+      "step": 4039
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 39.14182172292222,
+      "learning_rate": 7.88024925797557e-06,
+      "loss": 0.4403,
+      "step": 4040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.679977555245296,
+      "learning_rate": 7.879182632972742e-06,
+      "loss": 0.4705,
+      "step": 4041
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.082376604161601,
+      "learning_rate": 7.878115811910307e-06,
+      "loss": 0.3871,
+      "step": 4042
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.590363766992771,
+      "learning_rate": 7.87704879486091e-06,
+      "loss": 0.4747,
+      "step": 4043
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.645774282370095,
+      "learning_rate": 7.875981581897212e-06,
+      "loss": 0.6234,
+      "step": 4044
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.748370271484157,
+      "learning_rate": 7.874914173091884e-06,
+      "loss": 0.4048,
+      "step": 4045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.61668506861456,
+      "learning_rate": 7.87384656851761e-06,
+      "loss": 0.5661,
+      "step": 4046
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.136219313250537,
+      "learning_rate": 7.872778768247094e-06,
+      "loss": 0.6016,
+      "step": 4047
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.775316700822403,
+      "learning_rate": 7.871710772353044e-06,
+      "loss": 0.6213,
+      "step": 4048
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.998901170192656,
+      "learning_rate": 7.870642580908189e-06,
+      "loss": 0.6395,
+      "step": 4049
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.04024028605908,
+      "learning_rate": 7.869574193985265e-06,
+      "loss": 0.5003,
+      "step": 4050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.421503754531498,
+      "learning_rate": 7.868505611657028e-06,
+      "loss": 0.3868,
+      "step": 4051
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.5228920941743604,
+      "learning_rate": 7.86743683399624e-06,
+      "loss": 0.5297,
+      "step": 4052
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.1290351076597345,
+      "learning_rate": 7.866367861075684e-06,
+      "loss": 0.5526,
+      "step": 4053
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.013353906420884,
+      "learning_rate": 7.865298692968149e-06,
+      "loss": 0.5216,
+      "step": 4054
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.289765302812977,
+      "learning_rate": 7.864229329746441e-06,
+      "loss": 0.5509,
+      "step": 4055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.208574852492193,
+      "learning_rate": 7.863159771483379e-06,
+      "loss": 0.4686,
+      "step": 4056
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.987074403687751,
+      "learning_rate": 7.862090018251798e-06,
+      "loss": 0.441,
+      "step": 4057
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.814340947554589,
+      "learning_rate": 7.861020070124541e-06,
+      "loss": 0.4135,
+      "step": 4058
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.104303099301054,
+      "learning_rate": 7.859949927174468e-06,
+      "loss": 0.604,
+      "step": 4059
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.59682304837195,
+      "learning_rate": 7.858879589474448e-06,
+      "loss": 0.5248,
+      "step": 4060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7155359791621567,
+      "learning_rate": 7.85780905709737e-06,
+      "loss": 0.8268,
+      "step": 4061
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.152943311567023,
+      "learning_rate": 7.856738330116129e-06,
+      "loss": 0.4715,
+      "step": 4062
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.224237565643305,
+      "learning_rate": 7.85566740860364e-06,
+      "loss": 0.441,
+      "step": 4063
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.113718099620097,
+      "learning_rate": 7.854596292632827e-06,
+      "loss": 0.6182,
+      "step": 4064
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1275502749615893,
+      "learning_rate": 7.853524982276629e-06,
+      "loss": 0.7726,
+      "step": 4065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.118048141093473,
+      "learning_rate": 7.852453477607994e-06,
+      "loss": 0.5453,
+      "step": 4066
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.649475173933046,
+      "learning_rate": 7.85138177869989e-06,
+      "loss": 0.7458,
+      "step": 4067
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.764469088349507,
+      "learning_rate": 7.850309885625294e-06,
+      "loss": 0.5722,
+      "step": 4068
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 19.098151514638666,
+      "learning_rate": 7.849237798457197e-06,
+      "loss": 0.4035,
+      "step": 4069
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.169926212989057,
+      "learning_rate": 7.848165517268604e-06,
+      "loss": 0.4909,
+      "step": 4070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.129286025329911,
+      "learning_rate": 7.847093042132532e-06,
+      "loss": 0.4511,
+      "step": 4071
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 41.81628330847653,
+      "learning_rate": 7.846020373122013e-06,
+      "loss": 0.4664,
+      "step": 4072
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 32.93182738965246,
+      "learning_rate": 7.84494751031009e-06,
+      "loss": 0.4909,
+      "step": 4073
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.590189083453147,
+      "learning_rate": 7.843874453769819e-06,
+      "loss": 0.5281,
+      "step": 4074
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.120488096993062,
+      "learning_rate": 7.842801203574274e-06,
+      "loss": 0.4001,
+      "step": 4075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.450757972471422,
+      "learning_rate": 7.841727759796536e-06,
+      "loss": 0.5476,
+      "step": 4076
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.29443331608717,
+      "learning_rate": 7.840654122509699e-06,
+      "loss": 0.5162,
+      "step": 4077
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.960228205166006,
+      "learning_rate": 7.83958029178688e-06,
+      "loss": 0.4571,
+      "step": 4078
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.827014633771618,
+      "learning_rate": 7.838506267701196e-06,
+      "loss": 0.6167,
+      "step": 4079
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 18.62373881621914,
+      "learning_rate": 7.837432050325788e-06,
+      "loss": 0.4823,
+      "step": 4080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.580411789939719,
+      "learning_rate": 7.836357639733801e-06,
+      "loss": 0.5364,
+      "step": 4081
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.367999712369837,
+      "learning_rate": 7.835283035998398e-06,
+      "loss": 0.5303,
+      "step": 4082
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.771701591139204,
+      "learning_rate": 7.83420823919276e-06,
+      "loss": 0.484,
+      "step": 4083
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.22545591517937,
+      "learning_rate": 7.83313324939007e-06,
+      "loss": 0.5436,
+      "step": 4084
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.998131684584878,
+      "learning_rate": 7.832058066663532e-06,
+      "loss": 0.5253,
+      "step": 4085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.755717993127089,
+      "learning_rate": 7.830982691086361e-06,
+      "loss": 0.4722,
+      "step": 4086
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.176136970486652,
+      "learning_rate": 7.829907122731788e-06,
+      "loss": 0.4451,
+      "step": 4087
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.16528145876642,
+      "learning_rate": 7.82883136167305e-06,
+      "loss": 0.5318,
+      "step": 4088
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.158383426866603,
+      "learning_rate": 7.827755407983406e-06,
+      "loss": 0.5864,
+      "step": 4089
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.282629374474652,
+      "learning_rate": 7.826679261736122e-06,
+      "loss": 0.4636,
+      "step": 4090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.027614757501738,
+      "learning_rate": 7.825602923004476e-06,
+      "loss": 0.479,
+      "step": 4091
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 31.604605485942646,
+      "learning_rate": 7.824526391861764e-06,
+      "loss": 0.4201,
+      "step": 4092
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.585377358630415,
+      "learning_rate": 7.823449668381296e-06,
+      "loss": 0.491,
+      "step": 4093
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.106730044017482,
+      "learning_rate": 7.822372752636387e-06,
+      "loss": 0.4066,
+      "step": 4094
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.97693307911921,
+      "learning_rate": 7.821295644700373e-06,
+      "loss": 0.4556,
+      "step": 4095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.62639041951133,
+      "learning_rate": 7.8202183446466e-06,
+      "loss": 0.499,
+      "step": 4096
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.372236884629658,
+      "learning_rate": 7.819140852548426e-06,
+      "loss": 0.483,
+      "step": 4097
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.26314942191242,
+      "learning_rate": 7.818063168479227e-06,
+      "loss": 0.3725,
+      "step": 4098
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.05467728005092,
+      "learning_rate": 7.816985292512385e-06,
+      "loss": 0.536,
+      "step": 4099
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.953104967211745,
+      "learning_rate": 7.8159072247213e-06,
+      "loss": 0.4966,
+      "step": 4100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.9185168374587995,
+      "learning_rate": 7.814828965179383e-06,
+      "loss": 0.5315,
+      "step": 4101
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 20.756413468181584,
+      "learning_rate": 7.81375051396006e-06,
+      "loss": 0.4798,
+      "step": 4102
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 30.70826941336905,
+      "learning_rate": 7.812671871136767e-06,
+      "loss": 0.4435,
+      "step": 4103
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.872270021722912,
+      "learning_rate": 7.811593036782957e-06,
+      "loss": 0.4341,
+      "step": 4104
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.324042394901413,
+      "learning_rate": 7.81051401097209e-06,
+      "loss": 0.467,
+      "step": 4105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.506704442665612,
+      "learning_rate": 7.809434793777646e-06,
+      "loss": 0.4529,
+      "step": 4106
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 26.73458026068241,
+      "learning_rate": 7.808355385273116e-06,
+      "loss": 0.5872,
+      "step": 4107
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 33.82780836969727,
+      "learning_rate": 7.807275785532e-06,
+      "loss": 0.4811,
+      "step": 4108
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.578766126528265,
+      "learning_rate": 7.806195994627816e-06,
+      "loss": 0.4709,
+      "step": 4109
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.397385925523029,
+      "learning_rate": 7.80511601263409e-06,
+      "loss": 0.4937,
+      "step": 4110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.801113458318113,
+      "learning_rate": 7.80403583962437e-06,
+      "loss": 0.4458,
+      "step": 4111
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 37.30204596190142,
+      "learning_rate": 7.802955475672205e-06,
+      "loss": 0.4468,
+      "step": 4112
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.71611553798248,
+      "learning_rate": 7.801874920851167e-06,
+      "loss": 0.4067,
+      "step": 4113
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 28.848263901070304,
+      "learning_rate": 7.800794175234834e-06,
+      "loss": 0.5548,
+      "step": 4114
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.70812416716223,
+      "learning_rate": 7.7997132388968e-06,
+      "loss": 0.6159,
+      "step": 4115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 29.028067628308094,
+      "learning_rate": 7.798632111910673e-06,
+      "loss": 0.3749,
+      "step": 4116
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.019728165677641,
+      "learning_rate": 7.797550794350074e-06,
+      "loss": 0.4818,
+      "step": 4117
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.05714845530537,
+      "learning_rate": 7.796469286288638e-06,
+      "loss": 0.4847,
+      "step": 4118
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.443890306090521,
+      "learning_rate": 7.795387587800006e-06,
+      "loss": 0.4673,
+      "step": 4119
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.285562278948357,
+      "learning_rate": 7.794305698957839e-06,
+      "loss": 0.575,
+      "step": 4120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.126943275783354,
+      "learning_rate": 7.793223619835808e-06,
+      "loss": 0.5086,
+      "step": 4121
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 15.546062330425341,
+      "learning_rate": 7.792141350507599e-06,
+      "loss": 0.4975,
+      "step": 4122
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.989097946592583,
+      "learning_rate": 7.79105889104691e-06,
+      "loss": 0.5293,
+      "step": 4123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.560794939402738,
+      "learning_rate": 7.789976241527451e-06,
+      "loss": 0.4398,
+      "step": 4124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 19.00385529604857,
+      "learning_rate": 7.788893402022945e-06,
+      "loss": 0.5234,
+      "step": 4125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.476586111562897,
+      "learning_rate": 7.78781037260713e-06,
+      "loss": 0.5019,
+      "step": 4126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.034874293233091,
+      "learning_rate": 7.786727153353755e-06,
+      "loss": 0.4967,
+      "step": 4127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.512922496792179,
+      "learning_rate": 7.785643744336582e-06,
+      "loss": 0.4878,
+      "step": 4128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8906136393541029,
+      "learning_rate": 7.784560145629386e-06,
+      "loss": 0.7613,
+      "step": 4129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.01707046968446,
+      "learning_rate": 7.783476357305957e-06,
+      "loss": 0.49,
+      "step": 4130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.288975573710422,
+      "learning_rate": 7.782392379440093e-06,
+      "loss": 0.4364,
+      "step": 4131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 22.54142923610065,
+      "learning_rate": 7.781308212105612e-06,
+      "loss": 0.5651,
+      "step": 4132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.12493014898308,
+      "learning_rate": 7.780223855376339e-06,
+      "loss": 0.4098,
+      "step": 4133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.993690956480332,
+      "learning_rate": 7.779139309326114e-06,
+      "loss": 0.625,
+      "step": 4134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.333336235415057,
+      "learning_rate": 7.778054574028789e-06,
+      "loss": 0.3681,
+      "step": 4135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.4191371920105444,
+      "learning_rate": 7.776969649558231e-06,
+      "loss": 0.4858,
+      "step": 4136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.277046285516569,
+      "learning_rate": 7.775884535988318e-06,
+      "loss": 0.4848,
+      "step": 4137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.488051001779276,
+      "learning_rate": 7.774799233392942e-06,
+      "loss": 0.581,
+      "step": 4138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.16830281396762,
+      "learning_rate": 7.773713741846008e-06,
+      "loss": 0.4365,
+      "step": 4139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 14.486997654652415,
+      "learning_rate": 7.77262806142143e-06,
+      "loss": 0.5226,
+      "step": 4140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.30593589241496,
+      "learning_rate": 7.771542192193141e-06,
+      "loss": 0.6014,
+      "step": 4141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.1792598454956345,
+      "learning_rate": 7.770456134235082e-06,
+      "loss": 0.5905,
+      "step": 4142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.381489648634956,
+      "learning_rate": 7.769369887621212e-06,
+      "loss": 0.4651,
+      "step": 4143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.8725190332027815,
+      "learning_rate": 7.768283452425496e-06,
+      "loss": 0.4682,
+      "step": 4144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.7155804864656465,
+      "learning_rate": 7.767196828721918e-06,
+      "loss": 0.3894,
+      "step": 4145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.430447133698429,
+      "learning_rate": 7.766110016584467e-06,
+      "loss": 0.406,
+      "step": 4146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.243036877699527,
+      "learning_rate": 7.765023016087157e-06,
+      "loss": 0.5178,
+      "step": 4147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.3078092265118375,
+      "learning_rate": 7.763935827304004e-06,
+      "loss": 0.4361,
+      "step": 4148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.248861251833672,
+      "learning_rate": 7.762848450309043e-06,
+      "loss": 0.5336,
+      "step": 4149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 16.24101055554693,
+      "learning_rate": 7.761760885176318e-06,
+      "loss": 0.475,
+      "step": 4150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.986522727886057,
+      "learning_rate": 7.760673131979888e-06,
+      "loss": 0.5732,
+      "step": 4151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.974671406071769,
+      "learning_rate": 7.759585190793824e-06,
+      "loss": 0.7366,
+      "step": 4152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.384951286542625,
+      "learning_rate": 7.75849706169221e-06,
+      "loss": 0.5077,
+      "step": 4153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.045256103171747,
+      "learning_rate": 7.757408744749141e-06,
+      "loss": 0.4344,
+      "step": 4154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.008784478087392,
+      "learning_rate": 7.756320240038731e-06,
+      "loss": 0.4715,
+      "step": 4155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.920619558322575,
+      "learning_rate": 7.755231547635096e-06,
+      "loss": 0.5035,
+      "step": 4156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.396004151249258,
+      "learning_rate": 7.754142667612376e-06,
+      "loss": 0.4395,
+      "step": 4157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.61078897906087,
+      "learning_rate": 7.75305360004472e-06,
+      "loss": 0.5097,
+      "step": 4158
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.613771151901615,
+      "learning_rate": 7.751964345006284e-06,
+      "loss": 0.4996,
+      "step": 4159
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.128146424537801,
+      "learning_rate": 7.750874902571245e-06,
+      "loss": 0.5214,
+      "step": 4160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.687551105301235,
+      "learning_rate": 7.749785272813788e-06,
+      "loss": 0.5317,
+      "step": 4161
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.180030695682657,
+      "learning_rate": 7.748695455808111e-06,
+      "loss": 0.3698,
+      "step": 4162
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.5700009550155,
+      "learning_rate": 7.747605451628425e-06,
+      "loss": 0.6512,
+      "step": 4163
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.219121862340644,
+      "learning_rate": 7.74651526034896e-06,
+      "loss": 0.5203,
+      "step": 4164
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.07398652099986,
+      "learning_rate": 7.745424882043949e-06,
+      "loss": 0.569,
+      "step": 4165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.18110061683274,
+      "learning_rate": 7.74433431678764e-06,
+      "loss": 0.4499,
+      "step": 4166
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.392514975237832,
+      "learning_rate": 7.743243564654299e-06,
+      "loss": 0.4112,
+      "step": 4167
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.788614458227595,
+      "learning_rate": 7.7421526257182e-06,
+      "loss": 0.4671,
+      "step": 4168
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.523072227919053,
+      "learning_rate": 7.741061500053633e-06,
+      "loss": 0.5879,
+      "step": 4169
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.844240806249585,
+      "learning_rate": 7.739970187734895e-06,
+      "loss": 0.61,
+      "step": 4170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.748562151373546,
+      "learning_rate": 7.738878688836303e-06,
+      "loss": 0.4256,
+      "step": 4171
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.40624155318709,
+      "learning_rate": 7.737787003432183e-06,
+      "loss": 0.4375,
+      "step": 4172
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 34.82906287412024,
+      "learning_rate": 7.736695131596871e-06,
+      "loss": 0.4469,
+      "step": 4173
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.109486549363996,
+      "learning_rate": 7.735603073404723e-06,
+      "loss": 0.6054,
+      "step": 4174
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.961908188160633,
+      "learning_rate": 7.7345108289301e-06,
+      "loss": 0.5806,
+      "step": 4175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.365822256838663,
+      "learning_rate": 7.73341839824738e-06,
+      "loss": 0.4885,
+      "step": 4176
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.516546832957653,
+      "learning_rate": 7.732325781430952e-06,
+      "loss": 0.5962,
+      "step": 4177
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.997178410596604,
+      "learning_rate": 7.731232978555221e-06,
+      "loss": 0.4913,
+      "step": 4178
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.708791515391263,
+      "learning_rate": 7.730139989694598e-06,
+      "loss": 0.4404,
+      "step": 4179
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.051382098560548,
+      "learning_rate": 7.729046814923515e-06,
+      "loss": 0.4695,
+      "step": 4180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.55448443510757,
+      "learning_rate": 7.72795345431641e-06,
+      "loss": 0.4269,
+      "step": 4181
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 27.81342476220695,
+      "learning_rate": 7.726859907947735e-06,
+      "loss": 0.4917,
+      "step": 4182
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.570315431290458,
+      "learning_rate": 7.725766175891958e-06,
+      "loss": 0.7919,
+      "step": 4183
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.283761044704137,
+      "learning_rate": 7.724672258223556e-06,
+      "loss": 0.4888,
+      "step": 4184
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.032650025417697,
+      "learning_rate": 7.72357815501702e-06,
+      "loss": 0.4622,
+      "step": 4185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.386520426933671,
+      "learning_rate": 7.722483866346855e-06,
+      "loss": 0.4604,
+      "step": 4186
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.403053562455826,
+      "learning_rate": 7.721389392287574e-06,
+      "loss": 0.4595,
+      "step": 4187
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.906077693847882,
+      "learning_rate": 7.72029473291371e-06,
+      "loss": 0.3744,
+      "step": 4188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.044587182558708,
+      "learning_rate": 7.719199888299799e-06,
+      "loss": 0.5473,
+      "step": 4189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.58442060615517,
+      "learning_rate": 7.718104858520402e-06,
+      "loss": 0.413,
+      "step": 4190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.477573748416255,
+      "learning_rate": 7.717009643650081e-06,
+      "loss": 0.4243,
+      "step": 4191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.27899278977525,
+      "learning_rate": 7.715914243763416e-06,
+      "loss": 0.5064,
+      "step": 4192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.207778708413743,
+      "learning_rate": 7.714818658934999e-06,
+      "loss": 0.4611,
+      "step": 4193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.779518337344322,
+      "learning_rate": 7.713722889239437e-06,
+      "loss": 0.4608,
+      "step": 4194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.521810091827312,
+      "learning_rate": 7.712626934751343e-06,
+      "loss": 0.4727,
+      "step": 4195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.009899203704922,
+      "learning_rate": 7.71153079554535e-06,
+      "loss": 0.4555,
+      "step": 4196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.881533921952323,
+      "learning_rate": 7.710434471696095e-06,
+      "loss": 0.5165,
+      "step": 4197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.64465214155606,
+      "learning_rate": 7.70933796327824e-06,
+      "loss": 0.5171,
+      "step": 4198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.460973080087464,
+      "learning_rate": 7.708241270366448e-06,
+      "loss": 0.432,
+      "step": 4199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.447802812195214,
+      "learning_rate": 7.7071443930354e-06,
+      "loss": 0.4545,
+      "step": 4200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.949681682840498,
+      "learning_rate": 7.706047331359788e-06,
+      "loss": 0.3705,
+      "step": 4201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 28.54618552830326,
+      "learning_rate": 7.704950085414316e-06,
+      "loss": 0.4802,
+      "step": 4202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.696261060689764,
+      "learning_rate": 7.703852655273705e-06,
+      "loss": 0.5785,
+      "step": 4203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.19876927124139,
+      "learning_rate": 7.702755041012684e-06,
+      "loss": 0.5685,
+      "step": 4204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.449017937012093,
+      "learning_rate": 7.701657242705993e-06,
+      "loss": 0.4341,
+      "step": 4205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.290412176864393,
+      "learning_rate": 7.700559260428389e-06,
+      "loss": 0.5541,
+      "step": 4206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.250095801822074,
+      "learning_rate": 7.699461094254641e-06,
+      "loss": 0.5338,
+      "step": 4207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 41.84284630375101,
+      "learning_rate": 7.698362744259526e-06,
+      "loss": 0.4628,
+      "step": 4208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.997571001352428,
+      "learning_rate": 7.697264210517842e-06,
+      "loss": 0.5171,
+      "step": 4209
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.562414563769874,
+      "learning_rate": 7.696165493104388e-06,
+      "loss": 0.5672,
+      "step": 4210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.929304732502658,
+      "learning_rate": 7.695066592093988e-06,
+      "loss": 0.6478,
+      "step": 4211
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.685818548127033,
+      "learning_rate": 7.693967507561465e-06,
+      "loss": 0.488,
+      "step": 4212
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.394122000591842,
+      "learning_rate": 7.69286823958167e-06,
+      "loss": 0.5017,
+      "step": 4213
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.942400671078701,
+      "learning_rate": 7.691768788229452e-06,
+      "loss": 0.455,
+      "step": 4214
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.226506661701315,
+      "learning_rate": 7.690669153579682e-06,
+      "loss": 0.4896,
+      "step": 4215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.557666216545817,
+      "learning_rate": 7.689569335707238e-06,
+      "loss": 0.5028,
+      "step": 4216
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.644267144703523,
+      "learning_rate": 7.688469334687015e-06,
+      "loss": 0.4647,
+      "step": 4217
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.809099487700237,
+      "learning_rate": 7.687369150593917e-06,
+      "loss": 0.5396,
+      "step": 4218
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.26286537152816,
+      "learning_rate": 7.686268783502863e-06,
+      "loss": 0.5463,
+      "step": 4219
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.285377819277352,
+      "learning_rate": 7.685168233488782e-06,
+      "loss": 0.538,
+      "step": 4220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.84950856511389,
+      "learning_rate": 7.684067500626616e-06,
+      "loss": 0.4993,
+      "step": 4221
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.685988455421956,
+      "learning_rate": 7.68296658499132e-06,
+      "loss": 0.5308,
+      "step": 4222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.10978642769295,
+      "learning_rate": 7.681865486657863e-06,
+      "loss": 0.4879,
+      "step": 4223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8954976153027496,
+      "learning_rate": 7.680764205701224e-06,
+      "loss": 0.7546,
+      "step": 4224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 72.31599491670806,
+      "learning_rate": 7.679662742196396e-06,
+      "loss": 0.6367,
+      "step": 4225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.616344871845664,
+      "learning_rate": 7.678561096218382e-06,
+      "loss": 0.5059,
+      "step": 4226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 19.87462548191197,
+      "learning_rate": 7.677459267842201e-06,
+      "loss": 0.4728,
+      "step": 4227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.52287523061753,
+      "learning_rate": 7.676357257142881e-06,
+      "loss": 0.5123,
+      "step": 4228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.181406093170319,
+      "learning_rate": 7.675255064195468e-06,
+      "loss": 0.3942,
+      "step": 4229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.841841135593334,
+      "learning_rate": 7.674152689075011e-06,
+      "loss": 0.4928,
+      "step": 4230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.101010887368824,
+      "learning_rate": 7.673050131856581e-06,
+      "loss": 0.4203,
+      "step": 4231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.014293283825609,
+      "learning_rate": 7.671947392615257e-06,
+      "loss": 0.4991,
+      "step": 4232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.137278679877056,
+      "learning_rate": 7.670844471426126e-06,
+      "loss": 0.5702,
+      "step": 4233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.07998763334762,
+      "learning_rate": 7.6697413683643e-06,
+      "loss": 0.5019,
+      "step": 4234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.35450381196739,
+      "learning_rate": 7.668638083504887e-06,
+      "loss": 0.4591,
+      "step": 4235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.426466117577506,
+      "learning_rate": 7.667534616923022e-06,
+      "loss": 0.4874,
+      "step": 4236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.167959534169654,
+      "learning_rate": 7.666430968693842e-06,
+      "loss": 0.5933,
+      "step": 4237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.978717054070785,
+      "learning_rate": 7.665327138892505e-06,
+      "loss": 0.5059,
+      "step": 4238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 39.55488909034561,
+      "learning_rate": 7.664223127594173e-06,
+      "loss": 0.5812,
+      "step": 4239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.097147699404967,
+      "learning_rate": 7.663118934874025e-06,
+      "loss": 0.42,
+      "step": 4240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.681368122370163,
+      "learning_rate": 7.662014560807254e-06,
+      "loss": 0.5433,
+      "step": 4241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.899010870380614,
+      "learning_rate": 7.66091000546906e-06,
+      "loss": 0.5011,
+      "step": 4242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.90965457877598,
+      "learning_rate": 7.659805268934661e-06,
+      "loss": 0.4483,
+      "step": 4243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.887514260928086,
+      "learning_rate": 7.658700351279283e-06,
+      "loss": 0.484,
+      "step": 4244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.594159100933787,
+      "learning_rate": 7.657595252578166e-06,
+      "loss": 0.4355,
+      "step": 4245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.48597205458202,
+      "learning_rate": 7.656489972906565e-06,
+      "loss": 0.4707,
+      "step": 4246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.156153992432745,
+      "learning_rate": 7.655384512339739e-06,
+      "loss": 0.4973,
+      "step": 4247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 20.056273423106383,
+      "learning_rate": 7.65427887095297e-06,
+      "loss": 0.5595,
+      "step": 4248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 15.104060969158747,
+      "learning_rate": 7.653173048821545e-06,
+      "loss": 0.5909,
+      "step": 4249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 29.898289235847813,
+      "learning_rate": 7.652067046020768e-06,
+      "loss": 0.6308,
+      "step": 4250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.03597895680034,
+      "learning_rate": 7.650960862625949e-06,
+      "loss": 0.4578,
+      "step": 4251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.163209492961808,
+      "learning_rate": 7.649854498712417e-06,
+      "loss": 0.6232,
+      "step": 4252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.795689532071691,
+      "learning_rate": 7.648747954355512e-06,
+      "loss": 0.5506,
+      "step": 4253
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.923276779471346,
+      "learning_rate": 7.647641229630578e-06,
+      "loss": 0.5096,
+      "step": 4254
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.027229797647093,
+      "learning_rate": 7.646534324612987e-06,
+      "loss": 0.4852,
+      "step": 4255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 28.220470614709555,
+      "learning_rate": 7.645427239378108e-06,
+      "loss": 0.4792,
+      "step": 4256
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.424888129114958,
+      "learning_rate": 7.64431997400133e-06,
+      "loss": 0.5202,
+      "step": 4257
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.377871042539999,
+      "learning_rate": 7.643212528558055e-06,
+      "loss": 0.4962,
+      "step": 4258
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.429365969941628,
+      "learning_rate": 7.642104903123695e-06,
+      "loss": 0.4733,
+      "step": 4259
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.316196334736745,
+      "learning_rate": 7.64099709777367e-06,
+      "loss": 0.4564,
+      "step": 4260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.072168895647872,
+      "learning_rate": 7.63988911258342e-06,
+      "loss": 0.5628,
+      "step": 4261
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.669455830324384,
+      "learning_rate": 7.638780947628397e-06,
+      "loss": 0.4504,
+      "step": 4262
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.22614386250134,
+      "learning_rate": 7.637672602984056e-06,
+      "loss": 0.4699,
+      "step": 4263
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.515308408787831,
+      "learning_rate": 7.636564078725874e-06,
+      "loss": 0.4761,
+      "step": 4264
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 39.0856601989589,
+      "learning_rate": 7.635455374929334e-06,
+      "loss": 0.5055,
+      "step": 4265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.82918764380701,
+      "learning_rate": 7.634346491669937e-06,
+      "loss": 0.4219,
+      "step": 4266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.52684031067075,
+      "learning_rate": 7.63323742902319e-06,
+      "loss": 0.5843,
+      "step": 4267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.171103019934336,
+      "learning_rate": 7.632128187064618e-06,
+      "loss": 0.4717,
+      "step": 4268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.374989415523428,
+      "learning_rate": 7.631018765869755e-06,
+      "loss": 0.5828,
+      "step": 4269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7402183950171695,
+      "learning_rate": 7.629909165514147e-06,
+      "loss": 0.7206,
+      "step": 4270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.247277989582704,
+      "learning_rate": 7.628799386073352e-06,
+      "loss": 0.4826,
+      "step": 4271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.759381912722084,
+      "learning_rate": 7.627689427622942e-06,
+      "loss": 0.5528,
+      "step": 4272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.064720910175435,
+      "learning_rate": 7.626579290238501e-06,
+      "loss": 0.5734,
+      "step": 4273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.904542690635738,
+      "learning_rate": 7.625468973995622e-06,
+      "loss": 0.5217,
+      "step": 4274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 24.295133744699545,
+      "learning_rate": 7.624358478969915e-06,
+      "loss": 0.4458,
+      "step": 4275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.082194981305975,
+      "learning_rate": 7.623247805237e-06,
+      "loss": 0.3911,
+      "step": 4276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.793042503386358,
+      "learning_rate": 7.622136952872508e-06,
+      "loss": 0.4034,
+      "step": 4277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 20.80433090663261,
+      "learning_rate": 7.621025921952081e-06,
+      "loss": 0.5782,
+      "step": 4278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9131040887044455,
+      "learning_rate": 7.619914712551379e-06,
+      "loss": 0.78,
+      "step": 4279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.0887068168416,
+      "learning_rate": 7.618803324746069e-06,
+      "loss": 0.5544,
+      "step": 4280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.020143767169577,
+      "learning_rate": 7.617691758611831e-06,
+      "loss": 0.4454,
+      "step": 4281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 22.305181226236286,
+      "learning_rate": 7.616580014224359e-06,
+      "loss": 0.4999,
+      "step": 4282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.661859382241694,
+      "learning_rate": 7.615468091659357e-06,
+      "loss": 0.5062,
+      "step": 4283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.751614471862634,
+      "learning_rate": 7.614355990992542e-06,
+      "loss": 0.5105,
+      "step": 4284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.150183151741675,
+      "learning_rate": 7.613243712299645e-06,
+      "loss": 0.5404,
+      "step": 4285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 26.801880725062524,
+      "learning_rate": 7.612131255656403e-06,
+      "loss": 0.4929,
+      "step": 4286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.885959035004115,
+      "learning_rate": 7.611018621138573e-06,
+      "loss": 0.5896,
+      "step": 4287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.587015570157665,
+      "learning_rate": 7.609905808821918e-06,
+      "loss": 0.4949,
+      "step": 4288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.692157533592628,
+      "learning_rate": 7.608792818782218e-06,
+      "loss": 0.433,
+      "step": 4289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.860818401327679,
+      "learning_rate": 7.607679651095262e-06,
+      "loss": 0.3447,
+      "step": 4290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.310276919639533,
+      "learning_rate": 7.606566305836851e-06,
+      "loss": 0.4517,
+      "step": 4291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 65.0437180444407,
+      "learning_rate": 7.605452783082801e-06,
+      "loss": 0.55,
+      "step": 4292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.341090187002118,
+      "learning_rate": 7.6043390829089336e-06,
+      "loss": 0.4835,
+      "step": 4293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.885634918359541,
+      "learning_rate": 7.603225205391091e-06,
+      "loss": 0.502,
+      "step": 4294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.577533828200485,
+      "learning_rate": 7.6021111506051225e-06,
+      "loss": 0.5506,
+      "step": 4295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6116260701295977,
+      "learning_rate": 7.6009969186268906e-06,
+      "loss": 0.7806,
+      "step": 4296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.08652193334389,
+      "learning_rate": 7.599882509532267e-06,
+      "loss": 0.5351,
+      "step": 4297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.66762792965455,
+      "learning_rate": 7.598767923397139e-06,
+      "loss": 0.3985,
+      "step": 4298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.37197881850713,
+      "learning_rate": 7.5976531602974064e-06,
+      "loss": 0.5999,
+      "step": 4299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.74808895586608,
+      "learning_rate": 7.59653822030898e-06,
+      "loss": 0.5124,
+      "step": 4300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.839608126675706,
+      "learning_rate": 7.595423103507782e-06,
+      "loss": 0.7993,
+      "step": 4301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.083996869514742,
+      "learning_rate": 7.5943078099697444e-06,
+      "loss": 0.5346,
+      "step": 4302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.503643834749557,
+      "learning_rate": 7.593192339770816e-06,
+      "loss": 0.5603,
+      "step": 4303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.64519741260756,
+      "learning_rate": 7.592076692986956e-06,
+      "loss": 0.5251,
+      "step": 4304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 26.969183920431295,
+      "learning_rate": 7.5909608696941326e-06,
+      "loss": 0.5906,
+      "step": 4305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 22.44569631127385,
+      "learning_rate": 7.58984486996833e-06,
+      "loss": 0.5146,
+      "step": 4306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.74739015383114,
+      "learning_rate": 7.588728693885543e-06,
+      "loss": 0.5963,
+      "step": 4307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.114459353111076,
+      "learning_rate": 7.587612341521778e-06,
+      "loss": 0.5323,
+      "step": 4308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.320502205425744,
+      "learning_rate": 7.586495812953053e-06,
+      "loss": 0.4032,
+      "step": 4309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 24.11315922135648,
+      "learning_rate": 7.585379108255401e-06,
+      "loss": 0.5685,
+      "step": 4310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.45979678335941,
+      "learning_rate": 7.58426222750486e-06,
+      "loss": 0.5056,
+      "step": 4311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.61250188557507,
+      "learning_rate": 7.58314517077749e-06,
+      "loss": 0.4781,
+      "step": 4312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.116773389805605,
+      "learning_rate": 7.582027938149355e-06,
+      "loss": 0.481,
+      "step": 4313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.929152175438343,
+      "learning_rate": 7.580910529696532e-06,
+      "loss": 0.5072,
+      "step": 4314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 31.817316115985445,
+      "learning_rate": 7.5797929454951144e-06,
+      "loss": 0.5227,
+      "step": 4315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.855116349800088,
+      "learning_rate": 7.578675185621203e-06,
+      "loss": 0.4507,
+      "step": 4316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7804875106714098,
+      "learning_rate": 7.577557250150914e-06,
+      "loss": 0.7577,
+      "step": 4317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.27755248862274,
+      "learning_rate": 7.576439139160372e-06,
+      "loss": 0.5333,
+      "step": 4318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.972382469494762,
+      "learning_rate": 7.575320852725716e-06,
+      "loss": 0.4125,
+      "step": 4319
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.959560002219604,
+      "learning_rate": 7.574202390923098e-06,
+      "loss": 0.5285,
+      "step": 4320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.431180082514533,
+      "learning_rate": 7.573083753828678e-06,
+      "loss": 0.4992,
+      "step": 4321
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.520350382201304,
+      "learning_rate": 7.57196494151863e-06,
+      "loss": 0.3964,
+      "step": 4322
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9818123483738799,
+      "learning_rate": 7.570845954069143e-06,
+      "loss": 0.8324,
+      "step": 4323
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.88928844838767,
+      "learning_rate": 7.569726791556413e-06,
+      "loss": 0.4607,
+      "step": 4324
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.78019653429973,
+      "learning_rate": 7.56860745405665e-06,
+      "loss": 0.4316,
+      "step": 4325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.510944829595176,
+      "learning_rate": 7.567487941646078e-06,
+      "loss": 0.4959,
+      "step": 4326
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.621215319495279,
+      "learning_rate": 7.5663682544009265e-06,
+      "loss": 0.4224,
+      "step": 4327
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 66.1313251212345,
+      "learning_rate": 7.565248392397446e-06,
+      "loss": 0.5338,
+      "step": 4328
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.139236584183394,
+      "learning_rate": 7.56412835571189e-06,
+      "loss": 0.5025,
+      "step": 4329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 93.58697929109586,
+      "learning_rate": 7.563008144420534e-06,
+      "loss": 0.5819,
+      "step": 4330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.925009461770593,
+      "learning_rate": 7.561887758599652e-06,
+      "loss": 0.4577,
+      "step": 4331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.948736815488342,
+      "learning_rate": 7.560767198325541e-06,
+      "loss": 0.5175,
+      "step": 4332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.837716412751938,
+      "learning_rate": 7.5596464636745085e-06,
+      "loss": 0.5562,
+      "step": 4333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.578390645754666,
+      "learning_rate": 7.558525554722867e-06,
+      "loss": 0.5552,
+      "step": 4334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.737222066765383,
+      "learning_rate": 7.55740447154695e-06,
+      "loss": 0.4994,
+      "step": 4335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.571785745373134,
+      "learning_rate": 7.556283214223094e-06,
+      "loss": 0.5453,
+      "step": 4336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.29430207276976,
+      "learning_rate": 7.555161782827654e-06,
+      "loss": 0.4588,
+      "step": 4337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 24.12866826020616,
+      "learning_rate": 7.554040177436994e-06,
+      "loss": 0.4763,
+      "step": 4338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 24.599951630441925,
+      "learning_rate": 7.552918398127491e-06,
+      "loss": 0.4703,
+      "step": 4339
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.40662076997364,
+      "learning_rate": 7.551796444975532e-06,
+      "loss": 0.4484,
+      "step": 4340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.758483402767673,
+      "learning_rate": 7.550674318057518e-06,
+      "loss": 0.4706,
+      "step": 4341
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.150212262401746,
+      "learning_rate": 7.549552017449861e-06,
+      "loss": 0.495,
+      "step": 4342
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 42.917120261823825,
+      "learning_rate": 7.548429543228984e-06,
+      "loss": 0.5643,
+      "step": 4343
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.834188614830682,
+      "learning_rate": 7.5473068954713246e-06,
+      "loss": 0.5101,
+      "step": 4344
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.431863739064593,
+      "learning_rate": 7.546184074253327e-06,
+      "loss": 0.553,
+      "step": 4345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.858493694886445,
+      "learning_rate": 7.5450610796514504e-06,
+      "loss": 0.4626,
+      "step": 4346
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.032896665865923,
+      "learning_rate": 7.543937911742169e-06,
+      "loss": 0.4878,
+      "step": 4347
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.054224039701847,
+      "learning_rate": 7.542814570601963e-06,
+      "loss": 0.4815,
+      "step": 4348
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.015141071571005,
+      "learning_rate": 7.541691056307327e-06,
+      "loss": 0.4525,
+      "step": 4349
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.873277473442911,
+      "learning_rate": 7.540567368934769e-06,
+      "loss": 0.521,
+      "step": 4350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.003453306673823,
+      "learning_rate": 7.539443508560806e-06,
+      "loss": 0.5313,
+      "step": 4351
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.97529884220925,
+      "learning_rate": 7.5383194752619686e-06,
+      "loss": 0.4351,
+      "step": 4352
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0446036943156702,
+      "learning_rate": 7.5371952691147985e-06,
+      "loss": 0.7748,
+      "step": 4353
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.44723922868187,
+      "learning_rate": 7.536070890195848e-06,
+      "loss": 0.469,
+      "step": 4354
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 57.112980318908065,
+      "learning_rate": 7.534946338581681e-06,
+      "loss": 0.6657,
+      "step": 4355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.864227542457654,
+      "learning_rate": 7.53382161434888e-06,
+      "loss": 0.5163,
+      "step": 4356
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.425554105860217,
+      "learning_rate": 7.532696717574028e-06,
+      "loss": 0.5117,
+      "step": 4357
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.788432601522025,
+      "learning_rate": 7.531571648333727e-06,
+      "loss": 0.4766,
+      "step": 4358
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.15050714673444,
+      "learning_rate": 7.5304464067045916e-06,
+      "loss": 0.3791,
+      "step": 4359
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.013770582287602,
+      "learning_rate": 7.529320992763245e-06,
+      "loss": 0.4629,
+      "step": 4360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 29.63488781133419,
+      "learning_rate": 7.52819540658632e-06,
+      "loss": 0.4512,
+      "step": 4361
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 14.833036134428529,
+      "learning_rate": 7.527069648250467e-06,
+      "loss": 0.5724,
+      "step": 4362
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.071021325764075,
+      "learning_rate": 7.525943717832345e-06,
+      "loss": 0.4484,
+      "step": 4363
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.310293416905914,
+      "learning_rate": 7.524817615408622e-06,
+      "loss": 0.5521,
+      "step": 4364
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.726664697932584,
+      "learning_rate": 7.523691341055985e-06,
+      "loss": 0.4138,
+      "step": 4365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.997613196552027,
+      "learning_rate": 7.5225648948511255e-06,
+      "loss": 0.4447,
+      "step": 4366
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.348286489442543,
+      "learning_rate": 7.521438276870749e-06,
+      "loss": 0.5055,
+      "step": 4367
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.838134595277182,
+      "learning_rate": 7.520311487191575e-06,
+      "loss": 0.5347,
+      "step": 4368
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 23.293460839789805,
+      "learning_rate": 7.519184525890333e-06,
+      "loss": 0.5552,
+      "step": 4369
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 25.038756237094546,
+      "learning_rate": 7.518057393043763e-06,
+      "loss": 0.6357,
+      "step": 4370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.143176497496805,
+      "learning_rate": 7.516930088728616e-06,
+      "loss": 0.446,
+      "step": 4371
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.862338311982215,
+      "learning_rate": 7.515802613021661e-06,
+      "loss": 0.5301,
+      "step": 4372
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.04308785234668,
+      "learning_rate": 7.514674965999671e-06,
+      "loss": 0.5132,
+      "step": 4373
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 44.943116775554515,
+      "learning_rate": 7.513547147739434e-06,
+      "loss": 0.4956,
+      "step": 4374
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 263.7197244625826,
+      "learning_rate": 7.51241915831775e-06,
+      "loss": 0.4398,
+      "step": 4375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 23.050362318525195,
+      "learning_rate": 7.511290997811429e-06,
+      "loss": 0.3978,
+      "step": 4376
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.056336755231296,
+      "learning_rate": 7.510162666297296e-06,
+      "loss": 0.5329,
+      "step": 4377
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.1630500213515,
+      "learning_rate": 7.509034163852182e-06,
+      "loss": 0.4992,
+      "step": 4378
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.3676507265494955,
+      "learning_rate": 7.5079054905529354e-06,
+      "loss": 0.4428,
+      "step": 4379
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.64726848249301,
+      "learning_rate": 7.506776646476415e-06,
+      "loss": 0.4878,
+      "step": 4380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 23.284888789835346,
+      "learning_rate": 7.5056476316994886e-06,
+      "loss": 0.4844,
+      "step": 4381
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9685197159111396,
+      "learning_rate": 7.504518446299036e-06,
+      "loss": 0.7941,
+      "step": 4382
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.239092303415658,
+      "learning_rate": 7.5033890903519516e-06,
+      "loss": 0.4686,
+      "step": 4383
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.402769157346885,
+      "learning_rate": 7.502259563935138e-06,
+      "loss": 0.4385,
+      "step": 4384
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.310824308368728,
+      "learning_rate": 7.501129867125513e-06,
+      "loss": 0.545,
+      "step": 4385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.022038648816729,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.4973,
+      "step": 4386
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.23684487509062,
+      "learning_rate": 7.498869962635544e-06,
+      "loss": 0.5255,
+      "step": 4387
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 35.622553879048155,
+      "learning_rate": 7.497739755109089e-06,
+      "loss": 0.4537,
+      "step": 4388
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.630785153826661,
+      "learning_rate": 7.496609377497602e-06,
+      "loss": 0.6256,
+      "step": 4389
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.987628292568215,
+      "learning_rate": 7.495478829878055e-06,
+      "loss": 0.4645,
+      "step": 4390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.606855916841916,
+      "learning_rate": 7.4943481123274346e-06,
+      "loss": 0.5032,
+      "step": 4391
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.119130696390199,
+      "learning_rate": 7.493217224922735e-06,
+      "loss": 0.6205,
+      "step": 4392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.226528873104476,
+      "learning_rate": 7.492086167740966e-06,
+      "loss": 0.5501,
+      "step": 4393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 20.749242934096436,
+      "learning_rate": 7.490954940859148e-06,
+      "loss": 0.4301,
+      "step": 4394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.226419808566789,
+      "learning_rate": 7.489823544354312e-06,
+      "loss": 0.4811,
+      "step": 4395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 21.291794055393225,
+      "learning_rate": 7.488691978303501e-06,
+      "loss": 0.6012,
+      "step": 4396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.339011640907634,
+      "learning_rate": 7.487560242783771e-06,
+      "loss": 0.4515,
+      "step": 4397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.10820164399523,
+      "learning_rate": 7.4864283378721845e-06,
+      "loss": 0.5092,
+      "step": 4398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.106408011188666,
+      "learning_rate": 7.485296263645825e-06,
+      "loss": 0.498,
+      "step": 4399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.102385705060834,
+      "learning_rate": 7.4841640201817764e-06,
+      "loss": 0.4908,
+      "step": 4400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.015244651984663,
+      "learning_rate": 7.483031607557143e-06,
+      "loss": 0.5182,
+      "step": 4401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.21005273159906,
+      "learning_rate": 7.481899025849035e-06,
+      "loss": 0.4435,
+      "step": 4402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 26.546292594822418,
+      "learning_rate": 7.480766275134577e-06,
+      "loss": 0.5371,
+      "step": 4403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.62234797208716,
+      "learning_rate": 7.479633355490905e-06,
+      "loss": 0.4948,
+      "step": 4404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 17.75495956361248,
+      "learning_rate": 7.478500266995163e-06,
+      "loss": 0.4229,
+      "step": 4405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.931349292832767,
+      "learning_rate": 7.4773670097245146e-06,
+      "loss": 0.5675,
+      "step": 4406
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.625798405008876,
+      "learning_rate": 7.476233583756124e-06,
+      "loss": 0.5143,
+      "step": 4407
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.669141130460117,
+      "learning_rate": 7.475099989167176e-06,
+      "loss": 0.6309,
+      "step": 4408
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.602198137505818,
+      "learning_rate": 7.473966226034861e-06,
+      "loss": 0.5234,
+      "step": 4409
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.462618432450903,
+      "learning_rate": 7.472832294436387e-06,
+      "loss": 0.5095,
+      "step": 4410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.639219186836032,
+      "learning_rate": 7.471698194448964e-06,
+      "loss": 0.4978,
+      "step": 4411
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.550645267315108,
+      "learning_rate": 7.470563926149824e-06,
+      "loss": 0.591,
+      "step": 4412
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.740323528057776,
+      "learning_rate": 7.469429489616204e-06,
+      "loss": 0.614,
+      "step": 4413
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.361482063642283,
+      "learning_rate": 7.468294884925355e-06,
+      "loss": 0.4009,
+      "step": 4414
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 21.09190789201292,
+      "learning_rate": 7.467160112154539e-06,
+      "loss": 0.5996,
+      "step": 4415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.69522004104232,
+      "learning_rate": 7.466025171381025e-06,
+      "loss": 0.3404,
+      "step": 4416
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.66260360945823,
+      "learning_rate": 7.4648900626821004e-06,
+      "loss": 0.4092,
+      "step": 4417
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.651386697515862,
+      "learning_rate": 7.4637547861350625e-06,
+      "loss": 0.6065,
+      "step": 4418
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.688879727989972,
+      "learning_rate": 7.462619341817217e-06,
+      "loss": 0.5706,
+      "step": 4419
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 28.174361015882408,
+      "learning_rate": 7.461483729805882e-06,
+      "loss": 0.5384,
+      "step": 4420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.449062445606092,
+      "learning_rate": 7.4603479501783886e-06,
+      "loss": 0.4848,
+      "step": 4421
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.036731605880811,
+      "learning_rate": 7.459212003012077e-06,
+      "loss": 0.4532,
+      "step": 4422
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.121537981280614,
+      "learning_rate": 7.458075888384303e-06,
+      "loss": 0.4662,
+      "step": 4423
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.83920020366548,
+      "learning_rate": 7.456939606372429e-06,
+      "loss": 0.7158,
+      "step": 4424
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.540760987499025,
+      "learning_rate": 7.45580315705383e-06,
+      "loss": 0.4673,
+      "step": 4425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.357674400873284,
+      "learning_rate": 7.454666540505895e-06,
+      "loss": 0.5053,
+      "step": 4426
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.322580972650344,
+      "learning_rate": 7.453529756806022e-06,
+      "loss": 0.513,
+      "step": 4427
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.92927365045403,
+      "learning_rate": 7.45239280603162e-06,
+      "loss": 0.5059,
+      "step": 4428
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.303994407353393,
+      "learning_rate": 7.4512556882601114e-06,
+      "loss": 0.5275,
+      "step": 4429
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.076037138109943,
+      "learning_rate": 7.450118403568929e-06,
+      "loss": 0.5342,
+      "step": 4430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8273409336891608,
+      "learning_rate": 7.448980952035516e-06,
+      "loss": 0.7831,
+      "step": 4431
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.405924455047013,
+      "learning_rate": 7.447843333737328e-06,
+      "loss": 0.5166,
+      "step": 4432
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.44743630138376,
+      "learning_rate": 7.446705548751833e-06,
+      "loss": 0.4787,
+      "step": 4433
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.533174417144748,
+      "learning_rate": 7.445567597156507e-06,
+      "loss": 0.4373,
+      "step": 4434
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.142105006165092,
+      "learning_rate": 7.444429479028841e-06,
+      "loss": 0.4445,
+      "step": 4435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.016887170062057,
+      "learning_rate": 7.443291194446336e-06,
+      "loss": 0.4567,
+      "step": 4436
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.692533431683108,
+      "learning_rate": 7.4421527434865026e-06,
+      "loss": 0.5361,
+      "step": 4437
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.816525026996781,
+      "learning_rate": 7.441014126226866e-06,
+      "loss": 0.786,
+      "step": 4438
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.031927279913067,
+      "learning_rate": 7.439875342744959e-06,
+      "loss": 0.4187,
+      "step": 4439
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.84960887547989,
+      "learning_rate": 7.4387363931183295e-06,
+      "loss": 0.596,
+      "step": 4440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 22.88130087535631,
+      "learning_rate": 7.437597277424535e-06,
+      "loss": 0.5567,
+      "step": 4441
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.401775374422144,
+      "learning_rate": 7.436457995741143e-06,
+      "loss": 0.5332,
+      "step": 4442
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.822567567994156,
+      "learning_rate": 7.435318548145736e-06,
+      "loss": 0.6016,
+      "step": 4443
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.059204911597808,
+      "learning_rate": 7.434178934715901e-06,
+      "loss": 0.6027,
+      "step": 4444
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.972367652598797,
+      "learning_rate": 7.4330391555292445e-06,
+      "loss": 0.499,
+      "step": 4445
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.823528886838002,
+      "learning_rate": 7.4318992106633795e-06,
+      "loss": 0.4893,
+      "step": 4446
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.75068767013798,
+      "learning_rate": 7.43075910019593e-06,
+      "loss": 0.3987,
+      "step": 4447
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 22.137128837127495,
+      "learning_rate": 7.429618824204533e-06,
+      "loss": 0.4847,
+      "step": 4448
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.079264013318202,
+      "learning_rate": 7.428478382766837e-06,
+      "loss": 0.562,
+      "step": 4449
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.913925327669395,
+      "learning_rate": 7.427337775960499e-06,
+      "loss": 0.5902,
+      "step": 4450
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.438668747789846,
+      "learning_rate": 7.426197003863192e-06,
+      "loss": 0.4907,
+      "step": 4451
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.45243532579385,
+      "learning_rate": 7.425056066552598e-06,
+      "loss": 0.4541,
+      "step": 4452
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.523956843038798,
+      "learning_rate": 7.423914964106407e-06,
+      "loss": 0.5359,
+      "step": 4453
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.690959641731929,
+      "learning_rate": 7.422773696602325e-06,
+      "loss": 0.4672,
+      "step": 4454
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.638620557408665,
+      "learning_rate": 7.421632264118064e-06,
+      "loss": 0.5737,
+      "step": 4455
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.973561787393699,
+      "learning_rate": 7.420490666731355e-06,
+      "loss": 0.5892,
+      "step": 4456
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.082846915574013,
+      "learning_rate": 7.419348904519933e-06,
+      "loss": 0.5125,
+      "step": 4457
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.348949814441902,
+      "learning_rate": 7.418206977561547e-06,
+      "loss": 0.4423,
+      "step": 4458
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.569612651288798,
+      "learning_rate": 7.417064885933959e-06,
+      "loss": 0.5006,
+      "step": 4459
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.263398119020806,
+      "learning_rate": 7.4159226297149376e-06,
+      "loss": 0.4822,
+      "step": 4460
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.9169485671226045,
+      "learning_rate": 7.414780208982269e-06,
+      "loss": 0.482,
+      "step": 4461
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.599230486586455,
+      "learning_rate": 7.413637623813744e-06,
+      "loss": 0.4826,
+      "step": 4462
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.928816778093689,
+      "learning_rate": 7.412494874287169e-06,
+      "loss": 0.4782,
+      "step": 4463
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.354799692558164,
+      "learning_rate": 7.411351960480359e-06,
+      "loss": 0.4756,
+      "step": 4464
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.888355628728052,
+      "learning_rate": 7.410208882471144e-06,
+      "loss": 0.3614,
+      "step": 4465
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.348769352204108,
+      "learning_rate": 7.4090656403373595e-06,
+      "loss": 0.492,
+      "step": 4466
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.258361584412002,
+      "learning_rate": 7.407922234156857e-06,
+      "loss": 0.495,
+      "step": 4467
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.159668819905207,
+      "learning_rate": 7.406778664007497e-06,
+      "loss": 0.4286,
+      "step": 4468
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.634751414752732,
+      "learning_rate": 7.40563492996715e-06,
+      "loss": 0.5255,
+      "step": 4469
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.780962322927586,
+      "learning_rate": 7.404491032113701e-06,
+      "loss": 0.3913,
+      "step": 4470
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.37909974764437,
+      "learning_rate": 7.403346970525045e-06,
+      "loss": 0.5538,
+      "step": 4471
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.905243947755956,
+      "learning_rate": 7.402202745279086e-06,
+      "loss": 0.4948,
+      "step": 4472
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.215179471328343,
+      "learning_rate": 7.401058356453741e-06,
+      "loss": 0.4389,
+      "step": 4473
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.77728034096194,
+      "learning_rate": 7.3999138041269395e-06,
+      "loss": 0.4231,
+      "step": 4474
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.539741291994101,
+      "learning_rate": 7.398769088376619e-06,
+      "loss": 0.416,
+      "step": 4475
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.128356218148332,
+      "learning_rate": 7.397624209280727e-06,
+      "loss": 0.54,
+      "step": 4476
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.655663808994232,
+      "learning_rate": 7.396479166917229e-06,
+      "loss": 0.4477,
+      "step": 4477
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.264663444575943,
+      "learning_rate": 7.395333961364096e-06,
+      "loss": 0.4775,
+      "step": 4478
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.848927630239777,
+      "learning_rate": 7.394188592699311e-06,
+      "loss": 0.4773,
+      "step": 4479
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.434896403048752,
+      "learning_rate": 7.393043061000869e-06,
+      "loss": 0.5165,
+      "step": 4480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.6088710092771,
+      "learning_rate": 7.391897366346774e-06,
+      "loss": 0.456,
+      "step": 4481
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.379695293883424,
+      "learning_rate": 7.390751508815045e-06,
+      "loss": 0.4919,
+      "step": 4482
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 29.302227900073806,
+      "learning_rate": 7.389605488483709e-06,
+      "loss": 0.4675,
+      "step": 4483
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.563130170611167,
+      "learning_rate": 7.388459305430804e-06,
+      "loss": 0.5663,
+      "step": 4484
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.710224365303638,
+      "learning_rate": 7.387312959734381e-06,
+      "loss": 0.4823,
+      "step": 4485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 24.499878635513603,
+      "learning_rate": 7.386166451472502e-06,
+      "loss": 0.5721,
+      "step": 4486
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.282995647525976,
+      "learning_rate": 7.385019780723237e-06,
+      "loss": 0.4126,
+      "step": 4487
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.96302084729856,
+      "learning_rate": 7.383872947564669e-06,
+      "loss": 0.4349,
+      "step": 4488
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.388740639967414,
+      "learning_rate": 7.382725952074895e-06,
+      "loss": 0.5522,
+      "step": 4489
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.265123396990328,
+      "learning_rate": 7.381578794332018e-06,
+      "loss": 0.4632,
+      "step": 4490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1468753760355592,
+      "learning_rate": 7.380431474414156e-06,
+      "loss": 0.8263,
+      "step": 4491
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.6945481044649,
+      "learning_rate": 7.3792839923994355e-06,
+      "loss": 0.4654,
+      "step": 4492
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.050743510571985,
+      "learning_rate": 7.378136348365995e-06,
+      "loss": 0.5169,
+      "step": 4493
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.475747205772684,
+      "learning_rate": 7.376988542391983e-06,
+      "loss": 0.5209,
+      "step": 4494
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 15.818178091812731,
+      "learning_rate": 7.375840574555563e-06,
+      "loss": 0.5348,
+      "step": 4495
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.033092440223484,
+      "learning_rate": 7.374692444934903e-06,
+      "loss": 0.4713,
+      "step": 4496
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.823161325224817,
+      "learning_rate": 7.373544153608188e-06,
+      "loss": 0.4472,
+      "step": 4497
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.145685175763018,
+      "learning_rate": 7.37239570065361e-06,
+      "loss": 0.4304,
+      "step": 4498
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.765272925144542,
+      "learning_rate": 7.371247086149372e-06,
+      "loss": 0.4675,
+      "step": 4499
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.737346417233972,
+      "learning_rate": 7.370098310173696e-06,
+      "loss": 0.5444,
+      "step": 4500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.936335529632343,
+      "learning_rate": 7.368949372804801e-06,
+      "loss": 0.6478,
+      "step": 4501
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.42134165686766,
+      "learning_rate": 7.367800274120931e-06,
+      "loss": 0.4637,
+      "step": 4502
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.629535040534973,
+      "learning_rate": 7.366651014200329e-06,
+      "loss": 0.4461,
+      "step": 4503
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.70070436095542,
+      "learning_rate": 7.365501593121258e-06,
+      "loss": 0.5532,
+      "step": 4504
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.774918105786875,
+      "learning_rate": 7.364352010961987e-06,
+      "loss": 0.3705,
+      "step": 4505
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.299485091386789,
+      "learning_rate": 7.3632022678007984e-06,
+      "loss": 0.476,
+      "step": 4506
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.745041131694816,
+      "learning_rate": 7.3620523637159845e-06,
+      "loss": 0.4613,
+      "step": 4507
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0527112512698227,
+      "learning_rate": 7.360902298785848e-06,
+      "loss": 0.8618,
+      "step": 4508
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.205198581546721,
+      "learning_rate": 7.3597520730887025e-06,
+      "loss": 0.5198,
+      "step": 4509
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.119729212936302,
+      "learning_rate": 7.358601686702875e-06,
+      "loss": 0.4402,
+      "step": 4510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.284955179321972,
+      "learning_rate": 7.357451139706702e-06,
+      "loss": 0.5779,
+      "step": 4511
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.153402479826644,
+      "learning_rate": 7.356300432178529e-06,
+      "loss": 0.4625,
+      "step": 4512
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.10712582901531,
+      "learning_rate": 7.355149564196714e-06,
+      "loss": 0.3906,
+      "step": 4513
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.860864624259037,
+      "learning_rate": 7.3539985358396296e-06,
+      "loss": 0.6834,
+      "step": 4514
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.909211415359861,
+      "learning_rate": 7.35284734718565e-06,
+      "loss": 0.3972,
+      "step": 4515
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.96330309151543,
+      "learning_rate": 7.3516959983131705e-06,
+      "loss": 0.4718,
+      "step": 4516
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.22957468422264,
+      "learning_rate": 7.350544489300592e-06,
+      "loss": 0.6305,
+      "step": 4517
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.643579679078071,
+      "learning_rate": 7.3493928202263265e-06,
+      "loss": 0.5263,
+      "step": 4518
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.757568408495068,
+      "learning_rate": 7.348240991168795e-06,
+      "loss": 0.4497,
+      "step": 4519
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.949928786249687,
+      "learning_rate": 7.347089002206438e-06,
+      "loss": 0.4426,
+      "step": 4520
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.976501494820937,
+      "learning_rate": 7.345936853417696e-06,
+      "loss": 0.6744,
+      "step": 4521
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.530011914116006,
+      "learning_rate": 7.344784544881026e-06,
+      "loss": 0.5155,
+      "step": 4522
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.549850521786697,
+      "learning_rate": 7.343632076674898e-06,
+      "loss": 0.5613,
+      "step": 4523
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.448933423881836,
+      "learning_rate": 7.342479448877787e-06,
+      "loss": 0.497,
+      "step": 4524
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.221627987664327,
+      "learning_rate": 7.3413266615681825e-06,
+      "loss": 0.4149,
+      "step": 4525
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.86815621039865,
+      "learning_rate": 7.340173714824583e-06,
+      "loss": 0.4847,
+      "step": 4526
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 23.016442207648694,
+      "learning_rate": 7.339020608725502e-06,
+      "loss": 0.689,
+      "step": 4527
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.801134629241704,
+      "learning_rate": 7.33786734334946e-06,
+      "loss": 0.5301,
+      "step": 4528
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.994562101712222,
+      "learning_rate": 7.336713918774987e-06,
+      "loss": 0.4388,
+      "step": 4529
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.140235096094841,
+      "learning_rate": 7.3355603350806286e-06,
+      "loss": 0.5801,
+      "step": 4530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.75469924917666,
+      "learning_rate": 7.334406592344938e-06,
+      "loss": 0.6169,
+      "step": 4531
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.967186981252167,
+      "learning_rate": 7.333252690646481e-06,
+      "loss": 0.5038,
+      "step": 4532
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 20.939964130944407,
+      "learning_rate": 7.33209863006383e-06,
+      "loss": 0.5271,
+      "step": 4533
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.991445780916145,
+      "learning_rate": 7.330944410675575e-06,
+      "loss": 0.493,
+      "step": 4534
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.847753775851642,
+      "learning_rate": 7.329790032560311e-06,
+      "loss": 0.5047,
+      "step": 4535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.951339248622388,
+      "learning_rate": 7.328635495796647e-06,
+      "loss": 0.4863,
+      "step": 4536
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.369058425975958,
+      "learning_rate": 7.327480800463204e-06,
+      "loss": 0.4252,
+      "step": 4537
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.55608520632619,
+      "learning_rate": 7.3263259466386065e-06,
+      "loss": 0.6233,
+      "step": 4538
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.772936157020374,
+      "learning_rate": 7.325170934401497e-06,
+      "loss": 0.5072,
+      "step": 4539
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.414238082091838,
+      "learning_rate": 7.3240157638305296e-06,
+      "loss": 0.4492,
+      "step": 4540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.266505614373867,
+      "learning_rate": 7.322860435004365e-06,
+      "loss": 0.3455,
+      "step": 4541
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.486484801456742,
+      "learning_rate": 7.321704948001673e-06,
+      "loss": 0.4628,
+      "step": 4542
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.030469591457942,
+      "learning_rate": 7.320549302901142e-06,
+      "loss": 0.5062,
+      "step": 4543
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.543638929883572,
+      "learning_rate": 7.319393499781462e-06,
+      "loss": 0.5333,
+      "step": 4544
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.48937516675614,
+      "learning_rate": 7.318237538721339e-06,
+      "loss": 0.4868,
+      "step": 4545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0308416020008364,
+      "learning_rate": 7.317081419799491e-06,
+      "loss": 0.7773,
+      "step": 4546
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.720540109307889,
+      "learning_rate": 7.315925143094642e-06,
+      "loss": 0.4771,
+      "step": 4547
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.990938943957497,
+      "learning_rate": 7.314768708685532e-06,
+      "loss": 0.4533,
+      "step": 4548
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.796420500903477,
+      "learning_rate": 7.313612116650906e-06,
+      "loss": 0.4401,
+      "step": 4549
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7869538737038368,
+      "learning_rate": 7.312455367069526e-06,
+      "loss": 0.8111,
+      "step": 4550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 30.184205590614305,
+      "learning_rate": 7.311298460020157e-06,
+      "loss": 0.6291,
+      "step": 4551
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.160438033919117,
+      "learning_rate": 7.3101413955815856e-06,
+      "loss": 0.5301,
+      "step": 4552
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.636535844893781,
+      "learning_rate": 7.3089841738325986e-06,
+      "loss": 0.5756,
+      "step": 4553
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.442305924998065,
+      "learning_rate": 7.307826794851997e-06,
+      "loss": 0.4892,
+      "step": 4554
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 28.883642546331927,
+      "learning_rate": 7.3066692587185964e-06,
+      "loss": 0.4952,
+      "step": 4555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.996729695896564,
+      "learning_rate": 7.305511565511217e-06,
+      "loss": 0.6121,
+      "step": 4556
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.25054987629699,
+      "learning_rate": 7.304353715308694e-06,
+      "loss": 0.3263,
+      "step": 4557
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.793791009460612,
+      "learning_rate": 7.303195708189871e-06,
+      "loss": 0.3651,
+      "step": 4558
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.01040897515619,
+      "learning_rate": 7.3020375442336044e-06,
+      "loss": 0.479,
+      "step": 4559
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.36106408051119,
+      "learning_rate": 7.300879223518757e-06,
+      "loss": 0.4855,
+      "step": 4560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.350549565942579,
+      "learning_rate": 7.29972074612421e-06,
+      "loss": 0.515,
+      "step": 4561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.748457787688643,
+      "learning_rate": 7.2985621121288495e-06,
+      "loss": 0.5456,
+      "step": 4562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 25.671454823849917,
+      "learning_rate": 7.297403321611572e-06,
+      "loss": 0.4821,
+      "step": 4563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.819870986596234,
+      "learning_rate": 7.2962443746512844e-06,
+      "loss": 0.5299,
+      "step": 4564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.163019625336426,
+      "learning_rate": 7.2950852713269084e-06,
+      "loss": 0.5638,
+      "step": 4565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.893052723960418,
+      "learning_rate": 7.2939260117173736e-06,
+      "loss": 0.419,
+      "step": 4566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 31.531885767791064,
+      "learning_rate": 7.292766595901618e-06,
+      "loss": 0.4263,
+      "step": 4567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.2125720672024,
+      "learning_rate": 7.291607023958596e-06,
+      "loss": 0.6148,
+      "step": 4568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.829592999823682,
+      "learning_rate": 7.290447295967269e-06,
+      "loss": 0.5349,
+      "step": 4569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.159824798215977,
+      "learning_rate": 7.289287412006605e-06,
+      "loss": 0.4653,
+      "step": 4570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.53330327047796,
+      "learning_rate": 7.2881273721555935e-06,
+      "loss": 0.6257,
+      "step": 4571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.805511046726146,
+      "learning_rate": 7.286967176493222e-06,
+      "loss": 0.5737,
+      "step": 4572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.329834090067538,
+      "learning_rate": 7.2858068250985e-06,
+      "loss": 0.5191,
+      "step": 4573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 42.34798605315762,
+      "learning_rate": 7.284646318050439e-06,
+      "loss": 0.371,
+      "step": 4574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.534912801175837,
+      "learning_rate": 7.283485655428064e-06,
+      "loss": 0.41,
+      "step": 4575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 20.96578323982638,
+      "learning_rate": 7.282324837310415e-06,
+      "loss": 0.5239,
+      "step": 4576
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.511410659657066,
+      "learning_rate": 7.2811638637765324e-06,
+      "loss": 0.4069,
+      "step": 4577
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 26.146507913294055,
+      "learning_rate": 7.2800027349054805e-06,
+      "loss": 0.5007,
+      "step": 4578
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.772488017693243,
+      "learning_rate": 7.278841450776319e-06,
+      "loss": 0.4271,
+      "step": 4579
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.106036587371262,
+      "learning_rate": 7.277680011468133e-06,
+      "loss": 0.6421,
+      "step": 4580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.920587266012534,
+      "learning_rate": 7.276518417060008e-06,
+      "loss": 0.6372,
+      "step": 4581
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.037673035113128,
+      "learning_rate": 7.275356667631046e-06,
+      "loss": 0.527,
+      "step": 4582
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 33.493798156669726,
+      "learning_rate": 7.274194763260354e-06,
+      "loss": 0.4778,
+      "step": 4583
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.682484606530785,
+      "learning_rate": 7.273032704027053e-06,
+      "loss": 0.5287,
+      "step": 4584
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.660548258827113,
+      "learning_rate": 7.271870490010278e-06,
+      "loss": 0.3263,
+      "step": 4585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.916761530705994,
+      "learning_rate": 7.270708121289165e-06,
+      "loss": 0.5378,
+      "step": 4586
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.990732401324088,
+      "learning_rate": 7.269545597942869e-06,
+      "loss": 0.5591,
+      "step": 4587
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.975459577634464,
+      "learning_rate": 7.268382920050553e-06,
+      "loss": 0.3713,
+      "step": 4588
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.844366329581367,
+      "learning_rate": 7.26722008769139e-06,
+      "loss": 0.6274,
+      "step": 4589
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.155693650783505,
+      "learning_rate": 7.2660571009445624e-06,
+      "loss": 0.4797,
+      "step": 4590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 27.45183368549548,
+      "learning_rate": 7.264893959889266e-06,
+      "loss": 0.4249,
+      "step": 4591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.88543818519557,
+      "learning_rate": 7.263730664604706e-06,
+      "loss": 0.434,
+      "step": 4592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.07169147298337,
+      "learning_rate": 7.262567215170095e-06,
+      "loss": 0.5468,
+      "step": 4593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.234502083889297,
+      "learning_rate": 7.261403611664663e-06,
+      "loss": 0.4359,
+      "step": 4594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.3996409370245,
+      "learning_rate": 7.260239854167641e-06,
+      "loss": 0.4379,
+      "step": 4595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.24533674537989,
+      "learning_rate": 7.25907594275828e-06,
+      "loss": 0.4342,
+      "step": 4596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.484971400229314,
+      "learning_rate": 7.257911877515836e-06,
+      "loss": 0.4747,
+      "step": 4597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.223656175452065,
+      "learning_rate": 7.256747658519575e-06,
+      "loss": 0.5954,
+      "step": 4598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.220019769398815,
+      "learning_rate": 7.255583285848778e-06,
+      "loss": 0.5498,
+      "step": 4599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.696460970723335,
+      "learning_rate": 7.254418759582731e-06,
+      "loss": 0.5979,
+      "step": 4600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.979443869341765,
+      "learning_rate": 7.253254079800735e-06,
+      "loss": 0.5332,
+      "step": 4601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5343979766680715,
+      "learning_rate": 7.252089246582098e-06,
+      "loss": 0.8404,
+      "step": 4602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 33.89212263960842,
+      "learning_rate": 7.2509242600061425e-06,
+      "loss": 0.457,
+      "step": 4603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.168497561868433,
+      "learning_rate": 7.249759120152195e-06,
+      "loss": 0.4858,
+      "step": 4604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7916268114385336,
+      "learning_rate": 7.248593827099602e-06,
+      "loss": 0.7337,
+      "step": 4605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.192266022270724,
+      "learning_rate": 7.247428380927708e-06,
+      "loss": 0.4872,
+      "step": 4606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7027702521057446,
+      "learning_rate": 7.246262781715879e-06,
+      "loss": 0.7757,
+      "step": 4607
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.022333745085772,
+      "learning_rate": 7.245097029543487e-06,
+      "loss": 0.4623,
+      "step": 4608
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.042930323437924,
+      "learning_rate": 7.243931124489914e-06,
+      "loss": 0.406,
+      "step": 4609
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.161391225948236,
+      "learning_rate": 7.242765066634552e-06,
+      "loss": 0.3819,
+      "step": 4610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 26.411646336241223,
+      "learning_rate": 7.241598856056805e-06,
+      "loss": 0.3838,
+      "step": 4611
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.026418318154576,
+      "learning_rate": 7.240432492836089e-06,
+      "loss": 0.4893,
+      "step": 4612
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.579177342653354,
+      "learning_rate": 7.239265977051825e-06,
+      "loss": 0.4869,
+      "step": 4613
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 72.93580991281702,
+      "learning_rate": 7.2380993087834485e-06,
+      "loss": 0.4634,
+      "step": 4614
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.288657657539968,
+      "learning_rate": 7.236932488110405e-06,
+      "loss": 0.5049,
+      "step": 4615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.17450379890601,
+      "learning_rate": 7.2357655151121495e-06,
+      "loss": 0.4968,
+      "step": 4616
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.795865882444662,
+      "learning_rate": 7.234598389868148e-06,
+      "loss": 0.4879,
+      "step": 4617
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.748456710689102,
+      "learning_rate": 7.233431112457876e-06,
+      "loss": 0.7419,
+      "step": 4618
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.281464811784522,
+      "learning_rate": 7.232263682960822e-06,
+      "loss": 0.4973,
+      "step": 4619
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.505782926128607,
+      "learning_rate": 7.231096101456477e-06,
+      "loss": 0.6623,
+      "step": 4620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.325932877544676,
+      "learning_rate": 7.229928368024356e-06,
+      "loss": 0.5162,
+      "step": 4621
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.567687313383837,
+      "learning_rate": 7.228760482743971e-06,
+      "loss": 0.5346,
+      "step": 4622
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 62.05217499676669,
+      "learning_rate": 7.227592445694853e-06,
+      "loss": 0.5033,
+      "step": 4623
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.422613225716751,
+      "learning_rate": 7.226424256956537e-06,
+      "loss": 0.4223,
+      "step": 4624
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.434124734248827,
+      "learning_rate": 7.2252559166085735e-06,
+      "loss": 0.5244,
+      "step": 4625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.262700913253722,
+      "learning_rate": 7.224087424730521e-06,
+      "loss": 0.5018,
+      "step": 4626
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 36.21904399603829,
+      "learning_rate": 7.222918781401949e-06,
+      "loss": 0.5046,
+      "step": 4627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.949300387833183,
+      "learning_rate": 7.2217499867024365e-06,
+      "loss": 0.536,
+      "step": 4628
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.065180626597789,
+      "learning_rate": 7.220581040711573e-06,
+      "loss": 0.4408,
+      "step": 4629
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 24.18229414678252,
+      "learning_rate": 7.219411943508958e-06,
+      "loss": 0.4999,
+      "step": 4630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.263876559792536,
+      "learning_rate": 7.2182426951742035e-06,
+      "loss": 0.4535,
+      "step": 4631
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 22.83050120772622,
+      "learning_rate": 7.217073295786929e-06,
+      "loss": 0.532,
+      "step": 4632
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.025473450766015,
+      "learning_rate": 7.215903745426767e-06,
+      "loss": 0.5386,
+      "step": 4633
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 18.410885863493466,
+      "learning_rate": 7.214734044173356e-06,
+      "loss": 0.5495,
+      "step": 4634
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 17.336237285836194,
+      "learning_rate": 7.2135641921063505e-06,
+      "loss": 0.562,
+      "step": 4635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.699430104895352,
+      "learning_rate": 7.212394189305409e-06,
+      "loss": 0.4884,
+      "step": 4636
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.107679291132193,
+      "learning_rate": 7.2112240358502065e-06,
+      "loss": 0.4738,
+      "step": 4637
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.6223141186976,
+      "learning_rate": 7.210053731820423e-06,
+      "loss": 0.5506,
+      "step": 4638
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.194075784105152,
+      "learning_rate": 7.208883277295752e-06,
+      "loss": 0.4848,
+      "step": 4639
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.268957950998532,
+      "learning_rate": 7.207712672355897e-06,
+      "loss": 0.4599,
+      "step": 4640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.346618871531433,
+      "learning_rate": 7.20654191708057e-06,
+      "loss": 0.4614,
+      "step": 4641
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.411702284308628,
+      "learning_rate": 7.2053710115494945e-06,
+      "loss": 0.4573,
+      "step": 4642
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.8372781399406,
+      "learning_rate": 7.204199955842403e-06,
+      "loss": 0.581,
+      "step": 4643
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.49946284692289,
+      "learning_rate": 7.203028750039043e-06,
+      "loss": 0.4957,
+      "step": 4644
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 25.26754557068293,
+      "learning_rate": 7.201857394219164e-06,
+      "loss": 0.5507,
+      "step": 4645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.706103898939885,
+      "learning_rate": 7.200685888462531e-06,
+      "loss": 0.54,
+      "step": 4646
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7956875495778613,
+      "learning_rate": 7.199514232848921e-06,
+      "loss": 0.75,
+      "step": 4647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.3096510969146875,
+      "learning_rate": 7.198342427458116e-06,
+      "loss": 0.4406,
+      "step": 4648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 31.039948932148633,
+      "learning_rate": 7.197170472369912e-06,
+      "loss": 0.4413,
+      "step": 4649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.594953615262378,
+      "learning_rate": 7.195998367664112e-06,
+      "loss": 0.415,
+      "step": 4650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.165857243499426,
+      "learning_rate": 7.194826113420534e-06,
+      "loss": 0.4739,
+      "step": 4651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.548802553334383,
+      "learning_rate": 7.193653709719002e-06,
+      "loss": 0.6301,
+      "step": 4652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.593282456145369,
+      "learning_rate": 7.192481156639352e-06,
+      "loss": 0.5107,
+      "step": 4653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.039485117871147,
+      "learning_rate": 7.191308454261429e-06,
+      "loss": 0.5285,
+      "step": 4654
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.845989252925744,
+      "learning_rate": 7.190135602665088e-06,
+      "loss": 0.4659,
+      "step": 4655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 22.082289782187576,
+      "learning_rate": 7.1889626019301975e-06,
+      "loss": 0.4522,
+      "step": 4656
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.938927698539677,
+      "learning_rate": 7.187789452136631e-06,
+      "loss": 0.5962,
+      "step": 4657
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.733592359512452,
+      "learning_rate": 7.186616153364276e-06,
+      "loss": 0.4069,
+      "step": 4658
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1957370082569834,
+      "learning_rate": 7.18544270569303e-06,
+      "loss": 0.7064,
+      "step": 4659
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.455269733077026,
+      "learning_rate": 7.184269109202797e-06,
+      "loss": 0.4969,
+      "step": 4660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 33.43649950065253,
+      "learning_rate": 7.183095363973496e-06,
+      "loss": 0.5744,
+      "step": 4661
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.062704724884435,
+      "learning_rate": 7.181921470085055e-06,
+      "loss": 0.5445,
+      "step": 4662
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.320811325039946,
+      "learning_rate": 7.180747427617407e-06,
+      "loss": 0.5295,
+      "step": 4663
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 24.915195198183056,
+      "learning_rate": 7.179573236650501e-06,
+      "loss": 0.5322,
+      "step": 4664
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.827718379535622,
+      "learning_rate": 7.178398897264296e-06,
+      "loss": 0.3853,
+      "step": 4665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.666215687843243,
+      "learning_rate": 7.177224409538757e-06,
+      "loss": 0.5862,
+      "step": 4666
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.493994534701283,
+      "learning_rate": 7.1760497735538634e-06,
+      "loss": 0.4263,
+      "step": 4667
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.393553509271218,
+      "learning_rate": 7.1748749893896e-06,
+      "loss": 0.386,
+      "step": 4668
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.25458156400112,
+      "learning_rate": 7.173700057125965e-06,
+      "loss": 0.4733,
+      "step": 4669
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.304181253647316,
+      "learning_rate": 7.172524976842969e-06,
+      "loss": 0.5515,
+      "step": 4670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.583818322898473,
+      "learning_rate": 7.1713497486206286e-06,
+      "loss": 0.4641,
+      "step": 4671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.989461595263306,
+      "learning_rate": 7.170174372538968e-06,
+      "loss": 0.4965,
+      "step": 4672
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.480056574976008,
+      "learning_rate": 7.168998848678029e-06,
+      "loss": 0.4441,
+      "step": 4673
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1830114425101708,
+      "learning_rate": 7.167823177117861e-06,
+      "loss": 0.8008,
+      "step": 4674
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.43460853014367,
+      "learning_rate": 7.166647357938517e-06,
+      "loss": 0.5419,
+      "step": 4675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.4921019976691055,
+      "learning_rate": 7.165471391220069e-06,
+      "loss": 0.4451,
+      "step": 4676
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 22.27777229514903,
+      "learning_rate": 7.164295277042592e-06,
+      "loss": 0.5454,
+      "step": 4677
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 80.13106901552081,
+      "learning_rate": 7.1631190154861776e-06,
+      "loss": 0.5674,
+      "step": 4678
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.359120415990061,
+      "learning_rate": 7.1619426066309216e-06,
+      "loss": 0.553,
+      "step": 4679
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.87519913715308,
+      "learning_rate": 7.160766050556932e-06,
+      "loss": 0.5519,
+      "step": 4680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.463277958116502,
+      "learning_rate": 7.15958934734433e-06,
+      "loss": 0.4277,
+      "step": 4681
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.605129086288956,
+      "learning_rate": 7.15841249707324e-06,
+      "loss": 0.3683,
+      "step": 4682
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.31841257078823,
+      "learning_rate": 7.157235499823805e-06,
+      "loss": 0.4375,
+      "step": 4683
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.024689043922612,
+      "learning_rate": 7.156058355676168e-06,
+      "loss": 0.6091,
+      "step": 4684
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.016926355108861,
+      "learning_rate": 7.154881064710493e-06,
+      "loss": 0.558,
+      "step": 4685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.902324018408304,
+      "learning_rate": 7.153703627006943e-06,
+      "loss": 0.4766,
+      "step": 4686
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.959707403639692,
+      "learning_rate": 7.152526042645697e-06,
+      "loss": 0.5276,
+      "step": 4687
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.775337095224042,
+      "learning_rate": 7.151348311706948e-06,
+      "loss": 0.459,
+      "step": 4688
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.417475303002195,
+      "learning_rate": 7.150170434270889e-06,
+      "loss": 0.4722,
+      "step": 4689
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.18440583161062,
+      "learning_rate": 7.148992410417731e-06,
+      "loss": 0.4776,
+      "step": 4690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7630393422376098,
+      "learning_rate": 7.1478142402276905e-06,
+      "loss": 0.7881,
+      "step": 4691
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.630244731868538,
+      "learning_rate": 7.146635923780999e-06,
+      "loss": 0.5143,
+      "step": 4692
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.758607913580788,
+      "learning_rate": 7.1454574611578905e-06,
+      "loss": 0.503,
+      "step": 4693
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.420765897818079,
+      "learning_rate": 7.144278852438615e-06,
+      "loss": 0.5003,
+      "step": 4694
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.643675672026078,
+      "learning_rate": 7.143100097703433e-06,
+      "loss": 0.7465,
+      "step": 4695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 22.392737560657736,
+      "learning_rate": 7.141921197032606e-06,
+      "loss": 0.5734,
+      "step": 4696
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.394009660647956,
+      "learning_rate": 7.140742150506419e-06,
+      "loss": 0.4788,
+      "step": 4697
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.775561835457543,
+      "learning_rate": 7.1395629582051555e-06,
+      "loss": 0.4846,
+      "step": 4698
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.488461787986836,
+      "learning_rate": 7.138383620209115e-06,
+      "loss": 0.4624,
+      "step": 4699
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.204299568143888,
+      "learning_rate": 7.137204136598603e-06,
+      "loss": 0.4037,
+      "step": 4700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6065025847328416,
+      "learning_rate": 7.13602450745394e-06,
+      "loss": 0.6957,
+      "step": 4701
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.521365558375693,
+      "learning_rate": 7.134844732855452e-06,
+      "loss": 0.5092,
+      "step": 4702
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.7435644649725,
+      "learning_rate": 7.133664812883478e-06,
+      "loss": 0.4232,
+      "step": 4703
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.632919487018729,
+      "learning_rate": 7.132484747618364e-06,
+      "loss": 0.5798,
+      "step": 4704
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.865226476866989,
+      "learning_rate": 7.1313045371404665e-06,
+      "loss": 0.5415,
+      "step": 4705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 22.760330030355227,
+      "learning_rate": 7.130124181530155e-06,
+      "loss": 0.5057,
+      "step": 4706
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9640393975111292,
+      "learning_rate": 7.1289436808678036e-06,
+      "loss": 0.7658,
+      "step": 4707
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.885718861616673,
+      "learning_rate": 7.127763035233801e-06,
+      "loss": 0.4597,
+      "step": 4708
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.805466791211217,
+      "learning_rate": 7.126582244708544e-06,
+      "loss": 0.4506,
+      "step": 4709
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.34137698710562,
+      "learning_rate": 7.125401309372437e-06,
+      "loss": 0.5359,
+      "step": 4710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.375695039075726,
+      "learning_rate": 7.124220229305899e-06,
+      "loss": 0.5339,
+      "step": 4711
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.092780196341447,
+      "learning_rate": 7.123039004589356e-06,
+      "loss": 0.4987,
+      "step": 4712
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.520907304232086,
+      "learning_rate": 7.121857635303244e-06,
+      "loss": 0.5449,
+      "step": 4713
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 33.334055285555685,
+      "learning_rate": 7.120676121528008e-06,
+      "loss": 0.5802,
+      "step": 4714
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.118401382235236,
+      "learning_rate": 7.1194944633441056e-06,
+      "loss": 0.5876,
+      "step": 4715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.614496102218467,
+      "learning_rate": 7.1183126608320005e-06,
+      "loss": 0.4909,
+      "step": 4716
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.121417588962807,
+      "learning_rate": 7.117130714072169e-06,
+      "loss": 0.4327,
+      "step": 4717
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.434520696606373,
+      "learning_rate": 7.115948623145098e-06,
+      "loss": 0.4352,
+      "step": 4718
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.002432372828403,
+      "learning_rate": 7.1147663881312805e-06,
+      "loss": 0.6202,
+      "step": 4719
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.521565673483346,
+      "learning_rate": 7.113584009111223e-06,
+      "loss": 0.3481,
+      "step": 4720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.320680992277206,
+      "learning_rate": 7.112401486165437e-06,
+      "loss": 0.513,
+      "step": 4721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.388385175080879,
+      "learning_rate": 7.1112188193744525e-06,
+      "loss": 0.471,
+      "step": 4722
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 25.30986964669905,
+      "learning_rate": 7.1100360088188e-06,
+      "loss": 0.6219,
+      "step": 4723
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.463835365962403,
+      "learning_rate": 7.108853054579026e-06,
+      "loss": 0.5709,
+      "step": 4724
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.37815226723635,
+      "learning_rate": 7.107669956735681e-06,
+      "loss": 0.4826,
+      "step": 4725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.644331261464044,
+      "learning_rate": 7.106486715369332e-06,
+      "loss": 0.4961,
+      "step": 4726
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 36.839658419094825,
+      "learning_rate": 7.105303330560552e-06,
+      "loss": 0.4804,
+      "step": 4727
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.220219627723381,
+      "learning_rate": 7.104119802389923e-06,
+      "loss": 0.5725,
+      "step": 4728
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 32.95585642607746,
+      "learning_rate": 7.10293613093804e-06,
+      "loss": 0.5117,
+      "step": 4729
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 27.05341475239536,
+      "learning_rate": 7.101752316285503e-06,
+      "loss": 0.5023,
+      "step": 4730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.471892027651306,
+      "learning_rate": 7.100568358512926e-06,
+      "loss": 0.5244,
+      "step": 4731
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.34885441360084,
+      "learning_rate": 7.099384257700931e-06,
+      "loss": 0.4893,
+      "step": 4732
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.928976962592884,
+      "learning_rate": 7.098200013930152e-06,
+      "loss": 0.4688,
+      "step": 4733
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.38436903177863,
+      "learning_rate": 7.0970156272812285e-06,
+      "loss": 0.5138,
+      "step": 4734
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 23.228706220399506,
+      "learning_rate": 7.095831097834812e-06,
+      "loss": 0.4314,
+      "step": 4735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.748920527391235,
+      "learning_rate": 7.094646425671565e-06,
+      "loss": 0.6075,
+      "step": 4736
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.766448304781779,
+      "learning_rate": 7.093461610872158e-06,
+      "loss": 0.5186,
+      "step": 4737
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.585681099069426,
+      "learning_rate": 7.092276653517272e-06,
+      "loss": 0.4053,
+      "step": 4738
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 69.46669881982785,
+      "learning_rate": 7.091091553687595e-06,
+      "loss": 0.5234,
+      "step": 4739
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 22.347137076090288,
+      "learning_rate": 7.08990631146383e-06,
+      "loss": 0.5542,
+      "step": 4740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 21.577226061996555,
+      "learning_rate": 7.088720926926684e-06,
+      "loss": 0.5046,
+      "step": 4741
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.16540374455338,
+      "learning_rate": 7.08753540015688e-06,
+      "loss": 0.4269,
+      "step": 4742
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.519132493335114,
+      "learning_rate": 7.086349731235145e-06,
+      "loss": 0.539,
+      "step": 4743
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 19.625083198312428,
+      "learning_rate": 7.0851639202422175e-06,
+      "loss": 0.4568,
+      "step": 4744
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.080510918298675,
+      "learning_rate": 7.083977967258847e-06,
+      "loss": 0.5524,
+      "step": 4745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.58561904496669,
+      "learning_rate": 7.082791872365791e-06,
+      "loss": 0.4721,
+      "step": 4746
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 30.359941536747467,
+      "learning_rate": 7.081605635643818e-06,
+      "loss": 0.3655,
+      "step": 4747
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.971682278377182,
+      "learning_rate": 7.080419257173704e-06,
+      "loss": 0.3907,
+      "step": 4748
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.712907038934617,
+      "learning_rate": 7.0792327370362365e-06,
+      "loss": 0.5726,
+      "step": 4749
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 46.1754361129663,
+      "learning_rate": 7.078046075312215e-06,
+      "loss": 0.5081,
+      "step": 4750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8605408208856733,
+      "learning_rate": 7.076859272082441e-06,
+      "loss": 0.8099,
+      "step": 4751
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.183025310917992,
+      "learning_rate": 7.075672327427735e-06,
+      "loss": 0.5122,
+      "step": 4752
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 67.82751636936288,
+      "learning_rate": 7.074485241428921e-06,
+      "loss": 0.4082,
+      "step": 4753
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 32.37673882964795,
+      "learning_rate": 7.073298014166834e-06,
+      "loss": 0.5812,
+      "step": 4754
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.999360126668867,
+      "learning_rate": 7.072110645722318e-06,
+      "loss": 0.5588,
+      "step": 4755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.754947609649886,
+      "learning_rate": 7.070923136176232e-06,
+      "loss": 0.4118,
+      "step": 4756
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.441018695299915,
+      "learning_rate": 7.069735485609434e-06,
+      "loss": 0.5019,
+      "step": 4757
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 29.716771275042312,
+      "learning_rate": 7.068547694102802e-06,
+      "loss": 0.442,
+      "step": 4758
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.137373730838403,
+      "learning_rate": 7.067359761737218e-06,
+      "loss": 0.3831,
+      "step": 4759
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.978248355067969,
+      "learning_rate": 7.066171688593575e-06,
+      "loss": 0.3985,
+      "step": 4760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.784545935523763,
+      "learning_rate": 7.064983474752774e-06,
+      "loss": 0.5378,
+      "step": 4761
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 25.82466846671606,
+      "learning_rate": 7.06379512029573e-06,
+      "loss": 0.4364,
+      "step": 4762
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.4558183100456,
+      "learning_rate": 7.062606625303363e-06,
+      "loss": 0.4858,
+      "step": 4763
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.12930935499367,
+      "learning_rate": 7.061417989856605e-06,
+      "loss": 0.4076,
+      "step": 4764
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 17.2948410354761,
+      "learning_rate": 7.060229214036396e-06,
+      "loss": 0.4559,
+      "step": 4765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.990884477840984,
+      "learning_rate": 7.059040297923687e-06,
+      "loss": 0.4421,
+      "step": 4766
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.303331786230983,
+      "learning_rate": 7.057851241599437e-06,
+      "loss": 0.5128,
+      "step": 4767
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.40629045124149,
+      "learning_rate": 7.056662045144617e-06,
+      "loss": 0.486,
+      "step": 4768
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 29.56667205067409,
+      "learning_rate": 7.055472708640205e-06,
+      "loss": 0.5423,
+      "step": 4769
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 22.581636662021893,
+      "learning_rate": 7.05428323216719e-06,
+      "loss": 0.5354,
+      "step": 4770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.306488635493466,
+      "learning_rate": 7.053093615806567e-06,
+      "loss": 0.5173,
+      "step": 4771
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.17189022318604,
+      "learning_rate": 7.05190385963935e-06,
+      "loss": 0.4341,
+      "step": 4772
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 20.64303962241694,
+      "learning_rate": 7.05071396374655e-06,
+      "loss": 0.609,
+      "step": 4773
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.768726399282873,
+      "learning_rate": 7.049523928209197e-06,
+      "loss": 0.4625,
+      "step": 4774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.643587007053567,
+      "learning_rate": 7.048333753108329e-06,
+      "loss": 0.4536,
+      "step": 4775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 29.93889869830261,
+      "learning_rate": 7.047143438524986e-06,
+      "loss": 0.5156,
+      "step": 4776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.103374837231913,
+      "learning_rate": 7.045952984540227e-06,
+      "loss": 0.4766,
+      "step": 4777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.09615677443988,
+      "learning_rate": 7.044762391235116e-06,
+      "loss": 0.4395,
+      "step": 4778
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.793937534232818,
+      "learning_rate": 7.0435716586907275e-06,
+      "loss": 0.4751,
+      "step": 4779
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.1070764340137,
+      "learning_rate": 7.042380786988144e-06,
+      "loss": 0.513,
+      "step": 4780
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7843108193621502,
+      "learning_rate": 7.041189776208458e-06,
+      "loss": 0.7677,
+      "step": 4781
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.417192098615907,
+      "learning_rate": 7.039998626432775e-06,
+      "loss": 0.3561,
+      "step": 4782
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.713949103733226,
+      "learning_rate": 7.038807337742204e-06,
+      "loss": 0.4493,
+      "step": 4783
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.19681096713445,
+      "learning_rate": 7.037615910217868e-06,
+      "loss": 0.5839,
+      "step": 4784
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.29871277259435,
+      "learning_rate": 7.0364243439408994e-06,
+      "loss": 0.443,
+      "step": 4785
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.041577135335395,
+      "learning_rate": 7.035232638992436e-06,
+      "loss": 0.4792,
+      "step": 4786
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 20.832753403762457,
+      "learning_rate": 7.034040795453628e-06,
+      "loss": 0.4933,
+      "step": 4787
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.046524351603367,
+      "learning_rate": 7.032848813405635e-06,
+      "loss": 0.5483,
+      "step": 4788
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.563522059661924,
+      "learning_rate": 7.031656692929628e-06,
+      "loss": 0.495,
+      "step": 4789
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.770372424395536,
+      "learning_rate": 7.030464434106781e-06,
+      "loss": 0.4397,
+      "step": 4790
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 26.344677403545617,
+      "learning_rate": 7.029272037018284e-06,
+      "loss": 0.4506,
+      "step": 4791
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.751044646235309,
+      "learning_rate": 7.028079501745334e-06,
+      "loss": 0.5054,
+      "step": 4792
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.696548418801974,
+      "learning_rate": 7.0268868283691375e-06,
+      "loss": 0.4326,
+      "step": 4793
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 22.54758081821395,
+      "learning_rate": 7.025694016970909e-06,
+      "loss": 0.5001,
+      "step": 4794
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 27.722488518477334,
+      "learning_rate": 7.024501067631875e-06,
+      "loss": 0.5493,
+      "step": 4795
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.94484422283585,
+      "learning_rate": 7.023307980433269e-06,
+      "loss": 0.4724,
+      "step": 4796
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.05545668590893,
+      "learning_rate": 7.022114755456335e-06,
+      "loss": 0.4538,
+      "step": 4797
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 34.11969302336773,
+      "learning_rate": 7.0209213927823294e-06,
+      "loss": 0.4312,
+      "step": 4798
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.14709420963477,
+      "learning_rate": 7.01972789249251e-06,
+      "loss": 0.4611,
+      "step": 4799
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 37.167142870450824,
+      "learning_rate": 7.018534254668152e-06,
+      "loss": 0.5144,
+      "step": 4800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 20.822366775704353,
+      "learning_rate": 7.017340479390535e-06,
+      "loss": 0.4527,
+      "step": 4801
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.965904787195829,
+      "learning_rate": 7.0161465667409524e-06,
+      "loss": 0.4734,
+      "step": 4802
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.063994047495562,
+      "learning_rate": 7.014952516800702e-06,
+      "loss": 0.5692,
+      "step": 4803
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.923657970101749,
+      "learning_rate": 7.013758329651096e-06,
+      "loss": 0.5511,
+      "step": 4804
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.12379426206981,
+      "learning_rate": 7.0125640053734485e-06,
+      "loss": 0.5582,
+      "step": 4805
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.94288777923747,
+      "learning_rate": 7.011369544049092e-06,
+      "loss": 0.5576,
+      "step": 4806
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.61653429094788,
+      "learning_rate": 7.0101749457593645e-06,
+      "loss": 0.5699,
+      "step": 4807
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7373097907648205,
+      "learning_rate": 7.00898021058561e-06,
+      "loss": 0.7929,
+      "step": 4808
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.415417917790872,
+      "learning_rate": 7.007785338609185e-06,
+      "loss": 0.5847,
+      "step": 4809
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.263258531891566,
+      "learning_rate": 7.006590329911456e-06,
+      "loss": 0.4909,
+      "step": 4810
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 20.906184577267158,
+      "learning_rate": 7.005395184573798e-06,
+      "loss": 0.5034,
+      "step": 4811
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.012447396926257,
+      "learning_rate": 7.004199902677594e-06,
+      "loss": 0.4422,
+      "step": 4812
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.731486156889632,
+      "learning_rate": 7.0030044843042384e-06,
+      "loss": 0.479,
+      "step": 4813
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.204875882668635,
+      "learning_rate": 7.001808929535134e-06,
+      "loss": 0.5133,
+      "step": 4814
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 20.195219677148177,
+      "learning_rate": 7.0006132384516925e-06,
+      "loss": 0.4119,
+      "step": 4815
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.3206176413025,
+      "learning_rate": 6.9994174111353355e-06,
+      "loss": 0.6021,
+      "step": 4816
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 34.48875133512773,
+      "learning_rate": 6.998221447667492e-06,
+      "loss": 0.4377,
+      "step": 4817
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.800434639685218,
+      "learning_rate": 6.997025348129605e-06,
+      "loss": 0.5051,
+      "step": 4818
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 20.220610785444933,
+      "learning_rate": 6.9958291126031185e-06,
+      "loss": 0.4775,
+      "step": 4819
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 26.71534306455965,
+      "learning_rate": 6.9946327411694955e-06,
+      "loss": 0.4587,
+      "step": 4820
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.77618532248779,
+      "learning_rate": 6.9934362339102016e-06,
+      "loss": 0.4995,
+      "step": 4821
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.542933923957833,
+      "learning_rate": 6.992239590906713e-06,
+      "loss": 0.3622,
+      "step": 4822
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.441486829556588,
+      "learning_rate": 6.99104281224052e-06,
+      "loss": 0.4194,
+      "step": 4823
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.350540801055654,
+      "learning_rate": 6.989845897993113e-06,
+      "loss": 0.4925,
+      "step": 4824
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 44.16348049191988,
+      "learning_rate": 6.988648848245998e-06,
+      "loss": 0.4216,
+      "step": 4825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.35211641995628,
+      "learning_rate": 6.98745166308069e-06,
+      "loss": 0.4914,
+      "step": 4826
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 37.114803082169686,
+      "learning_rate": 6.986254342578711e-06,
+      "loss": 0.4224,
+      "step": 4827
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.5611144156088,
+      "learning_rate": 6.985056886821593e-06,
+      "loss": 0.5919,
+      "step": 4828
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.079983096615338,
+      "learning_rate": 6.983859295890878e-06,
+      "loss": 0.4793,
+      "step": 4829
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.21455220539513,
+      "learning_rate": 6.982661569868116e-06,
+      "loss": 0.3985,
+      "step": 4830
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.043922814520055,
+      "learning_rate": 6.981463708834867e-06,
+      "loss": 0.5443,
+      "step": 4831
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.834664845731615,
+      "learning_rate": 6.980265712872702e-06,
+      "loss": 0.7652,
+      "step": 4832
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.464786660691507,
+      "learning_rate": 6.979067582063197e-06,
+      "loss": 0.498,
+      "step": 4833
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.844527183344674,
+      "learning_rate": 6.97786931648794e-06,
+      "loss": 0.5009,
+      "step": 4834
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.000175644207518,
+      "learning_rate": 6.976670916228527e-06,
+      "loss": 0.5438,
+      "step": 4835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.76753636220333,
+      "learning_rate": 6.9754723813665646e-06,
+      "loss": 0.5405,
+      "step": 4836
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 24.85826562100113,
+      "learning_rate": 6.974273711983668e-06,
+      "loss": 0.4385,
+      "step": 4837
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.75707491492848,
+      "learning_rate": 6.97307490816146e-06,
+      "loss": 0.5955,
+      "step": 4838
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 53.399613785138996,
+      "learning_rate": 6.971875969981576e-06,
+      "loss": 0.5477,
+      "step": 4839
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.61155005474755,
+      "learning_rate": 6.970676897525656e-06,
+      "loss": 0.4872,
+      "step": 4840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 36.60252741932084,
+      "learning_rate": 6.969477690875354e-06,
+      "loss": 0.5314,
+      "step": 4841
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 26.06400413071724,
+      "learning_rate": 6.968278350112326e-06,
+      "loss": 0.4763,
+      "step": 4842
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.36120227252667,
+      "learning_rate": 6.9670788753182475e-06,
+      "loss": 0.4954,
+      "step": 4843
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.502123590130811,
+      "learning_rate": 6.9658792665747945e-06,
+      "loss": 0.4611,
+      "step": 4844
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.764207418023382,
+      "learning_rate": 6.964679523963654e-06,
+      "loss": 0.455,
+      "step": 4845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.02935907277024,
+      "learning_rate": 6.963479647566527e-06,
+      "loss": 0.5925,
+      "step": 4846
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 46.41975488087757,
+      "learning_rate": 6.962279637465117e-06,
+      "loss": 0.4438,
+      "step": 4847
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.08686385940896,
+      "learning_rate": 6.96107949374114e-06,
+      "loss": 0.4887,
+      "step": 4848
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.76852688464956,
+      "learning_rate": 6.959879216476319e-06,
+      "loss": 0.458,
+      "step": 4849
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.249936891900095,
+      "learning_rate": 6.958678805752389e-06,
+      "loss": 0.4593,
+      "step": 4850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 299.44819637663653,
+      "learning_rate": 6.957478261651094e-06,
+      "loss": 0.5104,
+      "step": 4851
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 27.570999645043152,
+      "learning_rate": 6.956277584254183e-06,
+      "loss": 0.5671,
+      "step": 4852
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 22.04790222795693,
+      "learning_rate": 6.955076773643418e-06,
+      "loss": 0.4328,
+      "step": 4853
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 40.887997610128664,
+      "learning_rate": 6.953875829900568e-06,
+      "loss": 0.4707,
+      "step": 4854
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.553736495768778,
+      "learning_rate": 6.952674753107415e-06,
+      "loss": 0.4574,
+      "step": 4855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 40.85683291378138,
+      "learning_rate": 6.951473543345745e-06,
+      "loss": 0.4945,
+      "step": 4856
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 29.333676379835943,
+      "learning_rate": 6.950272200697355e-06,
+      "loss": 0.3348,
+      "step": 4857
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 142.71470953726455,
+      "learning_rate": 6.949070725244049e-06,
+      "loss": 0.5872,
+      "step": 4858
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.385165680942086,
+      "learning_rate": 6.947869117067647e-06,
+      "loss": 0.5547,
+      "step": 4859
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.508048177153718,
+      "learning_rate": 6.946667376249969e-06,
+      "loss": 0.5177,
+      "step": 4860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.37733830882154,
+      "learning_rate": 6.94546550287285e-06,
+      "loss": 0.5604,
+      "step": 4861
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 83.79228105542684,
+      "learning_rate": 6.944263497018132e-06,
+      "loss": 0.493,
+      "step": 4862
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 36.68856378685893,
+      "learning_rate": 6.943061358767666e-06,
+      "loss": 0.424,
+      "step": 4863
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 51.90161960303694,
+      "learning_rate": 6.941859088203314e-06,
+      "loss": 0.5246,
+      "step": 4864
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 50.1146497062552,
+      "learning_rate": 6.940656685406943e-06,
+      "loss": 0.4181,
+      "step": 4865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 23.33589334025303,
+      "learning_rate": 6.939454150460434e-06,
+      "loss": 0.531,
+      "step": 4866
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.20167664732198,
+      "learning_rate": 6.93825148344567e-06,
+      "loss": 0.4987,
+      "step": 4867
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.670035241999333,
+      "learning_rate": 6.937048684444551e-06,
+      "loss": 0.3915,
+      "step": 4868
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 16.616053778049295,
+      "learning_rate": 6.9358457535389815e-06,
+      "loss": 0.4898,
+      "step": 4869
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 31.955812692593085,
+      "learning_rate": 6.934642690810874e-06,
+      "loss": 0.5422,
+      "step": 4870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7171331992052543,
+      "learning_rate": 6.933439496342156e-06,
+      "loss": 0.7662,
+      "step": 4871
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.37299213536648,
+      "learning_rate": 6.932236170214753e-06,
+      "loss": 0.516,
+      "step": 4872
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 23.287042489444293,
+      "learning_rate": 6.931032712510614e-06,
+      "loss": 0.5227,
+      "step": 4873
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 23.26251731282772,
+      "learning_rate": 6.929829123311682e-06,
+      "loss": 0.5224,
+      "step": 4874
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.408564047429778,
+      "learning_rate": 6.9286254026999224e-06,
+      "loss": 0.4622,
+      "step": 4875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.14569370623185,
+      "learning_rate": 6.927421550757299e-06,
+      "loss": 0.5105,
+      "step": 4876
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 26.11086612402754,
+      "learning_rate": 6.926217567565788e-06,
+      "loss": 0.6091,
+      "step": 4877
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.54096757098868,
+      "learning_rate": 6.92501345320738e-06,
+      "loss": 0.4233,
+      "step": 4878
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.359914531068393,
+      "learning_rate": 6.923809207764065e-06,
+      "loss": 0.4094,
+      "step": 4879
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 68.02733550614299,
+      "learning_rate": 6.9226048313178515e-06,
+      "loss": 0.6458,
+      "step": 4880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 35.548159028193936,
+      "learning_rate": 6.921400323950749e-06,
+      "loss": 0.4561,
+      "step": 4881
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.438309192154527,
+      "learning_rate": 6.920195685744778e-06,
+      "loss": 0.3582,
+      "step": 4882
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 30.161318538700073,
+      "learning_rate": 6.918990916781971e-06,
+      "loss": 0.5544,
+      "step": 4883
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 36.779874961969384,
+      "learning_rate": 6.917786017144369e-06,
+      "loss": 0.4718,
+      "step": 4884
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 64.88564048731165,
+      "learning_rate": 6.916580986914019e-06,
+      "loss": 0.5809,
+      "step": 4885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.8429185179219,
+      "learning_rate": 6.915375826172976e-06,
+      "loss": 0.3773,
+      "step": 4886
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 44.29804759365776,
+      "learning_rate": 6.9141705350033095e-06,
+      "loss": 0.3702,
+      "step": 4887
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.427840718622189,
+      "learning_rate": 6.912965113487092e-06,
+      "loss": 0.448,
+      "step": 4888
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 18.03763011739873,
+      "learning_rate": 6.911759561706409e-06,
+      "loss": 0.4889,
+      "step": 4889
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.681828958683113,
+      "learning_rate": 6.910553879743352e-06,
+      "loss": 0.5006,
+      "step": 4890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.776185680744474,
+      "learning_rate": 6.909348067680024e-06,
+      "loss": 0.4601,
+      "step": 4891
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 25.25145860077523,
+      "learning_rate": 6.908142125598534e-06,
+      "loss": 0.5606,
+      "step": 4892
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 26.735350161318134,
+      "learning_rate": 6.906936053581001e-06,
+      "loss": 0.4979,
+      "step": 4893
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 21.111285313679982,
+      "learning_rate": 6.9057298517095576e-06,
+      "loss": 0.5139,
+      "step": 4894
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 41.68956399680393,
+      "learning_rate": 6.904523520066334e-06,
+      "loss": 0.4076,
+      "step": 4895
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 33.54055363172813,
+      "learning_rate": 6.903317058733481e-06,
+      "loss": 0.4483,
+      "step": 4896
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 36.35892645260604,
+      "learning_rate": 6.902110467793152e-06,
+      "loss": 0.6145,
+      "step": 4897
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 23.248581885319364,
+      "learning_rate": 6.900903747327508e-06,
+      "loss": 0.5326,
+      "step": 4898
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 38.78605587512912,
+      "learning_rate": 6.899696897418726e-06,
+      "loss": 0.4479,
+      "step": 4899
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 30.930724910781432,
+      "learning_rate": 6.898489918148985e-06,
+      "loss": 0.5211,
+      "step": 4900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.77507924294102,
+      "learning_rate": 6.897282809600474e-06,
+      "loss": 0.5052,
+      "step": 4901
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.62117532903126,
+      "learning_rate": 6.896075571855391e-06,
+      "loss": 0.4308,
+      "step": 4902
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.297938987680528,
+      "learning_rate": 6.894868204995946e-06,
+      "loss": 0.5229,
+      "step": 4903
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.582019966589666,
+      "learning_rate": 6.893660709104354e-06,
+      "loss": 0.4918,
+      "step": 4904
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 18.539726381476658,
+      "learning_rate": 6.892453084262842e-06,
+      "loss": 0.4963,
+      "step": 4905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 37.838945619986475,
+      "learning_rate": 6.891245330553641e-06,
+      "loss": 0.4172,
+      "step": 4906
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 26.22386482973563,
+      "learning_rate": 6.8900374480589946e-06,
+      "loss": 0.4797,
+      "step": 4907
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.335067742683718,
+      "learning_rate": 6.888829436861155e-06,
+      "loss": 0.5206,
+      "step": 4908
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 28.130535976703673,
+      "learning_rate": 6.8876212970423826e-06,
+      "loss": 0.4692,
+      "step": 4909
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 260.5216726811766,
+      "learning_rate": 6.8864130286849465e-06,
+      "loss": 0.4262,
+      "step": 4910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 335.2618112662547,
+      "learning_rate": 6.885204631871123e-06,
+      "loss": 0.4332,
+      "step": 4911
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 20.728154270368993,
+      "learning_rate": 6.8839961066831995e-06,
+      "loss": 0.474,
+      "step": 4912
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 26.112740272073356,
+      "learning_rate": 6.88278745320347e-06,
+      "loss": 0.4715,
+      "step": 4913
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.494430084951357,
+      "learning_rate": 6.881578671514242e-06,
+      "loss": 0.5405,
+      "step": 4914
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 41.061344527783824,
+      "learning_rate": 6.880369761697824e-06,
+      "loss": 0.5027,
+      "step": 4915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.273069767950012,
+      "learning_rate": 6.87916072383654e-06,
+      "loss": 0.5154,
+      "step": 4916
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.635020476922946,
+      "learning_rate": 6.877951558012719e-06,
+      "loss": 0.5382,
+      "step": 4917
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6170243609301729,
+      "learning_rate": 6.8767422643087e-06,
+      "loss": 0.7796,
+      "step": 4918
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 37.89664084415061,
+      "learning_rate": 6.875532842806831e-06,
+      "loss": 0.4469,
+      "step": 4919
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.804221788414477,
+      "learning_rate": 6.874323293589469e-06,
+      "loss": 0.5106,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 54.044953013640374,
+      "learning_rate": 6.873113616738976e-06,
+      "loss": 0.5113,
+      "step": 4921
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 80.20059968499869,
+      "learning_rate": 6.871903812337728e-06,
+      "loss": 0.5508,
+      "step": 4922
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.682483749270077,
+      "learning_rate": 6.870693880468109e-06,
+      "loss": 0.5704,
+      "step": 4923
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 31.246851914179654,
+      "learning_rate": 6.869483821212506e-06,
+      "loss": 0.5127,
+      "step": 4924
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.051523561622265,
+      "learning_rate": 6.86827363465332e-06,
+      "loss": 0.5681,
+      "step": 4925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.570771698722565,
+      "learning_rate": 6.867063320872962e-06,
+      "loss": 0.7058,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.23230593853571,
+      "learning_rate": 6.865852879953845e-06,
+      "loss": 0.5237,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 59.554137037572175,
+      "learning_rate": 6.864642311978399e-06,
+      "loss": 0.451,
+      "step": 4928
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 35.90669302620372,
+      "learning_rate": 6.863431617029054e-06,
+      "loss": 0.3659,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.97180070154394,
+      "learning_rate": 6.862220795188255e-06,
+      "loss": 0.4831,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.15402650635392,
+      "learning_rate": 6.861009846538456e-06,
+      "loss": 0.4061,
+      "step": 4931
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.93056367044174,
+      "learning_rate": 6.859798771162111e-06,
+      "loss": 0.4945,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 30.057221904266857,
+      "learning_rate": 6.858587569141695e-06,
+      "loss": 0.6147,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 24.817925516829696,
+      "learning_rate": 6.857376240559683e-06,
+      "loss": 0.4377,
+      "step": 4934
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.936686042393617,
+      "learning_rate": 6.856164785498561e-06,
+      "loss": 0.451,
+      "step": 4935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 29.671862771011437,
+      "learning_rate": 6.854953204040824e-06,
+      "loss": 0.5428,
+      "step": 4936
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.671955266131665,
+      "learning_rate": 6.853741496268978e-06,
+      "loss": 0.5523,
+      "step": 4937
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.32025580623945,
+      "learning_rate": 6.852529662265528e-06,
+      "loss": 0.5307,
+      "step": 4938
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.166411607357425,
+      "learning_rate": 6.851317702113001e-06,
+      "loss": 0.5799,
+      "step": 4939
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 28.93764830942758,
+      "learning_rate": 6.850105615893925e-06,
+      "loss": 0.5608,
+      "step": 4940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.697911478951562,
+      "learning_rate": 6.848893403690835e-06,
+      "loss": 0.4517,
+      "step": 4941
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 43.33469158070088,
+      "learning_rate": 6.847681065586279e-06,
+      "loss": 0.4805,
+      "step": 4942
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 81.01105703593949,
+      "learning_rate": 6.846468601662812e-06,
+      "loss": 0.485,
+      "step": 4943
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8070991144145874,
+      "learning_rate": 6.845256012002999e-06,
+      "loss": 0.8261,
+      "step": 4944
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 17.672908760724248,
+      "learning_rate": 6.844043296689409e-06,
+      "loss": 0.4698,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.620288677903696,
+      "learning_rate": 6.8428304558046245e-06,
+      "loss": 0.4898,
+      "step": 4946
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.971108252512552,
+      "learning_rate": 6.841617489431234e-06,
+      "loss": 0.3871,
+      "step": 4947
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 66.88207787043963,
+      "learning_rate": 6.840404397651833e-06,
+      "loss": 0.566,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 61.94081754167107,
+      "learning_rate": 6.839191180549032e-06,
+      "loss": 0.5809,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.442644549338663,
+      "learning_rate": 6.837977838205443e-06,
+      "loss": 0.4554,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 53.69669511841583,
+      "learning_rate": 6.83676437070369e-06,
+      "loss": 0.5751,
+      "step": 4951
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 51.370696384412575,
+      "learning_rate": 6.835550778126403e-06,
+      "loss": 0.5591,
+      "step": 4952
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.754445929824627,
+      "learning_rate": 6.8343370605562255e-06,
+      "loss": 0.412,
+      "step": 4953
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 48.93721997861718,
+      "learning_rate": 6.8331232180758035e-06,
+      "loss": 0.3732,
+      "step": 4954
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 116.04265746858317,
+      "learning_rate": 6.8319092507677965e-06,
+      "loss": 0.4268,
+      "step": 4955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 53.04222052449719,
+      "learning_rate": 6.830695158714869e-06,
+      "loss": 0.4865,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 92.26045629762675,
+      "learning_rate": 6.829480941999694e-06,
+      "loss": 0.6032,
+      "step": 4957
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.583092628621566,
+      "learning_rate": 6.828266600704958e-06,
+      "loss": 0.4912,
+      "step": 4958
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 52.707706473987386,
+      "learning_rate": 6.827052134913349e-06,
+      "loss": 0.4049,
+      "step": 4959
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 45.47065541969783,
+      "learning_rate": 6.825837544707569e-06,
+      "loss": 0.5139,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 38.06610879492803,
+      "learning_rate": 6.824622830170325e-06,
+      "loss": 0.5708,
+      "step": 4961
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 48.26333207180999,
+      "learning_rate": 6.823407991384331e-06,
+      "loss": 0.516,
+      "step": 4962
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 32.55066372908863,
+      "learning_rate": 6.8221930284323165e-06,
+      "loss": 0.5695,
+      "step": 4963
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 47.767477040692704,
+      "learning_rate": 6.820977941397014e-06,
+      "loss": 0.4274,
+      "step": 4964
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 72.94560611722221,
+      "learning_rate": 6.819762730361166e-06,
+      "loss": 0.4487,
+      "step": 4965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 110.07426500943747,
+      "learning_rate": 6.8185473954075195e-06,
+      "loss": 0.4398,
+      "step": 4966
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 151.57676341945555,
+      "learning_rate": 6.817331936618839e-06,
+      "loss": 0.4764,
+      "step": 4967
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.47353342054602,
+      "learning_rate": 6.816116354077886e-06,
+      "loss": 0.4762,
+      "step": 4968
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 45.26174327094636,
+      "learning_rate": 6.81490064786744e-06,
+      "loss": 0.485,
+      "step": 4969
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 84.11397601354369,
+      "learning_rate": 6.813684818070285e-06,
+      "loss": 0.5567,
+      "step": 4970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 32.81149809735091,
+      "learning_rate": 6.812468864769213e-06,
+      "loss": 0.5093,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.95286479554115,
+      "learning_rate": 6.8112527880470245e-06,
+      "loss": 0.5665,
+      "step": 4972
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.294775434158947,
+      "learning_rate": 6.810036587986531e-06,
+      "loss": 0.6192,
+      "step": 4973
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 42.42805520866898,
+      "learning_rate": 6.808820264670547e-06,
+      "loss": 0.573,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 51.5526420633163,
+      "learning_rate": 6.807603818181902e-06,
+      "loss": 0.5289,
+      "step": 4975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 32.63452726120556,
+      "learning_rate": 6.806387248603429e-06,
+      "loss": 0.3731,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 30.91236420568391,
+      "learning_rate": 6.805170556017971e-06,
+      "loss": 0.4834,
+      "step": 4977
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 38.51228842918579,
+      "learning_rate": 6.803953740508378e-06,
+      "loss": 0.5097,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 44.14557233420398,
+      "learning_rate": 6.802736802157514e-06,
+      "loss": 0.408,
+      "step": 4979
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 47.940667145764024,
+      "learning_rate": 6.801519741048242e-06,
+      "loss": 0.4896,
+      "step": 4980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.867521706707285,
+      "learning_rate": 6.800302557263443e-06,
+      "loss": 0.4672,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.29650034684226,
+      "learning_rate": 6.799085250885999e-06,
+      "loss": 0.5432,
+      "step": 4982
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 22.750139327023216,
+      "learning_rate": 6.797867821998803e-06,
+      "loss": 0.6078,
+      "step": 4983
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 139.01725506642344,
+      "learning_rate": 6.796650270684759e-06,
+      "loss": 0.4638,
+      "step": 4984
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9628172643536062,
+      "learning_rate": 6.795432597026776e-06,
+      "loss": 0.8204,
+      "step": 4985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 45.57313127406531,
+      "learning_rate": 6.7942148011077704e-06,
+      "loss": 0.4821,
+      "step": 4986
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 124.87925059441069,
+      "learning_rate": 6.79299688301067e-06,
+      "loss": 0.4634,
+      "step": 4987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.01341159451641,
+      "learning_rate": 6.791778842818409e-06,
+      "loss": 0.4586,
+      "step": 4988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8119729883347704,
+      "learning_rate": 6.790560680613932e-06,
+      "loss": 0.7749,
+      "step": 4989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.512826862412094,
+      "learning_rate": 6.789342396480189e-06,
+      "loss": 0.5259,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.404718649433732,
+      "learning_rate": 6.788123990500142e-06,
+      "loss": 0.4107,
+      "step": 4991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.95526275118313,
+      "learning_rate": 6.786905462756754e-06,
+      "loss": 0.3538,
+      "step": 4992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 30.10896147536327,
+      "learning_rate": 6.785686813333006e-06,
+      "loss": 0.516,
+      "step": 4993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 36.607679025371084,
+      "learning_rate": 6.784468042311883e-06,
+      "loss": 0.6074,
+      "step": 4994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.979325458163707,
+      "learning_rate": 6.783249149776375e-06,
+      "loss": 0.4827,
+      "step": 4995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 31.482914620447772,
+      "learning_rate": 6.782030135809484e-06,
+      "loss": 0.5489,
+      "step": 4996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 26.309418662666157,
+      "learning_rate": 6.780811000494222e-06,
+      "loss": 0.4126,
+      "step": 4997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.154612625254007,
+      "learning_rate": 6.779591743913603e-06,
+      "loss": 0.5069,
+      "step": 4998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 25.449752250373415,
+      "learning_rate": 6.778372366150655e-06,
+      "loss": 0.5344,
+      "step": 4999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.16679170983966,
+      "learning_rate": 6.777152867288412e-06,
+      "loss": 0.5882,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 170.8158992811127,
+      "learning_rate": 6.775933247409915e-06,
+      "loss": 0.6231,
+      "step": 5001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 54.87394527826737,
+      "learning_rate": 6.774713506598218e-06,
+      "loss": 0.5733,
+      "step": 5002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.487907323470388,
+      "learning_rate": 6.773493644936375e-06,
+      "loss": 0.4282,
+      "step": 5003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 24.96844585568688,
+      "learning_rate": 6.772273662507461e-06,
+      "loss": 0.432,
+      "step": 5004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 56.6924789112057,
+      "learning_rate": 6.771053559394543e-06,
+      "loss": 0.531,
+      "step": 5005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8480555966329795,
+      "learning_rate": 6.769833335680709e-06,
+      "loss": 0.7548,
+      "step": 5006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.011348468340335,
+      "learning_rate": 6.76861299144905e-06,
+      "loss": 0.6067,
+      "step": 5007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 26.37822822223767,
+      "learning_rate": 6.767392526782667e-06,
+      "loss": 0.5269,
+      "step": 5008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 27.634077020246327,
+      "learning_rate": 6.766171941764667e-06,
+      "loss": 0.5434,
+      "step": 5009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.17147482229825,
+      "learning_rate": 6.7649512364781655e-06,
+      "loss": 0.4385,
+      "step": 5010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 45.80306407525706,
+      "learning_rate": 6.7637304110062905e-06,
+      "loss": 0.4401,
+      "step": 5011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.953522968499657,
+      "learning_rate": 6.762509465432172e-06,
+      "loss": 0.4188,
+      "step": 5012
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.178224529512567,
+      "learning_rate": 6.76128839983895e-06,
+      "loss": 0.4877,
+      "step": 5013
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.70596745397901,
+      "learning_rate": 6.7600672143097775e-06,
+      "loss": 0.5308,
+      "step": 5014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 18.61974912713037,
+      "learning_rate": 6.758845908927809e-06,
+      "loss": 0.477,
+      "step": 5015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.001243833238615,
+      "learning_rate": 6.757624483776212e-06,
+      "loss": 0.4984,
+      "step": 5016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 21.589582741708025,
+      "learning_rate": 6.756402938938157e-06,
+      "loss": 0.4941,
+      "step": 5017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 16.66940682840448,
+      "learning_rate": 6.7551812744968295e-06,
+      "loss": 0.5091,
+      "step": 5018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.54155972130823,
+      "learning_rate": 6.7539594905354166e-06,
+      "loss": 0.5003,
+      "step": 5019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 18.144264299451248,
+      "learning_rate": 6.752737587137118e-06,
+      "loss": 0.5447,
+      "step": 5020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 15.705832525582263,
+      "learning_rate": 6.751515564385138e-06,
+      "loss": 0.5432,
+      "step": 5021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.34571366545852,
+      "learning_rate": 6.750293422362695e-06,
+      "loss": 0.4907,
+      "step": 5022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.65540386436675,
+      "learning_rate": 6.749071161153006e-06,
+      "loss": 0.5307,
+      "step": 5023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.490955738341109,
+      "learning_rate": 6.747848780839306e-06,
+      "loss": 0.3879,
+      "step": 5024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 20.996618238073548,
+      "learning_rate": 6.7466262815048315e-06,
+      "loss": 0.6033,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.319394918367976,
+      "learning_rate": 6.745403663232829e-06,
+      "loss": 0.5429,
+      "step": 5026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 23.092026393491484,
+      "learning_rate": 6.744180926106556e-06,
+      "loss": 0.4348,
+      "step": 5027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 25.701004537612228,
+      "learning_rate": 6.742958070209273e-06,
+      "loss": 0.4477,
+      "step": 5028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 42.43991072807157,
+      "learning_rate": 6.741735095624252e-06,
+      "loss": 0.5026,
+      "step": 5029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.114731119342537,
+      "learning_rate": 6.740512002434771e-06,
+      "loss": 0.5872,
+      "step": 5030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.781862937768175,
+      "learning_rate": 6.739288790724121e-06,
+      "loss": 0.4117,
+      "step": 5031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 27.504402880432117,
+      "learning_rate": 6.7380654605755915e-06,
+      "loss": 0.4651,
+      "step": 5032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.330248035230532,
+      "learning_rate": 6.73684201207249e-06,
+      "loss": 0.4832,
+      "step": 5033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.651936876698045,
+      "learning_rate": 6.735618445298126e-06,
+      "loss": 0.5398,
+      "step": 5034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.034034654507455,
+      "learning_rate": 6.7343947603358205e-06,
+      "loss": 0.4914,
+      "step": 5035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.334749571480906,
+      "learning_rate": 6.733170957268902e-06,
+      "loss": 0.5399,
+      "step": 5036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 88.76019781219121,
+      "learning_rate": 6.731947036180702e-06,
+      "loss": 0.5338,
+      "step": 5037
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.826534119717664,
+      "learning_rate": 6.730722997154568e-06,
+      "loss": 0.4426,
+      "step": 5038
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 40.21277884310467,
+      "learning_rate": 6.729498840273849e-06,
+      "loss": 0.4909,
+      "step": 5039
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.79575181794563,
+      "learning_rate": 6.728274565621907e-06,
+      "loss": 0.491,
+      "step": 5040
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.59825703387802,
+      "learning_rate": 6.727050173282109e-06,
+      "loss": 0.4295,
+      "step": 5041
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 35.524531176774865,
+      "learning_rate": 6.725825663337829e-06,
+      "loss": 0.4584,
+      "step": 5042
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.316516996825841,
+      "learning_rate": 6.724601035872452e-06,
+      "loss": 0.4395,
+      "step": 5043
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 25.451895731024766,
+      "learning_rate": 6.72337629096937e-06,
+      "loss": 0.4305,
+      "step": 5044
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.111055620414987,
+      "learning_rate": 6.722151428711983e-06,
+      "loss": 0.4997,
+      "step": 5045
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.566372320116892,
+      "learning_rate": 6.720926449183698e-06,
+      "loss": 0.4303,
+      "step": 5046
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 55.708624686438746,
+      "learning_rate": 6.719701352467931e-06,
+      "loss": 0.5113,
+      "step": 5047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.451240957921662,
+      "learning_rate": 6.7184761386481045e-06,
+      "loss": 0.4748,
+      "step": 5048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 24.653077691248566,
+      "learning_rate": 6.717250807807652e-06,
+      "loss": 0.4761,
+      "step": 5049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 24.500114908693522,
+      "learning_rate": 6.7160253600300115e-06,
+      "loss": 0.4409,
+      "step": 5050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 25.731738454086393,
+      "learning_rate": 6.714799795398632e-06,
+      "loss": 0.4378,
+      "step": 5051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.53786331796245,
+      "learning_rate": 6.713574113996969e-06,
+      "loss": 0.5399,
+      "step": 5052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 41.742608484220604,
+      "learning_rate": 6.712348315908483e-06,
+      "loss": 0.4132,
+      "step": 5053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 29.144623163139038,
+      "learning_rate": 6.71112240121665e-06,
+      "loss": 0.3957,
+      "step": 5054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.965365076393432,
+      "learning_rate": 6.709896370004946e-06,
+      "loss": 0.5558,
+      "step": 5055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.695287474079159,
+      "learning_rate": 6.70867022235686e-06,
+      "loss": 0.485,
+      "step": 5056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.56414825019872,
+      "learning_rate": 6.707443958355886e-06,
+      "loss": 0.4552,
+      "step": 5057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.121010690343116,
+      "learning_rate": 6.706217578085528e-06,
+      "loss": 0.4294,
+      "step": 5058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.013826981750396,
+      "learning_rate": 6.7049910816292975e-06,
+      "loss": 0.4915,
+      "step": 5059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.906110415772854,
+      "learning_rate": 6.703764469070713e-06,
+      "loss": 0.4881,
+      "step": 5060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.802072991744406,
+      "learning_rate": 6.7025377404933e-06,
+      "loss": 0.3615,
+      "step": 5061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.582391240686178,
+      "learning_rate": 6.701310895980596e-06,
+      "loss": 0.4728,
+      "step": 5062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.62565736672444,
+      "learning_rate": 6.7000839356161405e-06,
+      "loss": 0.4564,
+      "step": 5063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.634677228898923,
+      "learning_rate": 6.698856859483487e-06,
+      "loss": 0.5104,
+      "step": 5064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.30287829219067,
+      "learning_rate": 6.6976296676661925e-06,
+      "loss": 0.4534,
+      "step": 5065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.75440636684261,
+      "learning_rate": 6.696402360247825e-06,
+      "loss": 0.4844,
+      "step": 5066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 60.71707617681646,
+      "learning_rate": 6.695174937311955e-06,
+      "loss": 0.5481,
+      "step": 5067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 34.65351905426145,
+      "learning_rate": 6.693947398942169e-06,
+      "loss": 0.3727,
+      "step": 5068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.837086245357,
+      "learning_rate": 6.692719745222054e-06,
+      "loss": 0.4644,
+      "step": 5069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.703096110938162,
+      "learning_rate": 6.691491976235209e-06,
+      "loss": 0.5255,
+      "step": 5070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 28.970343264173064,
+      "learning_rate": 6.690264092065238e-06,
+      "loss": 0.576,
+      "step": 5071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 54.60395901112345,
+      "learning_rate": 6.689036092795757e-06,
+      "loss": 0.4986,
+      "step": 5072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.79368282426744,
+      "learning_rate": 6.687807978510385e-06,
+      "loss": 0.5432,
+      "step": 5073
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.557422089037438,
+      "learning_rate": 6.686579749292753e-06,
+      "loss": 0.442,
+      "step": 5074
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 372.8980560471404,
+      "learning_rate": 6.685351405226498e-06,
+      "loss": 0.4494,
+      "step": 5075
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 43.845711477841164,
+      "learning_rate": 6.684122946395263e-06,
+      "loss": 0.5092,
+      "step": 5076
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.919939963653015,
+      "learning_rate": 6.682894372882702e-06,
+      "loss": 0.5864,
+      "step": 5077
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.325303358341507,
+      "learning_rate": 6.681665684772475e-06,
+      "loss": 0.461,
+      "step": 5078
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.957155724776355,
+      "learning_rate": 6.6804368821482504e-06,
+      "loss": 0.5693,
+      "step": 5079
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 30.29931909674975,
+      "learning_rate": 6.679207965093703e-06,
+      "loss": 0.5418,
+      "step": 5080
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 78.79453537104348,
+      "learning_rate": 6.677978933692518e-06,
+      "loss": 0.4283,
+      "step": 5081
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.669568770092198,
+      "learning_rate": 6.676749788028388e-06,
+      "loss": 0.4636,
+      "step": 5082
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.088700522877222,
+      "learning_rate": 6.67552052818501e-06,
+      "loss": 0.4847,
+      "step": 5083
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.292115321834714,
+      "learning_rate": 6.674291154246093e-06,
+      "loss": 0.5469,
+      "step": 5084
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.880245552946974,
+      "learning_rate": 6.67306166629535e-06,
+      "loss": 0.4686,
+      "step": 5085
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.819436293010078,
+      "learning_rate": 6.671832064416507e-06,
+      "loss": 0.5577,
+      "step": 5086
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 58.01667425900393,
+      "learning_rate": 6.670602348693291e-06,
+      "loss": 0.5148,
+      "step": 5087
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6661588938818779,
+      "learning_rate": 6.669372519209441e-06,
+      "loss": 0.7639,
+      "step": 5088
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.907269414909624,
+      "learning_rate": 6.668142576048705e-06,
+      "loss": 0.4432,
+      "step": 5089
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8268760244625488,
+      "learning_rate": 6.666912519294833e-06,
+      "loss": 0.8049,
+      "step": 5090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 25.28682932410093,
+      "learning_rate": 6.665682349031592e-06,
+      "loss": 0.5914,
+      "step": 5091
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 35.10539972160337,
+      "learning_rate": 6.664452065342745e-06,
+      "loss": 0.4843,
+      "step": 5092
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.71828333167138,
+      "learning_rate": 6.663221668312072e-06,
+      "loss": 0.453,
+      "step": 5093
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.642106243294982,
+      "learning_rate": 6.661991158023358e-06,
+      "loss": 0.4692,
+      "step": 5094
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 78.6902626425506,
+      "learning_rate": 6.6607605345603945e-06,
+      "loss": 0.4345,
+      "step": 5095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 37.26179769841436,
+      "learning_rate": 6.659529798006981e-06,
+      "loss": 0.4745,
+      "step": 5096
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.623080608882724,
+      "learning_rate": 6.658298948446925e-06,
+      "loss": 0.461,
+      "step": 5097
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.655439691396012,
+      "learning_rate": 6.657067985964045e-06,
+      "loss": 0.488,
+      "step": 5098
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.281481602603527,
+      "learning_rate": 6.65583691064216e-06,
+      "loss": 0.8384,
+      "step": 5099
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 27.561862620319246,
+      "learning_rate": 6.654605722565104e-06,
+      "loss": 0.5344,
+      "step": 5100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.654744504591207,
+      "learning_rate": 6.6533744218167125e-06,
+      "loss": 0.4169,
+      "step": 5101
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.643946756528235,
+      "learning_rate": 6.652143008480836e-06,
+      "loss": 0.3927,
+      "step": 5102
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.131169353300644,
+      "learning_rate": 6.650911482641321e-06,
+      "loss": 0.549,
+      "step": 5103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.310922417357776,
+      "learning_rate": 6.6496798443820375e-06,
+      "loss": 0.5015,
+      "step": 5104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 19.189999369310733,
+      "learning_rate": 6.6484480937868475e-06,
+      "loss": 0.5674,
+      "step": 5105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.57609068184319,
+      "learning_rate": 6.6472162309396326e-06,
+      "loss": 0.4414,
+      "step": 5106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 15.549352754244781,
+      "learning_rate": 6.645984255924276e-06,
+      "loss": 0.4669,
+      "step": 5107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.692041807950762,
+      "learning_rate": 6.644752168824668e-06,
+      "loss": 0.4954,
+      "step": 5108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.267830054480607,
+      "learning_rate": 6.643519969724711e-06,
+      "loss": 0.4809,
+      "step": 5109
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.269632019505778,
+      "learning_rate": 6.64228765870831e-06,
+      "loss": 0.4665,
+      "step": 5110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 25.38870029769778,
+      "learning_rate": 6.6410552358593795e-06,
+      "loss": 0.3999,
+      "step": 5111
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 50.99765816464764,
+      "learning_rate": 6.639822701261845e-06,
+      "loss": 0.4795,
+      "step": 5112
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 52.41793026035574,
+      "learning_rate": 6.638590054999634e-06,
+      "loss": 0.4136,
+      "step": 5113
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 22.071147646323347,
+      "learning_rate": 6.637357297156685e-06,
+      "loss": 0.4838,
+      "step": 5114
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.349739749943218,
+      "learning_rate": 6.6361244278169434e-06,
+      "loss": 0.4907,
+      "step": 5115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.531699719030245,
+      "learning_rate": 6.634891447064364e-06,
+      "loss": 0.5011,
+      "step": 5116
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.729729635620307,
+      "learning_rate": 6.633658354982904e-06,
+      "loss": 0.449,
+      "step": 5117
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.191450284595078,
+      "learning_rate": 6.632425151656534e-06,
+      "loss": 0.4669,
+      "step": 5118
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 17.279228615804353,
+      "learning_rate": 6.631191837169229e-06,
+      "loss": 0.553,
+      "step": 5119
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.545043523221675,
+      "learning_rate": 6.6299584116049705e-06,
+      "loss": 0.5093,
+      "step": 5120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 20.575836218279566,
+      "learning_rate": 6.628724875047754e-06,
+      "loss": 0.3893,
+      "step": 5121
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.172692617885275,
+      "learning_rate": 6.627491227581573e-06,
+      "loss": 0.5278,
+      "step": 5122
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 43.21582269555956,
+      "learning_rate": 6.626257469290436e-06,
+      "loss": 0.5556,
+      "step": 5123
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.838042661712556,
+      "learning_rate": 6.625023600258354e-06,
+      "loss": 0.4863,
+      "step": 5124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 43.062061327129676,
+      "learning_rate": 6.623789620569352e-06,
+      "loss": 0.5455,
+      "step": 5125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.65932226005602,
+      "learning_rate": 6.622555530307454e-06,
+      "loss": 0.4608,
+      "step": 5126
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.110366108384145,
+      "learning_rate": 6.6213213295567e-06,
+      "loss": 0.3904,
+      "step": 5127
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 44.9505211179757,
+      "learning_rate": 6.6200870184011324e-06,
+      "loss": 0.4379,
+      "step": 5128
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.85627140996626,
+      "learning_rate": 6.618852596924801e-06,
+      "loss": 0.4098,
+      "step": 5129
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 47.24334786266866,
+      "learning_rate": 6.6176180652117665e-06,
+      "loss": 0.4936,
+      "step": 5130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 18.201686244940934,
+      "learning_rate": 6.616383423346093e-06,
+      "loss": 0.4712,
+      "step": 5131
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.775864855951724,
+      "learning_rate": 6.615148671411856e-06,
+      "loss": 0.475,
+      "step": 5132
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 31.73358356076133,
+      "learning_rate": 6.613913809493136e-06,
+      "loss": 0.5368,
+      "step": 5133
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.50794010938095,
+      "learning_rate": 6.612678837674019e-06,
+      "loss": 0.429,
+      "step": 5134
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.06661912826307,
+      "learning_rate": 6.611443756038605e-06,
+      "loss": 0.5147,
+      "step": 5135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.05435242556648,
+      "learning_rate": 6.610208564670997e-06,
+      "loss": 0.555,
+      "step": 5136
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 30.332427397125365,
+      "learning_rate": 6.608973263655305e-06,
+      "loss": 0.456,
+      "step": 5137
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 29.313032981726362,
+      "learning_rate": 6.607737853075646e-06,
+      "loss": 0.5364,
+      "step": 5138
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.065127088347165,
+      "learning_rate": 6.6065023330161505e-06,
+      "loss": 0.598,
+      "step": 5139
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.509577189283725,
+      "learning_rate": 6.605266703560948e-06,
+      "loss": 0.4984,
+      "step": 5140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 29.491946801602573,
+      "learning_rate": 6.6040309647941805e-06,
+      "loss": 0.4441,
+      "step": 5141
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.026097739769973,
+      "learning_rate": 6.602795116799996e-06,
+      "loss": 0.505,
+      "step": 5142
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.769305698873008,
+      "learning_rate": 6.60155915966255e-06,
+      "loss": 0.4473,
+      "step": 5143
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 39.241165752403646,
+      "learning_rate": 6.600323093466008e-06,
+      "loss": 0.5331,
+      "step": 5144
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 30.028854080685054,
+      "learning_rate": 6.599086918294538e-06,
+      "loss": 0.4983,
+      "step": 5145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 26.180006079943652,
+      "learning_rate": 6.59785063423232e-06,
+      "loss": 0.4752,
+      "step": 5146
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 44.502866991792985,
+      "learning_rate": 6.596614241363539e-06,
+      "loss": 0.5128,
+      "step": 5147
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 23.064141948941373,
+      "learning_rate": 6.5953777397723876e-06,
+      "loss": 0.5469,
+      "step": 5148
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 11.578390610326071,
+      "learning_rate": 6.594141129543065e-06,
+      "loss": 0.5637,
+      "step": 5149
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 21.650608164785677,
+      "learning_rate": 6.592904410759781e-06,
+      "loss": 0.5607,
+      "step": 5150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 27.617662310478707,
+      "learning_rate": 6.591667583506748e-06,
+      "loss": 0.6076,
+      "step": 5151
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.134339958391358,
+      "learning_rate": 6.590430647868191e-06,
+      "loss": 0.5179,
+      "step": 5152
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.892957074615124,
+      "learning_rate": 6.589193603928339e-06,
+      "loss": 0.4703,
+      "step": 5153
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2152301801473344,
+      "learning_rate": 6.587956451771428e-06,
+      "loss": 0.8052,
+      "step": 5154
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.41579145402381,
+      "learning_rate": 6.586719191481705e-06,
+      "loss": 0.3911,
+      "step": 5155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.97764840040192,
+      "learning_rate": 6.58548182314342e-06,
+      "loss": 0.5746,
+      "step": 5156
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 27.78540016197237,
+      "learning_rate": 6.584244346840834e-06,
+      "loss": 0.3843,
+      "step": 5157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 52.71776485976877,
+      "learning_rate": 6.583006762658212e-06,
+      "loss": 0.5483,
+      "step": 5158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 37.76955167945157,
+      "learning_rate": 6.581769070679828e-06,
+      "loss": 0.4578,
+      "step": 5159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.143291675399375,
+      "learning_rate": 6.580531270989965e-06,
+      "loss": 0.4154,
+      "step": 5160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.761468465059085,
+      "learning_rate": 6.57929336367291e-06,
+      "loss": 0.4662,
+      "step": 5161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.848793189886884,
+      "learning_rate": 6.57805534881296e-06,
+      "loss": 0.5391,
+      "step": 5162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 28.85901508839016,
+      "learning_rate": 6.576817226494419e-06,
+      "loss": 0.6125,
+      "step": 5163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 36.80476566126683,
+      "learning_rate": 6.5755789968015945e-06,
+      "loss": 0.4732,
+      "step": 5164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 28.922849571702322,
+      "learning_rate": 6.574340659818807e-06,
+      "loss": 0.4344,
+      "step": 5165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 26.17123576372821,
+      "learning_rate": 6.573102215630385e-06,
+      "loss": 0.4783,
+      "step": 5166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.067793527700056,
+      "learning_rate": 6.571863664320654e-06,
+      "loss": 0.4653,
+      "step": 5167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 91.52309122773295,
+      "learning_rate": 6.5706250059739585e-06,
+      "loss": 0.5572,
+      "step": 5168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.401633683567916,
+      "learning_rate": 6.569386240674644e-06,
+      "loss": 0.5398,
+      "step": 5169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 20.06228218615895,
+      "learning_rate": 6.568147368507066e-06,
+      "loss": 0.465,
+      "step": 5170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 43.639280610196906,
+      "learning_rate": 6.566908389555586e-06,
+      "loss": 0.4925,
+      "step": 5171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8670145432882645,
+      "learning_rate": 6.565669303904572e-06,
+      "loss": 0.8095,
+      "step": 5172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 63.309283042904624,
+      "learning_rate": 6.5644301116384e-06,
+      "loss": 0.5771,
+      "step": 5173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 28.557591253500753,
+      "learning_rate": 6.5631908128414555e-06,
+      "loss": 0.509,
+      "step": 5174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.724746182588746,
+      "learning_rate": 6.561951407598129e-06,
+      "loss": 0.7682,
+      "step": 5175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 29.332255882018686,
+      "learning_rate": 6.560711895992818e-06,
+      "loss": 0.5576,
+      "step": 5176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.43676258040485,
+      "learning_rate": 6.559472278109926e-06,
+      "loss": 0.358,
+      "step": 5177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 36.315194804494546,
+      "learning_rate": 6.558232554033867e-06,
+      "loss": 0.515,
+      "step": 5178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 132.77473623287037,
+      "learning_rate": 6.556992723849062e-06,
+      "loss": 0.4072,
+      "step": 5179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 53.872491018348974,
+      "learning_rate": 6.555752787639936e-06,
+      "loss": 0.581,
+      "step": 5180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 24.31102074685067,
+      "learning_rate": 6.554512745490924e-06,
+      "loss": 0.4792,
+      "step": 5181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.063784960157943,
+      "learning_rate": 6.553272597486466e-06,
+      "loss": 0.4921,
+      "step": 5182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.586418611012764,
+      "learning_rate": 6.552032343711014e-06,
+      "loss": 0.4117,
+      "step": 5183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 46.57299389281683,
+      "learning_rate": 6.55079198424902e-06,
+      "loss": 0.4757,
+      "step": 5184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.728195557105845,
+      "learning_rate": 6.549551519184949e-06,
+      "loss": 0.6101,
+      "step": 5185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 41.5151894264206,
+      "learning_rate": 6.548310948603271e-06,
+      "loss": 0.4605,
+      "step": 5186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 28.98769548478009,
+      "learning_rate": 6.547070272588465e-06,
+      "loss": 0.5152,
+      "step": 5187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.075649446328168,
+      "learning_rate": 6.5458294912250105e-06,
+      "loss": 0.5415,
+      "step": 5188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.751471067022019,
+      "learning_rate": 6.544588604597405e-06,
+      "loss": 0.4601,
+      "step": 5189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.359212367041035,
+      "learning_rate": 6.543347612790143e-06,
+      "loss": 0.495,
+      "step": 5190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.48359131486582,
+      "learning_rate": 6.542106515887734e-06,
+      "loss": 0.3869,
+      "step": 5191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 28.41763673289064,
+      "learning_rate": 6.5408653139746895e-06,
+      "loss": 0.5225,
+      "step": 5192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.69355181466168,
+      "learning_rate": 6.539624007135529e-06,
+      "loss": 0.3981,
+      "step": 5193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.32220226630471,
+      "learning_rate": 6.53838259545478e-06,
+      "loss": 0.4768,
+      "step": 5194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 39.17438012066886,
+      "learning_rate": 6.53714107901698e-06,
+      "loss": 0.4855,
+      "step": 5195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.03363821365312,
+      "learning_rate": 6.53589945790667e-06,
+      "loss": 0.4771,
+      "step": 5196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 41.584935390948544,
+      "learning_rate": 6.5346577322083945e-06,
+      "loss": 0.4868,
+      "step": 5197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.939902659003998,
+      "learning_rate": 6.5334159020067156e-06,
+      "loss": 0.5771,
+      "step": 5198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 73.59840027461944,
+      "learning_rate": 6.532173967386192e-06,
+      "loss": 0.4498,
+      "step": 5199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.364933379833715,
+      "learning_rate": 6.530931928431397e-06,
+      "loss": 0.4993,
+      "step": 5200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 24.297278413734265,
+      "learning_rate": 6.529689785226906e-06,
+      "loss": 0.4576,
+      "step": 5201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.07543776658867,
+      "learning_rate": 6.528447537857303e-06,
+      "loss": 0.4434,
+      "step": 5202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.905968834531727,
+      "learning_rate": 6.527205186407183e-06,
+      "loss": 0.4608,
+      "step": 5203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 31.21877432835484,
+      "learning_rate": 6.525962730961139e-06,
+      "loss": 0.4937,
+      "step": 5204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 32.45329943111932,
+      "learning_rate": 6.524720171603784e-06,
+      "loss": 0.4161,
+      "step": 5205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.068424867063879,
+      "learning_rate": 6.523477508419723e-06,
+      "loss": 0.5162,
+      "step": 5206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.769928601340222,
+      "learning_rate": 6.522234741493582e-06,
+      "loss": 0.4728,
+      "step": 5207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 20.57865341317555,
+      "learning_rate": 6.520991870909985e-06,
+      "loss": 0.4702,
+      "step": 5208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 48.84107022445608,
+      "learning_rate": 6.519748896753566e-06,
+      "loss": 0.4041,
+      "step": 5209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.897330222739614,
+      "learning_rate": 6.518505819108969e-06,
+      "loss": 0.5569,
+      "step": 5210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.499476746846911,
+      "learning_rate": 6.517262638060837e-06,
+      "loss": 0.5059,
+      "step": 5211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 25.47192097486,
+      "learning_rate": 6.51601935369383e-06,
+      "loss": 0.4968,
+      "step": 5212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.567749024516987,
+      "learning_rate": 6.514775966092605e-06,
+      "loss": 0.4074,
+      "step": 5213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 24.08150786311464,
+      "learning_rate": 6.513532475341837e-06,
+      "loss": 0.5075,
+      "step": 5214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.671901900353927,
+      "learning_rate": 6.512288881526199e-06,
+      "loss": 0.4407,
+      "step": 5215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.580373766703989,
+      "learning_rate": 6.511045184730373e-06,
+      "loss": 0.5149,
+      "step": 5216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.378962405771233,
+      "learning_rate": 6.509801385039053e-06,
+      "loss": 0.8681,
+      "step": 5217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.886302126370433,
+      "learning_rate": 6.508557482536932e-06,
+      "loss": 0.6697,
+      "step": 5218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.044674843095876,
+      "learning_rate": 6.507313477308719e-06,
+      "loss": 0.5049,
+      "step": 5219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 29.337427931603173,
+      "learning_rate": 6.506069369439122e-06,
+      "loss": 0.4939,
+      "step": 5220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.065118061911594,
+      "learning_rate": 6.504825159012858e-06,
+      "loss": 0.4164,
+      "step": 5221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.735246600974408,
+      "learning_rate": 6.503580846114657e-06,
+      "loss": 0.645,
+      "step": 5222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 29.59013316786341,
+      "learning_rate": 6.502336430829247e-06,
+      "loss": 0.5686,
+      "step": 5223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.536096647051153,
+      "learning_rate": 6.501091913241369e-06,
+      "loss": 0.5673,
+      "step": 5224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.905060210916766,
+      "learning_rate": 6.499847293435767e-06,
+      "loss": 0.4596,
+      "step": 5225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.831519500334586,
+      "learning_rate": 6.498602571497198e-06,
+      "loss": 0.4855,
+      "step": 5226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.928684391227947,
+      "learning_rate": 6.497357747510419e-06,
+      "loss": 0.4146,
+      "step": 5227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 32.27229404668733,
+      "learning_rate": 6.496112821560199e-06,
+      "loss": 0.4886,
+      "step": 5228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.841653280100976,
+      "learning_rate": 6.494867793731309e-06,
+      "loss": 0.481,
+      "step": 5229
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.571881561733495,
+      "learning_rate": 6.493622664108532e-06,
+      "loss": 0.5081,
+      "step": 5230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.535512718563403,
+      "learning_rate": 6.492377432776657e-06,
+      "loss": 0.5142,
+      "step": 5231
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 28.542904046205784,
+      "learning_rate": 6.491132099820476e-06,
+      "loss": 0.5028,
+      "step": 5232
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9914422437783863,
+      "learning_rate": 6.489886665324794e-06,
+      "loss": 0.7512,
+      "step": 5233
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.81901612025334,
+      "learning_rate": 6.488641129374414e-06,
+      "loss": 0.5704,
+      "step": 5234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.466387201038303,
+      "learning_rate": 6.4873954920541585e-06,
+      "loss": 0.4458,
+      "step": 5235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.015572046460829,
+      "learning_rate": 6.486149753448845e-06,
+      "loss": 0.4367,
+      "step": 5236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.88496646229703,
+      "learning_rate": 6.484903913643304e-06,
+      "loss": 0.5367,
+      "step": 5237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.253362953159584,
+      "learning_rate": 6.4836579727223724e-06,
+      "loss": 0.5437,
+      "step": 5238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 26.415502254396234,
+      "learning_rate": 6.482411930770892e-06,
+      "loss": 0.4454,
+      "step": 5239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.889238522766727,
+      "learning_rate": 6.481165787873714e-06,
+      "loss": 0.5242,
+      "step": 5240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.576250785720728,
+      "learning_rate": 6.479919544115695e-06,
+      "loss": 0.5106,
+      "step": 5241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 25.269515711855888,
+      "learning_rate": 6.4786731995817e-06,
+      "loss": 0.5693,
+      "step": 5242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.229014694726544,
+      "learning_rate": 6.477426754356595e-06,
+      "loss": 0.4634,
+      "step": 5243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.778038033033255,
+      "learning_rate": 6.476180208525261e-06,
+      "loss": 0.4784,
+      "step": 5244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 27.403347624069852,
+      "learning_rate": 6.474933562172583e-06,
+      "loss": 0.4981,
+      "step": 5245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 104.02468439596517,
+      "learning_rate": 6.473686815383451e-06,
+      "loss": 0.4646,
+      "step": 5246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 35.07971294823995,
+      "learning_rate": 6.472439968242761e-06,
+      "loss": 0.4962,
+      "step": 5247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.502898731760762,
+      "learning_rate": 6.47119302083542e-06,
+      "loss": 0.4687,
+      "step": 5248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.341845806235472,
+      "learning_rate": 6.4699459732463395e-06,
+      "loss": 0.485,
+      "step": 5249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.44318991628492,
+      "learning_rate": 6.4686988255604375e-06,
+      "loss": 0.4861,
+      "step": 5250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.767930760296624,
+      "learning_rate": 6.46745157786264e-06,
+      "loss": 0.5986,
+      "step": 5251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.0010323122205,
+      "learning_rate": 6.466204230237877e-06,
+      "loss": 0.427,
+      "step": 5252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 84.34649512906167,
+      "learning_rate": 6.464956782771089e-06,
+      "loss": 0.4578,
+      "step": 5253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.39634262944746,
+      "learning_rate": 6.463709235547222e-06,
+      "loss": 0.473,
+      "step": 5254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5613060729621615,
+      "learning_rate": 6.462461588651226e-06,
+      "loss": 0.719,
+      "step": 5255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.99673564118382,
+      "learning_rate": 6.461213842168065e-06,
+      "loss": 0.5531,
+      "step": 5256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.472156663591102,
+      "learning_rate": 6.4599659961827e-06,
+      "loss": 0.4786,
+      "step": 5257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.93784740777339,
+      "learning_rate": 6.458718050780106e-06,
+      "loss": 0.3958,
+      "step": 5258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 41.24544810715005,
+      "learning_rate": 6.457470006045263e-06,
+      "loss": 0.5695,
+      "step": 5259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.44344340629926,
+      "learning_rate": 6.4562218620631565e-06,
+      "loss": 0.4673,
+      "step": 5260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.169356855528461,
+      "learning_rate": 6.45497361891878e-06,
+      "loss": 0.5247,
+      "step": 5261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.708015424331707,
+      "learning_rate": 6.453725276697132e-06,
+      "loss": 0.5566,
+      "step": 5262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6963588161185021,
+      "learning_rate": 6.4524768354832225e-06,
+      "loss": 0.8003,
+      "step": 5263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.4071205610665,
+      "learning_rate": 6.451228295362061e-06,
+      "loss": 0.5504,
+      "step": 5264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.17479698344927,
+      "learning_rate": 6.449979656418668e-06,
+      "loss": 0.5001,
+      "step": 5265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.630017228687235,
+      "learning_rate": 6.4487309187380735e-06,
+      "loss": 0.549,
+      "step": 5266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5220945948877427,
+      "learning_rate": 6.447482082405308e-06,
+      "loss": 0.6836,
+      "step": 5267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 29.05857781099308,
+      "learning_rate": 6.446233147505413e-06,
+      "loss": 0.4101,
+      "step": 5268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 19.45138753042776,
+      "learning_rate": 6.444984114123436e-06,
+      "loss": 0.4065,
+      "step": 5269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 28.522512473730266,
+      "learning_rate": 6.443734982344427e-06,
+      "loss": 0.6017,
+      "step": 5270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 23.15815509166064,
+      "learning_rate": 6.4424857522534515e-06,
+      "loss": 0.5333,
+      "step": 5271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.546611903544742,
+      "learning_rate": 6.441236423935574e-06,
+      "loss": 0.5243,
+      "step": 5272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 36.49022638481092,
+      "learning_rate": 6.439986997475868e-06,
+      "loss": 0.4813,
+      "step": 5273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 21.107664843857897,
+      "learning_rate": 6.438737472959414e-06,
+      "loss": 0.5655,
+      "step": 5274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 27.954457973358725,
+      "learning_rate": 6.437487850471298e-06,
+      "loss": 0.4538,
+      "step": 5275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.243403752770533,
+      "learning_rate": 6.436238130096618e-06,
+      "loss": 0.5777,
+      "step": 5276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.211338551639763,
+      "learning_rate": 6.4349883119204706e-06,
+      "loss": 0.3907,
+      "step": 5277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 25.213342259280168,
+      "learning_rate": 6.433738396027963e-06,
+      "loss": 0.5328,
+      "step": 5278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.205214559574408,
+      "learning_rate": 6.432488382504211e-06,
+      "loss": 0.5032,
+      "step": 5279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7465994126890587,
+      "learning_rate": 6.431238271434332e-06,
+      "loss": 0.7262,
+      "step": 5280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.261087818143128,
+      "learning_rate": 6.429988062903458e-06,
+      "loss": 0.4264,
+      "step": 5281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.524834942145215,
+      "learning_rate": 6.4287377569967165e-06,
+      "loss": 0.4553,
+      "step": 5282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 27.678559890986698,
+      "learning_rate": 6.4274873537992536e-06,
+      "loss": 0.568,
+      "step": 5283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 21.19490006966482,
+      "learning_rate": 6.426236853396212e-06,
+      "loss": 0.4452,
+      "step": 5284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.88834840174053,
+      "learning_rate": 6.4249862558727464e-06,
+      "loss": 0.5481,
+      "step": 5285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.74313831102038,
+      "learning_rate": 6.423735561314018e-06,
+      "loss": 0.405,
+      "step": 5286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7690597169068731,
+      "learning_rate": 6.42248476980519e-06,
+      "loss": 0.7817,
+      "step": 5287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 30.260480196657454,
+      "learning_rate": 6.421233881431443e-06,
+      "loss": 0.5057,
+      "step": 5288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.753891569412538,
+      "learning_rate": 6.41998289627795e-06,
+      "loss": 0.516,
+      "step": 5289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 32.44757778370053,
+      "learning_rate": 6.418731814429902e-06,
+      "loss": 0.5651,
+      "step": 5290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.353781758979107,
+      "learning_rate": 6.417480635972488e-06,
+      "loss": 0.5574,
+      "step": 5291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 43.19488282216346,
+      "learning_rate": 6.416229360990911e-06,
+      "loss": 0.623,
+      "step": 5292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.590446305315275,
+      "learning_rate": 6.414977989570378e-06,
+      "loss": 0.4392,
+      "step": 5293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.880210570526284,
+      "learning_rate": 6.4137265217960975e-06,
+      "loss": 0.4082,
+      "step": 5294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.343988128536028,
+      "learning_rate": 6.412474957753293e-06,
+      "loss": 0.4163,
+      "step": 5295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.517800668996937,
+      "learning_rate": 6.411223297527188e-06,
+      "loss": 0.443,
+      "step": 5296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.035660035891112,
+      "learning_rate": 6.409971541203018e-06,
+      "loss": 0.4867,
+      "step": 5297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 26.421990671422083,
+      "learning_rate": 6.408719688866019e-06,
+      "loss": 0.5103,
+      "step": 5298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 82.65075935566807,
+      "learning_rate": 6.407467740601439e-06,
+      "loss": 0.5429,
+      "step": 5299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.832126473407778,
+      "learning_rate": 6.4062156964945266e-06,
+      "loss": 0.4245,
+      "step": 5300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.141940311191762,
+      "learning_rate": 6.404963556630543e-06,
+      "loss": 0.4523,
+      "step": 5301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.554932345219711,
+      "learning_rate": 6.403711321094757e-06,
+      "loss": 0.4491,
+      "step": 5302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 25.105239189327236,
+      "learning_rate": 6.402458989972432e-06,
+      "loss": 0.4765,
+      "step": 5303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.310520643917467,
+      "learning_rate": 6.401206563348854e-06,
+      "loss": 0.5145,
+      "step": 5304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 40.211795603596244,
+      "learning_rate": 6.399954041309301e-06,
+      "loss": 0.3809,
+      "step": 5305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 31.79904150631571,
+      "learning_rate": 6.39870142393907e-06,
+      "loss": 0.4078,
+      "step": 5306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 64.14542121821964,
+      "learning_rate": 6.397448711323455e-06,
+      "loss": 0.5753,
+      "step": 5307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.100700130099103,
+      "learning_rate": 6.396195903547764e-06,
+      "loss": 0.5547,
+      "step": 5308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 21.769511711255998,
+      "learning_rate": 6.394943000697303e-06,
+      "loss": 0.5529,
+      "step": 5309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.707332176205895,
+      "learning_rate": 6.3936900028573916e-06,
+      "loss": 0.4594,
+      "step": 5310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.484480004462934,
+      "learning_rate": 6.392436910113354e-06,
+      "loss": 0.415,
+      "step": 5311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.537193278832712,
+      "learning_rate": 6.391183722550518e-06,
+      "loss": 0.4921,
+      "step": 5312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 32.3791714485812,
+      "learning_rate": 6.3899304402542224e-06,
+      "loss": 0.4907,
+      "step": 5313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 21.465066348465005,
+      "learning_rate": 6.38867706330981e-06,
+      "loss": 0.4696,
+      "step": 5314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.105359090010374,
+      "learning_rate": 6.387423591802627e-06,
+      "loss": 0.4383,
+      "step": 5315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.84715707336062,
+      "learning_rate": 6.386170025818034e-06,
+      "loss": 0.542,
+      "step": 5316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.794147447437332,
+      "learning_rate": 6.384916365441392e-06,
+      "loss": 0.486,
+      "step": 5317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 29.958138002869923,
+      "learning_rate": 6.383662610758067e-06,
+      "loss": 0.4373,
+      "step": 5318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.176040541053432,
+      "learning_rate": 6.382408761853436e-06,
+      "loss": 0.4717,
+      "step": 5319
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.979916002477644,
+      "learning_rate": 6.3811548188128825e-06,
+      "loss": 0.3851,
+      "step": 5320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.25553498230641,
+      "learning_rate": 6.3799007817217905e-06,
+      "loss": 0.5049,
+      "step": 5321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.849986030540952,
+      "learning_rate": 6.378646650665558e-06,
+      "loss": 0.4437,
+      "step": 5322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.071803729149588,
+      "learning_rate": 6.377392425729583e-06,
+      "loss": 0.5251,
+      "step": 5323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.31206688521181,
+      "learning_rate": 6.376138106999273e-06,
+      "loss": 0.5036,
+      "step": 5324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 28.0146159832172,
+      "learning_rate": 6.374883694560045e-06,
+      "loss": 0.5189,
+      "step": 5325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 28.86432697213625,
+      "learning_rate": 6.373629188497315e-06,
+      "loss": 0.522,
+      "step": 5326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.875544241258382,
+      "learning_rate": 6.372374588896511e-06,
+      "loss": 0.3848,
+      "step": 5327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.7556125299067,
+      "learning_rate": 6.371119895843064e-06,
+      "loss": 0.5269,
+      "step": 5328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 27.04447838562003,
+      "learning_rate": 6.369865109422417e-06,
+      "loss": 0.4946,
+      "step": 5329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.6481167318306,
+      "learning_rate": 6.368610229720011e-06,
+      "loss": 0.5749,
+      "step": 5330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.84022055042212,
+      "learning_rate": 6.3673552568213015e-06,
+      "loss": 0.5511,
+      "step": 5331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 23.040727570007423,
+      "learning_rate": 6.366100190811743e-06,
+      "loss": 0.4848,
+      "step": 5332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.97910560139186,
+      "learning_rate": 6.3648450317768e-06,
+      "loss": 0.4671,
+      "step": 5333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.77809286297529,
+      "learning_rate": 6.363589779801949e-06,
+      "loss": 0.4632,
+      "step": 5334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.486691287085163,
+      "learning_rate": 6.362334434972659e-06,
+      "loss": 0.4798,
+      "step": 5335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 46.296516081831186,
+      "learning_rate": 6.361078997374421e-06,
+      "loss": 0.4132,
+      "step": 5336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.67149850561241,
+      "learning_rate": 6.359823467092718e-06,
+      "loss": 0.7012,
+      "step": 5337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.171028800241663,
+      "learning_rate": 6.358567844213053e-06,
+      "loss": 0.4196,
+      "step": 5338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 27.000655253157262,
+      "learning_rate": 6.3573121288209215e-06,
+      "loss": 0.4491,
+      "step": 5339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 59.75141006866929,
+      "learning_rate": 6.356056321001837e-06,
+      "loss": 0.4545,
+      "step": 5340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.441018764117274,
+      "learning_rate": 6.354800420841313e-06,
+      "loss": 0.4268,
+      "step": 5341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.220093146546924,
+      "learning_rate": 6.353544428424869e-06,
+      "loss": 0.4981,
+      "step": 5342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.133316089025964,
+      "learning_rate": 6.352288343838036e-06,
+      "loss": 0.4393,
+      "step": 5343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 38.26900442945354,
+      "learning_rate": 6.351032167166345e-06,
+      "loss": 0.6051,
+      "step": 5344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 26.27163754619335,
+      "learning_rate": 6.349775898495335e-06,
+      "loss": 0.5302,
+      "step": 5345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 35.87341303181566,
+      "learning_rate": 6.348519537910556e-06,
+      "loss": 0.5449,
+      "step": 5346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.857475839771544,
+      "learning_rate": 6.347263085497559e-06,
+      "loss": 0.5025,
+      "step": 5347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.632146031021524,
+      "learning_rate": 6.346006541341903e-06,
+      "loss": 0.5121,
+      "step": 5348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 34.36355223453446,
+      "learning_rate": 6.344749905529151e-06,
+      "loss": 0.4866,
+      "step": 5349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.51780854143369,
+      "learning_rate": 6.343493178144879e-06,
+      "loss": 0.5031,
+      "step": 5350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.867169209767166,
+      "learning_rate": 6.342236359274661e-06,
+      "loss": 0.4652,
+      "step": 5351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 21.926544493230047,
+      "learning_rate": 6.340979449004081e-06,
+      "loss": 0.3994,
+      "step": 5352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.378153100445907,
+      "learning_rate": 6.339722447418729e-06,
+      "loss": 0.4241,
+      "step": 5353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.235891675882776,
+      "learning_rate": 6.338465354604204e-06,
+      "loss": 0.521,
+      "step": 5354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.656197299326912,
+      "learning_rate": 6.337208170646103e-06,
+      "loss": 0.5854,
+      "step": 5355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.800380149786616,
+      "learning_rate": 6.335950895630041e-06,
+      "loss": 0.4695,
+      "step": 5356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 170.494643002451,
+      "learning_rate": 6.334693529641628e-06,
+      "loss": 0.5711,
+      "step": 5357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.1344189946957,
+      "learning_rate": 6.333436072766487e-06,
+      "loss": 0.432,
+      "step": 5358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.053000556738267,
+      "learning_rate": 6.332178525090247e-06,
+      "loss": 0.5305,
+      "step": 5359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 29.549515416360553,
+      "learning_rate": 6.3309208866985375e-06,
+      "loss": 0.546,
+      "step": 5360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.799283871194978,
+      "learning_rate": 6.329663157677002e-06,
+      "loss": 0.5613,
+      "step": 5361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.716218424540996,
+      "learning_rate": 6.328405338111283e-06,
+      "loss": 0.7261,
+      "step": 5362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.013895687509976,
+      "learning_rate": 6.327147428087034e-06,
+      "loss": 0.4137,
+      "step": 5363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 59.75025212709786,
+      "learning_rate": 6.325889427689914e-06,
+      "loss": 0.5125,
+      "step": 5364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.444541849636144,
+      "learning_rate": 6.324631337005585e-06,
+      "loss": 0.4605,
+      "step": 5365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.910167887677645,
+      "learning_rate": 6.3233731561197196e-06,
+      "loss": 0.4909,
+      "step": 5366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.993820525648472,
+      "learning_rate": 6.322114885117994e-06,
+      "loss": 0.5601,
+      "step": 5367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.086162677401894,
+      "learning_rate": 6.32085652408609e-06,
+      "loss": 0.428,
+      "step": 5368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.402149144820047,
+      "learning_rate": 6.319598073109696e-06,
+      "loss": 0.4294,
+      "step": 5369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 28.433653004626375,
+      "learning_rate": 6.31833953227451e-06,
+      "loss": 0.4984,
+      "step": 5370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.515625772090944,
+      "learning_rate": 6.31708090166623e-06,
+      "loss": 0.6067,
+      "step": 5371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.114861029421002,
+      "learning_rate": 6.315822181370562e-06,
+      "loss": 0.5972,
+      "step": 5372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.603846056279178,
+      "learning_rate": 6.314563371473224e-06,
+      "loss": 0.6129,
+      "step": 5373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 34.30259896047753,
+      "learning_rate": 6.313304472059931e-06,
+      "loss": 0.4195,
+      "step": 5374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7553605078924912,
+      "learning_rate": 6.31204548321641e-06,
+      "loss": 0.7852,
+      "step": 5375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.965431072333496,
+      "learning_rate": 6.310786405028393e-06,
+      "loss": 0.5025,
+      "step": 5376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.1722615444211,
+      "learning_rate": 6.3095272375816185e-06,
+      "loss": 0.4954,
+      "step": 5377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 21.720685222390955,
+      "learning_rate": 6.308267980961828e-06,
+      "loss": 0.4789,
+      "step": 5378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 21.47733216000079,
+      "learning_rate": 6.307008635254773e-06,
+      "loss": 0.4534,
+      "step": 5379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5837057883747907,
+      "learning_rate": 6.3057492005462086e-06,
+      "loss": 0.6836,
+      "step": 5380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.640641641395105,
+      "learning_rate": 6.304489676921897e-06,
+      "loss": 0.4468,
+      "step": 5381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 49.91724171446624,
+      "learning_rate": 6.303230064467607e-06,
+      "loss": 0.4828,
+      "step": 5382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.201899224055348,
+      "learning_rate": 6.30197036326911e-06,
+      "loss": 0.4725,
+      "step": 5383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.075540665988069,
+      "learning_rate": 6.30071057341219e-06,
+      "loss": 0.5976,
+      "step": 5384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.770652657753715,
+      "learning_rate": 6.29945069498263e-06,
+      "loss": 0.4681,
+      "step": 5385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 13.002910384532937,
+      "learning_rate": 6.298190728066221e-06,
+      "loss": 0.5869,
+      "step": 5386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 46.136001801884504,
+      "learning_rate": 6.296930672748765e-06,
+      "loss": 0.5157,
+      "step": 5387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.976610014823116,
+      "learning_rate": 6.2956705291160645e-06,
+      "loss": 0.4974,
+      "step": 5388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 54.489839267388085,
+      "learning_rate": 6.294410297253931e-06,
+      "loss": 0.4016,
+      "step": 5389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.117930320327595,
+      "learning_rate": 6.293149977248178e-06,
+      "loss": 0.4762,
+      "step": 5390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 33.879386077010544,
+      "learning_rate": 6.291889569184631e-06,
+      "loss": 0.5866,
+      "step": 5391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.42220155392883,
+      "learning_rate": 6.290629073149114e-06,
+      "loss": 0.4692,
+      "step": 5392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.44994558101095,
+      "learning_rate": 6.2893684892274654e-06,
+      "loss": 0.3991,
+      "step": 5393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.215831308603356,
+      "learning_rate": 6.288107817505522e-06,
+      "loss": 0.4938,
+      "step": 5394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 28.101562229463575,
+      "learning_rate": 6.286847058069132e-06,
+      "loss": 0.4432,
+      "step": 5395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 78.31757508296566,
+      "learning_rate": 6.285586211004147e-06,
+      "loss": 0.5521,
+      "step": 5396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 16.193297560845583,
+      "learning_rate": 6.284325276396425e-06,
+      "loss": 0.4909,
+      "step": 5397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 10.744677239401032,
+      "learning_rate": 6.283064254331831e-06,
+      "loss": 0.5495,
+      "step": 5398
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 27.52631204608955,
+      "learning_rate": 6.281803144896234e-06,
+      "loss": 0.4013,
+      "step": 5399
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 22.99656664931258,
+      "learning_rate": 6.2805419481755115e-06,
+      "loss": 0.4625,
+      "step": 5400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.944107905288085,
+      "learning_rate": 6.2792806642555426e-06,
+      "loss": 0.4621,
+      "step": 5401
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 36.90198823653752,
+      "learning_rate": 6.278019293222217e-06,
+      "loss": 0.4699,
+      "step": 5402
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.217401702334946,
+      "learning_rate": 6.276757835161429e-06,
+      "loss": 0.4705,
+      "step": 5403
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.278734405360314,
+      "learning_rate": 6.2754962901590755e-06,
+      "loss": 0.4842,
+      "step": 5404
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 24.250563324501847,
+      "learning_rate": 6.274234658301067e-06,
+      "loss": 0.5289,
+      "step": 5405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.644905364586986,
+      "learning_rate": 6.27297293967331e-06,
+      "loss": 0.4262,
+      "step": 5406
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.94134195135299,
+      "learning_rate": 6.271711134361726e-06,
+      "loss": 0.4925,
+      "step": 5407
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.187826146703188,
+      "learning_rate": 6.270449242452234e-06,
+      "loss": 0.422,
+      "step": 5408
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.076837690672374,
+      "learning_rate": 6.269187264030768e-06,
+      "loss": 0.4184,
+      "step": 5409
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 26.522102274759053,
+      "learning_rate": 6.26792519918326e-06,
+      "loss": 0.4551,
+      "step": 5410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.795568748000186,
+      "learning_rate": 6.2666630479956525e-06,
+      "loss": 0.504,
+      "step": 5411
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.674436885454169,
+      "learning_rate": 6.265400810553893e-06,
+      "loss": 0.5226,
+      "step": 5412
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.422275490256663,
+      "learning_rate": 6.264138486943932e-06,
+      "loss": 0.5834,
+      "step": 5413
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7432460314656202,
+      "learning_rate": 6.262876077251729e-06,
+      "loss": 0.8185,
+      "step": 5414
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 30.8371172121442,
+      "learning_rate": 6.261613581563251e-06,
+      "loss": 0.4643,
+      "step": 5415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 113.92081665106996,
+      "learning_rate": 6.260350999964463e-06,
+      "loss": 0.454,
+      "step": 5416
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.543128804643208,
+      "learning_rate": 6.259088332541345e-06,
+      "loss": 0.4667,
+      "step": 5417
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 96.39364852855957,
+      "learning_rate": 6.257825579379882e-06,
+      "loss": 0.4969,
+      "step": 5418
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.915106806621825,
+      "learning_rate": 6.2565627405660555e-06,
+      "loss": 0.494,
+      "step": 5419
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.50870898048096,
+      "learning_rate": 6.255299816185862e-06,
+      "loss": 0.4801,
+      "step": 5420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.954362171158985,
+      "learning_rate": 6.2540368063253035e-06,
+      "loss": 0.6205,
+      "step": 5421
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.394352384106186,
+      "learning_rate": 6.252773711070381e-06,
+      "loss": 0.5065,
+      "step": 5422
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.671845605401911,
+      "learning_rate": 6.25151053050711e-06,
+      "loss": 0.4376,
+      "step": 5423
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.612399751105198,
+      "learning_rate": 6.250247264721503e-06,
+      "loss": 0.5505,
+      "step": 5424
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.206699991248454,
+      "learning_rate": 6.248983913799587e-06,
+      "loss": 0.3457,
+      "step": 5425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.447355886859299,
+      "learning_rate": 6.247720477827387e-06,
+      "loss": 0.6073,
+      "step": 5426
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.820984074559497,
+      "learning_rate": 6.246456956890941e-06,
+      "loss": 0.4631,
+      "step": 5427
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.50165953170762,
+      "learning_rate": 6.245193351076286e-06,
+      "loss": 0.5683,
+      "step": 5428
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.327359510714379,
+      "learning_rate": 6.24392966046947e-06,
+      "loss": 0.4108,
+      "step": 5429
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.536907420534641,
+      "learning_rate": 6.242665885156545e-06,
+      "loss": 0.4094,
+      "step": 5430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.481231227670603,
+      "learning_rate": 6.241402025223565e-06,
+      "loss": 0.4091,
+      "step": 5431
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.346752939521686,
+      "learning_rate": 6.240138080756599e-06,
+      "loss": 0.6074,
+      "step": 5432
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.717995897277175,
+      "learning_rate": 6.2388740518417116e-06,
+      "loss": 0.3763,
+      "step": 5433
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.955447645593965,
+      "learning_rate": 6.237609938564978e-06,
+      "loss": 0.4491,
+      "step": 5434
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.172765180068744,
+      "learning_rate": 6.236345741012481e-06,
+      "loss": 0.4897,
+      "step": 5435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.124860921857657,
+      "learning_rate": 6.235081459270305e-06,
+      "loss": 0.5778,
+      "step": 5436
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4859786740750298,
+      "learning_rate": 6.233817093424541e-06,
+      "loss": 0.6731,
+      "step": 5437
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.113222432262985,
+      "learning_rate": 6.2325526435612895e-06,
+      "loss": 0.4514,
+      "step": 5438
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7756519033716358,
+      "learning_rate": 6.231288109766654e-06,
+      "loss": 0.7434,
+      "step": 5439
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 29.747217199971892,
+      "learning_rate": 6.23002349212674e-06,
+      "loss": 0.3993,
+      "step": 5440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.364593575203443,
+      "learning_rate": 6.228758790727667e-06,
+      "loss": 0.4455,
+      "step": 5441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.287817265204913,
+      "learning_rate": 6.227494005655551e-06,
+      "loss": 0.5016,
+      "step": 5442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 26.86239532550636,
+      "learning_rate": 6.226229136996522e-06,
+      "loss": 0.4932,
+      "step": 5443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.636376454874275,
+      "learning_rate": 6.22496418483671e-06,
+      "loss": 0.3776,
+      "step": 5444
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.873112679139298,
+      "learning_rate": 6.223699149262254e-06,
+      "loss": 0.5229,
+      "step": 5445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.095713615387435,
+      "learning_rate": 6.222434030359294e-06,
+      "loss": 0.6271,
+      "step": 5446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.664040914665426,
+      "learning_rate": 6.221168828213984e-06,
+      "loss": 0.3633,
+      "step": 5447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.778255993729756,
+      "learning_rate": 6.2199035429124754e-06,
+      "loss": 0.5688,
+      "step": 5448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.5892527912422,
+      "learning_rate": 6.218638174540929e-06,
+      "loss": 0.4903,
+      "step": 5449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.203702814676777,
+      "learning_rate": 6.217372723185512e-06,
+      "loss": 0.5385,
+      "step": 5450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.93341924605143,
+      "learning_rate": 6.2161071889323934e-06,
+      "loss": 0.6146,
+      "step": 5451
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 24.971181158481112,
+      "learning_rate": 6.2148415718677525e-06,
+      "loss": 0.4623,
+      "step": 5452
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.485082814731918,
+      "learning_rate": 6.2135758720777735e-06,
+      "loss": 0.4523,
+      "step": 5453
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.923732563529619,
+      "learning_rate": 6.212310089648642e-06,
+      "loss": 0.4449,
+      "step": 5454
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.757602214992557,
+      "learning_rate": 6.211044224666555e-06,
+      "loss": 0.3785,
+      "step": 5455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.94808164565035,
+      "learning_rate": 6.209778277217709e-06,
+      "loss": 0.3992,
+      "step": 5456
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.301440189485994,
+      "learning_rate": 6.208512247388312e-06,
+      "loss": 0.4764,
+      "step": 5457
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.762297684756298,
+      "learning_rate": 6.207246135264574e-06,
+      "loss": 0.4679,
+      "step": 5458
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.445060021634363,
+      "learning_rate": 6.205979940932713e-06,
+      "loss": 0.5025,
+      "step": 5459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.166051528588655,
+      "learning_rate": 6.204713664478949e-06,
+      "loss": 0.5457,
+      "step": 5460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 26.507402042843378,
+      "learning_rate": 6.203447305989512e-06,
+      "loss": 0.4381,
+      "step": 5461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 28.345134260281327,
+      "learning_rate": 6.202180865550636e-06,
+      "loss": 0.4931,
+      "step": 5462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.684886251558002,
+      "learning_rate": 6.200914343248555e-06,
+      "loss": 0.4502,
+      "step": 5463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.104434155056282,
+      "learning_rate": 6.19964773916952e-06,
+      "loss": 0.5402,
+      "step": 5464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.617861210459052,
+      "learning_rate": 6.198381053399775e-06,
+      "loss": 0.4773,
+      "step": 5465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 42.64138939309219,
+      "learning_rate": 6.197114286025582e-06,
+      "loss": 0.5195,
+      "step": 5466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 17.015592944506736,
+      "learning_rate": 6.195847437133197e-06,
+      "loss": 0.4111,
+      "step": 5467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 22.993058063110563,
+      "learning_rate": 6.1945805068088895e-06,
+      "loss": 0.448,
+      "step": 5468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.082186546885525,
+      "learning_rate": 6.193313495138932e-06,
+      "loss": 0.4412,
+      "step": 5469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.2377004941484415,
+      "learning_rate": 6.1920464022096006e-06,
+      "loss": 0.5504,
+      "step": 5470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.451731248227766,
+      "learning_rate": 6.190779228107181e-06,
+      "loss": 0.5914,
+      "step": 5471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.02245373677639,
+      "learning_rate": 6.18951197291796e-06,
+      "loss": 0.4343,
+      "step": 5472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5636433908191432,
+      "learning_rate": 6.188244636728233e-06,
+      "loss": 0.6704,
+      "step": 5473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.44120475854962,
+      "learning_rate": 6.1869772196243e-06,
+      "loss": 0.4756,
+      "step": 5474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.747362771017665,
+      "learning_rate": 6.185709721692466e-06,
+      "loss": 0.8278,
+      "step": 5475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.784834040137824,
+      "learning_rate": 6.1844421430190435e-06,
+      "loss": 0.4173,
+      "step": 5476
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.941099329531761,
+      "learning_rate": 6.183174483690348e-06,
+      "loss": 0.653,
+      "step": 5477
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.538495230333914,
+      "learning_rate": 6.181906743792703e-06,
+      "loss": 0.5446,
+      "step": 5478
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.625942883169248,
+      "learning_rate": 6.180638923412432e-06,
+      "loss": 0.4988,
+      "step": 5479
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.209144705086214,
+      "learning_rate": 6.179371022635872e-06,
+      "loss": 0.4183,
+      "step": 5480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 61.12597849078918,
+      "learning_rate": 6.178103041549358e-06,
+      "loss": 0.4724,
+      "step": 5481
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.812260522313174,
+      "learning_rate": 6.176834980239237e-06,
+      "loss": 0.4738,
+      "step": 5482
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.43227408729951,
+      "learning_rate": 6.175566838791858e-06,
+      "loss": 0.4222,
+      "step": 5483
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.676575027362372,
+      "learning_rate": 6.174298617293574e-06,
+      "loss": 0.475,
+      "step": 5484
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.578362358173687,
+      "learning_rate": 6.173030315830747e-06,
+      "loss": 0.4745,
+      "step": 5485
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.680511516032045,
+      "learning_rate": 6.171761934489739e-06,
+      "loss": 0.5569,
+      "step": 5486
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.050825631907193,
+      "learning_rate": 6.170493473356929e-06,
+      "loss": 0.3709,
+      "step": 5487
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 19.790837405996694,
+      "learning_rate": 6.169224932518684e-06,
+      "loss": 0.5315,
+      "step": 5488
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.894039690891603,
+      "learning_rate": 6.167956312061395e-06,
+      "loss": 0.5603,
+      "step": 5489
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 41.04494273705162,
+      "learning_rate": 6.166687612071444e-06,
+      "loss": 0.4224,
+      "step": 5490
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.11637047891603,
+      "learning_rate": 6.165418832635224e-06,
+      "loss": 0.467,
+      "step": 5491
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.79603837344482,
+      "learning_rate": 6.164149973839135e-06,
+      "loss": 0.5237,
+      "step": 5492
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.876832241497908,
+      "learning_rate": 6.162881035769579e-06,
+      "loss": 0.4463,
+      "step": 5493
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.182739022533635,
+      "learning_rate": 6.161612018512968e-06,
+      "loss": 0.4295,
+      "step": 5494
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.676246394565511,
+      "learning_rate": 6.160342922155713e-06,
+      "loss": 0.7396,
+      "step": 5495
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.41970588794909,
+      "learning_rate": 6.159073746784235e-06,
+      "loss": 0.4196,
+      "step": 5496
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 16.11502800560984,
+      "learning_rate": 6.15780449248496e-06,
+      "loss": 0.4077,
+      "step": 5497
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.62743670394671,
+      "learning_rate": 6.15653515934432e-06,
+      "loss": 0.4484,
+      "step": 5498
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.722313757012135,
+      "learning_rate": 6.155265747448748e-06,
+      "loss": 0.5,
+      "step": 5499
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.174682985955203,
+      "learning_rate": 6.153996256884686e-06,
+      "loss": 0.39,
+      "step": 5500
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.86713028408222,
+      "learning_rate": 6.152726687738582e-06,
+      "loss": 0.5474,
+      "step": 5501
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.240536151778771,
+      "learning_rate": 6.1514570400968865e-06,
+      "loss": 0.5778,
+      "step": 5502
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 14.74419282006103,
+      "learning_rate": 6.150187314046059e-06,
+      "loss": 0.5071,
+      "step": 5503
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.62934339838071,
+      "learning_rate": 6.14891750967256e-06,
+      "loss": 0.3806,
+      "step": 5504
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.543212606372812,
+      "learning_rate": 6.147647627062859e-06,
+      "loss": 0.4422,
+      "step": 5505
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 35.03243092104436,
+      "learning_rate": 6.146377666303427e-06,
+      "loss": 0.4784,
+      "step": 5506
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 42.469761486679694,
+      "learning_rate": 6.145107627480747e-06,
+      "loss": 0.4162,
+      "step": 5507
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.86693621535183,
+      "learning_rate": 6.143837510681299e-06,
+      "loss": 0.5475,
+      "step": 5508
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 21.82121522014584,
+      "learning_rate": 6.1425673159915744e-06,
+      "loss": 0.508,
+      "step": 5509
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.321660806287943,
+      "learning_rate": 6.141297043498069e-06,
+      "loss": 0.4325,
+      "step": 5510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.527707775891692,
+      "learning_rate": 6.1400266932872795e-06,
+      "loss": 0.5107,
+      "step": 5511
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.024101051606012,
+      "learning_rate": 6.1387562654457135e-06,
+      "loss": 0.5196,
+      "step": 5512
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.151771475076693,
+      "learning_rate": 6.13748576005988e-06,
+      "loss": 0.5504,
+      "step": 5513
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.358781484768766,
+      "learning_rate": 6.136215177216296e-06,
+      "loss": 0.3864,
+      "step": 5514
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 44.83999311134331,
+      "learning_rate": 6.134944517001482e-06,
+      "loss": 0.5673,
+      "step": 5515
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.400671030846736,
+      "learning_rate": 6.133673779501964e-06,
+      "loss": 0.5252,
+      "step": 5516
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 30.91616975641728,
+      "learning_rate": 6.132402964804275e-06,
+      "loss": 0.6407,
+      "step": 5517
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.577742908201962,
+      "learning_rate": 6.131132072994949e-06,
+      "loss": 0.4757,
+      "step": 5518
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.481868695600545,
+      "learning_rate": 6.129861104160534e-06,
+      "loss": 0.5298,
+      "step": 5519
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 23.75107100505546,
+      "learning_rate": 6.12859005838757e-06,
+      "loss": 0.429,
+      "step": 5520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.51467550652253,
+      "learning_rate": 6.1273189357626155e-06,
+      "loss": 0.4081,
+      "step": 5521
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.912107265287785,
+      "learning_rate": 6.126047736372223e-06,
+      "loss": 0.4642,
+      "step": 5522
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.733128004979857,
+      "learning_rate": 6.12477646030296e-06,
+      "loss": 0.4265,
+      "step": 5523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.773638114209655,
+      "learning_rate": 6.123505107641394e-06,
+      "loss": 0.4386,
+      "step": 5524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.354571815896678,
+      "learning_rate": 6.122233678474096e-06,
+      "loss": 0.5207,
+      "step": 5525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 39.28125055092799,
+      "learning_rate": 6.120962172887649e-06,
+      "loss": 0.535,
+      "step": 5526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.039590184147965,
+      "learning_rate": 6.119690590968631e-06,
+      "loss": 0.5021,
+      "step": 5527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.815683374959367,
+      "learning_rate": 6.118418932803637e-06,
+      "loss": 0.5189,
+      "step": 5528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.24846389415697,
+      "learning_rate": 6.117147198479258e-06,
+      "loss": 0.537,
+      "step": 5529
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 50.6534793853881,
+      "learning_rate": 6.115875388082093e-06,
+      "loss": 0.4947,
+      "step": 5530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 28.38107628536905,
+      "learning_rate": 6.114603501698749e-06,
+      "loss": 0.5541,
+      "step": 5531
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 56.138789512562724,
+      "learning_rate": 6.1133315394158345e-06,
+      "loss": 0.5067,
+      "step": 5532
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.282280749414559,
+      "learning_rate": 6.112059501319966e-06,
+      "loss": 0.375,
+      "step": 5533
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.83665537958439,
+      "learning_rate": 6.110787387497761e-06,
+      "loss": 0.4445,
+      "step": 5534
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.348104533607938,
+      "learning_rate": 6.109515198035847e-06,
+      "loss": 0.4577,
+      "step": 5535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.173239571308564,
+      "learning_rate": 6.108242933020852e-06,
+      "loss": 0.4256,
+      "step": 5536
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.415887063153346,
+      "learning_rate": 6.106970592539414e-06,
+      "loss": 0.451,
+      "step": 5537
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.774441219546954,
+      "learning_rate": 6.105698176678173e-06,
+      "loss": 0.5103,
+      "step": 5538
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5197741892241483,
+      "learning_rate": 6.104425685523775e-06,
+      "loss": 0.7672,
+      "step": 5539
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.027004036890037,
+      "learning_rate": 6.103153119162873e-06,
+      "loss": 0.4964,
+      "step": 5540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.310205264959468,
+      "learning_rate": 6.101880477682118e-06,
+      "loss": 0.5333,
+      "step": 5541
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.094382204379134,
+      "learning_rate": 6.100607761168177e-06,
+      "loss": 0.5277,
+      "step": 5542
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.767281956295287,
+      "learning_rate": 6.099334969707712e-06,
+      "loss": 0.6083,
+      "step": 5543
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.758770087165491,
+      "learning_rate": 6.098062103387398e-06,
+      "loss": 0.4133,
+      "step": 5544
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 78.94193587050262,
+      "learning_rate": 6.096789162293908e-06,
+      "loss": 0.4284,
+      "step": 5545
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 25.254538582076865,
+      "learning_rate": 6.095516146513926e-06,
+      "loss": 0.4813,
+      "step": 5546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 25.78976199601607,
+      "learning_rate": 6.094243056134138e-06,
+      "loss": 0.4651,
+      "step": 5547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.639609964432525,
+      "learning_rate": 6.092969891241236e-06,
+      "loss": 0.395,
+      "step": 5548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.974974341558166,
+      "learning_rate": 6.091696651921918e-06,
+      "loss": 0.5471,
+      "step": 5549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 59.535195485689954,
+      "learning_rate": 6.090423338262883e-06,
+      "loss": 0.5561,
+      "step": 5550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.16917117936389,
+      "learning_rate": 6.089149950350842e-06,
+      "loss": 0.4599,
+      "step": 5551
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 29.999452505487163,
+      "learning_rate": 6.087876488272504e-06,
+      "loss": 0.4478,
+      "step": 5552
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.365170792529305,
+      "learning_rate": 6.086602952114587e-06,
+      "loss": 0.525,
+      "step": 5553
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.52846282765215,
+      "learning_rate": 6.085329341963814e-06,
+      "loss": 0.4376,
+      "step": 5554
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 23.778422225987665,
+      "learning_rate": 6.084055657906911e-06,
+      "loss": 0.457,
+      "step": 5555
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 35.814557984050396,
+      "learning_rate": 6.082781900030612e-06,
+      "loss": 0.5557,
+      "step": 5556
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.462428920352416,
+      "learning_rate": 6.08150806842165e-06,
+      "loss": 0.4576,
+      "step": 5557
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.137788698347858,
+      "learning_rate": 6.0802341631667725e-06,
+      "loss": 0.4431,
+      "step": 5558
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5681763264536643,
+      "learning_rate": 6.078960184352725e-06,
+      "loss": 0.7362,
+      "step": 5559
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.226248330371664,
+      "learning_rate": 6.0776861320662595e-06,
+      "loss": 0.4582,
+      "step": 5560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.91395125640563,
+      "learning_rate": 6.076412006394133e-06,
+      "loss": 0.5803,
+      "step": 5561
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 38.5156383884162,
+      "learning_rate": 6.0751378074231075e-06,
+      "loss": 0.3973,
+      "step": 5562
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.87048618454261,
+      "learning_rate": 6.0738635352399525e-06,
+      "loss": 0.5281,
+      "step": 5563
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.247366340191718,
+      "learning_rate": 6.072589189931437e-06,
+      "loss": 0.4864,
+      "step": 5564
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.965657380781156,
+      "learning_rate": 6.071314771584343e-06,
+      "loss": 0.4772,
+      "step": 5565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9427348100691741,
+      "learning_rate": 6.070040280285447e-06,
+      "loss": 0.7612,
+      "step": 5566
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.684148345801713,
+      "learning_rate": 6.068765716121538e-06,
+      "loss": 0.5679,
+      "step": 5567
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 31.018792422613835,
+      "learning_rate": 6.067491079179412e-06,
+      "loss": 0.5217,
+      "step": 5568
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.722762154941007,
+      "learning_rate": 6.066216369545863e-06,
+      "loss": 0.4882,
+      "step": 5569
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.099612944073215,
+      "learning_rate": 6.064941587307692e-06,
+      "loss": 0.4191,
+      "step": 5570
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.013087806049313,
+      "learning_rate": 6.063666732551708e-06,
+      "loss": 0.4392,
+      "step": 5571
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 25.68237258161551,
+      "learning_rate": 6.062391805364724e-06,
+      "loss": 0.5813,
+      "step": 5572
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 40.36576082061511,
+      "learning_rate": 6.0611168058335534e-06,
+      "loss": 0.433,
+      "step": 5573
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.40720696165754,
+      "learning_rate": 6.059841734045022e-06,
+      "loss": 0.4657,
+      "step": 5574
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 35.716514160247876,
+      "learning_rate": 6.0585665900859536e-06,
+      "loss": 0.4485,
+      "step": 5575
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 36.61178746214247,
+      "learning_rate": 6.057291374043179e-06,
+      "loss": 0.4449,
+      "step": 5576
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.067772777136803,
+      "learning_rate": 6.056016086003539e-06,
+      "loss": 0.5207,
+      "step": 5577
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.263251630275438,
+      "learning_rate": 6.054740726053872e-06,
+      "loss": 0.5294,
+      "step": 5578
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.051400862899397,
+      "learning_rate": 6.053465294281025e-06,
+      "loss": 0.5472,
+      "step": 5579
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.161608999187127,
+      "learning_rate": 6.0521897907718485e-06,
+      "loss": 0.447,
+      "step": 5580
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 48.498429488483765,
+      "learning_rate": 6.050914215613201e-06,
+      "loss": 0.582,
+      "step": 5581
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.802867579731387,
+      "learning_rate": 6.0496385688919416e-06,
+      "loss": 0.5313,
+      "step": 5582
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.706748690441486,
+      "learning_rate": 6.0483628506949364e-06,
+      "loss": 0.503,
+      "step": 5583
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.015650257824674,
+      "learning_rate": 6.047087061109056e-06,
+      "loss": 0.6157,
+      "step": 5584
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.349344396063962,
+      "learning_rate": 6.045811200221177e-06,
+      "loss": 0.577,
+      "step": 5585
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 323.0257087248853,
+      "learning_rate": 6.044535268118179e-06,
+      "loss": 0.5569,
+      "step": 5586
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.63031503022889,
+      "learning_rate": 6.043259264886946e-06,
+      "loss": 0.4088,
+      "step": 5587
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.442188794112184,
+      "learning_rate": 6.041983190614371e-06,
+      "loss": 0.5197,
+      "step": 5588
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 23.03793918024161,
+      "learning_rate": 6.040707045387349e-06,
+      "loss": 0.5807,
+      "step": 5589
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 27.955928562125983,
+      "learning_rate": 6.039430829292777e-06,
+      "loss": 0.4554,
+      "step": 5590
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 35.885257309529,
+      "learning_rate": 6.038154542417562e-06,
+      "loss": 0.516,
+      "step": 5591
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.93632669106511,
+      "learning_rate": 6.036878184848614e-06,
+      "loss": 0.5189,
+      "step": 5592
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.866678923977705,
+      "learning_rate": 6.035601756672844e-06,
+      "loss": 0.407,
+      "step": 5593
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 35.32758086980519,
+      "learning_rate": 6.034325257977174e-06,
+      "loss": 0.6289,
+      "step": 5594
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.34346881362312,
+      "learning_rate": 6.033048688848528e-06,
+      "loss": 0.3945,
+      "step": 5595
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.586286997379673,
+      "learning_rate": 6.031772049373833e-06,
+      "loss": 0.4943,
+      "step": 5596
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 27.657250124465282,
+      "learning_rate": 6.0304953396400234e-06,
+      "loss": 0.5514,
+      "step": 5597
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.90036724082484,
+      "learning_rate": 6.029218559734037e-06,
+      "loss": 0.5271,
+      "step": 5598
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.47203720593892,
+      "learning_rate": 6.02794170974282e-06,
+      "loss": 0.442,
+      "step": 5599
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.963551385066328,
+      "learning_rate": 6.026664789753313e-06,
+      "loss": 0.5643,
+      "step": 5600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9359697332549806,
+      "learning_rate": 6.025387799852477e-06,
+      "loss": 0.8135,
+      "step": 5601
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.129613355355557,
+      "learning_rate": 6.024110740127265e-06,
+      "loss": 0.4799,
+      "step": 5602
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.53118604904051,
+      "learning_rate": 6.022833610664638e-06,
+      "loss": 0.3981,
+      "step": 5603
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 36.5345172243807,
+      "learning_rate": 6.0215564115515654e-06,
+      "loss": 0.508,
+      "step": 5604
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 63.72547531734111,
+      "learning_rate": 6.020279142875019e-06,
+      "loss": 0.4878,
+      "step": 5605
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.379415660309775,
+      "learning_rate": 6.019001804721973e-06,
+      "loss": 0.434,
+      "step": 5606
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.409565690988556,
+      "learning_rate": 6.017724397179409e-06,
+      "loss": 0.2741,
+      "step": 5607
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 150.9513680472549,
+      "learning_rate": 6.016446920334313e-06,
+      "loss": 0.4854,
+      "step": 5608
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 15.676764745216087,
+      "learning_rate": 6.015169374273677e-06,
+      "loss": 0.5405,
+      "step": 5609
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.612989783594653,
+      "learning_rate": 6.0138917590844945e-06,
+      "loss": 0.3594,
+      "step": 5610
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 79.33953910952478,
+      "learning_rate": 6.012614074853768e-06,
+      "loss": 0.4539,
+      "step": 5611
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 27.852038695052144,
+      "learning_rate": 6.011336321668498e-06,
+      "loss": 0.4146,
+      "step": 5612
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 20.96394855482594,
+      "learning_rate": 6.010058499615699e-06,
+      "loss": 0.4681,
+      "step": 5613
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.513152939215827,
+      "learning_rate": 6.008780608782379e-06,
+      "loss": 0.5768,
+      "step": 5614
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 30.072654573016084,
+      "learning_rate": 6.007502649255561e-06,
+      "loss": 0.4758,
+      "step": 5615
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 35.60352398692751,
+      "learning_rate": 6.006224621122267e-06,
+      "loss": 0.5073,
+      "step": 5616
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.6944744061225,
+      "learning_rate": 6.004946524469525e-06,
+      "loss": 0.5004,
+      "step": 5617
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.943782094797719,
+      "learning_rate": 6.003668359384368e-06,
+      "loss": 0.4783,
+      "step": 5618
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.21992064016248,
+      "learning_rate": 6.0023901259538345e-06,
+      "loss": 0.4497,
+      "step": 5619
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.633723695295767,
+      "learning_rate": 6.001111824264964e-06,
+      "loss": 0.5295,
+      "step": 5620
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 16.760122095897586,
+      "learning_rate": 5.999833454404806e-06,
+      "loss": 0.5101,
+      "step": 5621
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 19.11935536348808,
+      "learning_rate": 5.99855501646041e-06,
+      "loss": 0.5097,
+      "step": 5622
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.724007009842484,
+      "learning_rate": 5.997276510518832e-06,
+      "loss": 0.5327,
+      "step": 5623
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.168862255727609,
+      "learning_rate": 5.9959979366671325e-06,
+      "loss": 0.5136,
+      "step": 5624
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5012988686681599,
+      "learning_rate": 5.994719294992379e-06,
+      "loss": 0.7291,
+      "step": 5625
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 22.88259810663682,
+      "learning_rate": 5.993440585581639e-06,
+      "loss": 0.4472,
+      "step": 5626
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.90406126369076,
+      "learning_rate": 5.992161808521987e-06,
+      "loss": 0.5307,
+      "step": 5627
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 33.774664640175395,
+      "learning_rate": 5.9908829639005026e-06,
+      "loss": 0.4321,
+      "step": 5628
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.910417618331977,
+      "learning_rate": 5.989604051804271e-06,
+      "loss": 0.4873,
+      "step": 5629
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.693503599668862,
+      "learning_rate": 5.988325072320377e-06,
+      "loss": 0.4493,
+      "step": 5630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 25.999279797702616,
+      "learning_rate": 5.987046025535916e-06,
+      "loss": 0.5786,
+      "step": 5631
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 27.847817527276835,
+      "learning_rate": 5.985766911537986e-06,
+      "loss": 0.5184,
+      "step": 5632
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.414464098023448,
+      "learning_rate": 5.984487730413686e-06,
+      "loss": 0.4449,
+      "step": 5633
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 21.555036019510577,
+      "learning_rate": 5.983208482250126e-06,
+      "loss": 0.5095,
+      "step": 5634
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 34.79923486525435,
+      "learning_rate": 5.981929167134415e-06,
+      "loss": 0.4589,
+      "step": 5635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 13.252575008131519,
+      "learning_rate": 5.980649785153669e-06,
+      "loss": 0.445,
+      "step": 5636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 53.949743797193406,
+      "learning_rate": 5.9793703363950086e-06,
+      "loss": 0.4281,
+      "step": 5637
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.34656483325276,
+      "learning_rate": 5.978090820945559e-06,
+      "loss": 0.5676,
+      "step": 5638
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 27.951978317729925,
+      "learning_rate": 5.976811238892448e-06,
+      "loss": 0.4653,
+      "step": 5639
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 33.74511030110443,
+      "learning_rate": 5.975531590322812e-06,
+      "loss": 0.4983,
+      "step": 5640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.95963665246352,
+      "learning_rate": 5.9742518753237865e-06,
+      "loss": 0.5552,
+      "step": 5641
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 11.119072474908336,
+      "learning_rate": 5.972972093982516e-06,
+      "loss": 0.4776,
+      "step": 5642
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 28.693048100766575,
+      "learning_rate": 5.9716922463861504e-06,
+      "loss": 0.4032,
+      "step": 5643
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 33.49711267450135,
+      "learning_rate": 5.970412332621836e-06,
+      "loss": 0.3413,
+      "step": 5644
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 26.93337306426732,
+      "learning_rate": 5.969132352776735e-06,
+      "loss": 0.3877,
+      "step": 5645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 14.35959678003401,
+      "learning_rate": 5.967852306938005e-06,
+      "loss": 0.414,
+      "step": 5646
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 18.13114692826754,
+      "learning_rate": 5.96657219519281e-06,
+      "loss": 0.4618,
+      "step": 5647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.110079609385433,
+      "learning_rate": 5.9652920176283236e-06,
+      "loss": 0.3986,
+      "step": 5648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.365991832266605,
+      "learning_rate": 5.964011774331718e-06,
+      "loss": 0.5489,
+      "step": 5649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.68615790363059,
+      "learning_rate": 5.962731465390176e-06,
+      "loss": 0.4315,
+      "step": 5650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.25907672957274,
+      "learning_rate": 5.961451090890875e-06,
+      "loss": 0.4795,
+      "step": 5651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.761946986369658,
+      "learning_rate": 5.960170650921007e-06,
+      "loss": 0.4312,
+      "step": 5652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 77.6573539358925,
+      "learning_rate": 5.958890145567762e-06,
+      "loss": 0.612,
+      "step": 5653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.313289438292355,
+      "learning_rate": 5.957609574918339e-06,
+      "loss": 0.4842,
+      "step": 5654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.979812106437027,
+      "learning_rate": 5.956328939059937e-06,
+      "loss": 0.5054,
+      "step": 5655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 36.335532318197,
+      "learning_rate": 5.955048238079762e-06,
+      "loss": 0.4499,
+      "step": 5656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 28.55357236298007,
+      "learning_rate": 5.953767472065026e-06,
+      "loss": 0.5512,
+      "step": 5657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 27.615779528284477,
+      "learning_rate": 5.952486641102941e-06,
+      "loss": 0.5772,
+      "step": 5658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 72.86578480939718,
+      "learning_rate": 5.9512057452807294e-06,
+      "loss": 0.3866,
+      "step": 5659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 36.51852367654336,
+      "learning_rate": 5.94992478468561e-06,
+      "loss": 0.6142,
+      "step": 5660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.490964987712548,
+      "learning_rate": 5.948643759404814e-06,
+      "loss": 0.4247,
+      "step": 5661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.537086277496206,
+      "learning_rate": 5.947362669525571e-06,
+      "loss": 0.4443,
+      "step": 5662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.30512071755876,
+      "learning_rate": 5.94608151513512e-06,
+      "loss": 0.4348,
+      "step": 5663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.30650544379677,
+      "learning_rate": 5.9448002963207015e-06,
+      "loss": 0.6221,
+      "step": 5664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 57.05987935761666,
+      "learning_rate": 5.943519013169558e-06,
+      "loss": 0.365,
+      "step": 5665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 94.41128315666678,
+      "learning_rate": 5.942237665768945e-06,
+      "loss": 0.5017,
+      "step": 5666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 50.88314326128499,
+      "learning_rate": 5.9409562542061105e-06,
+      "loss": 0.5367,
+      "step": 5667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 28.62323080327222,
+      "learning_rate": 5.939674778568315e-06,
+      "loss": 0.4751,
+      "step": 5668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9649205358953112,
+      "learning_rate": 5.938393238942822e-06,
+      "loss": 0.7023,
+      "step": 5669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.42993213474764,
+      "learning_rate": 5.937111635416901e-06,
+      "loss": 0.3717,
+      "step": 5670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 26.491218176681844,
+      "learning_rate": 5.935829968077819e-06,
+      "loss": 0.4453,
+      "step": 5671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 206.507753978563,
+      "learning_rate": 5.934548237012854e-06,
+      "loss": 0.5057,
+      "step": 5672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 31.14347710542717,
+      "learning_rate": 5.933266442309287e-06,
+      "loss": 0.5253,
+      "step": 5673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.802902005557446,
+      "learning_rate": 5.9319845840544e-06,
+      "loss": 0.56,
+      "step": 5674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.82385379511782,
+      "learning_rate": 5.930702662335485e-06,
+      "loss": 0.4926,
+      "step": 5675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9879569991642367,
+      "learning_rate": 5.929420677239833e-06,
+      "loss": 0.7808,
+      "step": 5676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.825605454501737,
+      "learning_rate": 5.928138628854743e-06,
+      "loss": 0.5917,
+      "step": 5677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.456056074224465,
+      "learning_rate": 5.926856517267514e-06,
+      "loss": 0.4634,
+      "step": 5678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 36.735990795263824,
+      "learning_rate": 5.925574342565456e-06,
+      "loss": 0.5871,
+      "step": 5679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.04624769298401,
+      "learning_rate": 5.924292104835878e-06,
+      "loss": 0.5077,
+      "step": 5680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.817216162432032,
+      "learning_rate": 5.923009804166093e-06,
+      "loss": 0.4771,
+      "step": 5681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.469962120791102,
+      "learning_rate": 5.921727440643423e-06,
+      "loss": 0.5804,
+      "step": 5682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 24.967634649612645,
+      "learning_rate": 5.92044501435519e-06,
+      "loss": 0.5244,
+      "step": 5683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.471609874791955,
+      "learning_rate": 5.91916252538872e-06,
+      "loss": 0.4903,
+      "step": 5684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.996573442772387,
+      "learning_rate": 5.9178799738313465e-06,
+      "loss": 0.4385,
+      "step": 5685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.477016039130445,
+      "learning_rate": 5.916597359770404e-06,
+      "loss": 0.4067,
+      "step": 5686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.496312842824638,
+      "learning_rate": 5.915314683293236e-06,
+      "loss": 0.5331,
+      "step": 5687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.58892728995629,
+      "learning_rate": 5.914031944487183e-06,
+      "loss": 0.4872,
+      "step": 5688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 38.60051152147192,
+      "learning_rate": 5.912749143439597e-06,
+      "loss": 0.4784,
+      "step": 5689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.71977721151928,
+      "learning_rate": 5.91146628023783e-06,
+      "loss": 0.5018,
+      "step": 5690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.921761308070723,
+      "learning_rate": 5.910183354969239e-06,
+      "loss": 0.4712,
+      "step": 5691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.67432700015851,
+      "learning_rate": 5.908900367721186e-06,
+      "loss": 0.4624,
+      "step": 5692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.7107793980465,
+      "learning_rate": 5.907617318581037e-06,
+      "loss": 0.467,
+      "step": 5693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 88.65903325015142,
+      "learning_rate": 5.906334207636162e-06,
+      "loss": 0.3868,
+      "step": 5694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.485168594406243,
+      "learning_rate": 5.905051034973933e-06,
+      "loss": 0.4991,
+      "step": 5695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.051160103650975,
+      "learning_rate": 5.903767800681732e-06,
+      "loss": 0.5327,
+      "step": 5696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.106122306025256,
+      "learning_rate": 5.902484504846939e-06,
+      "loss": 0.4844,
+      "step": 5697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.77735060173498,
+      "learning_rate": 5.901201147556941e-06,
+      "loss": 0.5395,
+      "step": 5698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.810207386372758,
+      "learning_rate": 5.89991772889913e-06,
+      "loss": 0.4523,
+      "step": 5699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.916785458743421,
+      "learning_rate": 5.898634248960901e-06,
+      "loss": 0.529,
+      "step": 5700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.080320762556687,
+      "learning_rate": 5.897350707829652e-06,
+      "loss": 0.4665,
+      "step": 5701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 27.169837593862265,
+      "learning_rate": 5.896067105592789e-06,
+      "loss": 0.4664,
+      "step": 5702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.952681572624147,
+      "learning_rate": 5.894783442337716e-06,
+      "loss": 0.4919,
+      "step": 5703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.558036810745447,
+      "learning_rate": 5.893499718151847e-06,
+      "loss": 0.4258,
+      "step": 5704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.46625274303331,
+      "learning_rate": 5.892215933122601e-06,
+      "loss": 0.519,
+      "step": 5705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.519783718469743,
+      "learning_rate": 5.890932087337391e-06,
+      "loss": 0.4986,
+      "step": 5706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.106321593542914,
+      "learning_rate": 5.889648180883647e-06,
+      "loss": 0.4193,
+      "step": 5707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.879507213422438,
+      "learning_rate": 5.888364213848794e-06,
+      "loss": 0.7626,
+      "step": 5708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9770291078589657,
+      "learning_rate": 5.887080186320267e-06,
+      "loss": 0.7604,
+      "step": 5709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 35.90268338293074,
+      "learning_rate": 5.885796098385501e-06,
+      "loss": 0.3882,
+      "step": 5710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.500687795845598,
+      "learning_rate": 5.884511950131938e-06,
+      "loss": 0.5106,
+      "step": 5711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.581189247040527,
+      "learning_rate": 5.883227741647022e-06,
+      "loss": 0.5321,
+      "step": 5712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 38.73268391281208,
+      "learning_rate": 5.8819434730182e-06,
+      "loss": 0.4729,
+      "step": 5713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.320406177625642,
+      "learning_rate": 5.88065914433293e-06,
+      "loss": 0.5062,
+      "step": 5714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.852897243021498,
+      "learning_rate": 5.879374755678664e-06,
+      "loss": 0.4248,
+      "step": 5715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 27.984050915286993,
+      "learning_rate": 5.878090307142867e-06,
+      "loss": 0.5098,
+      "step": 5716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.281239674600812,
+      "learning_rate": 5.876805798813e-06,
+      "loss": 0.4676,
+      "step": 5717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 27.449730809672428,
+      "learning_rate": 5.875521230776536e-06,
+      "loss": 0.523,
+      "step": 5718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.250690821725073,
+      "learning_rate": 5.874236603120947e-06,
+      "loss": 0.5109,
+      "step": 5719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 25.47923151187043,
+      "learning_rate": 5.872951915933711e-06,
+      "loss": 0.4291,
+      "step": 5720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 28.689841340053807,
+      "learning_rate": 5.87166716930231e-06,
+      "loss": 0.5106,
+      "step": 5721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 37.0888246892715,
+      "learning_rate": 5.8703823633142274e-06,
+      "loss": 0.5096,
+      "step": 5722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.78955499230004,
+      "learning_rate": 5.869097498056956e-06,
+      "loss": 0.5245,
+      "step": 5723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 17.544629778366083,
+      "learning_rate": 5.867812573617986e-06,
+      "loss": 0.3782,
+      "step": 5724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 32.06990420442564,
+      "learning_rate": 5.8665275900848185e-06,
+      "loss": 0.526,
+      "step": 5725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 32.39258497355955,
+      "learning_rate": 5.865242547544952e-06,
+      "loss": 0.4809,
+      "step": 5726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.741201424078245,
+      "learning_rate": 5.8639574460858935e-06,
+      "loss": 0.5264,
+      "step": 5727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.310218147493437,
+      "learning_rate": 5.862672285795152e-06,
+      "loss": 0.5713,
+      "step": 5728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.711028913670276,
+      "learning_rate": 5.861387066760245e-06,
+      "loss": 0.346,
+      "step": 5729
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 23.714464264547356,
+      "learning_rate": 5.860101789068685e-06,
+      "loss": 0.4939,
+      "step": 5730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.153509103894358,
+      "learning_rate": 5.858816452807997e-06,
+      "loss": 0.3729,
+      "step": 5731
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.33033399303297,
+      "learning_rate": 5.857531058065706e-06,
+      "loss": 0.5367,
+      "step": 5732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.344491559267267,
+      "learning_rate": 5.8562456049293414e-06,
+      "loss": 0.4037,
+      "step": 5733
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.177667073767934,
+      "learning_rate": 5.854960093486436e-06,
+      "loss": 0.4275,
+      "step": 5734
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.717300909149591,
+      "learning_rate": 5.853674523824529e-06,
+      "loss": 0.431,
+      "step": 5735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 27.892808506489374,
+      "learning_rate": 5.852388896031161e-06,
+      "loss": 0.5049,
+      "step": 5736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 38.71338750291582,
+      "learning_rate": 5.85110321019388e-06,
+      "loss": 0.5115,
+      "step": 5737
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7182529565836144,
+      "learning_rate": 5.849817466400231e-06,
+      "loss": 0.7811,
+      "step": 5738
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.483557623772793,
+      "learning_rate": 5.848531664737771e-06,
+      "loss": 0.5058,
+      "step": 5739
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 57.777526368967074,
+      "learning_rate": 5.8472458052940565e-06,
+      "loss": 0.4636,
+      "step": 5740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7478910024577043,
+      "learning_rate": 5.845959888156649e-06,
+      "loss": 0.7808,
+      "step": 5741
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.160639989110736,
+      "learning_rate": 5.844673913413112e-06,
+      "loss": 0.4707,
+      "step": 5742
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 26.436443601462173,
+      "learning_rate": 5.843387881151017e-06,
+      "loss": 0.4394,
+      "step": 5743
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9330156279883082,
+      "learning_rate": 5.842101791457938e-06,
+      "loss": 0.7968,
+      "step": 5744
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 24.28241975989588,
+      "learning_rate": 5.8408156444214485e-06,
+      "loss": 0.4737,
+      "step": 5745
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 37.14563135675197,
+      "learning_rate": 5.8395294401291325e-06,
+      "loss": 0.4652,
+      "step": 5746
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 25.948472248103872,
+      "learning_rate": 5.838243178668572e-06,
+      "loss": 0.461,
+      "step": 5747
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.585839834651626,
+      "learning_rate": 5.836956860127357e-06,
+      "loss": 0.5272,
+      "step": 5748
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 27.756888543912986,
+      "learning_rate": 5.835670484593081e-06,
+      "loss": 0.4695,
+      "step": 5749
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.889933508954892,
+      "learning_rate": 5.8343840521533415e-06,
+      "loss": 0.4964,
+      "step": 5750
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 43.862520049067975,
+      "learning_rate": 5.833097562895735e-06,
+      "loss": 0.5386,
+      "step": 5751
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.569653736564458,
+      "learning_rate": 5.831811016907869e-06,
+      "loss": 0.567,
+      "step": 5752
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 19.05035029938852,
+      "learning_rate": 5.830524414277351e-06,
+      "loss": 0.3877,
+      "step": 5753
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.024083290395843,
+      "learning_rate": 5.829237755091792e-06,
+      "loss": 0.4229,
+      "step": 5754
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 15.19504480937777,
+      "learning_rate": 5.827951039438809e-06,
+      "loss": 0.4301,
+      "step": 5755
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 31.189858447916585,
+      "learning_rate": 5.826664267406019e-06,
+      "loss": 0.46,
+      "step": 5756
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 43.65139764211973,
+      "learning_rate": 5.825377439081049e-06,
+      "loss": 0.4603,
+      "step": 5757
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 21.70888957424251,
+      "learning_rate": 5.8240905545515244e-06,
+      "loss": 0.5953,
+      "step": 5758
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.08425112432749,
+      "learning_rate": 5.8228036139050766e-06,
+      "loss": 0.511,
+      "step": 5759
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.770492534468079,
+      "learning_rate": 5.8215166172293415e-06,
+      "loss": 0.4746,
+      "step": 5760
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 26.78903929283302,
+      "learning_rate": 5.820229564611956e-06,
+      "loss": 0.5223,
+      "step": 5761
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 175.25192466580236,
+      "learning_rate": 5.818942456140565e-06,
+      "loss": 0.456,
+      "step": 5762
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.125578226877842,
+      "learning_rate": 5.8176552919028125e-06,
+      "loss": 0.4954,
+      "step": 5763
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 14.093809893481636,
+      "learning_rate": 5.816368071986351e-06,
+      "loss": 0.4938,
+      "step": 5764
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.487489994366939,
+      "learning_rate": 5.815080796478833e-06,
+      "loss": 0.4872,
+      "step": 5765
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 30.512852125184647,
+      "learning_rate": 5.813793465467915e-06,
+      "loss": 0.4253,
+      "step": 5766
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.004987881033637,
+      "learning_rate": 5.812506079041263e-06,
+      "loss": 0.5641,
+      "step": 5767
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 26.892794293076854,
+      "learning_rate": 5.811218637286535e-06,
+      "loss": 0.534,
+      "step": 5768
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 94.17763719251555,
+      "learning_rate": 5.809931140291408e-06,
+      "loss": 0.4586,
+      "step": 5769
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 16.63267200047627,
+      "learning_rate": 5.808643588143551e-06,
+      "loss": 0.4444,
+      "step": 5770
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 20.66087652273546,
+      "learning_rate": 5.8073559809306415e-06,
+      "loss": 0.4606,
+      "step": 5771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 31.559977423702797,
+      "learning_rate": 5.8060683187403574e-06,
+      "loss": 0.4172,
+      "step": 5772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.641109857465914,
+      "learning_rate": 5.804780601660388e-06,
+      "loss": 0.4464,
+      "step": 5773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.062757228915649,
+      "learning_rate": 5.803492829778415e-06,
+      "loss": 0.4132,
+      "step": 5774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.910969696809188,
+      "learning_rate": 5.802205003182134e-06,
+      "loss": 0.4644,
+      "step": 5775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.789155747708238,
+      "learning_rate": 5.80091712195924e-06,
+      "loss": 0.4581,
+      "step": 5776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.567394732428777,
+      "learning_rate": 5.79962918619743e-06,
+      "loss": 0.438,
+      "step": 5777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.7362891933261,
+      "learning_rate": 5.798341195984407e-06,
+      "loss": 0.3659,
+      "step": 5778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.426367333889948,
+      "learning_rate": 5.79705315140788e-06,
+      "loss": 0.4534,
+      "step": 5779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.18904336111527,
+      "learning_rate": 5.795765052555557e-06,
+      "loss": 0.5006,
+      "step": 5780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 47.586129815804284,
+      "learning_rate": 5.794476899515153e-06,
+      "loss": 0.4867,
+      "step": 5781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 57.23615556809289,
+      "learning_rate": 5.793188692374384e-06,
+      "loss": 0.4438,
+      "step": 5782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.163051090888718,
+      "learning_rate": 5.791900431220974e-06,
+      "loss": 0.459,
+      "step": 5783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.92754088068561,
+      "learning_rate": 5.790612116142643e-06,
+      "loss": 0.431,
+      "step": 5784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.11645458455111,
+      "learning_rate": 5.789323747227126e-06,
+      "loss": 0.5148,
+      "step": 5785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.59716407965029,
+      "learning_rate": 5.788035324562149e-06,
+      "loss": 0.5558,
+      "step": 5786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.813615381343714,
+      "learning_rate": 5.786746848235452e-06,
+      "loss": 0.4237,
+      "step": 5787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 30.73500867814401,
+      "learning_rate": 5.7854583183347725e-06,
+      "loss": 0.3467,
+      "step": 5788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 26.68138327574717,
+      "learning_rate": 5.784169734947855e-06,
+      "loss": 0.4631,
+      "step": 5789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 26.732398948648807,
+      "learning_rate": 5.782881098162444e-06,
+      "loss": 0.5117,
+      "step": 5790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.39681647154442,
+      "learning_rate": 5.781592408066293e-06,
+      "loss": 0.4566,
+      "step": 5791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.11798940480814,
+      "learning_rate": 5.780303664747155e-06,
+      "loss": 0.3808,
+      "step": 5792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.268097042596404,
+      "learning_rate": 5.779014868292787e-06,
+      "loss": 0.4592,
+      "step": 5793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.88141011141419,
+      "learning_rate": 5.7777260187909525e-06,
+      "loss": 0.6106,
+      "step": 5794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 26.611603694351494,
+      "learning_rate": 5.7764371163294135e-06,
+      "loss": 0.5305,
+      "step": 5795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.895426790424324,
+      "learning_rate": 5.775148160995941e-06,
+      "loss": 0.5267,
+      "step": 5796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.23353714661784,
+      "learning_rate": 5.773859152878304e-06,
+      "loss": 0.4203,
+      "step": 5797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 28.469049034893654,
+      "learning_rate": 5.77257009206428e-06,
+      "loss": 0.4742,
+      "step": 5798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.585726148300523,
+      "learning_rate": 5.77128097864165e-06,
+      "loss": 0.487,
+      "step": 5799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.16483590339656,
+      "learning_rate": 5.769991812698195e-06,
+      "loss": 0.45,
+      "step": 5800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.295449075055988,
+      "learning_rate": 5.7687025943217025e-06,
+      "loss": 0.5155,
+      "step": 5801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 26.80036172556736,
+      "learning_rate": 5.767413323599962e-06,
+      "loss": 0.3353,
+      "step": 5802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.084239171301098,
+      "learning_rate": 5.766124000620767e-06,
+      "loss": 0.522,
+      "step": 5803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 18.547854068504208,
+      "learning_rate": 5.764834625471916e-06,
+      "loss": 0.4944,
+      "step": 5804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.659588041907895,
+      "learning_rate": 5.763545198241207e-06,
+      "loss": 0.4292,
+      "step": 5805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.646315165883003,
+      "learning_rate": 5.7622557190164475e-06,
+      "loss": 0.5293,
+      "step": 5806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 33.71773500200797,
+      "learning_rate": 5.760966187885443e-06,
+      "loss": 0.455,
+      "step": 5807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.118234127279974,
+      "learning_rate": 5.7596766049360065e-06,
+      "loss": 0.4428,
+      "step": 5808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.260248387216516,
+      "learning_rate": 5.758386970255949e-06,
+      "loss": 0.4908,
+      "step": 5809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.632143605113647,
+      "learning_rate": 5.7570972839330965e-06,
+      "loss": 0.5624,
+      "step": 5810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 41.70866019938723,
+      "learning_rate": 5.755807546055263e-06,
+      "loss": 0.5713,
+      "step": 5811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 25.838121487089172,
+      "learning_rate": 5.75451775671028e-06,
+      "loss": 0.5098,
+      "step": 5812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.216984161369714,
+      "learning_rate": 5.753227915985972e-06,
+      "loss": 0.3697,
+      "step": 5813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.46102881427275,
+      "learning_rate": 5.751938023970174e-06,
+      "loss": 0.5268,
+      "step": 5814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.2521908233429,
+      "learning_rate": 5.7506480807507215e-06,
+      "loss": 0.529,
+      "step": 5815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 38.18544737433413,
+      "learning_rate": 5.749358086415453e-06,
+      "loss": 0.5322,
+      "step": 5816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 54.05129799104052,
+      "learning_rate": 5.7480680410522125e-06,
+      "loss": 0.6618,
+      "step": 5817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 15.944800392745831,
+      "learning_rate": 5.746777944748846e-06,
+      "loss": 0.4641,
+      "step": 5818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6845333907202158,
+      "learning_rate": 5.7454877975932026e-06,
+      "loss": 0.7573,
+      "step": 5819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.856109554131905,
+      "learning_rate": 5.744197599673136e-06,
+      "loss": 0.5025,
+      "step": 5820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.831737171724729,
+      "learning_rate": 5.7429073510765055e-06,
+      "loss": 0.4683,
+      "step": 5821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 31.929699060070995,
+      "learning_rate": 5.741617051891166e-06,
+      "loss": 0.4264,
+      "step": 5822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.537740127515493,
+      "learning_rate": 5.740326702204986e-06,
+      "loss": 0.5671,
+      "step": 5823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 23.60662458171216,
+      "learning_rate": 5.739036302105832e-06,
+      "loss": 0.5116,
+      "step": 5824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.863740547547998,
+      "learning_rate": 5.737745851681572e-06,
+      "loss": 0.5356,
+      "step": 5825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 24.09672192599432,
+      "learning_rate": 5.736455351020083e-06,
+      "loss": 0.4411,
+      "step": 5826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 18.064121658579,
+      "learning_rate": 5.735164800209238e-06,
+      "loss": 0.5387,
+      "step": 5827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.769849783696177,
+      "learning_rate": 5.733874199336922e-06,
+      "loss": 0.508,
+      "step": 5828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.35805315429012,
+      "learning_rate": 5.732583548491017e-06,
+      "loss": 0.5134,
+      "step": 5829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.020324445486974,
+      "learning_rate": 5.731292847759413e-06,
+      "loss": 0.5511,
+      "step": 5830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 18.64378335062785,
+      "learning_rate": 5.73000209723e-06,
+      "loss": 0.4032,
+      "step": 5831
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 40.82544208224531,
+      "learning_rate": 5.72871129699067e-06,
+      "loss": 0.5266,
+      "step": 5832
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 18.206069523407702,
+      "learning_rate": 5.727420447129324e-06,
+      "loss": 0.4865,
+      "step": 5833
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.949612235617508,
+      "learning_rate": 5.726129547733861e-06,
+      "loss": 0.4993,
+      "step": 5834
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.637429593256789,
+      "learning_rate": 5.724838598892188e-06,
+      "loss": 0.4661,
+      "step": 5835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 26.486048479748266,
+      "learning_rate": 5.723547600692208e-06,
+      "loss": 0.4856,
+      "step": 5836
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 25.557409696372506,
+      "learning_rate": 5.722256553221838e-06,
+      "loss": 0.4899,
+      "step": 5837
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.172608495188772,
+      "learning_rate": 5.72096545656899e-06,
+      "loss": 0.5171,
+      "step": 5838
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.016908653140817,
+      "learning_rate": 5.71967431082158e-06,
+      "loss": 0.4114,
+      "step": 5839
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.20589093735394,
+      "learning_rate": 5.718383116067534e-06,
+      "loss": 0.4847,
+      "step": 5840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.964640115812106,
+      "learning_rate": 5.7170918723947724e-06,
+      "loss": 0.4118,
+      "step": 5841
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.581179992124538,
+      "learning_rate": 5.715800579891226e-06,
+      "loss": 0.4523,
+      "step": 5842
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.479365513739715,
+      "learning_rate": 5.714509238644824e-06,
+      "loss": 0.3883,
+      "step": 5843
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 30.845103197908212,
+      "learning_rate": 5.7132178487435045e-06,
+      "loss": 0.4688,
+      "step": 5844
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.663626391438363,
+      "learning_rate": 5.7119264102752e-06,
+      "loss": 0.477,
+      "step": 5845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 53.59864549637048,
+      "learning_rate": 5.710634923327856e-06,
+      "loss": 0.4733,
+      "step": 5846
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.036828809275766,
+      "learning_rate": 5.7093433879894154e-06,
+      "loss": 0.4454,
+      "step": 5847
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.533636717906283,
+      "learning_rate": 5.708051804347826e-06,
+      "loss": 0.4477,
+      "step": 5848
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 28.293750156072942,
+      "learning_rate": 5.70676017249104e-06,
+      "loss": 0.444,
+      "step": 5849
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.97569438957179,
+      "learning_rate": 5.7054684925070105e-06,
+      "loss": 0.4668,
+      "step": 5850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.829428054393839,
+      "learning_rate": 5.704176764483698e-06,
+      "loss": 0.5096,
+      "step": 5851
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.849381047902703,
+      "learning_rate": 5.7028849885090586e-06,
+      "loss": 0.3721,
+      "step": 5852
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.96380281652589,
+      "learning_rate": 5.70159316467106e-06,
+      "loss": 0.612,
+      "step": 5853
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.603549903204346,
+      "learning_rate": 5.700301293057671e-06,
+      "loss": 0.4509,
+      "step": 5854
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.684951687193136,
+      "learning_rate": 5.699009373756858e-06,
+      "loss": 0.4015,
+      "step": 5855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 53.69870983138704,
+      "learning_rate": 5.697717406856598e-06,
+      "loss": 0.4985,
+      "step": 5856
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.848318550835312,
+      "learning_rate": 5.696425392444868e-06,
+      "loss": 0.5362,
+      "step": 5857
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.398783934765499,
+      "learning_rate": 5.695133330609648e-06,
+      "loss": 0.4762,
+      "step": 5858
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1144590786098614,
+      "learning_rate": 5.69384122143892e-06,
+      "loss": 0.6844,
+      "step": 5859
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.785439721556733,
+      "learning_rate": 5.692549065020674e-06,
+      "loss": 0.5415,
+      "step": 5860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 39.97515881383588,
+      "learning_rate": 5.691256861442898e-06,
+      "loss": 0.5698,
+      "step": 5861
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.068507011536859,
+      "learning_rate": 5.689964610793586e-06,
+      "loss": 0.5589,
+      "step": 5862
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.051415692475224,
+      "learning_rate": 5.688672313160736e-06,
+      "loss": 0.5919,
+      "step": 5863
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.502887579850896,
+      "learning_rate": 5.687379968632345e-06,
+      "loss": 0.5072,
+      "step": 5864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.792698090230495,
+      "learning_rate": 5.68608757729642e-06,
+      "loss": 0.4149,
+      "step": 5865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.429827931496114,
+      "learning_rate": 5.684795139240962e-06,
+      "loss": 0.4881,
+      "step": 5866
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.24887488810741,
+      "learning_rate": 5.683502654553983e-06,
+      "loss": 0.6136,
+      "step": 5867
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.564476467411449,
+      "learning_rate": 5.6822101233234946e-06,
+      "loss": 0.5314,
+      "step": 5868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 19.43005126683465,
+      "learning_rate": 5.680917545637514e-06,
+      "loss": 0.4617,
+      "step": 5869
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.205862452913552,
+      "learning_rate": 5.67962492158406e-06,
+      "loss": 0.5142,
+      "step": 5870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 26.481351666455215,
+      "learning_rate": 5.678332251251153e-06,
+      "loss": 0.4779,
+      "step": 5871
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.587810578304419,
+      "learning_rate": 5.677039534726819e-06,
+      "loss": 0.3864,
+      "step": 5872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.874908217632779,
+      "learning_rate": 5.675746772099086e-06,
+      "loss": 0.5468,
+      "step": 5873
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.24287910348663,
+      "learning_rate": 5.674453963455987e-06,
+      "loss": 0.5547,
+      "step": 5874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.646952939495,
+      "learning_rate": 5.673161108885554e-06,
+      "loss": 0.4652,
+      "step": 5875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.018433916792118,
+      "learning_rate": 5.671868208475825e-06,
+      "loss": 0.4675,
+      "step": 5876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.097661112914107,
+      "learning_rate": 5.670575262314844e-06,
+      "loss": 0.6026,
+      "step": 5877
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.961908005704995,
+      "learning_rate": 5.66928227049065e-06,
+      "loss": 0.4426,
+      "step": 5878
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.318157631804617,
+      "learning_rate": 5.667989233091293e-06,
+      "loss": 0.4183,
+      "step": 5879
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 34.64478442094014,
+      "learning_rate": 5.666696150204824e-06,
+      "loss": 0.5849,
+      "step": 5880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.467833385392538,
+      "learning_rate": 5.665403021919293e-06,
+      "loss": 0.5133,
+      "step": 5881
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.155911828245806,
+      "learning_rate": 5.66410984832276e-06,
+      "loss": 0.4394,
+      "step": 5882
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 30.773898274045695,
+      "learning_rate": 5.662816629503283e-06,
+      "loss": 0.4586,
+      "step": 5883
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.837438562340061,
+      "learning_rate": 5.661523365548923e-06,
+      "loss": 0.4053,
+      "step": 5884
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.996493472802005,
+      "learning_rate": 5.660230056547747e-06,
+      "loss": 0.5197,
+      "step": 5885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.227862310568863,
+      "learning_rate": 5.658936702587824e-06,
+      "loss": 0.556,
+      "step": 5886
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 36.060461327177165,
+      "learning_rate": 5.6576433037572245e-06,
+      "loss": 0.5141,
+      "step": 5887
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 18.57664644749339,
+      "learning_rate": 5.656349860144024e-06,
+      "loss": 0.5383,
+      "step": 5888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.637743606716532,
+      "learning_rate": 5.655056371836298e-06,
+      "loss": 0.406,
+      "step": 5889
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.93129550611275,
+      "learning_rate": 5.653762838922132e-06,
+      "loss": 0.4983,
+      "step": 5890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.144418634184518,
+      "learning_rate": 5.652469261489606e-06,
+      "loss": 0.4717,
+      "step": 5891
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.424711047938493,
+      "learning_rate": 5.651175639626809e-06,
+      "loss": 0.5288,
+      "step": 5892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.278625769533239,
+      "learning_rate": 5.649881973421829e-06,
+      "loss": 0.7755,
+      "step": 5893
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 21.321270618177223,
+      "learning_rate": 5.64858826296276e-06,
+      "loss": 0.4695,
+      "step": 5894
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 22.32707947616129,
+      "learning_rate": 5.6472945083376994e-06,
+      "loss": 0.3149,
+      "step": 5895
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 72.19311863548907,
+      "learning_rate": 5.646000709634743e-06,
+      "loss": 0.5486,
+      "step": 5896
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.327995998004615,
+      "learning_rate": 5.644706866941997e-06,
+      "loss": 0.5737,
+      "step": 5897
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.62372373440045,
+      "learning_rate": 5.643412980347562e-06,
+      "loss": 0.4736,
+      "step": 5898
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.630874774240738,
+      "learning_rate": 5.642119049939548e-06,
+      "loss": 0.5582,
+      "step": 5899
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.270314300274693,
+      "learning_rate": 5.640825075806066e-06,
+      "loss": 0.4826,
+      "step": 5900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.438455748921816,
+      "learning_rate": 5.6395310580352296e-06,
+      "loss": 0.4649,
+      "step": 5901
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.582056978486065,
+      "learning_rate": 5.638236996715156e-06,
+      "loss": 0.3962,
+      "step": 5902
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 53.46946998459515,
+      "learning_rate": 5.636942891933966e-06,
+      "loss": 0.4424,
+      "step": 5903
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8571008533327178,
+      "learning_rate": 5.635648743779779e-06,
+      "loss": 0.7239,
+      "step": 5904
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.32064166944387,
+      "learning_rate": 5.634354552340724e-06,
+      "loss": 0.3034,
+      "step": 5905
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.05939699341803,
+      "learning_rate": 5.633060317704931e-06,
+      "loss": 0.4996,
+      "step": 5906
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.05126124240775,
+      "learning_rate": 5.631766039960525e-06,
+      "loss": 0.7686,
+      "step": 5907
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.62690004242662,
+      "learning_rate": 5.630471719195648e-06,
+      "loss": 0.4788,
+      "step": 5908
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.603706696558627,
+      "learning_rate": 5.629177355498433e-06,
+      "loss": 0.4104,
+      "step": 5909
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.022075421406521,
+      "learning_rate": 5.627882948957023e-06,
+      "loss": 0.4069,
+      "step": 5910
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.599067143450249,
+      "learning_rate": 5.626588499659561e-06,
+      "loss": 0.5374,
+      "step": 5911
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.83292821023461,
+      "learning_rate": 5.625294007694191e-06,
+      "loss": 0.5033,
+      "step": 5912
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.242348768944309,
+      "learning_rate": 5.623999473149065e-06,
+      "loss": 0.4944,
+      "step": 5913
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.620300041011868,
+      "learning_rate": 5.622704896112333e-06,
+      "loss": 0.4588,
+      "step": 5914
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 40.270417638895516,
+      "learning_rate": 5.6214102766721525e-06,
+      "loss": 0.5161,
+      "step": 5915
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.260857789437257,
+      "learning_rate": 5.620115614916678e-06,
+      "loss": 0.4456,
+      "step": 5916
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.807172818992766,
+      "learning_rate": 5.618820910934072e-06,
+      "loss": 0.4764,
+      "step": 5917
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 17.0797194566387,
+      "learning_rate": 5.617526164812499e-06,
+      "loss": 0.4755,
+      "step": 5918
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.306806673584877,
+      "learning_rate": 5.616231376640123e-06,
+      "loss": 0.5068,
+      "step": 5919
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.262311637086283,
+      "learning_rate": 5.614936546505117e-06,
+      "loss": 0.5448,
+      "step": 5920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.825700458007458,
+      "learning_rate": 5.613641674495649e-06,
+      "loss": 0.7895,
+      "step": 5921
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.999437004642278,
+      "learning_rate": 5.612346760699898e-06,
+      "loss": 0.5534,
+      "step": 5922
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.350405501755965,
+      "learning_rate": 5.611051805206039e-06,
+      "loss": 0.4475,
+      "step": 5923
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.27984445434619,
+      "learning_rate": 5.609756808102254e-06,
+      "loss": 0.4154,
+      "step": 5924
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.67474684162296,
+      "learning_rate": 5.608461769476727e-06,
+      "loss": 0.4145,
+      "step": 5925
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 31.318201078208613,
+      "learning_rate": 5.607166689417642e-06,
+      "loss": 0.4584,
+      "step": 5926
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.8326370282329,
+      "learning_rate": 5.605871568013193e-06,
+      "loss": 0.438,
+      "step": 5927
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5198273216058629,
+      "learning_rate": 5.6045764053515665e-06,
+      "loss": 0.7353,
+      "step": 5928
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.589893017107732,
+      "learning_rate": 5.603281201520962e-06,
+      "loss": 0.481,
+      "step": 5929
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.296383990318372,
+      "learning_rate": 5.601985956609572e-06,
+      "loss": 0.6002,
+      "step": 5930
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.613256616919067,
+      "learning_rate": 5.600690670705602e-06,
+      "loss": 0.4691,
+      "step": 5931
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.618466329932629,
+      "learning_rate": 5.5993953438972525e-06,
+      "loss": 0.4219,
+      "step": 5932
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.667696352793148,
+      "learning_rate": 5.598099976272731e-06,
+      "loss": 0.4733,
+      "step": 5933
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.942222662993155,
+      "learning_rate": 5.596804567920245e-06,
+      "loss": 0.744,
+      "step": 5934
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.86829636820146,
+      "learning_rate": 5.595509118928008e-06,
+      "loss": 0.53,
+      "step": 5935
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.218812488822785,
+      "learning_rate": 5.594213629384232e-06,
+      "loss": 0.4718,
+      "step": 5936
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.770764469555287,
+      "learning_rate": 5.592918099377135e-06,
+      "loss": 0.5305,
+      "step": 5937
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.048561863939279,
+      "learning_rate": 5.591622528994936e-06,
+      "loss": 0.4746,
+      "step": 5938
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.424032400524872,
+      "learning_rate": 5.590326918325861e-06,
+      "loss": 0.5231,
+      "step": 5939
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 28.021809573719242,
+      "learning_rate": 5.5890312674581315e-06,
+      "loss": 0.5212,
+      "step": 5940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.769424152964413,
+      "learning_rate": 5.5877355764799765e-06,
+      "loss": 0.4457,
+      "step": 5941
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 54.29313055714654,
+      "learning_rate": 5.586439845479629e-06,
+      "loss": 0.5523,
+      "step": 5942
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.030304705141033,
+      "learning_rate": 5.58514407454532e-06,
+      "loss": 0.4478,
+      "step": 5943
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.124451201993345,
+      "learning_rate": 5.583848263765287e-06,
+      "loss": 0.4887,
+      "step": 5944
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.02429166011633,
+      "learning_rate": 5.582552413227769e-06,
+      "loss": 0.5044,
+      "step": 5945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.487707254975671,
+      "learning_rate": 5.581256523021007e-06,
+      "loss": 0.4579,
+      "step": 5946
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.246375995720866,
+      "learning_rate": 5.5799605932332445e-06,
+      "loss": 0.573,
+      "step": 5947
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.264728239771562,
+      "learning_rate": 5.578664623952732e-06,
+      "loss": 0.4615,
+      "step": 5948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.978800020141087,
+      "learning_rate": 5.577368615267716e-06,
+      "loss": 0.3946,
+      "step": 5949
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.109868101088416,
+      "learning_rate": 5.5760725672664494e-06,
+      "loss": 0.5203,
+      "step": 5950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.525873457171993,
+      "learning_rate": 5.57477648003719e-06,
+      "loss": 0.5292,
+      "step": 5951
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.19103411585924,
+      "learning_rate": 5.573480353668192e-06,
+      "loss": 0.4734,
+      "step": 5952
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 27.859284388687364,
+      "learning_rate": 5.572184188247718e-06,
+      "loss": 0.5449,
+      "step": 5953
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.437502020140974,
+      "learning_rate": 5.570887983864032e-06,
+      "loss": 0.5242,
+      "step": 5954
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.493882253733483,
+      "learning_rate": 5.569591740605397e-06,
+      "loss": 0.4948,
+      "step": 5955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.377836872061458,
+      "learning_rate": 5.568295458560083e-06,
+      "loss": 0.5033,
+      "step": 5956
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.651320587567676,
+      "learning_rate": 5.566999137816361e-06,
+      "loss": 0.6048,
+      "step": 5957
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.993920666726698,
+      "learning_rate": 5.565702778462505e-06,
+      "loss": 0.3506,
+      "step": 5958
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.789926426822955,
+      "learning_rate": 5.564406380586792e-06,
+      "loss": 0.5557,
+      "step": 5959
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.06006294018052,
+      "learning_rate": 5.563109944277497e-06,
+      "loss": 0.58,
+      "step": 5960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.09116053531565,
+      "learning_rate": 5.5618134696229075e-06,
+      "loss": 0.5304,
+      "step": 5961
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.093061971467238,
+      "learning_rate": 5.560516956711302e-06,
+      "loss": 0.4961,
+      "step": 5962
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 15.18522426556895,
+      "learning_rate": 5.559220405630974e-06,
+      "loss": 0.52,
+      "step": 5963
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.805005495119156,
+      "learning_rate": 5.5579238164702055e-06,
+      "loss": 0.4284,
+      "step": 5964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.883705296573348,
+      "learning_rate": 5.556627189317293e-06,
+      "loss": 0.4718,
+      "step": 5965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.057852490639267,
+      "learning_rate": 5.5553305242605306e-06,
+      "loss": 0.5342,
+      "step": 5966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 18.5861182860255,
+      "learning_rate": 5.554033821388214e-06,
+      "loss": 0.5083,
+      "step": 5967
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.625802085835429,
+      "learning_rate": 5.552737080788645e-06,
+      "loss": 0.4935,
+      "step": 5968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.149419262068081,
+      "learning_rate": 5.551440302550124e-06,
+      "loss": 0.5021,
+      "step": 5969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.688189762555197,
+      "learning_rate": 5.550143486760956e-06,
+      "loss": 0.5405,
+      "step": 5970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.142623720032933,
+      "learning_rate": 5.548846633509448e-06,
+      "loss": 0.5037,
+      "step": 5971
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.119708364498433,
+      "learning_rate": 5.547549742883913e-06,
+      "loss": 0.3956,
+      "step": 5972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.81190599202035,
+      "learning_rate": 5.546252814972661e-06,
+      "loss": 0.3785,
+      "step": 5973
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 41.38738367324692,
+      "learning_rate": 5.544955849864008e-06,
+      "loss": 0.4225,
+      "step": 5974
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.539012828218068,
+      "learning_rate": 5.54365884764627e-06,
+      "loss": 0.5577,
+      "step": 5975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.594546046827224,
+      "learning_rate": 5.542361808407769e-06,
+      "loss": 0.5075,
+      "step": 5976
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.93193376508858,
+      "learning_rate": 5.541064732236827e-06,
+      "loss": 0.5425,
+      "step": 5977
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.005904569910765,
+      "learning_rate": 5.539767619221769e-06,
+      "loss": 0.4433,
+      "step": 5978
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.777002431271802,
+      "learning_rate": 5.538470469450921e-06,
+      "loss": 0.506,
+      "step": 5979
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.364166928064478,
+      "learning_rate": 5.537173283012616e-06,
+      "loss": 0.4833,
+      "step": 5980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.053001962056486,
+      "learning_rate": 5.535876059995185e-06,
+      "loss": 0.4596,
+      "step": 5981
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.828997307431855,
+      "learning_rate": 5.534578800486966e-06,
+      "loss": 0.4222,
+      "step": 5982
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.965381346424159,
+      "learning_rate": 5.533281504576292e-06,
+      "loss": 0.4603,
+      "step": 5983
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 19.663566847433557,
+      "learning_rate": 5.531984172351506e-06,
+      "loss": 0.3956,
+      "step": 5984
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.951305873888174,
+      "learning_rate": 5.530686803900951e-06,
+      "loss": 0.5281,
+      "step": 5985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.82618632728959,
+      "learning_rate": 5.5293893993129695e-06,
+      "loss": 0.4764,
+      "step": 5986
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7086719071115553,
+      "learning_rate": 5.5280919586759114e-06,
+      "loss": 0.7902,
+      "step": 5987
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.52089389332474,
+      "learning_rate": 5.526794482078126e-06,
+      "loss": 0.5402,
+      "step": 5988
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.325377977012406,
+      "learning_rate": 5.5254969696079666e-06,
+      "loss": 0.5828,
+      "step": 5989
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 24.169120447946817,
+      "learning_rate": 5.524199421353784e-06,
+      "loss": 0.4727,
+      "step": 5990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 15.488490315945432,
+      "learning_rate": 5.522901837403941e-06,
+      "loss": 0.6454,
+      "step": 5991
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.277715352132924,
+      "learning_rate": 5.521604217846793e-06,
+      "loss": 0.414,
+      "step": 5992
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.361720597888608,
+      "learning_rate": 5.520306562770706e-06,
+      "loss": 0.5375,
+      "step": 5993
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.728130107983823,
+      "learning_rate": 5.519008872264041e-06,
+      "loss": 0.4651,
+      "step": 5994
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.407263466147226,
+      "learning_rate": 5.517711146415167e-06,
+      "loss": 0.5381,
+      "step": 5995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.692483872775846,
+      "learning_rate": 5.516413385312454e-06,
+      "loss": 0.4489,
+      "step": 5996
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.907272238012974,
+      "learning_rate": 5.515115589044272e-06,
+      "loss": 0.4763,
+      "step": 5997
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.671290190972354,
+      "learning_rate": 5.513817757698997e-06,
+      "loss": 0.4236,
+      "step": 5998
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.888660650434168,
+      "learning_rate": 5.512519891365002e-06,
+      "loss": 0.4501,
+      "step": 5999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.842835789514129,
+      "learning_rate": 5.511221990130671e-06,
+      "loss": 0.4372,
+      "step": 6000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.46844375786068,
+      "learning_rate": 5.509924054084381e-06,
+      "loss": 0.4244,
+      "step": 6001
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.754278817048071,
+      "learning_rate": 5.5086260833145185e-06,
+      "loss": 0.4952,
+      "step": 6002
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 13.932675402943874,
+      "learning_rate": 5.5073280779094685e-06,
+      "loss": 0.4156,
+      "step": 6003
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.655482065871583,
+      "learning_rate": 5.506030037957619e-06,
+      "loss": 0.502,
+      "step": 6004
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.9806268017075395,
+      "learning_rate": 5.504731963547363e-06,
+      "loss": 0.4046,
+      "step": 6005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.078439390497064,
+      "learning_rate": 5.50343385476709e-06,
+      "loss": 0.4744,
+      "step": 6006
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.566786806854063,
+      "learning_rate": 5.502135711705199e-06,
+      "loss": 0.4851,
+      "step": 6007
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 11.282636448630386,
+      "learning_rate": 5.500837534450084e-06,
+      "loss": 0.4589,
+      "step": 6008
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7494205178103097,
+      "learning_rate": 5.499539323090148e-06,
+      "loss": 0.7672,
+      "step": 6009
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.31454545136136,
+      "learning_rate": 5.498241077713794e-06,
+      "loss": 0.4893,
+      "step": 6010
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.506441729360887,
+      "learning_rate": 5.4969427984094255e-06,
+      "loss": 0.4226,
+      "step": 6011
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.006824637161963,
+      "learning_rate": 5.495644485265449e-06,
+      "loss": 0.5095,
+      "step": 6012
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.115225892542643,
+      "learning_rate": 5.494346138370275e-06,
+      "loss": 0.4481,
+      "step": 6013
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 20.716042689983386,
+      "learning_rate": 5.493047757812314e-06,
+      "loss": 0.4391,
+      "step": 6014
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.053794536573228,
+      "learning_rate": 5.4917493436799805e-06,
+      "loss": 0.4853,
+      "step": 6015
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 16.007871687351805,
+      "learning_rate": 5.490450896061692e-06,
+      "loss": 0.4473,
+      "step": 6016
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.486993652076462,
+      "learning_rate": 5.489152415045865e-06,
+      "loss": 0.542,
+      "step": 6017
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 9.869782408101472,
+      "learning_rate": 5.487853900720922e-06,
+      "loss": 0.5399,
+      "step": 6018
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.363602471699819,
+      "learning_rate": 5.486555353175286e-06,
+      "loss": 0.5229,
+      "step": 6019
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 14.56216344712661,
+      "learning_rate": 5.4852567724973796e-06,
+      "loss": 0.4643,
+      "step": 6020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.9579487072602175,
+      "learning_rate": 5.483958158775634e-06,
+      "loss": 0.4855,
+      "step": 6021
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.243616949970987,
+      "learning_rate": 5.4826595120984775e-06,
+      "loss": 0.3882,
+      "step": 6022
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.386571944828221,
+      "learning_rate": 5.481360832554342e-06,
+      "loss": 0.5563,
+      "step": 6023
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.338924738993729,
+      "learning_rate": 5.480062120231662e-06,
+      "loss": 0.5272,
+      "step": 6024
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.812048516386763,
+      "learning_rate": 5.478763375218875e-06,
+      "loss": 0.4316,
+      "step": 6025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6158351462713896,
+      "learning_rate": 5.477464597604418e-06,
+      "loss": 0.7416,
+      "step": 6026
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5010375784448082,
+      "learning_rate": 5.476165787476734e-06,
+      "loss": 0.6955,
+      "step": 6027
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.55200191121485,
+      "learning_rate": 5.474866944924266e-06,
+      "loss": 0.4609,
+      "step": 6028
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.003133378134754,
+      "learning_rate": 5.473568070035455e-06,
+      "loss": 0.6323,
+      "step": 6029
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.279815605093606,
+      "learning_rate": 5.472269162898756e-06,
+      "loss": 0.4697,
+      "step": 6030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.436273220753292,
+      "learning_rate": 5.470970223602613e-06,
+      "loss": 0.4928,
+      "step": 6031
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 37.002837559155914,
+      "learning_rate": 5.469671252235481e-06,
+      "loss": 0.545,
+      "step": 6032
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.785807089467218,
+      "learning_rate": 5.468372248885813e-06,
+      "loss": 0.4322,
+      "step": 6033
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.249306508194742,
+      "learning_rate": 5.467073213642067e-06,
+      "loss": 0.5333,
+      "step": 6034
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.896789986332099,
+      "learning_rate": 5.465774146592698e-06,
+      "loss": 0.4363,
+      "step": 6035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.108824082692225,
+      "learning_rate": 5.464475047826171e-06,
+      "loss": 0.51,
+      "step": 6036
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.392524654218608,
+      "learning_rate": 5.463175917430946e-06,
+      "loss": 0.5316,
+      "step": 6037
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.882610291657596,
+      "learning_rate": 5.461876755495489e-06,
+      "loss": 0.5749,
+      "step": 6038
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.627597648359279,
+      "learning_rate": 5.460577562108268e-06,
+      "loss": 0.4012,
+      "step": 6039
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.696021740569318,
+      "learning_rate": 5.4592783373577495e-06,
+      "loss": 0.4888,
+      "step": 6040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.924881076503828,
+      "learning_rate": 5.457979081332409e-06,
+      "loss": 0.3847,
+      "step": 6041
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.582579547218558,
+      "learning_rate": 5.456679794120716e-06,
+      "loss": 0.4662,
+      "step": 6042
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.658263928766095,
+      "learning_rate": 5.455380475811149e-06,
+      "loss": 0.4954,
+      "step": 6043
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.054245154700187,
+      "learning_rate": 5.454081126492187e-06,
+      "loss": 0.4413,
+      "step": 6044
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.587629191541138,
+      "learning_rate": 5.452781746252305e-06,
+      "loss": 0.606,
+      "step": 6045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.669519469814558,
+      "learning_rate": 5.4514823351799905e-06,
+      "loss": 0.5345,
+      "step": 6046
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.958402539068238,
+      "learning_rate": 5.450182893363723e-06,
+      "loss": 0.5242,
+      "step": 6047
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.3650733734145115,
+      "learning_rate": 5.448883420891993e-06,
+      "loss": 0.4947,
+      "step": 6048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.64618743812488,
+      "learning_rate": 5.447583917853285e-06,
+      "loss": 0.4208,
+      "step": 6049
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.689502459018412,
+      "learning_rate": 5.4462843843360925e-06,
+      "loss": 0.4876,
+      "step": 6050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.796441654528122,
+      "learning_rate": 5.444984820428906e-06,
+      "loss": 0.4872,
+      "step": 6051
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.079102950772386,
+      "learning_rate": 5.44368522622022e-06,
+      "loss": 0.515,
+      "step": 6052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.002524840930798,
+      "learning_rate": 5.442385601798533e-06,
+      "loss": 0.5096,
+      "step": 6053
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.910745641749775,
+      "learning_rate": 5.4410859472523415e-06,
+      "loss": 0.5016,
+      "step": 6054
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.727936221283294,
+      "learning_rate": 5.439786262670149e-06,
+      "loss": 0.4483,
+      "step": 6055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.181704398882256,
+      "learning_rate": 5.438486548140455e-06,
+      "loss": 0.5495,
+      "step": 6056
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.02349419689455,
+      "learning_rate": 5.437186803751766e-06,
+      "loss": 0.4614,
+      "step": 6057
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 20.036719608186143,
+      "learning_rate": 5.4358870295925895e-06,
+      "loss": 0.4796,
+      "step": 6058
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.843576593956252,
+      "learning_rate": 5.434587225751432e-06,
+      "loss": 0.3752,
+      "step": 6059
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.028338245051307,
+      "learning_rate": 5.433287392316807e-06,
+      "loss": 0.4916,
+      "step": 6060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.667687823612816,
+      "learning_rate": 5.431987529377225e-06,
+      "loss": 0.4197,
+      "step": 6061
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.32239644928921,
+      "learning_rate": 5.430687637021205e-06,
+      "loss": 0.4402,
+      "step": 6062
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.362064430122825,
+      "learning_rate": 5.4293877153372585e-06,
+      "loss": 0.4374,
+      "step": 6063
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.378400523028853,
+      "learning_rate": 5.428087764413909e-06,
+      "loss": 0.4871,
+      "step": 6064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 16.214150814519943,
+      "learning_rate": 5.426787784339675e-06,
+      "loss": 0.499,
+      "step": 6065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.764455344729786,
+      "learning_rate": 5.425487775203079e-06,
+      "loss": 0.4358,
+      "step": 6066
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.711297992614139,
+      "learning_rate": 5.424187737092648e-06,
+      "loss": 0.4607,
+      "step": 6067
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.35980904108652,
+      "learning_rate": 5.422887670096907e-06,
+      "loss": 0.4309,
+      "step": 6068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.14655667069176,
+      "learning_rate": 5.4215875743043855e-06,
+      "loss": 0.5739,
+      "step": 6069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 15.506752270543602,
+      "learning_rate": 5.420287449803614e-06,
+      "loss": 0.3969,
+      "step": 6070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.131779568049827,
+      "learning_rate": 5.418987296683125e-06,
+      "loss": 0.464,
+      "step": 6071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.335738330993184,
+      "learning_rate": 5.4176871150314535e-06,
+      "loss": 0.5145,
+      "step": 6072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.628335423862127,
+      "learning_rate": 5.416386904937139e-06,
+      "loss": 0.5202,
+      "step": 6073
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.47832445639394,
+      "learning_rate": 5.415086666488715e-06,
+      "loss": 0.5267,
+      "step": 6074
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.64278855152265,
+      "learning_rate": 5.413786399774724e-06,
+      "loss": 0.4658,
+      "step": 6075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.341968189412915,
+      "learning_rate": 5.412486104883709e-06,
+      "loss": 0.4582,
+      "step": 6076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 22.7193768756787,
+      "learning_rate": 5.411185781904214e-06,
+      "loss": 0.4019,
+      "step": 6077
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.459822948566032,
+      "learning_rate": 5.4098854309247876e-06,
+      "loss": 0.5864,
+      "step": 6078
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.46845295587613,
+      "learning_rate": 5.408585052033973e-06,
+      "loss": 0.5085,
+      "step": 6079
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.07127538276098,
+      "learning_rate": 5.407284645320323e-06,
+      "loss": 0.5142,
+      "step": 6080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 21.39901771333456,
+      "learning_rate": 5.40598421087239e-06,
+      "loss": 0.5946,
+      "step": 6081
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.912455940942982,
+      "learning_rate": 5.404683748778728e-06,
+      "loss": 0.5088,
+      "step": 6082
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.910007245005107,
+      "learning_rate": 5.403383259127891e-06,
+      "loss": 0.5023,
+      "step": 6083
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.062652473167706,
+      "learning_rate": 5.402082742008438e-06,
+      "loss": 0.5083,
+      "step": 6084
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8091866241752932,
+      "learning_rate": 5.4007821975089285e-06,
+      "loss": 0.7399,
+      "step": 6085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.123826249096709,
+      "learning_rate": 5.3994816257179226e-06,
+      "loss": 0.6519,
+      "step": 6086
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.85761538448238,
+      "learning_rate": 5.398181026723985e-06,
+      "loss": 0.5573,
+      "step": 6087
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.608614182654936,
+      "learning_rate": 5.3968804006156794e-06,
+      "loss": 0.4208,
+      "step": 6088
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.027758944147394,
+      "learning_rate": 5.395579747481573e-06,
+      "loss": 0.484,
+      "step": 6089
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.41210397608173,
+      "learning_rate": 5.394279067410236e-06,
+      "loss": 0.5278,
+      "step": 6090
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.040099751753747,
+      "learning_rate": 5.3929783604902366e-06,
+      "loss": 0.5313,
+      "step": 6091
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.403787865992028,
+      "learning_rate": 5.391677626810149e-06,
+      "loss": 0.482,
+      "step": 6092
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.410410638284416,
+      "learning_rate": 5.390376866458546e-06,
+      "loss": 0.4963,
+      "step": 6093
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.30715019245391,
+      "learning_rate": 5.389076079524006e-06,
+      "loss": 0.4307,
+      "step": 6094
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.01951559924173,
+      "learning_rate": 5.387775266095104e-06,
+      "loss": 0.4702,
+      "step": 6095
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.382246084527373,
+      "learning_rate": 5.386474426260423e-06,
+      "loss": 0.5087,
+      "step": 6096
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.894686574743377,
+      "learning_rate": 5.38517356010854e-06,
+      "loss": 0.5248,
+      "step": 6097
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.57214131773344,
+      "learning_rate": 5.38387266772804e-06,
+      "loss": 0.482,
+      "step": 6098
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.835188749919586,
+      "learning_rate": 5.382571749207511e-06,
+      "loss": 0.4272,
+      "step": 6099
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 19.052846771184473,
+      "learning_rate": 5.3812708046355365e-06,
+      "loss": 0.3996,
+      "step": 6100
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.670006282657415,
+      "learning_rate": 5.379969834100705e-06,
+      "loss": 0.5875,
+      "step": 6101
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.074759680532347,
+      "learning_rate": 5.3786688376916085e-06,
+      "loss": 0.4793,
+      "step": 6102
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.9584805171826005,
+      "learning_rate": 5.377367815496839e-06,
+      "loss": 0.4786,
+      "step": 6103
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.718609374163044,
+      "learning_rate": 5.376066767604989e-06,
+      "loss": 0.3943,
+      "step": 6104
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.23364654605861,
+      "learning_rate": 5.374765694104655e-06,
+      "loss": 0.4204,
+      "step": 6105
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.52252551056972,
+      "learning_rate": 5.3734645950844355e-06,
+      "loss": 0.4641,
+      "step": 6106
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.764178775977086,
+      "learning_rate": 5.3721634706329266e-06,
+      "loss": 0.3841,
+      "step": 6107
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.703679456602755,
+      "learning_rate": 5.370862320838732e-06,
+      "loss": 0.5742,
+      "step": 6108
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.0128453588116715,
+      "learning_rate": 5.369561145790453e-06,
+      "loss": 0.4488,
+      "step": 6109
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.512301543892976,
+      "learning_rate": 5.368259945576694e-06,
+      "loss": 0.4238,
+      "step": 6110
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.819155940588775,
+      "learning_rate": 5.366958720286058e-06,
+      "loss": 0.4677,
+      "step": 6111
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.879011419336908,
+      "learning_rate": 5.365657470007159e-06,
+      "loss": 0.5051,
+      "step": 6112
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.0614170586061835,
+      "learning_rate": 5.364356194828602e-06,
+      "loss": 0.4336,
+      "step": 6113
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9806048370367633,
+      "learning_rate": 5.363054894838999e-06,
+      "loss": 0.7894,
+      "step": 6114
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 12.041756157140622,
+      "learning_rate": 5.361753570126964e-06,
+      "loss": 0.5724,
+      "step": 6115
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.283326026085547,
+      "learning_rate": 5.36045222078111e-06,
+      "loss": 0.5272,
+      "step": 6116
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.496580646127994,
+      "learning_rate": 5.359150846890053e-06,
+      "loss": 0.5245,
+      "step": 6117
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.203842565778624,
+      "learning_rate": 5.3578494485424115e-06,
+      "loss": 0.4877,
+      "step": 6118
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.952494225118746,
+      "learning_rate": 5.356548025826805e-06,
+      "loss": 0.4295,
+      "step": 6119
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.315048766243329,
+      "learning_rate": 5.355246578831854e-06,
+      "loss": 0.5476,
+      "step": 6120
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.544511683774305,
+      "learning_rate": 5.353945107646181e-06,
+      "loss": 0.4142,
+      "step": 6121
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.828805199053949,
+      "learning_rate": 5.3526436123584114e-06,
+      "loss": 0.4385,
+      "step": 6122
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.955060569552572,
+      "learning_rate": 5.351342093057171e-06,
+      "loss": 0.5172,
+      "step": 6123
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.531740212865179,
+      "learning_rate": 5.350040549831089e-06,
+      "loss": 0.5384,
+      "step": 6124
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.264831072285281,
+      "learning_rate": 5.348738982768792e-06,
+      "loss": 0.4964,
+      "step": 6125
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 22.531931042195183,
+      "learning_rate": 5.347437391958913e-06,
+      "loss": 0.4012,
+      "step": 6126
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.88031361714173,
+      "learning_rate": 5.346135777490082e-06,
+      "loss": 0.475,
+      "step": 6127
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.046997403721633,
+      "learning_rate": 5.344834139450936e-06,
+      "loss": 0.6095,
+      "step": 6128
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.419444645273991,
+      "learning_rate": 5.343532477930111e-06,
+      "loss": 0.5314,
+      "step": 6129
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6338491712268906,
+      "learning_rate": 5.342230793016242e-06,
+      "loss": 0.7866,
+      "step": 6130
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 13.172420530426642,
+      "learning_rate": 5.340929084797969e-06,
+      "loss": 0.4624,
+      "step": 6131
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 14.54874122848959,
+      "learning_rate": 5.339627353363932e-06,
+      "loss": 0.4811,
+      "step": 6132
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.204657627062426,
+      "learning_rate": 5.338325598802776e-06,
+      "loss": 0.4187,
+      "step": 6133
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.410793910016524,
+      "learning_rate": 5.337023821203141e-06,
+      "loss": 0.55,
+      "step": 6134
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.243323540117103,
+      "learning_rate": 5.335722020653675e-06,
+      "loss": 0.4865,
+      "step": 6135
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.22246526992235,
+      "learning_rate": 5.3344201972430235e-06,
+      "loss": 0.5027,
+      "step": 6136
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.043878514696182,
+      "learning_rate": 5.3331183510598336e-06,
+      "loss": 0.4338,
+      "step": 6137
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.874206901621806,
+      "learning_rate": 5.33181648219276e-06,
+      "loss": 0.5183,
+      "step": 6138
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.298905751371334,
+      "learning_rate": 5.330514590730448e-06,
+      "loss": 0.5279,
+      "step": 6139
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6671388736740622,
+      "learning_rate": 5.329212676761556e-06,
+      "loss": 0.7374,
+      "step": 6140
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7792878738566396,
+      "learning_rate": 5.327910740374733e-06,
+      "loss": 0.7771,
+      "step": 6141
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.080579615546013,
+      "learning_rate": 5.326608781658642e-06,
+      "loss": 0.5547,
+      "step": 6142
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.091362899541675,
+      "learning_rate": 5.325306800701935e-06,
+      "loss": 0.4753,
+      "step": 6143
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 17.843652362916448,
+      "learning_rate": 5.324004797593275e-06,
+      "loss": 0.5125,
+      "step": 6144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.286405319087047,
+      "learning_rate": 5.3227027724213195e-06,
+      "loss": 0.3652,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.292818491016812,
+      "learning_rate": 5.321400725274732e-06,
+      "loss": 0.5228,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.874377148533195,
+      "learning_rate": 5.320098656242177e-06,
+      "loss": 0.4759,
+      "step": 6147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.036680399326002,
+      "learning_rate": 5.318796565412318e-06,
+      "loss": 0.4267,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0030396654074947,
+      "learning_rate": 5.317494452873824e-06,
+      "loss": 0.7283,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.32698582127824,
+      "learning_rate": 5.3161923187153595e-06,
+      "loss": 0.4019,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6303054713358642,
+      "learning_rate": 5.314890163025597e-06,
+      "loss": 0.7033,
+      "step": 6151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 32.384419391783425,
+      "learning_rate": 5.313587985893206e-06,
+      "loss": 0.4355,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.565460721767646,
+      "learning_rate": 5.312285787406859e-06,
+      "loss": 0.521,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.02639168502577,
+      "learning_rate": 5.310983567655233e-06,
+      "loss": 0.5242,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.401166893764872,
+      "learning_rate": 5.3096813267269995e-06,
+      "loss": 0.4612,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 138.00696210337048,
+      "learning_rate": 5.308379064710838e-06,
+      "loss": 0.4609,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0560413556157493,
+      "learning_rate": 5.307076781695425e-06,
+      "loss": 0.825,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.385548973137922,
+      "learning_rate": 5.305774477769441e-06,
+      "loss": 0.46,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.713578778519878,
+      "learning_rate": 5.304472153021567e-06,
+      "loss": 0.5004,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.790214930182131,
+      "learning_rate": 5.303169807540485e-06,
+      "loss": 0.4775,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.031797902254166,
+      "learning_rate": 5.301867441414879e-06,
+      "loss": 0.4236,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.387176425514738,
+      "learning_rate": 5.300565054733437e-06,
+      "loss": 0.4757,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.928742760859839,
+      "learning_rate": 5.299262647584843e-06,
+      "loss": 0.4068,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.536234659205391,
+      "learning_rate": 5.2979602200577855e-06,
+      "loss": 0.5285,
+      "step": 6164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.751338748227795,
+      "learning_rate": 5.296657772240955e-06,
+      "loss": 0.5365,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.626152758396223,
+      "learning_rate": 5.295355304223041e-06,
+      "loss": 0.5189,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.281009481429558,
+      "learning_rate": 5.294052816092737e-06,
+      "loss": 0.5418,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.760311230741653,
+      "learning_rate": 5.292750307938737e-06,
+      "loss": 0.4552,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.602366784408607,
+      "learning_rate": 5.291447779849735e-06,
+      "loss": 0.4745,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 38.5861831931482,
+      "learning_rate": 5.290145231914428e-06,
+      "loss": 0.4853,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.961256940883229,
+      "learning_rate": 5.288842664221511e-06,
+      "loss": 0.5585,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.327871710304526,
+      "learning_rate": 5.287540076859688e-06,
+      "loss": 0.5772,
+      "step": 6172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.4563032272552,
+      "learning_rate": 5.2862374699176575e-06,
+      "loss": 0.5202,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.8391686898798545,
+      "learning_rate": 5.284934843484121e-06,
+      "loss": 0.5076,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.551020257116127,
+      "learning_rate": 5.28363219764778e-06,
+      "loss": 0.4069,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.664135335922635,
+      "learning_rate": 5.2823295324973415e-06,
+      "loss": 0.3983,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7866092461961924,
+      "learning_rate": 5.281026848121511e-06,
+      "loss": 0.8246,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.600632348978905,
+      "learning_rate": 5.279724144608993e-06,
+      "loss": 0.5239,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8507220063267862,
+      "learning_rate": 5.278421422048499e-06,
+      "loss": 0.8075,
+      "step": 6179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.05734186324629,
+      "learning_rate": 5.277118680528736e-06,
+      "loss": 0.4632,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5829003404477404,
+      "learning_rate": 5.275815920138416e-06,
+      "loss": 0.8062,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 19.341547207504615,
+      "learning_rate": 5.274513140966251e-06,
+      "loss": 0.6805,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 18.42778884607876,
+      "learning_rate": 5.273210343100956e-06,
+      "loss": 0.6123,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.137540377627968,
+      "learning_rate": 5.271907526631244e-06,
+      "loss": 0.3977,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.705975013986501,
+      "learning_rate": 5.270604691645832e-06,
+      "loss": 0.4626,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.140070756039657,
+      "learning_rate": 5.269301838233438e-06,
+      "loss": 0.5375,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.616999025753575,
+      "learning_rate": 5.267998966482778e-06,
+      "loss": 0.361,
+      "step": 6187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.5420638039004,
+      "learning_rate": 5.266696076482575e-06,
+      "loss": 0.536,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.5201539941970585,
+      "learning_rate": 5.265393168321547e-06,
+      "loss": 0.5288,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.317853568594225,
+      "learning_rate": 5.264090242088418e-06,
+      "loss": 0.477,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.428300411909168,
+      "learning_rate": 5.262787297871911e-06,
+      "loss": 0.585,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.138829945513603,
+      "learning_rate": 5.261484335760751e-06,
+      "loss": 0.5475,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.145764396241544,
+      "learning_rate": 5.260181355843662e-06,
+      "loss": 0.4062,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.643950042470486,
+      "learning_rate": 5.258878358209375e-06,
+      "loss": 0.549,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.930049266852203,
+      "learning_rate": 5.257575342946617e-06,
+      "loss": 0.4679,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.968602007253523,
+      "learning_rate": 5.2562723101441156e-06,
+      "loss": 0.4876,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.96357256271058,
+      "learning_rate": 5.254969259890604e-06,
+      "loss": 0.5285,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.008025261758972,
+      "learning_rate": 5.2536661922748134e-06,
+      "loss": 0.4227,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.808621792536654,
+      "learning_rate": 5.252363107385475e-06,
+      "loss": 0.4272,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.174980943674894,
+      "learning_rate": 5.251060005311327e-06,
+      "loss": 0.5129,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.700167958915337,
+      "learning_rate": 5.249756886141101e-06,
+      "loss": 0.5134,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.858298896821795,
+      "learning_rate": 5.248453749963536e-06,
+      "loss": 0.4624,
+      "step": 6202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.107203882161444,
+      "learning_rate": 5.247150596867369e-06,
+      "loss": 0.5969,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.167205621327234,
+      "learning_rate": 5.24584742694134e-06,
+      "loss": 0.5999,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.569019480774212,
+      "learning_rate": 5.244544240274186e-06,
+      "loss": 0.4002,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.132551919524947,
+      "learning_rate": 5.243241036954653e-06,
+      "loss": 0.4605,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.213650447332196,
+      "learning_rate": 5.241937817071479e-06,
+      "loss": 0.5103,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.814513110959737,
+      "learning_rate": 5.240634580713409e-06,
+      "loss": 0.5333,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.537195342856704,
+      "learning_rate": 5.23933132796919e-06,
+      "loss": 0.3838,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.99526805688073,
+      "learning_rate": 5.238028058927564e-06,
+      "loss": 0.5535,
+      "step": 6210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.776048330943699,
+      "learning_rate": 5.236724773677282e-06,
+      "loss": 0.5791,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.224478769756219,
+      "learning_rate": 5.235421472307087e-06,
+      "loss": 0.6255,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 20.236792552827602,
+      "learning_rate": 5.2341181549057315e-06,
+      "loss": 0.4249,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.195555449106562,
+      "learning_rate": 5.232814821561965e-06,
+      "loss": 0.4904,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.078645313131968,
+      "learning_rate": 5.23151147236454e-06,
+      "loss": 0.4329,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.992529772354322,
+      "learning_rate": 5.2302081074022045e-06,
+      "loss": 0.4582,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.881874185837088,
+      "learning_rate": 5.228904726763716e-06,
+      "loss": 0.4567,
+      "step": 6217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.357350004751801,
+      "learning_rate": 5.22760133053783e-06,
+      "loss": 0.5446,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.513086977545465,
+      "learning_rate": 5.226297918813297e-06,
+      "loss": 0.487,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.022845119537772,
+      "learning_rate": 5.224994491678876e-06,
+      "loss": 0.5102,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.4191712380481,
+      "learning_rate": 5.223691049223326e-06,
+      "loss": 0.5099,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2439896666585106,
+      "learning_rate": 5.222387591535404e-06,
+      "loss": 0.7781,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.960401188596487,
+      "learning_rate": 5.221084118703871e-06,
+      "loss": 0.4927,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 35.79383636625562,
+      "learning_rate": 5.219780630817486e-06,
+      "loss": 0.5422,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.438930305757255,
+      "learning_rate": 5.218477127965013e-06,
+      "loss": 0.5077,
+      "step": 6225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.825717653516156,
+      "learning_rate": 5.217173610235212e-06,
+      "loss": 0.4818,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.420471168441713,
+      "learning_rate": 5.215870077716852e-06,
+      "loss": 0.5277,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.288391913413008,
+      "learning_rate": 5.21456653049869e-06,
+      "loss": 0.5372,
+      "step": 6228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.065660348038415,
+      "learning_rate": 5.213262968669498e-06,
+      "loss": 0.5846,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.909975898217256,
+      "learning_rate": 5.21195939231804e-06,
+      "loss": 0.5348,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.456675431110831,
+      "learning_rate": 5.2106558015330844e-06,
+      "loss": 0.4645,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.652670141110011,
+      "learning_rate": 5.209352196403401e-06,
+      "loss": 0.5018,
+      "step": 6232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.532115999014819,
+      "learning_rate": 5.208048577017758e-06,
+      "loss": 0.518,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.750701515130304,
+      "learning_rate": 5.206744943464928e-06,
+      "loss": 0.5141,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.416313940811423,
+      "learning_rate": 5.20544129583368e-06,
+      "loss": 0.4535,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.49016958011095,
+      "learning_rate": 5.204137634212789e-06,
+      "loss": 0.4011,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 17.99613981218858,
+      "learning_rate": 5.202833958691027e-06,
+      "loss": 0.464,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.28032798844463,
+      "learning_rate": 5.20153026935717e-06,
+      "loss": 0.5039,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.385829914738618,
+      "learning_rate": 5.200226566299992e-06,
+      "loss": 0.3888,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.128405046540003,
+      "learning_rate": 5.19892284960827e-06,
+      "loss": 0.6674,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.349427433646865,
+      "learning_rate": 5.197619119370782e-06,
+      "loss": 0.6063,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.908919223679301,
+      "learning_rate": 5.196315375676304e-06,
+      "loss": 0.4793,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.895902643359822,
+      "learning_rate": 5.19501161861362e-06,
+      "loss": 0.5484,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.087377496907699,
+      "learning_rate": 5.193707848271505e-06,
+      "loss": 0.514,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.745662425183123,
+      "learning_rate": 5.192404064738743e-06,
+      "loss": 0.5236,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 40.160756428864595,
+      "learning_rate": 5.191100268104115e-06,
+      "loss": 0.4528,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.369403639866697,
+      "learning_rate": 5.189796458456402e-06,
+      "loss": 0.5383,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.484577722663195,
+      "learning_rate": 5.188492635884391e-06,
+      "loss": 0.3899,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 11.193611720170106,
+      "learning_rate": 5.187188800476864e-06,
+      "loss": 0.4723,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.85061223709142,
+      "learning_rate": 5.185884952322609e-06,
+      "loss": 0.5886,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 14.109035492027088,
+      "learning_rate": 5.184581091510409e-06,
+      "loss": 0.5417,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.687000342213382,
+      "learning_rate": 5.183277218129052e-06,
+      "loss": 0.5161,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.056456361848365,
+      "learning_rate": 5.181973332267329e-06,
+      "loss": 0.471,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.948793425654184,
+      "learning_rate": 5.180669434014025e-06,
+      "loss": 0.536,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.564527932210989,
+      "learning_rate": 5.179365523457932e-06,
+      "loss": 0.4474,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.76339080395018,
+      "learning_rate": 5.178061600687839e-06,
+      "loss": 0.3966,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 19.29697079750918,
+      "learning_rate": 5.17675766579254e-06,
+      "loss": 0.3824,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.804996428592279,
+      "learning_rate": 5.175453718860824e-06,
+      "loss": 0.4788,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.412028898287723,
+      "learning_rate": 5.1741497599814875e-06,
+      "loss": 0.4403,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.97387832774475,
+      "learning_rate": 5.17284578924332e-06,
+      "loss": 0.4758,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.014554948133066,
+      "learning_rate": 5.171541806735119e-06,
+      "loss": 0.5041,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.0054796701881115,
+      "learning_rate": 5.170237812545679e-06,
+      "loss": 0.4597,
+      "step": 6262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.409756758099164,
+      "learning_rate": 5.168933806763798e-06,
+      "loss": 0.4197,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.82290287751571,
+      "learning_rate": 5.167629789478272e-06,
+      "loss": 0.5022,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.928399663872685,
+      "learning_rate": 5.1663257607778975e-06,
+      "loss": 0.5361,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.742481730123526,
+      "learning_rate": 5.1650217207514755e-06,
+      "loss": 0.4344,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.395614098913828,
+      "learning_rate": 5.163717669487802e-06,
+      "loss": 0.4001,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.479124110196597,
+      "learning_rate": 5.1624136070756805e-06,
+      "loss": 0.532,
+      "step": 6268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.037667109143037,
+      "learning_rate": 5.161109533603909e-06,
+      "loss": 0.4478,
+      "step": 6269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.427161274856547,
+      "learning_rate": 5.159805449161291e-06,
+      "loss": 0.4316,
+      "step": 6270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.52141556635946,
+      "learning_rate": 5.1585013538366305e-06,
+      "loss": 0.4623,
+      "step": 6271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.902276601697029,
+      "learning_rate": 5.157197247718725e-06,
+      "loss": 0.3832,
+      "step": 6272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.910476200284617,
+      "learning_rate": 5.155893130896385e-06,
+      "loss": 0.5401,
+      "step": 6273
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.06687675483873,
+      "learning_rate": 5.15458900345841e-06,
+      "loss": 0.562,
+      "step": 6274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.946815632618959,
+      "learning_rate": 5.153284865493609e-06,
+      "loss": 0.5343,
+      "step": 6275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.006627161262419,
+      "learning_rate": 5.151980717090785e-06,
+      "loss": 0.4417,
+      "step": 6276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.041054064182978,
+      "learning_rate": 5.150676558338748e-06,
+      "loss": 0.4908,
+      "step": 6277
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 19.310150113237007,
+      "learning_rate": 5.149372389326302e-06,
+      "loss": 0.4898,
+      "step": 6278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.6522035424481105,
+      "learning_rate": 5.148068210142257e-06,
+      "loss": 0.4299,
+      "step": 6279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.980213246698888,
+      "learning_rate": 5.146764020875423e-06,
+      "loss": 0.3918,
+      "step": 6280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7850232512767934,
+      "learning_rate": 5.145459821614607e-06,
+      "loss": 0.7283,
+      "step": 6281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.96599981372958,
+      "learning_rate": 5.14415561244862e-06,
+      "loss": 0.4295,
+      "step": 6282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.95779104750984,
+      "learning_rate": 5.142851393466273e-06,
+      "loss": 0.4867,
+      "step": 6283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.120892462572078,
+      "learning_rate": 5.141547164756379e-06,
+      "loss": 0.5517,
+      "step": 6284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.531369798383555,
+      "learning_rate": 5.140242926407747e-06,
+      "loss": 0.5137,
+      "step": 6285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.42017515895616,
+      "learning_rate": 5.138938678509195e-06,
+      "loss": 0.4361,
+      "step": 6286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.941268294316506,
+      "learning_rate": 5.13763442114953e-06,
+      "loss": 0.5775,
+      "step": 6287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.035113992758546,
+      "learning_rate": 5.1363301544175705e-06,
+      "loss": 0.4501,
+      "step": 6288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.7818922882974,
+      "learning_rate": 5.13502587840213e-06,
+      "loss": 0.4104,
+      "step": 6289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.856080374344831,
+      "learning_rate": 5.1337215931920244e-06,
+      "loss": 0.3743,
+      "step": 6290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.448971911695445,
+      "learning_rate": 5.132417298876071e-06,
+      "loss": 0.4706,
+      "step": 6291
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.098824926428984,
+      "learning_rate": 5.131112995543082e-06,
+      "loss": 0.5249,
+      "step": 6292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.292094215209731,
+      "learning_rate": 5.1298086832818805e-06,
+      "loss": 0.4764,
+      "step": 6293
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.528149999418225,
+      "learning_rate": 5.12850436218128e-06,
+      "loss": 0.5717,
+      "step": 6294
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.672049060153094,
+      "learning_rate": 5.1272000323301005e-06,
+      "loss": 0.4815,
+      "step": 6295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.957770170641762,
+      "learning_rate": 5.125895693817162e-06,
+      "loss": 0.4491,
+      "step": 6296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.101739576980169,
+      "learning_rate": 5.124591346731284e-06,
+      "loss": 0.4635,
+      "step": 6297
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.69514776354777,
+      "learning_rate": 5.123286991161286e-06,
+      "loss": 0.4952,
+      "step": 6298
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.468719158289995,
+      "learning_rate": 5.121982627195988e-06,
+      "loss": 0.4456,
+      "step": 6299
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.341504380254325,
+      "learning_rate": 5.120678254924213e-06,
+      "loss": 0.4351,
+      "step": 6300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.021690060434498,
+      "learning_rate": 5.119373874434782e-06,
+      "loss": 0.4332,
+      "step": 6301
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.282762717128763,
+      "learning_rate": 5.118069485816518e-06,
+      "loss": 0.5066,
+      "step": 6302
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 13.503693686423675,
+      "learning_rate": 5.116765089158244e-06,
+      "loss": 0.4508,
+      "step": 6303
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.842207232619035,
+      "learning_rate": 5.115460684548784e-06,
+      "loss": 0.5192,
+      "step": 6304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.398097886171469,
+      "learning_rate": 5.114156272076962e-06,
+      "loss": 0.5676,
+      "step": 6305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.429684526274368,
+      "learning_rate": 5.112851851831603e-06,
+      "loss": 0.4352,
+      "step": 6306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.593859885255535,
+      "learning_rate": 5.111547423901532e-06,
+      "loss": 0.4815,
+      "step": 6307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.674148378565628,
+      "learning_rate": 5.110242988375573e-06,
+      "loss": 0.4694,
+      "step": 6308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.548203067597923,
+      "learning_rate": 5.108938545342555e-06,
+      "loss": 0.4916,
+      "step": 6309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.07132074391085,
+      "learning_rate": 5.107634094891304e-06,
+      "loss": 0.41,
+      "step": 6310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.079254917664878,
+      "learning_rate": 5.106329637110647e-06,
+      "loss": 0.4527,
+      "step": 6311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.618710420512345,
+      "learning_rate": 5.105025172089411e-06,
+      "loss": 0.3992,
+      "step": 6312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 24.78069470690949,
+      "learning_rate": 5.103720699916425e-06,
+      "loss": 0.4288,
+      "step": 6313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.643504419816241,
+      "learning_rate": 5.102416220680519e-06,
+      "loss": 0.5467,
+      "step": 6314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.989336720310388,
+      "learning_rate": 5.101111734470521e-06,
+      "loss": 0.5124,
+      "step": 6315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 25.705614650873315,
+      "learning_rate": 5.099807241375261e-06,
+      "loss": 0.4598,
+      "step": 6316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.17888369355004,
+      "learning_rate": 5.098502741483569e-06,
+      "loss": 0.5382,
+      "step": 6317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.731759392608176,
+      "learning_rate": 5.0971982348842755e-06,
+      "loss": 0.5583,
+      "step": 6318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.469878714049249,
+      "learning_rate": 5.095893721666212e-06,
+      "loss": 0.5011,
+      "step": 6319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.773205544097657,
+      "learning_rate": 5.0945892019182096e-06,
+      "loss": 0.4962,
+      "step": 6320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.611237226394228,
+      "learning_rate": 5.0932846757291035e-06,
+      "loss": 0.4897,
+      "step": 6321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.063150979001676,
+      "learning_rate": 5.091980143187719e-06,
+      "loss": 0.4332,
+      "step": 6322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.123598171889999,
+      "learning_rate": 5.090675604382896e-06,
+      "loss": 0.5453,
+      "step": 6323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.3304270569267524,
+      "learning_rate": 5.089371059403465e-06,
+      "loss": 0.539,
+      "step": 6324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.98086598219569,
+      "learning_rate": 5.088066508338261e-06,
+      "loss": 0.4981,
+      "step": 6325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.55591192047073,
+      "learning_rate": 5.086761951276116e-06,
+      "loss": 0.4704,
+      "step": 6326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.292856534006251,
+      "learning_rate": 5.085457388305865e-06,
+      "loss": 0.4777,
+      "step": 6327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.234193435759577,
+      "learning_rate": 5.084152819516346e-06,
+      "loss": 0.4647,
+      "step": 6328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.724856142162293,
+      "learning_rate": 5.08284824499639e-06,
+      "loss": 0.4146,
+      "step": 6329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.871849693175367,
+      "learning_rate": 5.081543664834836e-06,
+      "loss": 0.4898,
+      "step": 6330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.534982139201281,
+      "learning_rate": 5.080239079120519e-06,
+      "loss": 0.3432,
+      "step": 6331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 21.55993438315232,
+      "learning_rate": 5.078934487942274e-06,
+      "loss": 0.4793,
+      "step": 6332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.51334592063618,
+      "learning_rate": 5.07762989138894e-06,
+      "loss": 0.569,
+      "step": 6333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.094731707374617,
+      "learning_rate": 5.076325289549356e-06,
+      "loss": 0.5888,
+      "step": 6334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.275632883356112,
+      "learning_rate": 5.075020682512354e-06,
+      "loss": 0.3963,
+      "step": 6335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.388805672074197,
+      "learning_rate": 5.0737160703667766e-06,
+      "loss": 0.5655,
+      "step": 6336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.725808473292457,
+      "learning_rate": 5.0724114532014616e-06,
+      "loss": 0.556,
+      "step": 6337
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.512826797752188,
+      "learning_rate": 5.071106831105245e-06,
+      "loss": 0.5174,
+      "step": 6338
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.281119214089518,
+      "learning_rate": 5.069802204166969e-06,
+      "loss": 0.5215,
+      "step": 6339
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.564312527234936,
+      "learning_rate": 5.06849757247547e-06,
+      "loss": 0.4841,
+      "step": 6340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.087443527386473,
+      "learning_rate": 5.06719293611959e-06,
+      "loss": 0.5678,
+      "step": 6341
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5142504329008877,
+      "learning_rate": 5.06588829518817e-06,
+      "loss": 0.8004,
+      "step": 6342
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.42609432765421,
+      "learning_rate": 5.064583649770044e-06,
+      "loss": 0.5647,
+      "step": 6343
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.781126075788679,
+      "learning_rate": 5.063278999954061e-06,
+      "loss": 0.3721,
+      "step": 6344
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.23947096089713,
+      "learning_rate": 5.061974345829056e-06,
+      "loss": 0.5329,
+      "step": 6345
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.367234538528507,
+      "learning_rate": 5.060669687483874e-06,
+      "loss": 0.4271,
+      "step": 6346
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.68077602937966,
+      "learning_rate": 5.059365025007352e-06,
+      "loss": 0.4227,
+      "step": 6347
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.811988322961136,
+      "learning_rate": 5.058060358488336e-06,
+      "loss": 0.4349,
+      "step": 6348
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7011566408317167,
+      "learning_rate": 5.056755688015666e-06,
+      "loss": 0.758,
+      "step": 6349
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.211645273768989,
+      "learning_rate": 5.055451013678183e-06,
+      "loss": 0.5889,
+      "step": 6350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.74265098717469,
+      "learning_rate": 5.0541463355647335e-06,
+      "loss": 0.5124,
+      "step": 6351
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.285222631486386,
+      "learning_rate": 5.052841653764157e-06,
+      "loss": 0.4109,
+      "step": 6352
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.346143200922829,
+      "learning_rate": 5.0515369683652975e-06,
+      "loss": 0.4942,
+      "step": 6353
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.701682842887056,
+      "learning_rate": 5.0502322794569966e-06,
+      "loss": 0.5233,
+      "step": 6354
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.172254700363093,
+      "learning_rate": 5.048927587128103e-06,
+      "loss": 0.4703,
+      "step": 6355
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.258722885120436,
+      "learning_rate": 5.0476228914674545e-06,
+      "loss": 0.5446,
+      "step": 6356
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.463416879727657,
+      "learning_rate": 5.0463181925639e-06,
+      "loss": 0.5489,
+      "step": 6357
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6750919683831014,
+      "learning_rate": 5.045013490506279e-06,
+      "loss": 0.7341,
+      "step": 6358
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.999472692982902,
+      "learning_rate": 5.043708785383439e-06,
+      "loss": 0.5273,
+      "step": 6359
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.693862318447176,
+      "learning_rate": 5.042404077284224e-06,
+      "loss": 0.5886,
+      "step": 6360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.069250063313536,
+      "learning_rate": 5.041099366297478e-06,
+      "loss": 0.5312,
+      "step": 6361
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 21.089823608795015,
+      "learning_rate": 5.039794652512049e-06,
+      "loss": 0.4721,
+      "step": 6362
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.565945502415876,
+      "learning_rate": 5.038489936016777e-06,
+      "loss": 0.5331,
+      "step": 6363
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.466555731465542,
+      "learning_rate": 5.037185216900514e-06,
+      "loss": 0.4524,
+      "step": 6364
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.765861711644265,
+      "learning_rate": 5.0358804952521e-06,
+      "loss": 0.5154,
+      "step": 6365
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.368432591974561,
+      "learning_rate": 5.034575771160383e-06,
+      "loss": 0.4349,
+      "step": 6366
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.939124793980984,
+      "learning_rate": 5.033271044714211e-06,
+      "loss": 0.342,
+      "step": 6367
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.764860234186086,
+      "learning_rate": 5.0319663160024255e-06,
+      "loss": 0.6045,
+      "step": 6368
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.89086713622611,
+      "learning_rate": 5.030661585113877e-06,
+      "loss": 0.5444,
+      "step": 6369
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.037691220814333,
+      "learning_rate": 5.0293568521374085e-06,
+      "loss": 0.4914,
+      "step": 6370
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.261271348958507,
+      "learning_rate": 5.028052117161869e-06,
+      "loss": 0.6029,
+      "step": 6371
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.584831456133717,
+      "learning_rate": 5.026747380276105e-06,
+      "loss": 0.4162,
+      "step": 6372
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.970285103206409,
+      "learning_rate": 5.025442641568961e-06,
+      "loss": 0.3959,
+      "step": 6373
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 36.79897771471028,
+      "learning_rate": 5.024137901129286e-06,
+      "loss": 0.5725,
+      "step": 6374
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.069760051756333,
+      "learning_rate": 5.022833159045927e-06,
+      "loss": 0.4625,
+      "step": 6375
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.89472033058201,
+      "learning_rate": 5.0215284154077316e-06,
+      "loss": 0.5269,
+      "step": 6376
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8897755734410966,
+      "learning_rate": 5.020223670303545e-06,
+      "loss": 0.7555,
+      "step": 6377
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.865088280938792,
+      "learning_rate": 5.018918923822217e-06,
+      "loss": 0.4089,
+      "step": 6378
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.492014011726818,
+      "learning_rate": 5.017614176052593e-06,
+      "loss": 0.5662,
+      "step": 6379
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.859872892329099,
+      "learning_rate": 5.016309427083522e-06,
+      "loss": 0.3621,
+      "step": 6380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 15.44975722600289,
+      "learning_rate": 5.015004677003852e-06,
+      "loss": 0.5006,
+      "step": 6381
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.943297868606638,
+      "learning_rate": 5.013699925902428e-06,
+      "loss": 0.4019,
+      "step": 6382
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.77571245486872,
+      "learning_rate": 5.0123951738681e-06,
+      "loss": 0.5361,
+      "step": 6383
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6709409298024742,
+      "learning_rate": 5.011090420989716e-06,
+      "loss": 0.7662,
+      "step": 6384
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.202778830857032,
+      "learning_rate": 5.009785667356124e-06,
+      "loss": 0.4687,
+      "step": 6385
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.84929823097296,
+      "learning_rate": 5.008480913056171e-06,
+      "loss": 0.6081,
+      "step": 6386
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.3829924998222065,
+      "learning_rate": 5.007176158178705e-06,
+      "loss": 0.4781,
+      "step": 6387
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.938911703259881,
+      "learning_rate": 5.005871402812574e-06,
+      "loss": 0.4777,
+      "step": 6388
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.985807699760448,
+      "learning_rate": 5.004566647046627e-06,
+      "loss": 0.4533,
+      "step": 6389
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 52.88959897890417,
+      "learning_rate": 5.003261890969712e-06,
+      "loss": 0.5048,
+      "step": 6390
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.33035352265465,
+      "learning_rate": 5.001957134670675e-06,
+      "loss": 0.5147,
+      "step": 6391
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.086138176570531,
+      "learning_rate": 5.000652378238368e-06,
+      "loss": 0.3985,
+      "step": 6392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.300621317629728,
+      "learning_rate": 4.999347621761634e-06,
+      "loss": 0.4984,
+      "step": 6393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1245945516707643,
+      "learning_rate": 4.9980428653293275e-06,
+      "loss": 0.83,
+      "step": 6394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.489058549829156,
+      "learning_rate": 4.99673810903029e-06,
+      "loss": 0.5561,
+      "step": 6395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.0523486403435545,
+      "learning_rate": 4.995433352953375e-06,
+      "loss": 0.373,
+      "step": 6396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.237167495140476,
+      "learning_rate": 4.994128597187428e-06,
+      "loss": 0.4513,
+      "step": 6397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.230868865435315,
+      "learning_rate": 4.992823841821296e-06,
+      "loss": 0.4116,
+      "step": 6398
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.318202781429468,
+      "learning_rate": 4.991519086943831e-06,
+      "loss": 0.5159,
+      "step": 6399
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.881599578638916,
+      "learning_rate": 4.990214332643876e-06,
+      "loss": 0.5031,
+      "step": 6400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.228297194039834,
+      "learning_rate": 4.9889095790102854e-06,
+      "loss": 0.4734,
+      "step": 6401
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.691311635428924,
+      "learning_rate": 4.9876048261319e-06,
+      "loss": 0.4844,
+      "step": 6402
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.980715258389614,
+      "learning_rate": 4.986300074097573e-06,
+      "loss": 0.5131,
+      "step": 6403
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.184940554591244,
+      "learning_rate": 4.984995322996149e-06,
+      "loss": 0.422,
+      "step": 6404
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.963501855608326,
+      "learning_rate": 4.98369057291648e-06,
+      "loss": 0.495,
+      "step": 6405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.445787651477062,
+      "learning_rate": 4.982385823947408e-06,
+      "loss": 0.4463,
+      "step": 6406
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.653119562089174,
+      "learning_rate": 4.9810810761777855e-06,
+      "loss": 0.5626,
+      "step": 6407
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.242333975520995,
+      "learning_rate": 4.979776329696456e-06,
+      "loss": 0.5657,
+      "step": 6408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.211952739643376,
+      "learning_rate": 4.978471584592269e-06,
+      "loss": 0.4926,
+      "step": 6409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.534723974901707,
+      "learning_rate": 4.9771668409540746e-06,
+      "loss": 0.455,
+      "step": 6410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.570330134891577,
+      "learning_rate": 4.975862098870715e-06,
+      "loss": 0.3578,
+      "step": 6411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.361770932395801,
+      "learning_rate": 4.974557358431041e-06,
+      "loss": 0.4813,
+      "step": 6412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.158407650932794,
+      "learning_rate": 4.973252619723897e-06,
+      "loss": 0.2707,
+      "step": 6413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.336762944955536,
+      "learning_rate": 4.971947882838133e-06,
+      "loss": 0.5264,
+      "step": 6414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.166984867857105,
+      "learning_rate": 4.970643147862593e-06,
+      "loss": 0.4834,
+      "step": 6415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.60868898406371,
+      "learning_rate": 4.969338414886126e-06,
+      "loss": 0.5338,
+      "step": 6416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.000227647281537,
+      "learning_rate": 4.968033683997576e-06,
+      "loss": 0.6024,
+      "step": 6417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.799269902062818,
+      "learning_rate": 4.96672895528579e-06,
+      "loss": 0.4487,
+      "step": 6418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.198111062784867,
+      "learning_rate": 4.965424228839618e-06,
+      "loss": 0.4343,
+      "step": 6419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.243079661424913,
+      "learning_rate": 4.9641195047479005e-06,
+      "loss": 0.5212,
+      "step": 6420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.0944503535432,
+      "learning_rate": 4.9628147830994875e-06,
+      "loss": 0.4435,
+      "step": 6421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.204344607538061,
+      "learning_rate": 4.961510063983223e-06,
+      "loss": 0.4748,
+      "step": 6422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.714814451072101,
+      "learning_rate": 4.960205347487953e-06,
+      "loss": 0.5459,
+      "step": 6423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.267556449465784,
+      "learning_rate": 4.958900633702523e-06,
+      "loss": 0.4517,
+      "step": 6424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.849612905065628,
+      "learning_rate": 4.957595922715778e-06,
+      "loss": 0.5331,
+      "step": 6425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.512354938419294,
+      "learning_rate": 4.956291214616563e-06,
+      "loss": 0.5251,
+      "step": 6426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.826104986810969,
+      "learning_rate": 4.954986509493724e-06,
+      "loss": 0.7937,
+      "step": 6427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.201494477030159,
+      "learning_rate": 4.953681807436103e-06,
+      "loss": 0.3758,
+      "step": 6428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.699486213486084,
+      "learning_rate": 4.952377108532546e-06,
+      "loss": 0.4935,
+      "step": 6429
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.975178771369759,
+      "learning_rate": 4.951072412871899e-06,
+      "loss": 0.5224,
+      "step": 6430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.167294032013958,
+      "learning_rate": 4.949767720543003e-06,
+      "loss": 0.4194,
+      "step": 6431
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.35103235085094,
+      "learning_rate": 4.948463031634705e-06,
+      "loss": 0.475,
+      "step": 6432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7467574546428994,
+      "learning_rate": 4.947158346235845e-06,
+      "loss": 0.7038,
+      "step": 6433
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7862457800457898,
+      "learning_rate": 4.945853664435269e-06,
+      "loss": 0.828,
+      "step": 6434
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.469212688980101,
+      "learning_rate": 4.944548986321819e-06,
+      "loss": 0.441,
+      "step": 6435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.004826851970158,
+      "learning_rate": 4.943244311984337e-06,
+      "loss": 0.4541,
+      "step": 6436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.943371692455184,
+      "learning_rate": 4.941939641511666e-06,
+      "loss": 0.5623,
+      "step": 6437
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.340443842618306,
+      "learning_rate": 4.940634974992649e-06,
+      "loss": 0.4897,
+      "step": 6438
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.441351451520395,
+      "learning_rate": 4.939330312516129e-06,
+      "loss": 0.537,
+      "step": 6439
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 19.74851480448892,
+      "learning_rate": 4.938025654170945e-06,
+      "loss": 0.452,
+      "step": 6440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.835113017216655,
+      "learning_rate": 4.936721000045941e-06,
+      "loss": 0.4225,
+      "step": 6441
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.187150079133128,
+      "learning_rate": 4.935416350229956e-06,
+      "loss": 0.5364,
+      "step": 6442
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.346151521909575,
+      "learning_rate": 4.934111704811834e-06,
+      "loss": 0.4628,
+      "step": 6443
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.395241364883198,
+      "learning_rate": 4.9328070638804115e-06,
+      "loss": 0.387,
+      "step": 6444
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.695908282009734,
+      "learning_rate": 4.9315024275245325e-06,
+      "loss": 0.5644,
+      "step": 6445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.648555857818137,
+      "learning_rate": 4.930197795833033e-06,
+      "loss": 0.533,
+      "step": 6446
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.018374195572617,
+      "learning_rate": 4.9288931688947575e-06,
+      "loss": 0.4331,
+      "step": 6447
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.356983026160856,
+      "learning_rate": 4.927588546798541e-06,
+      "loss": 0.4607,
+      "step": 6448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 28.39421226868874,
+      "learning_rate": 4.926283929633224e-06,
+      "loss": 0.5348,
+      "step": 6449
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.737322498505085,
+      "learning_rate": 4.9249793174876476e-06,
+      "loss": 0.4598,
+      "step": 6450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.537222874861207,
+      "learning_rate": 4.923674710450647e-06,
+      "loss": 0.4822,
+      "step": 6451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.672677390819862,
+      "learning_rate": 4.9223701086110615e-06,
+      "loss": 0.4939,
+      "step": 6452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.951062528119605,
+      "learning_rate": 4.9210655120577274e-06,
+      "loss": 0.495,
+      "step": 6453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.691637164552158,
+      "learning_rate": 4.919760920879484e-06,
+      "loss": 0.446,
+      "step": 6454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8185429162222968,
+      "learning_rate": 4.918456335165165e-06,
+      "loss": 0.7533,
+      "step": 6455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 23.444576057577162,
+      "learning_rate": 4.917151755003612e-06,
+      "loss": 0.4698,
+      "step": 6456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 25.310208654930634,
+      "learning_rate": 4.915847180483656e-06,
+      "loss": 0.5324,
+      "step": 6457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0200225845159623,
+      "learning_rate": 4.914542611694135e-06,
+      "loss": 0.8103,
+      "step": 6458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.282430757406226,
+      "learning_rate": 4.913238048723886e-06,
+      "loss": 0.5021,
+      "step": 6459
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.32253236640783,
+      "learning_rate": 4.91193349166174e-06,
+      "loss": 0.5191,
+      "step": 6460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.04925663249555,
+      "learning_rate": 4.910628940596536e-06,
+      "loss": 0.4638,
+      "step": 6461
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.429429390824085,
+      "learning_rate": 4.9093243956171036e-06,
+      "loss": 0.3999,
+      "step": 6462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.351474418133629,
+      "learning_rate": 4.908019856812282e-06,
+      "loss": 0.517,
+      "step": 6463
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.114277641604302,
+      "learning_rate": 4.906715324270899e-06,
+      "loss": 0.5572,
+      "step": 6464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.236349688198144,
+      "learning_rate": 4.905410798081792e-06,
+      "loss": 0.4685,
+      "step": 6465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.056821095408498,
+      "learning_rate": 4.904106278333789e-06,
+      "loss": 0.5028,
+      "step": 6466
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.277401248341732,
+      "learning_rate": 4.902801765115727e-06,
+      "loss": 0.5587,
+      "step": 6467
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.388210476928048,
+      "learning_rate": 4.901497258516433e-06,
+      "loss": 0.593,
+      "step": 6468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.769228776325726,
+      "learning_rate": 4.9001927586247405e-06,
+      "loss": 0.4951,
+      "step": 6469
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.050952618753139,
+      "learning_rate": 4.89888826552948e-06,
+      "loss": 0.5255,
+      "step": 6470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.529062959250505,
+      "learning_rate": 4.897583779319481e-06,
+      "loss": 0.4696,
+      "step": 6471
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.631674327431447,
+      "learning_rate": 4.896279300083577e-06,
+      "loss": 0.3865,
+      "step": 6472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.460175886182511,
+      "learning_rate": 4.89497482791059e-06,
+      "loss": 0.369,
+      "step": 6473
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9348263271618542,
+      "learning_rate": 4.893670362889355e-06,
+      "loss": 0.8343,
+      "step": 6474
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.454813444005316,
+      "learning_rate": 4.8923659051086975e-06,
+      "loss": 0.5137,
+      "step": 6475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.89133945718215,
+      "learning_rate": 4.891061454657447e-06,
+      "loss": 0.4373,
+      "step": 6476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.571314046372324,
+      "learning_rate": 4.889757011624428e-06,
+      "loss": 0.4055,
+      "step": 6477
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.837802076513085,
+      "learning_rate": 4.888452576098471e-06,
+      "loss": 0.5524,
+      "step": 6478
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.238986274362839,
+      "learning_rate": 4.887148148168399e-06,
+      "loss": 0.4265,
+      "step": 6479
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.472917092203414,
+      "learning_rate": 4.885843727923038e-06,
+      "loss": 0.4959,
+      "step": 6480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.271364433815826,
+      "learning_rate": 4.884539315451218e-06,
+      "loss": 0.4413,
+      "step": 6481
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 59.682336454342334,
+      "learning_rate": 4.883234910841757e-06,
+      "loss": 0.4962,
+      "step": 6482
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.359718415189628,
+      "learning_rate": 4.881930514183484e-06,
+      "loss": 0.5358,
+      "step": 6483
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 18.6313035774451,
+      "learning_rate": 4.880626125565219e-06,
+      "loss": 0.4834,
+      "step": 6484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.999160770297719,
+      "learning_rate": 4.879321745075789e-06,
+      "loss": 0.5039,
+      "step": 6485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.179292848445447,
+      "learning_rate": 4.8780173728040135e-06,
+      "loss": 0.4906,
+      "step": 6486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.748307613565242,
+      "learning_rate": 4.876713008838717e-06,
+      "loss": 0.4347,
+      "step": 6487
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.49418711056626,
+      "learning_rate": 4.875408653268718e-06,
+      "loss": 0.5283,
+      "step": 6488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.387367729529746,
+      "learning_rate": 4.874104306182838e-06,
+      "loss": 0.4989,
+      "step": 6489
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.297151174111338,
+      "learning_rate": 4.872799967669901e-06,
+      "loss": 0.4964,
+      "step": 6490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.79977015141072,
+      "learning_rate": 4.871495637818721e-06,
+      "loss": 0.4857,
+      "step": 6491
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.352764513555206,
+      "learning_rate": 4.870191316718121e-06,
+      "loss": 0.4923,
+      "step": 6492
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.883823871208466,
+      "learning_rate": 4.868887004456918e-06,
+      "loss": 0.4109,
+      "step": 6493
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.605417279608247,
+      "learning_rate": 4.8675827011239315e-06,
+      "loss": 0.4713,
+      "step": 6494
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.66191329250306,
+      "learning_rate": 4.866278406807976e-06,
+      "loss": 0.4387,
+      "step": 6495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.118292272322966,
+      "learning_rate": 4.864974121597872e-06,
+      "loss": 0.4582,
+      "step": 6496
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.387718157454266,
+      "learning_rate": 4.863669845582431e-06,
+      "loss": 0.453,
+      "step": 6497
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.760338405269712,
+      "learning_rate": 4.862365578850473e-06,
+      "loss": 0.5618,
+      "step": 6498
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.206672801082862,
+      "learning_rate": 4.861061321490808e-06,
+      "loss": 0.4522,
+      "step": 6499
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.632665017841955,
+      "learning_rate": 4.859757073592253e-06,
+      "loss": 0.4781,
+      "step": 6500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.631847580967257,
+      "learning_rate": 4.858452835243623e-06,
+      "loss": 0.4094,
+      "step": 6501
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.4270679989764234,
+      "learning_rate": 4.857148606533727e-06,
+      "loss": 0.4046,
+      "step": 6502
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.453662320973873,
+      "learning_rate": 4.8558443875513815e-06,
+      "loss": 0.5335,
+      "step": 6503
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.848360084790857,
+      "learning_rate": 4.854540178385395e-06,
+      "loss": 0.4857,
+      "step": 6504
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.766041398736874,
+      "learning_rate": 4.85323597912458e-06,
+      "loss": 0.7095,
+      "step": 6505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 15.685397231018687,
+      "learning_rate": 4.851931789857744e-06,
+      "loss": 0.5677,
+      "step": 6506
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.5457063765836,
+      "learning_rate": 4.8506276106737e-06,
+      "loss": 0.3932,
+      "step": 6507
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.57545385009088,
+      "learning_rate": 4.8493234416612525e-06,
+      "loss": 0.4332,
+      "step": 6508
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.314170156349832,
+      "learning_rate": 4.848019282909214e-06,
+      "loss": 0.5395,
+      "step": 6509
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.334066357072503,
+      "learning_rate": 4.846715134506392e-06,
+      "loss": 0.5337,
+      "step": 6510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 17.751919441358957,
+      "learning_rate": 4.84541099654159e-06,
+      "loss": 0.5719,
+      "step": 6511
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.938317142241833,
+      "learning_rate": 4.844106869103617e-06,
+      "loss": 0.4917,
+      "step": 6512
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 12.236719611231967,
+      "learning_rate": 4.842802752281275e-06,
+      "loss": 0.5286,
+      "step": 6513
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.72565599985623,
+      "learning_rate": 4.841498646163373e-06,
+      "loss": 0.4641,
+      "step": 6514
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.723801462593868,
+      "learning_rate": 4.840194550838711e-06,
+      "loss": 0.5125,
+      "step": 6515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.800329366535266,
+      "learning_rate": 4.838890466396093e-06,
+      "loss": 0.4755,
+      "step": 6516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.22832517759844,
+      "learning_rate": 4.837586392924321e-06,
+      "loss": 0.4865,
+      "step": 6517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.44316504443142,
+      "learning_rate": 4.836282330512201e-06,
+      "loss": 0.5814,
+      "step": 6518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 38.07590105569483,
+      "learning_rate": 4.834978279248527e-06,
+      "loss": 0.5192,
+      "step": 6519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.872270367040537,
+      "learning_rate": 4.833674239222103e-06,
+      "loss": 0.476,
+      "step": 6520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.236990511901715,
+      "learning_rate": 4.83237021052173e-06,
+      "loss": 0.5466,
+      "step": 6521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.15737249626968,
+      "learning_rate": 4.831066193236203e-06,
+      "loss": 0.6472,
+      "step": 6522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.220914007460113,
+      "learning_rate": 4.8297621874543214e-06,
+      "loss": 0.5543,
+      "step": 6523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 29.631763232916917,
+      "learning_rate": 4.828458193264882e-06,
+      "loss": 0.579,
+      "step": 6524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.344624710453814,
+      "learning_rate": 4.827154210756682e-06,
+      "loss": 0.5348,
+      "step": 6525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.0802340089982,
+      "learning_rate": 4.825850240018514e-06,
+      "loss": 0.3873,
+      "step": 6526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 23.115045172290586,
+      "learning_rate": 4.824546281139178e-06,
+      "loss": 0.5209,
+      "step": 6527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.569092727357037,
+      "learning_rate": 4.8232423342074615e-06,
+      "loss": 0.4428,
+      "step": 6528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.557229349657387,
+      "learning_rate": 4.821938399312161e-06,
+      "loss": 0.5203,
+      "step": 6529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7555818668108938,
+      "learning_rate": 4.82063447654207e-06,
+      "loss": 0.7192,
+      "step": 6530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9617115358552129,
+      "learning_rate": 4.819330565985976e-06,
+      "loss": 0.752,
+      "step": 6531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 22.879157341211418,
+      "learning_rate": 4.818026667732674e-06,
+      "loss": 0.5009,
+      "step": 6532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.468328678060088,
+      "learning_rate": 4.8167227818709484e-06,
+      "loss": 0.3973,
+      "step": 6533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.35836966647172,
+      "learning_rate": 4.8154189084895944e-06,
+      "loss": 0.4254,
+      "step": 6534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.765710532796,
+      "learning_rate": 4.814115047677393e-06,
+      "loss": 0.5438,
+      "step": 6535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.56573456198208,
+      "learning_rate": 4.812811199523137e-06,
+      "loss": 0.505,
+      "step": 6536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.63337686207223,
+      "learning_rate": 4.81150736411561e-06,
+      "loss": 0.5413,
+      "step": 6537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.992155737739974,
+      "learning_rate": 4.8102035415436e-06,
+      "loss": 0.4898,
+      "step": 6538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.761911079786984,
+      "learning_rate": 4.808899731895887e-06,
+      "loss": 0.4515,
+      "step": 6539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.623956571388657,
+      "learning_rate": 4.8075959352612585e-06,
+      "loss": 0.4423,
+      "step": 6540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.604049114376741,
+      "learning_rate": 4.806292151728497e-06,
+      "loss": 0.5424,
+      "step": 6541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.003056603566385,
+      "learning_rate": 4.804988381386381e-06,
+      "loss": 0.5337,
+      "step": 6542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.244358688512847,
+      "learning_rate": 4.803684624323697e-06,
+      "loss": 0.415,
+      "step": 6543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.997630005620815,
+      "learning_rate": 4.8023808806292185e-06,
+      "loss": 0.5178,
+      "step": 6544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 17.125358697533525,
+      "learning_rate": 4.8010771503917315e-06,
+      "loss": 0.5549,
+      "step": 6545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.26288851253668,
+      "learning_rate": 4.79977343370001e-06,
+      "loss": 0.4289,
+      "step": 6546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.743601990494675,
+      "learning_rate": 4.798469730642833e-06,
+      "loss": 0.4121,
+      "step": 6547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.472377819700224,
+      "learning_rate": 4.797166041308975e-06,
+      "loss": 0.4547,
+      "step": 6548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.798561953785793,
+      "learning_rate": 4.795862365787213e-06,
+      "loss": 0.4657,
+      "step": 6549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.506971527884984,
+      "learning_rate": 4.7945587041663225e-06,
+      "loss": 0.4871,
+      "step": 6550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.281498087107504,
+      "learning_rate": 4.793255056535073e-06,
+      "loss": 0.3896,
+      "step": 6551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.852692784083228,
+      "learning_rate": 4.791951422982244e-06,
+      "loss": 0.5496,
+      "step": 6552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.537131399594614,
+      "learning_rate": 4.790647803596599e-06,
+      "loss": 0.4304,
+      "step": 6553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.608060686075635,
+      "learning_rate": 4.789344198466916e-06,
+      "loss": 0.433,
+      "step": 6554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.688949331445592,
+      "learning_rate": 4.788040607681962e-06,
+      "loss": 0.3963,
+      "step": 6555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.997955584013756,
+      "learning_rate": 4.786737031330504e-06,
+      "loss": 0.4838,
+      "step": 6556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.343693224292,
+      "learning_rate": 4.7854334695013115e-06,
+      "loss": 0.4961,
+      "step": 6557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.4638227261466135,
+      "learning_rate": 4.784129922283152e-06,
+      "loss": 0.4771,
+      "step": 6558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.785268741597365,
+      "learning_rate": 4.782826389764789e-06,
+      "loss": 0.5276,
+      "step": 6559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5778747450215207,
+      "learning_rate": 4.781522872034987e-06,
+      "loss": 0.7203,
+      "step": 6560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.44420069893517,
+      "learning_rate": 4.780219369182515e-06,
+      "loss": 0.5239,
+      "step": 6561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 22.762164719848446,
+      "learning_rate": 4.77891588129613e-06,
+      "loss": 0.4595,
+      "step": 6562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.935036203190903,
+      "learning_rate": 4.777612408464597e-06,
+      "loss": 0.5294,
+      "step": 6563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.719022640732031,
+      "learning_rate": 4.776308950776675e-06,
+      "loss": 0.4831,
+      "step": 6564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0759912133693827,
+      "learning_rate": 4.7750055083211255e-06,
+      "loss": 0.8033,
+      "step": 6565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.681224308137772,
+      "learning_rate": 4.7737020811867055e-06,
+      "loss": 0.7178,
+      "step": 6566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.073717610896841,
+      "learning_rate": 4.772398669462173e-06,
+      "loss": 0.4942,
+      "step": 6567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.36937474445863,
+      "learning_rate": 4.771095273236285e-06,
+      "loss": 0.577,
+      "step": 6568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.796360116853574,
+      "learning_rate": 4.769791892597795e-06,
+      "loss": 0.4227,
+      "step": 6569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.18493077329924,
+      "learning_rate": 4.768488527635462e-06,
+      "loss": 0.4735,
+      "step": 6570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 15.04594284910781,
+      "learning_rate": 4.767185178438035e-06,
+      "loss": 0.5134,
+      "step": 6571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 20.08632446036565,
+      "learning_rate": 4.765881845094269e-06,
+      "loss": 0.453,
+      "step": 6572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.53797941201844,
+      "learning_rate": 4.764578527692913e-06,
+      "loss": 0.4605,
+      "step": 6573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.21874910468902,
+      "learning_rate": 4.76327522632272e-06,
+      "loss": 0.5611,
+      "step": 6574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.423884581743652,
+      "learning_rate": 4.761971941072437e-06,
+      "loss": 0.5212,
+      "step": 6575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.34545243746879,
+      "learning_rate": 4.760668672030812e-06,
+      "loss": 0.4649,
+      "step": 6576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.658448948464176,
+      "learning_rate": 4.759365419286592e-06,
+      "loss": 0.7555,
+      "step": 6577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.071986857303438,
+      "learning_rate": 4.758062182928523e-06,
+      "loss": 0.4385,
+      "step": 6578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.582893071954057,
+      "learning_rate": 4.756758963045349e-06,
+      "loss": 0.4942,
+      "step": 6579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.975477862875586,
+      "learning_rate": 4.755455759725815e-06,
+      "loss": 0.4239,
+      "step": 6580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.298541644463505,
+      "learning_rate": 4.754152573058662e-06,
+      "loss": 0.4222,
+      "step": 6581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.489458672670965,
+      "learning_rate": 4.752849403132632e-06,
+      "loss": 0.4808,
+      "step": 6582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.592771727385738,
+      "learning_rate": 4.7515462500364665e-06,
+      "loss": 0.6431,
+      "step": 6583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.398932387082132,
+      "learning_rate": 4.750243113858901e-06,
+      "loss": 0.4826,
+      "step": 6584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.141936641533885,
+      "learning_rate": 4.748939994688676e-06,
+      "loss": 0.5084,
+      "step": 6585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.288294494145829,
+      "learning_rate": 4.747636892614526e-06,
+      "loss": 0.4678,
+      "step": 6586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.279859165554443,
+      "learning_rate": 4.74633380772519e-06,
+      "loss": 0.3572,
+      "step": 6587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 20.621211557473686,
+      "learning_rate": 4.745030740109397e-06,
+      "loss": 0.5822,
+      "step": 6588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.532609080360828,
+      "learning_rate": 4.743727689855885e-06,
+      "loss": 0.4141,
+      "step": 6589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 10.322361732271656,
+      "learning_rate": 4.742424657053385e-06,
+      "loss": 0.4392,
+      "step": 6590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.896214938634368,
+      "learning_rate": 4.741121641790625e-06,
+      "loss": 0.4354,
+      "step": 6591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.49313128776141,
+      "learning_rate": 4.73981864415634e-06,
+      "loss": 0.5793,
+      "step": 6592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.797990252737877,
+      "learning_rate": 4.738515664239252e-06,
+      "loss": 0.5489,
+      "step": 6593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 26.43173678774609,
+      "learning_rate": 4.737212702128092e-06,
+      "loss": 0.5199,
+      "step": 6594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.564742433000497,
+      "learning_rate": 4.735909757911584e-06,
+      "loss": 0.4157,
+      "step": 6595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.268741011677052,
+      "learning_rate": 4.734606831678457e-06,
+      "loss": 0.4721,
+      "step": 6596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.058035772311955,
+      "learning_rate": 4.733303923517427e-06,
+      "loss": 0.5271,
+      "step": 6597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 25.674449327125473,
+      "learning_rate": 4.732001033517224e-06,
+      "loss": 0.5417,
+      "step": 6598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.2284246164104315,
+      "learning_rate": 4.730698161766564e-06,
+      "loss": 0.4225,
+      "step": 6599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5967202222353374,
+      "learning_rate": 4.7293953083541685e-06,
+      "loss": 0.7515,
+      "step": 6600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.723856571104735,
+      "learning_rate": 4.7280924733687564e-06,
+      "loss": 0.5594,
+      "step": 6601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.217159554277467,
+      "learning_rate": 4.726789656899045e-06,
+      "loss": 0.3863,
+      "step": 6602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.968651101983532,
+      "learning_rate": 4.72548685903375e-06,
+      "loss": 0.4305,
+      "step": 6603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.542098346313498,
+      "learning_rate": 4.724184079861585e-06,
+      "loss": 0.4484,
+      "step": 6604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.992084801786561,
+      "learning_rate": 4.722881319471267e-06,
+      "loss": 0.3954,
+      "step": 6605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.663918934815579,
+      "learning_rate": 4.721578577951503e-06,
+      "loss": 0.4483,
+      "step": 6606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.362722112339904,
+      "learning_rate": 4.720275855391009e-06,
+      "loss": 0.5002,
+      "step": 6607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.23221102847849,
+      "learning_rate": 4.718973151878491e-06,
+      "loss": 0.5445,
+      "step": 6608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.713489913403727,
+      "learning_rate": 4.71767046750266e-06,
+      "loss": 0.5074,
+      "step": 6609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.612130242031056,
+      "learning_rate": 4.716367802352221e-06,
+      "loss": 0.4963,
+      "step": 6610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.884714438345543,
+      "learning_rate": 4.7150651565158805e-06,
+      "loss": 0.4371,
+      "step": 6611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.693706253495765,
+      "learning_rate": 4.713762530082344e-06,
+      "loss": 0.5322,
+      "step": 6612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6114762254537167,
+      "learning_rate": 4.712459923140312e-06,
+      "loss": 0.7245,
+      "step": 6613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.081377947966364,
+      "learning_rate": 4.71115733577849e-06,
+      "loss": 0.3812,
+      "step": 6614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.222287546707868,
+      "learning_rate": 4.709854768085574e-06,
+      "loss": 0.4263,
+      "step": 6615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.84158058438009,
+      "learning_rate": 4.708552220150268e-06,
+      "loss": 0.5137,
+      "step": 6616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.207559589980967,
+      "learning_rate": 4.707249692061265e-06,
+      "loss": 0.4525,
+      "step": 6617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 38.44571508112985,
+      "learning_rate": 4.705947183907265e-06,
+      "loss": 0.4726,
+      "step": 6618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6565777689734735,
+      "learning_rate": 4.704644695776961e-06,
+      "loss": 0.6437,
+      "step": 6619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 51.63332940539762,
+      "learning_rate": 4.703342227759047e-06,
+      "loss": 0.4562,
+      "step": 6620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.984647532868596,
+      "learning_rate": 4.702039779942217e-06,
+      "loss": 0.4674,
+      "step": 6621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.5496035260038985,
+      "learning_rate": 4.700737352415158e-06,
+      "loss": 0.425,
+      "step": 6622
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.685304391195883,
+      "learning_rate": 4.699434945266565e-06,
+      "loss": 0.5466,
+      "step": 6623
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.073546276794074,
+      "learning_rate": 4.698132558585121e-06,
+      "loss": 0.5437,
+      "step": 6624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 14.783321501334699,
+      "learning_rate": 4.696830192459517e-06,
+      "loss": 0.6136,
+      "step": 6625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.954440463583934,
+      "learning_rate": 4.695527846978435e-06,
+      "loss": 0.5115,
+      "step": 6626
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.200449287722661,
+      "learning_rate": 4.6942255222305614e-06,
+      "loss": 0.5066,
+      "step": 6627
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.384314202193938,
+      "learning_rate": 4.692923218304577e-06,
+      "loss": 0.4697,
+      "step": 6628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.719627143179458,
+      "learning_rate": 4.691620935289165e-06,
+      "loss": 0.4749,
+      "step": 6629
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 11.425663620544581,
+      "learning_rate": 4.690318673273002e-06,
+      "loss": 0.564,
+      "step": 6630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 21.456763862544417,
+      "learning_rate": 4.689016432344768e-06,
+      "loss": 0.5522,
+      "step": 6631
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.469255028172487,
+      "learning_rate": 4.687714212593142e-06,
+      "loss": 0.5536,
+      "step": 6632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.430355259593238,
+      "learning_rate": 4.686412014106795e-06,
+      "loss": 0.487,
+      "step": 6633
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5701085672226551,
+      "learning_rate": 4.6851098369744055e-06,
+      "loss": 0.718,
+      "step": 6634
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.871558866026296,
+      "learning_rate": 4.683807681284642e-06,
+      "loss": 0.4274,
+      "step": 6635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.873279346955075,
+      "learning_rate": 4.6825055471261795e-06,
+      "loss": 0.4423,
+      "step": 6636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 18.30593236032555,
+      "learning_rate": 4.681203434587683e-06,
+      "loss": 0.4928,
+      "step": 6637
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.636395551528398,
+      "learning_rate": 4.679901343757825e-06,
+      "loss": 0.5181,
+      "step": 6638
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.079705875707377,
+      "learning_rate": 4.6785992747252695e-06,
+      "loss": 0.5187,
+      "step": 6639
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 12.378400797157399,
+      "learning_rate": 4.677297227578681e-06,
+      "loss": 0.3852,
+      "step": 6640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.704250727785105,
+      "learning_rate": 4.6759952024067265e-06,
+      "loss": 0.5502,
+      "step": 6641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.945635031805326,
+      "learning_rate": 4.674693199298065e-06,
+      "loss": 0.4901,
+      "step": 6642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.890650618221324,
+      "learning_rate": 4.6733912183413595e-06,
+      "loss": 0.599,
+      "step": 6643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 23.149378706652143,
+      "learning_rate": 4.672089259625267e-06,
+      "loss": 0.5443,
+      "step": 6644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.971763196683203,
+      "learning_rate": 4.670787323238447e-06,
+      "loss": 0.479,
+      "step": 6645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.00003558028889,
+      "learning_rate": 4.669485409269553e-06,
+      "loss": 0.4261,
+      "step": 6646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.544422740961137,
+      "learning_rate": 4.6681835178072434e-06,
+      "loss": 0.5392,
+      "step": 6647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.253707204184224,
+      "learning_rate": 4.666881648940167e-06,
+      "loss": 0.453,
+      "step": 6648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.838070388687831,
+      "learning_rate": 4.665579802756979e-06,
+      "loss": 0.4417,
+      "step": 6649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.946476812123095,
+      "learning_rate": 4.664277979346326e-06,
+      "loss": 0.4783,
+      "step": 6650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.30923881347937,
+      "learning_rate": 4.662976178796859e-06,
+      "loss": 0.4085,
+      "step": 6651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.870704666466787,
+      "learning_rate": 4.661674401197225e-06,
+      "loss": 0.4687,
+      "step": 6652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.234362028506045,
+      "learning_rate": 4.660372646636068e-06,
+      "loss": 0.5024,
+      "step": 6653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.273802632643939,
+      "learning_rate": 4.659070915202032e-06,
+      "loss": 0.5115,
+      "step": 6654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.51936034685672,
+      "learning_rate": 4.657769206983759e-06,
+      "loss": 0.4337,
+      "step": 6655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.5733715913389466,
+      "learning_rate": 4.656467522069891e-06,
+      "loss": 0.5074,
+      "step": 6656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.141211410430778,
+      "learning_rate": 4.655165860549065e-06,
+      "loss": 0.5731,
+      "step": 6657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.809229644249463,
+      "learning_rate": 4.653864222509919e-06,
+      "loss": 0.5537,
+      "step": 6658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.880031489210014,
+      "learning_rate": 4.652562608041089e-06,
+      "loss": 0.4598,
+      "step": 6659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.375268949319414,
+      "learning_rate": 4.651261017231208e-06,
+      "loss": 0.4531,
+      "step": 6660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.065968695025383,
+      "learning_rate": 4.649959450168912e-06,
+      "loss": 0.3946,
+      "step": 6661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7378218705539459,
+      "learning_rate": 4.648657906942829e-06,
+      "loss": 0.7253,
+      "step": 6662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.611713322725699,
+      "learning_rate": 4.647356387641589e-06,
+      "loss": 0.4888,
+      "step": 6663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.84606088468913,
+      "learning_rate": 4.64605489235382e-06,
+      "loss": 0.3767,
+      "step": 6664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.71819631447058,
+      "learning_rate": 4.644753421168149e-06,
+      "loss": 0.5034,
+      "step": 6665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.242395485157044,
+      "learning_rate": 4.6434519741731955e-06,
+      "loss": 0.4898,
+      "step": 6666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.575769294602399,
+      "learning_rate": 4.642150551457591e-06,
+      "loss": 0.6242,
+      "step": 6667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.863855306915243,
+      "learning_rate": 4.640849153109948e-06,
+      "loss": 0.562,
+      "step": 6668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.718278654827012,
+      "learning_rate": 4.639547779218893e-06,
+      "loss": 0.5481,
+      "step": 6669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.790209824594364,
+      "learning_rate": 4.6382464298730365e-06,
+      "loss": 0.4435,
+      "step": 6670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.967007673438331,
+      "learning_rate": 4.636945105161001e-06,
+      "loss": 0.5611,
+      "step": 6671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.517011123762297,
+      "learning_rate": 4.6356438051713985e-06,
+      "loss": 0.5248,
+      "step": 6672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.4230988038341,
+      "learning_rate": 4.6343425299928415e-06,
+      "loss": 0.5028,
+      "step": 6673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.595128462066354,
+      "learning_rate": 4.633041279713943e-06,
+      "loss": 0.4583,
+      "step": 6674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.01558437448044,
+      "learning_rate": 4.631740054423308e-06,
+      "loss": 0.507,
+      "step": 6675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 26.661685678892933,
+      "learning_rate": 4.63043885420955e-06,
+      "loss": 0.4194,
+      "step": 6676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.135350494051167,
+      "learning_rate": 4.629137679161269e-06,
+      "loss": 0.4714,
+      "step": 6677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.590428563539279,
+      "learning_rate": 4.627836529367074e-06,
+      "loss": 0.5252,
+      "step": 6678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.183341845468016,
+      "learning_rate": 4.626535404915566e-06,
+      "loss": 0.5064,
+      "step": 6679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.212340495238318,
+      "learning_rate": 4.625234305895346e-06,
+      "loss": 0.49,
+      "step": 6680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.277834338427967,
+      "learning_rate": 4.623933232395012e-06,
+      "loss": 0.5454,
+      "step": 6681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.495221477338074,
+      "learning_rate": 4.622632184503162e-06,
+      "loss": 0.4618,
+      "step": 6682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.790839611330279,
+      "learning_rate": 4.621331162308393e-06,
+      "loss": 0.4512,
+      "step": 6683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.44555833823089,
+      "learning_rate": 4.620030165899296e-06,
+      "loss": 0.5634,
+      "step": 6684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.950340276121885,
+      "learning_rate": 4.618729195364466e-06,
+      "loss": 0.3308,
+      "step": 6685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.507836117636295,
+      "learning_rate": 4.61742825079249e-06,
+      "loss": 0.386,
+      "step": 6686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.918656297306402,
+      "learning_rate": 4.6161273322719605e-06,
+      "loss": 0.4153,
+      "step": 6687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.208480913559832,
+      "learning_rate": 4.614826439891462e-06,
+      "loss": 0.4787,
+      "step": 6688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.180740860295714,
+      "learning_rate": 4.61352557373958e-06,
+      "loss": 0.488,
+      "step": 6689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.723788924826677,
+      "learning_rate": 4.612224733904897e-06,
+      "loss": 0.4131,
+      "step": 6690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.259751888092278,
+      "learning_rate": 4.610923920475995e-06,
+      "loss": 0.4078,
+      "step": 6691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.485346278459854,
+      "learning_rate": 4.609623133541455e-06,
+      "loss": 0.6078,
+      "step": 6692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8435634690354552,
+      "learning_rate": 4.6083223731898516e-06,
+      "loss": 0.751,
+      "step": 6693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.68705507593497,
+      "learning_rate": 4.607021639509766e-06,
+      "loss": 0.4677,
+      "step": 6694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.703205736093189,
+      "learning_rate": 4.605720932589765e-06,
+      "loss": 0.446,
+      "step": 6695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.605187667383023,
+      "learning_rate": 4.604420252518428e-06,
+      "loss": 0.4764,
+      "step": 6696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.348702398992822,
+      "learning_rate": 4.603119599384322e-06,
+      "loss": 0.4549,
+      "step": 6697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.07491573925745,
+      "learning_rate": 4.601818973276017e-06,
+      "loss": 0.5828,
+      "step": 6698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.092371328322557,
+      "learning_rate": 4.600518374282079e-06,
+      "loss": 0.5646,
+      "step": 6699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.026627874308389,
+      "learning_rate": 4.5992178024910715e-06,
+      "loss": 0.4358,
+      "step": 6700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.832032654780042,
+      "learning_rate": 4.597917257991564e-06,
+      "loss": 0.587,
+      "step": 6701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.785681386865514,
+      "learning_rate": 4.596616740872109e-06,
+      "loss": 0.4422,
+      "step": 6702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.737820819628271,
+      "learning_rate": 4.5953162512212734e-06,
+      "loss": 0.516,
+      "step": 6703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.537728154758012,
+      "learning_rate": 4.5940157891276106e-06,
+      "loss": 0.3692,
+      "step": 6704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.083617672289169,
+      "learning_rate": 4.592715354679678e-06,
+      "loss": 0.5469,
+      "step": 6705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.361747338347175,
+      "learning_rate": 4.591414947966028e-06,
+      "loss": 0.3976,
+      "step": 6706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.897570954134872,
+      "learning_rate": 4.590114569075215e-06,
+      "loss": 0.3968,
+      "step": 6707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 19.904484189417193,
+      "learning_rate": 4.5888142180957864e-06,
+      "loss": 0.3793,
+      "step": 6708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.95364664422057,
+      "learning_rate": 4.587513895116293e-06,
+      "loss": 0.5256,
+      "step": 6709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.060479389343824,
+      "learning_rate": 4.586213600225277e-06,
+      "loss": 0.4346,
+      "step": 6710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.949940183865227,
+      "learning_rate": 4.584913333511286e-06,
+      "loss": 0.4618,
+      "step": 6711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9679777676547694,
+      "learning_rate": 4.583613095062864e-06,
+      "loss": 0.76,
+      "step": 6712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.163405873666234,
+      "learning_rate": 4.5823128849685465e-06,
+      "loss": 0.4799,
+      "step": 6713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.942559502616408,
+      "learning_rate": 4.581012703316876e-06,
+      "loss": 0.4716,
+      "step": 6714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.315904226033062,
+      "learning_rate": 4.5797125501963875e-06,
+      "loss": 0.5204,
+      "step": 6715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.03996718695781,
+      "learning_rate": 4.578412425695617e-06,
+      "loss": 0.6327,
+      "step": 6716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.539396263902898,
+      "learning_rate": 4.5771123299030954e-06,
+      "loss": 0.5266,
+      "step": 6717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 22.78460471960427,
+      "learning_rate": 4.5758122629073545e-06,
+      "loss": 0.4204,
+      "step": 6718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.119083285995983,
+      "learning_rate": 4.574512224796923e-06,
+      "loss": 0.476,
+      "step": 6719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.344567664979769,
+      "learning_rate": 4.573212215660326e-06,
+      "loss": 0.5194,
+      "step": 6720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.914043554131318,
+      "learning_rate": 4.571912235586092e-06,
+      "loss": 0.5617,
+      "step": 6721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.135260335872088,
+      "learning_rate": 4.5706122846627415e-06,
+      "loss": 0.4112,
+      "step": 6722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 15.274669871524683,
+      "learning_rate": 4.569312362978798e-06,
+      "loss": 0.4823,
+      "step": 6723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.12588766667191,
+      "learning_rate": 4.568012470622775e-06,
+      "loss": 0.4505,
+      "step": 6724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.916533181329077,
+      "learning_rate": 4.566712607683195e-06,
+      "loss": 0.522,
+      "step": 6725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 60.377730737923,
+      "learning_rate": 4.565412774248569e-06,
+      "loss": 0.4154,
+      "step": 6726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.72429719145818,
+      "learning_rate": 4.564112970407413e-06,
+      "loss": 0.7073,
+      "step": 6727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.192397102715415,
+      "learning_rate": 4.5628131962482355e-06,
+      "loss": 0.4162,
+      "step": 6728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.882028581050388,
+      "learning_rate": 4.561513451859548e-06,
+      "loss": 0.6096,
+      "step": 6729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.358970629630676,
+      "learning_rate": 4.5602137373298526e-06,
+      "loss": 0.4758,
+      "step": 6730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.302940702041177,
+      "learning_rate": 4.5589140527476584e-06,
+      "loss": 0.4863,
+      "step": 6731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.161949849596247,
+      "learning_rate": 4.557614398201469e-06,
+      "loss": 0.4719,
+      "step": 6732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7094027572807224,
+      "learning_rate": 4.556314773779781e-06,
+      "loss": 0.7591,
+      "step": 6733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.601843472999419,
+      "learning_rate": 4.555015179571096e-06,
+      "loss": 0.5373,
+      "step": 6734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.43839002259317,
+      "learning_rate": 4.553715615663909e-06,
+      "loss": 0.5467,
+      "step": 6735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.441238973505088,
+      "learning_rate": 4.552416082146717e-06,
+      "loss": 0.4829,
+      "step": 6736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.222479185063725,
+      "learning_rate": 4.551116579108008e-06,
+      "loss": 0.5315,
+      "step": 6737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.307366993458372,
+      "learning_rate": 4.549817106636279e-06,
+      "loss": 0.5137,
+      "step": 6738
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.25146671506976,
+      "learning_rate": 4.548517664820011e-06,
+      "loss": 0.5225,
+      "step": 6739
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.908933293552418,
+      "learning_rate": 4.547218253747696e-06,
+      "loss": 0.6192,
+      "step": 6740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.488036167120562,
+      "learning_rate": 4.545918873507815e-06,
+      "loss": 0.5961,
+      "step": 6741
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.027798533751897,
+      "learning_rate": 4.544619524188852e-06,
+      "loss": 0.5568,
+      "step": 6742
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.031761959259116,
+      "learning_rate": 4.543320205879285e-06,
+      "loss": 0.5243,
+      "step": 6743
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.325271728219634,
+      "learning_rate": 4.542020918667593e-06,
+      "loss": 0.5766,
+      "step": 6744
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.107341083208302,
+      "learning_rate": 4.540721662642252e-06,
+      "loss": 0.4776,
+      "step": 6745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 23.673064652381143,
+      "learning_rate": 4.5394224378917335e-06,
+      "loss": 0.3954,
+      "step": 6746
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.78047858126158,
+      "learning_rate": 4.538123244504513e-06,
+      "loss": 0.4868,
+      "step": 6747
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.599092942058835,
+      "learning_rate": 4.536824082569055e-06,
+      "loss": 0.4289,
+      "step": 6748
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.495462698409287,
+      "learning_rate": 4.5355249521738306e-06,
+      "loss": 0.5071,
+      "step": 6749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.962315394488664,
+      "learning_rate": 4.534225853407303e-06,
+      "loss": 0.4193,
+      "step": 6750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.163875538994943,
+      "learning_rate": 4.532926786357935e-06,
+      "loss": 0.5097,
+      "step": 6751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.015218294276055,
+      "learning_rate": 4.531627751114188e-06,
+      "loss": 0.4822,
+      "step": 6752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 30.035481572705237,
+      "learning_rate": 4.53032874776452e-06,
+      "loss": 0.4662,
+      "step": 6753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.476972079994509,
+      "learning_rate": 4.529029776397389e-06,
+      "loss": 0.5114,
+      "step": 6754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.340089872931953,
+      "learning_rate": 4.527730837101245e-06,
+      "loss": 0.542,
+      "step": 6755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.44456603057091,
+      "learning_rate": 4.5264319299645466e-06,
+      "loss": 0.4634,
+      "step": 6756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.087407167489696,
+      "learning_rate": 4.525133055075737e-06,
+      "loss": 0.523,
+      "step": 6757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.375370164068718,
+      "learning_rate": 4.523834212523268e-06,
+      "loss": 0.4435,
+      "step": 6758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.245223339494827,
+      "learning_rate": 4.522535402395583e-06,
+      "loss": 0.5349,
+      "step": 6759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6007861337033271,
+      "learning_rate": 4.521236624781127e-06,
+      "loss": 0.725,
+      "step": 6760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.21584484481426,
+      "learning_rate": 4.519937879768339e-06,
+      "loss": 0.5841,
+      "step": 6761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0574950600344004,
+      "learning_rate": 4.5186391674456585e-06,
+      "loss": 0.7567,
+      "step": 6762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.339414641747323,
+      "learning_rate": 4.517340487901525e-06,
+      "loss": 0.4676,
+      "step": 6763
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.4755862436269505,
+      "learning_rate": 4.516041841224366e-06,
+      "loss": 0.5088,
+      "step": 6764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.524114621497828,
+      "learning_rate": 4.514743227502623e-06,
+      "loss": 0.4295,
+      "step": 6765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.459552609132313,
+      "learning_rate": 4.5134446468247156e-06,
+      "loss": 0.4569,
+      "step": 6766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.407853583049561,
+      "learning_rate": 4.5121460992790796e-06,
+      "loss": 0.5097,
+      "step": 6767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.049458609558787,
+      "learning_rate": 4.510847584954136e-06,
+      "loss": 0.5674,
+      "step": 6768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.89456033192156,
+      "learning_rate": 4.50954910393831e-06,
+      "loss": 0.3987,
+      "step": 6769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.202726877436598,
+      "learning_rate": 4.50825065632002e-06,
+      "loss": 0.4558,
+      "step": 6770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.222038833215072,
+      "learning_rate": 4.506952242187687e-06,
+      "loss": 0.5713,
+      "step": 6771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9979868172067918,
+      "learning_rate": 4.505653861629728e-06,
+      "loss": 0.8322,
+      "step": 6772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.6192467725726924,
+      "learning_rate": 4.504355514734552e-06,
+      "loss": 0.5359,
+      "step": 6773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.977760115560727,
+      "learning_rate": 4.503057201590576e-06,
+      "loss": 0.4574,
+      "step": 6774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.309548973976382,
+      "learning_rate": 4.501758922286207e-06,
+      "loss": 0.4366,
+      "step": 6775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.316534855316105,
+      "learning_rate": 4.500460676909853e-06,
+      "loss": 0.3953,
+      "step": 6776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 31.884023610806295,
+      "learning_rate": 4.499162465549916e-06,
+      "loss": 0.4655,
+      "step": 6777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.294499790449901,
+      "learning_rate": 4.497864288294804e-06,
+      "loss": 0.4687,
+      "step": 6778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.365775345673683,
+      "learning_rate": 4.496566145232912e-06,
+      "loss": 0.4952,
+      "step": 6779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.301238190844088,
+      "learning_rate": 4.4952680364526405e-06,
+      "loss": 0.3863,
+      "step": 6780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 33.80718743534867,
+      "learning_rate": 4.493969962042382e-06,
+      "loss": 0.4803,
+      "step": 6781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 22.461270927300923,
+      "learning_rate": 4.4926719220905315e-06,
+      "loss": 0.4996,
+      "step": 6782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 20.904907846561827,
+      "learning_rate": 4.491373916685482e-06,
+      "loss": 0.4199,
+      "step": 6783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.285563985883782,
+      "learning_rate": 4.49007594591562e-06,
+      "loss": 0.4985,
+      "step": 6784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.24676457672553,
+      "learning_rate": 4.4887780098693305e-06,
+      "loss": 0.5163,
+      "step": 6785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.859767539937134,
+      "learning_rate": 4.487480108634999e-06,
+      "loss": 0.4887,
+      "step": 6786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.432491889636916,
+      "learning_rate": 4.486182242301006e-06,
+      "loss": 0.5382,
+      "step": 6787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.537379308058805,
+      "learning_rate": 4.484884410955729e-06,
+      "loss": 0.561,
+      "step": 6788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.587604618546208,
+      "learning_rate": 4.483586614687548e-06,
+      "loss": 0.5126,
+      "step": 6789
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.934548587502874,
+      "learning_rate": 4.482288853584834e-06,
+      "loss": 0.4874,
+      "step": 6790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.38086121707137,
+      "learning_rate": 4.480991127735959e-06,
+      "loss": 0.4579,
+      "step": 6791
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.359099621185283,
+      "learning_rate": 4.479693437229295e-06,
+      "loss": 0.41,
+      "step": 6792
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.59886648553632,
+      "learning_rate": 4.4783957821532074e-06,
+      "loss": 0.5043,
+      "step": 6793
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.948367198187261,
+      "learning_rate": 4.477098162596061e-06,
+      "loss": 0.536,
+      "step": 6794
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.419058414405834,
+      "learning_rate": 4.4758005786462164e-06,
+      "loss": 0.4862,
+      "step": 6795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.78817930429037,
+      "learning_rate": 4.474503030392037e-06,
+      "loss": 0.4985,
+      "step": 6796
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 17.02420789454278,
+      "learning_rate": 4.473205517921875e-06,
+      "loss": 0.5348,
+      "step": 6797
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8855014365469276,
+      "learning_rate": 4.47190804132409e-06,
+      "loss": 0.7405,
+      "step": 6798
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.44141761111403,
+      "learning_rate": 4.470610600687032e-06,
+      "loss": 0.5958,
+      "step": 6799
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.82395852348263,
+      "learning_rate": 4.469313196099052e-06,
+      "loss": 0.5543,
+      "step": 6800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.562224633188206,
+      "learning_rate": 4.4680158276484945e-06,
+      "loss": 0.4935,
+      "step": 6801
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.7983713278450955,
+      "learning_rate": 4.466718495423708e-06,
+      "loss": 0.4475,
+      "step": 6802
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.707108018721548,
+      "learning_rate": 4.465421199513035e-06,
+      "loss": 0.526,
+      "step": 6803
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.146226744965395,
+      "learning_rate": 4.464123940004815e-06,
+      "loss": 0.3727,
+      "step": 6804
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.550979955683564,
+      "learning_rate": 4.462826716987385e-06,
+      "loss": 0.4692,
+      "step": 6805
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.092091992225488,
+      "learning_rate": 4.4615295305490795e-06,
+      "loss": 0.4999,
+      "step": 6806
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.1312827218934345,
+      "learning_rate": 4.460232380778234e-06,
+      "loss": 0.471,
+      "step": 6807
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.689195136513922,
+      "learning_rate": 4.458935267763174e-06,
+      "loss": 0.5913,
+      "step": 6808
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.95210252811806,
+      "learning_rate": 4.457638191592233e-06,
+      "loss": 0.532,
+      "step": 6809
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.729606409497869,
+      "learning_rate": 4.4563411523537305e-06,
+      "loss": 0.4383,
+      "step": 6810
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.528676000263658,
+      "learning_rate": 4.455044150135993e-06,
+      "loss": 0.4937,
+      "step": 6811
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.899468486520759,
+      "learning_rate": 4.45374718502734e-06,
+      "loss": 0.4485,
+      "step": 6812
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 25.16971438635407,
+      "learning_rate": 4.452450257116088e-06,
+      "loss": 0.4542,
+      "step": 6813
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.98133169940329,
+      "learning_rate": 4.451153366490553e-06,
+      "loss": 0.4723,
+      "step": 6814
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.176998105369915,
+      "learning_rate": 4.449856513239046e-06,
+      "loss": 0.4212,
+      "step": 6815
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.811273082238725,
+      "learning_rate": 4.448559697449879e-06,
+      "loss": 0.5449,
+      "step": 6816
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.84810597240319,
+      "learning_rate": 4.447262919211356e-06,
+      "loss": 0.5604,
+      "step": 6817
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.165624128304072,
+      "learning_rate": 4.445966178611788e-06,
+      "loss": 0.5019,
+      "step": 6818
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.495331569078978,
+      "learning_rate": 4.44466947573947e-06,
+      "loss": 0.3912,
+      "step": 6819
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.312502011971835,
+      "learning_rate": 4.4433728106827085e-06,
+      "loss": 0.5237,
+      "step": 6820
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.54132651154162,
+      "learning_rate": 4.442076183529795e-06,
+      "loss": 0.3652,
+      "step": 6821
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 41.54845828375662,
+      "learning_rate": 4.440779594369028e-06,
+      "loss": 0.6594,
+      "step": 6822
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.497870704017595,
+      "learning_rate": 4.4394830432886985e-06,
+      "loss": 0.4988,
+      "step": 6823
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.310248016303639,
+      "learning_rate": 4.438186530377094e-06,
+      "loss": 0.5011,
+      "step": 6824
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.4051119022215515,
+      "learning_rate": 4.4368900557225046e-06,
+      "loss": 0.4576,
+      "step": 6825
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.097930443574279,
+      "learning_rate": 4.43559361941321e-06,
+      "loss": 0.4084,
+      "step": 6826
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.10476760754591,
+      "learning_rate": 4.4342972215374975e-06,
+      "loss": 0.5205,
+      "step": 6827
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.606499443110718,
+      "learning_rate": 4.43300086218364e-06,
+      "loss": 0.4946,
+      "step": 6828
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.100674559180788,
+      "learning_rate": 4.431704541439918e-06,
+      "loss": 0.4362,
+      "step": 6829
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.573597821811983,
+      "learning_rate": 4.430408259394604e-06,
+      "loss": 0.4376,
+      "step": 6830
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.337264103287177,
+      "learning_rate": 4.42911201613597e-06,
+      "loss": 0.4393,
+      "step": 6831
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.356951814776648,
+      "learning_rate": 4.427815811752283e-06,
+      "loss": 0.5478,
+      "step": 6832
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.327477910555457,
+      "learning_rate": 4.426519646331808e-06,
+      "loss": 0.4781,
+      "step": 6833
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9389284952941817,
+      "learning_rate": 4.425223519962812e-06,
+      "loss": 0.7871,
+      "step": 6834
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.898336523693099,
+      "learning_rate": 4.4239274327335505e-06,
+      "loss": 0.4911,
+      "step": 6835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.37890048128136,
+      "learning_rate": 4.422631384732286e-06,
+      "loss": 0.5098,
+      "step": 6836
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.160187549655578,
+      "learning_rate": 4.4213353760472694e-06,
+      "loss": 0.4272,
+      "step": 6837
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.505400259407112,
+      "learning_rate": 4.420039406766757e-06,
+      "loss": 0.5531,
+      "step": 6838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.39717728106245,
+      "learning_rate": 4.418743476978995e-06,
+      "loss": 0.4695,
+      "step": 6839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.375774097661069,
+      "learning_rate": 4.417447586772234e-06,
+      "loss": 0.4692,
+      "step": 6840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.80896486646702,
+      "learning_rate": 4.416151736234715e-06,
+      "loss": 0.4842,
+      "step": 6841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7953611307454522,
+      "learning_rate": 4.4148559254546805e-06,
+      "loss": 0.7712,
+      "step": 6842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.429566215100296,
+      "learning_rate": 4.4135601545203735e-06,
+      "loss": 0.4695,
+      "step": 6843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.955454544277067,
+      "learning_rate": 4.4122644235200235e-06,
+      "loss": 0.4528,
+      "step": 6844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.187264021161646,
+      "learning_rate": 4.41096873254187e-06,
+      "loss": 0.5099,
+      "step": 6845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.578838444196456,
+      "learning_rate": 4.40967308167414e-06,
+      "loss": 0.5353,
+      "step": 6846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.906804255593302,
+      "learning_rate": 4.4083774710050645e-06,
+      "loss": 0.4233,
+      "step": 6847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.050567429446104,
+      "learning_rate": 4.407081900622866e-06,
+      "loss": 0.4459,
+      "step": 6848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.29414630729959,
+      "learning_rate": 4.405786370615771e-06,
+      "loss": 0.489,
+      "step": 6849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.886503290825287,
+      "learning_rate": 4.404490881071995e-06,
+      "loss": 0.4866,
+      "step": 6850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6848546753296008,
+      "learning_rate": 4.403195432079755e-06,
+      "loss": 0.3293,
+      "step": 6851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.176152056653336,
+      "learning_rate": 4.401900023727271e-06,
+      "loss": 0.4697,
+      "step": 6852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.504422881914081,
+      "learning_rate": 4.4006046561027474e-06,
+      "loss": 0.4033,
+      "step": 6853
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.6856180534510585,
+      "learning_rate": 4.3993093292944e-06,
+      "loss": 0.4203,
+      "step": 6854
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7111417553554713,
+      "learning_rate": 4.398014043390429e-06,
+      "loss": 0.7335,
+      "step": 6855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 27.47556735563902,
+      "learning_rate": 4.396718798479041e-06,
+      "loss": 0.4413,
+      "step": 6856
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.749894243385797,
+      "learning_rate": 4.395423594648434e-06,
+      "loss": 0.4536,
+      "step": 6857
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2327813487449095,
+      "learning_rate": 4.39412843198681e-06,
+      "loss": 0.7568,
+      "step": 6858
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 22.203502819536517,
+      "learning_rate": 4.392833310582359e-06,
+      "loss": 0.5091,
+      "step": 6859
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.261444161568729,
+      "learning_rate": 4.391538230523276e-06,
+      "loss": 0.4748,
+      "step": 6860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.418762241127785,
+      "learning_rate": 4.390243191897748e-06,
+      "loss": 0.5555,
+      "step": 6861
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.828051575325097,
+      "learning_rate": 4.388948194793962e-06,
+      "loss": 0.4823,
+      "step": 6862
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.470148915546554,
+      "learning_rate": 4.387653239300103e-06,
+      "loss": 0.5097,
+      "step": 6863
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.168997350018426,
+      "learning_rate": 4.386358325504351e-06,
+      "loss": 0.5095,
+      "step": 6864
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.32856219916987,
+      "learning_rate": 4.385063453494885e-06,
+      "loss": 0.4757,
+      "step": 6865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.744898306037904,
+      "learning_rate": 4.383768623359878e-06,
+      "loss": 0.4808,
+      "step": 6866
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8157987177327946,
+      "learning_rate": 4.382473835187503e-06,
+      "loss": 0.7607,
+      "step": 6867
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.796883604930445,
+      "learning_rate": 4.381179089065929e-06,
+      "loss": 0.4873,
+      "step": 6868
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.5684518202512505,
+      "learning_rate": 4.379884385083324e-06,
+      "loss": 0.5777,
+      "step": 6869
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.59832459990529,
+      "learning_rate": 4.378589723327848e-06,
+      "loss": 0.4842,
+      "step": 6870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.4362759865302746,
+      "learning_rate": 4.377295103887669e-06,
+      "loss": 0.5248,
+      "step": 6871
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.9214544704763705,
+      "learning_rate": 4.376000526850936e-06,
+      "loss": 0.5089,
+      "step": 6872
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.785510301019564,
+      "learning_rate": 4.374705992305809e-06,
+      "loss": 0.5233,
+      "step": 6873
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.969812369175014,
+      "learning_rate": 4.3734115003404405e-06,
+      "loss": 0.5503,
+      "step": 6874
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.591171323080915,
+      "learning_rate": 4.372117051042977e-06,
+      "loss": 0.4747,
+      "step": 6875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.459939119281805,
+      "learning_rate": 4.370822644501568e-06,
+      "loss": 0.426,
+      "step": 6876
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.9038020905799105,
+      "learning_rate": 4.369528280804354e-06,
+      "loss": 0.444,
+      "step": 6877
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.356846088438778,
+      "learning_rate": 4.368233960039476e-06,
+      "loss": 0.5436,
+      "step": 6878
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.410168848471893,
+      "learning_rate": 4.366939682295072e-06,
+      "loss": 0.5049,
+      "step": 6879
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.732992038639128,
+      "learning_rate": 4.365645447659278e-06,
+      "loss": 0.4952,
+      "step": 6880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.932724329096782,
+      "learning_rate": 4.364351256220222e-06,
+      "loss": 0.3838,
+      "step": 6881
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.981896969244786,
+      "learning_rate": 4.363057108066035e-06,
+      "loss": 0.5101,
+      "step": 6882
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.134747814285452,
+      "learning_rate": 4.361763003284845e-06,
+      "loss": 0.404,
+      "step": 6883
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.001991783276126,
+      "learning_rate": 4.360468941964771e-06,
+      "loss": 0.6236,
+      "step": 6884
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.287194529259021,
+      "learning_rate": 4.359174924193936e-06,
+      "loss": 0.485,
+      "step": 6885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.46470101489878,
+      "learning_rate": 4.3578809500604535e-06,
+      "loss": 0.4637,
+      "step": 6886
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.227552698719298,
+      "learning_rate": 4.356587019652441e-06,
+      "loss": 0.4308,
+      "step": 6887
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.652391188037974,
+      "learning_rate": 4.355293133058005e-06,
+      "loss": 0.5253,
+      "step": 6888
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.05009360333492,
+      "learning_rate": 4.353999290365259e-06,
+      "loss": 0.4946,
+      "step": 6889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.089214957883536,
+      "learning_rate": 4.352705491662302e-06,
+      "loss": 0.4692,
+      "step": 6890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.667251631622247,
+      "learning_rate": 4.351411737037241e-06,
+      "loss": 0.502,
+      "step": 6891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.444947359010136,
+      "learning_rate": 4.350118026578172e-06,
+      "loss": 0.4685,
+      "step": 6892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.279934010752951,
+      "learning_rate": 4.348824360373193e-06,
+      "loss": 0.4903,
+      "step": 6893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.568953322442282,
+      "learning_rate": 4.347530738510396e-06,
+      "loss": 0.4975,
+      "step": 6894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.595996936640563,
+      "learning_rate": 4.34623716107787e-06,
+      "loss": 0.4025,
+      "step": 6895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.839451310877074,
+      "learning_rate": 4.3449436281637035e-06,
+      "loss": 0.5367,
+      "step": 6896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.926433975001585,
+      "learning_rate": 4.343650139855978e-06,
+      "loss": 0.485,
+      "step": 6897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.683755006744278,
+      "learning_rate": 4.342356696242779e-06,
+      "loss": 0.5258,
+      "step": 6898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.04648315626226,
+      "learning_rate": 4.341063297412177e-06,
+      "loss": 0.4491,
+      "step": 6899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 27.397200311024438,
+      "learning_rate": 4.339769943452254e-06,
+      "loss": 0.5283,
+      "step": 6900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.40134408465744,
+      "learning_rate": 4.338476634451079e-06,
+      "loss": 0.3615,
+      "step": 6901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.142046432758399,
+      "learning_rate": 4.337183370496718e-06,
+      "loss": 0.4931,
+      "step": 6902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.511531552692786,
+      "learning_rate": 4.335890151677241e-06,
+      "loss": 0.52,
+      "step": 6903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.64516051431456,
+      "learning_rate": 4.334596978080706e-06,
+      "loss": 0.4798,
+      "step": 6904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.409728130588078,
+      "learning_rate": 4.333303849795179e-06,
+      "loss": 0.4499,
+      "step": 6905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.885774755482752,
+      "learning_rate": 4.332010766908708e-06,
+      "loss": 0.479,
+      "step": 6906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.009000413839344,
+      "learning_rate": 4.3307177295093515e-06,
+      "loss": 0.4671,
+      "step": 6907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.254360227938396,
+      "learning_rate": 4.329424737685158e-06,
+      "loss": 0.3887,
+      "step": 6908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.24110206298192,
+      "learning_rate": 4.328131791524177e-06,
+      "loss": 0.5199,
+      "step": 6909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.033890915112285,
+      "learning_rate": 4.326838891114448e-06,
+      "loss": 0.5563,
+      "step": 6910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.926655701703446,
+      "learning_rate": 4.325546036544016e-06,
+      "loss": 0.3732,
+      "step": 6911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.459554870711507,
+      "learning_rate": 4.324253227900915e-06,
+      "loss": 0.4482,
+      "step": 6912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0157162707670024,
+      "learning_rate": 4.322960465273181e-06,
+      "loss": 0.7445,
+      "step": 6913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 14.633080687345776,
+      "learning_rate": 4.321667748748849e-06,
+      "loss": 0.4941,
+      "step": 6914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.991699682471679,
+      "learning_rate": 4.32037507841594e-06,
+      "loss": 0.5002,
+      "step": 6915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.081074983457707,
+      "learning_rate": 4.319082454362487e-06,
+      "loss": 0.493,
+      "step": 6916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.20303642310051,
+      "learning_rate": 4.317789876676506e-06,
+      "loss": 0.3844,
+      "step": 6917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.808130851598055,
+      "learning_rate": 4.316497345446019e-06,
+      "loss": 0.47,
+      "step": 6918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.863137609203708,
+      "learning_rate": 4.31520486075904e-06,
+      "loss": 0.3846,
+      "step": 6919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.559407055307925,
+      "learning_rate": 4.3139124227035835e-06,
+      "loss": 0.3747,
+      "step": 6920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.1691319839876035,
+      "learning_rate": 4.312620031367656e-06,
+      "loss": 0.3473,
+      "step": 6921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.404771518879166,
+      "learning_rate": 4.311327686839264e-06,
+      "loss": 0.695,
+      "step": 6922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.37317348814188,
+      "learning_rate": 4.310035389206415e-06,
+      "loss": 0.4158,
+      "step": 6923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6457245409790797,
+      "learning_rate": 4.308743138557102e-06,
+      "loss": 0.7009,
+      "step": 6924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.1248248274087365,
+      "learning_rate": 4.307450934979328e-06,
+      "loss": 0.5462,
+      "step": 6925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.685675049005481,
+      "learning_rate": 4.306158778561081e-06,
+      "loss": 0.467,
+      "step": 6926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.24365121232544,
+      "learning_rate": 4.304866669390355e-06,
+      "loss": 0.3782,
+      "step": 6927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.037822117964065,
+      "learning_rate": 4.303574607555134e-06,
+      "loss": 0.438,
+      "step": 6928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.043885952876276,
+      "learning_rate": 4.3022825931434045e-06,
+      "loss": 0.5056,
+      "step": 6929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.949387165765749,
+      "learning_rate": 4.300990626243144e-06,
+      "loss": 0.5255,
+      "step": 6930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.661708800134382,
+      "learning_rate": 4.2996987069423325e-06,
+      "loss": 0.5152,
+      "step": 6931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.306369875792734,
+      "learning_rate": 4.298406835328941e-06,
+      "loss": 0.4584,
+      "step": 6932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.12454626634901,
+      "learning_rate": 4.297115011490941e-06,
+      "loss": 0.4722,
+      "step": 6933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.05305220654738,
+      "learning_rate": 4.295823235516304e-06,
+      "loss": 0.465,
+      "step": 6934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.880272762679711,
+      "learning_rate": 4.2945315074929895e-06,
+      "loss": 0.5011,
+      "step": 6935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.968029409130447,
+      "learning_rate": 4.293239827508961e-06,
+      "loss": 0.5163,
+      "step": 6936
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.461134409077392,
+      "learning_rate": 4.291948195652175e-06,
+      "loss": 0.4681,
+      "step": 6937
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.649756689196055,
+      "learning_rate": 4.290656612010587e-06,
+      "loss": 0.4264,
+      "step": 6938
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.734168255188714,
+      "learning_rate": 4.289365076672146e-06,
+      "loss": 0.4647,
+      "step": 6939
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.50397501200413,
+      "learning_rate": 4.288073589724803e-06,
+      "loss": 0.4221,
+      "step": 6940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.359492286774907,
+      "learning_rate": 4.286782151256498e-06,
+      "loss": 0.4059,
+      "step": 6941
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.422691454935232,
+      "learning_rate": 4.285490761355176e-06,
+      "loss": 0.3888,
+      "step": 6942
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9092922360691893,
+      "learning_rate": 4.284199420108775e-06,
+      "loss": 0.5536,
+      "step": 6943
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.462657213192356,
+      "learning_rate": 4.282908127605228e-06,
+      "loss": 0.4741,
+      "step": 6944
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.396168175857172,
+      "learning_rate": 4.281616883932468e-06,
+      "loss": 0.5085,
+      "step": 6945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.9877618445549965,
+      "learning_rate": 4.2803256891784205e-06,
+      "loss": 0.5605,
+      "step": 6946
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.146512842841402,
+      "learning_rate": 4.279034543431012e-06,
+      "loss": 0.5749,
+      "step": 6947
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.401165581620918,
+      "learning_rate": 4.277743446778164e-06,
+      "loss": 0.437,
+      "step": 6948
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.71779328990394,
+      "learning_rate": 4.276452399307794e-06,
+      "loss": 0.5204,
+      "step": 6949
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.337941125072783,
+      "learning_rate": 4.275161401107814e-06,
+      "loss": 0.4642,
+      "step": 6950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.835729804336195,
+      "learning_rate": 4.273870452266142e-06,
+      "loss": 0.5522,
+      "step": 6951
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.355735002949066,
+      "learning_rate": 4.272579552870677e-06,
+      "loss": 0.4971,
+      "step": 6952
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.317869213537685,
+      "learning_rate": 4.27128870300933e-06,
+      "loss": 0.4889,
+      "step": 6953
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9534957664373276,
+      "learning_rate": 4.269997902770002e-06,
+      "loss": 0.3874,
+      "step": 6954
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.92317855911967,
+      "learning_rate": 4.268707152240587e-06,
+      "loss": 0.5035,
+      "step": 6955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.470442449434471,
+      "learning_rate": 4.2674164515089834e-06,
+      "loss": 0.4786,
+      "step": 6956
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.579018316538248,
+      "learning_rate": 4.266125800663079e-06,
+      "loss": 0.4651,
+      "step": 6957
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.1285223898754175,
+      "learning_rate": 4.264835199790764e-06,
+      "loss": 0.4377,
+      "step": 6958
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.638297496977939,
+      "learning_rate": 4.263544648979919e-06,
+      "loss": 0.4454,
+      "step": 6959
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.823872938332205,
+      "learning_rate": 4.26225414831843e-06,
+      "loss": 0.4908,
+      "step": 6960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.94522907248571,
+      "learning_rate": 4.260963697894169e-06,
+      "loss": 0.4508,
+      "step": 6961
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5533044434100585,
+      "learning_rate": 4.259673297795014e-06,
+      "loss": 0.6839,
+      "step": 6962
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 58.955797885352844,
+      "learning_rate": 4.2583829481088346e-06,
+      "loss": 0.4543,
+      "step": 6963
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.98313494241208,
+      "learning_rate": 4.257092648923496e-06,
+      "loss": 0.4807,
+      "step": 6964
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.661116853785761,
+      "learning_rate": 4.255802400326865e-06,
+      "loss": 0.5352,
+      "step": 6965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.721470984846038,
+      "learning_rate": 4.254512202406799e-06,
+      "loss": 0.5805,
+      "step": 6966
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.534443704506299,
+      "learning_rate": 4.253222055251157e-06,
+      "loss": 0.5017,
+      "step": 6967
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.651799747597346,
+      "learning_rate": 4.251931958947788e-06,
+      "loss": 0.5155,
+      "step": 6968
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.216316165392483,
+      "learning_rate": 4.25064191358455e-06,
+      "loss": 0.4561,
+      "step": 6969
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.482260324032279,
+      "learning_rate": 4.24935191924928e-06,
+      "loss": 0.5081,
+      "step": 6970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.182847272720395,
+      "learning_rate": 4.2480619760298284e-06,
+      "loss": 0.4502,
+      "step": 6971
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.883127185797365,
+      "learning_rate": 4.246772084014029e-06,
+      "loss": 0.412,
+      "step": 6972
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.758037975617671,
+      "learning_rate": 4.245482243289721e-06,
+      "loss": 0.3895,
+      "step": 6973
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.930767872967486,
+      "learning_rate": 4.244192453944739e-06,
+      "loss": 0.4225,
+      "step": 6974
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.42469259071729,
+      "learning_rate": 4.242902716066904e-06,
+      "loss": 0.4619,
+      "step": 6975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.305476818898299,
+      "learning_rate": 4.2416130297440514e-06,
+      "loss": 0.4991,
+      "step": 6976
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.295847268358351,
+      "learning_rate": 4.240323395063995e-06,
+      "loss": 0.4415,
+      "step": 6977
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.211545665690557,
+      "learning_rate": 4.2390338121145585e-06,
+      "loss": 0.4682,
+      "step": 6978
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.82112081144481,
+      "learning_rate": 4.237744280983554e-06,
+      "loss": 0.4551,
+      "step": 6979
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.640019433677708,
+      "learning_rate": 4.236454801758795e-06,
+      "loss": 0.4937,
+      "step": 6980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.220501141427818,
+      "learning_rate": 4.235165374528086e-06,
+      "loss": 0.4776,
+      "step": 6981
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.9731378948044425,
+      "learning_rate": 4.233875999379234e-06,
+      "loss": 0.5006,
+      "step": 6982
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.009816631563244,
+      "learning_rate": 4.2325866764000396e-06,
+      "loss": 0.415,
+      "step": 6983
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.357510476996474,
+      "learning_rate": 4.2312974056782975e-06,
+      "loss": 0.5082,
+      "step": 6984
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.650865302355333,
+      "learning_rate": 4.230008187301807e-06,
+      "loss": 0.5587,
+      "step": 6985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6976864705088952,
+      "learning_rate": 4.228719021358351e-06,
+      "loss": 0.7816,
+      "step": 6986
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5774635283127325,
+      "learning_rate": 4.227429907935721e-06,
+      "loss": 0.4913,
+      "step": 6987
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.606350807161223,
+      "learning_rate": 4.226140847121698e-06,
+      "loss": 0.4383,
+      "step": 6988
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6303798553918445,
+      "learning_rate": 4.224851839004063e-06,
+      "loss": 0.754,
+      "step": 6989
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.042834260927682,
+      "learning_rate": 4.223562883670588e-06,
+      "loss": 0.4991,
+      "step": 6990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.240774501103652,
+      "learning_rate": 4.22227398120905e-06,
+      "loss": 0.5048,
+      "step": 6991
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.7028247333641,
+      "learning_rate": 4.2209851317072135e-06,
+      "loss": 0.4886,
+      "step": 6992
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.253539787519493,
+      "learning_rate": 4.219696335252845e-06,
+      "loss": 0.4128,
+      "step": 6993
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.67846377911167,
+      "learning_rate": 4.218407591933708e-06,
+      "loss": 0.5433,
+      "step": 6994
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.302110616512543,
+      "learning_rate": 4.217118901837556e-06,
+      "loss": 0.4448,
+      "step": 6995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.22214909686521,
+      "learning_rate": 4.215830265052147e-06,
+      "loss": 0.3707,
+      "step": 6996
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.857150514281009,
+      "learning_rate": 4.214541681665229e-06,
+      "loss": 0.5583,
+      "step": 6997
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.752852024414916,
+      "learning_rate": 4.2132531517645505e-06,
+      "loss": 0.5322,
+      "step": 6998
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.020870232188548,
+      "learning_rate": 4.211964675437852e-06,
+      "loss": 0.454,
+      "step": 6999
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.53440876928294,
+      "learning_rate": 4.210676252772877e-06,
+      "loss": 0.4539,
+      "step": 7000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.42115174838003,
+      "learning_rate": 4.2093878838573575e-06,
+      "loss": 0.5078,
+      "step": 7001
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.682332115865283,
+      "learning_rate": 4.2080995687790295e-06,
+      "loss": 0.421,
+      "step": 7002
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.912806190956683,
+      "learning_rate": 4.206811307625617e-06,
+      "loss": 0.4443,
+      "step": 7003
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.896581359167085,
+      "learning_rate": 4.205523100484848e-06,
+      "loss": 0.5352,
+      "step": 7004
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.728468370236514,
+      "learning_rate": 4.2042349474444434e-06,
+      "loss": 0.3885,
+      "step": 7005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.80287489918061,
+      "learning_rate": 4.202946848592121e-06,
+      "loss": 0.485,
+      "step": 7006
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.828063309400184,
+      "learning_rate": 4.201658804015594e-06,
+      "loss": 0.5082,
+      "step": 7007
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 24.422440855588075,
+      "learning_rate": 4.200370813802572e-06,
+      "loss": 0.4118,
+      "step": 7008
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.125579460658972,
+      "learning_rate": 4.199082878040763e-06,
+      "loss": 0.4824,
+      "step": 7009
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.537754256809768,
+      "learning_rate": 4.197794996817868e-06,
+      "loss": 0.424,
+      "step": 7010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.342713462587062,
+      "learning_rate": 4.196507170221587e-06,
+      "loss": 0.4224,
+      "step": 7011
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.41110922690955,
+      "learning_rate": 4.195219398339614e-06,
+      "loss": 0.4442,
+      "step": 7012
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.815939676187123,
+      "learning_rate": 4.193931681259642e-06,
+      "loss": 0.5647,
+      "step": 7013
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0438646600095827,
+      "learning_rate": 4.19264401906936e-06,
+      "loss": 0.7576,
+      "step": 7014
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.955133869536751,
+      "learning_rate": 4.19135641185645e-06,
+      "loss": 0.4701,
+      "step": 7015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.377412265610721,
+      "learning_rate": 4.190068859708593e-06,
+      "loss": 0.5507,
+      "step": 7016
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.846716690375671,
+      "learning_rate": 4.188781362713465e-06,
+      "loss": 0.4407,
+      "step": 7017
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.874625393740973,
+      "learning_rate": 4.187493920958741e-06,
+      "loss": 0.5616,
+      "step": 7018
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.02154403059042,
+      "learning_rate": 4.186206534532087e-06,
+      "loss": 0.5514,
+      "step": 7019
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.859092779690583,
+      "learning_rate": 4.184919203521171e-06,
+      "loss": 0.4377,
+      "step": 7020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.332402285727994,
+      "learning_rate": 4.183631928013651e-06,
+      "loss": 0.3869,
+      "step": 7021
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.599993714045125,
+      "learning_rate": 4.18234470809719e-06,
+      "loss": 0.5206,
+      "step": 7022
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.044187897458291,
+      "learning_rate": 4.181057543859436e-06,
+      "loss": 0.4912,
+      "step": 7023
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.059997840260419,
+      "learning_rate": 4.179770435388044e-06,
+      "loss": 0.3895,
+      "step": 7024
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.965170891195098,
+      "learning_rate": 4.17848338277066e-06,
+      "loss": 0.4814,
+      "step": 7025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.580074043326992,
+      "learning_rate": 4.177196386094924e-06,
+      "loss": 0.4679,
+      "step": 7026
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.337662526633133,
+      "learning_rate": 4.175909445448477e-06,
+      "loss": 0.4394,
+      "step": 7027
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.719779688367066,
+      "learning_rate": 4.174622560918952e-06,
+      "loss": 0.449,
+      "step": 7028
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3922873961213202,
+      "learning_rate": 4.173335732593983e-06,
+      "loss": 0.4143,
+      "step": 7029
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.717355019594077,
+      "learning_rate": 4.172048960561192e-06,
+      "loss": 0.4261,
+      "step": 7030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.001428712494358,
+      "learning_rate": 4.1707622449082105e-06,
+      "loss": 0.4663,
+      "step": 7031
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.813135507195409,
+      "learning_rate": 4.169475585722651e-06,
+      "loss": 0.4809,
+      "step": 7032
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.601365772752325,
+      "learning_rate": 4.1681889830921315e-06,
+      "loss": 0.4921,
+      "step": 7033
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.597810642441188,
+      "learning_rate": 4.166902437104266e-06,
+      "loss": 0.4157,
+      "step": 7034
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.282118665145644,
+      "learning_rate": 4.16561594784666e-06,
+      "loss": 0.5123,
+      "step": 7035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.893096920266445,
+      "learning_rate": 4.16432951540692e-06,
+      "loss": 0.4898,
+      "step": 7036
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.07798028220421,
+      "learning_rate": 4.163043139872643e-06,
+      "loss": 0.459,
+      "step": 7037
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.338643565814403,
+      "learning_rate": 4.161756821331431e-06,
+      "loss": 0.383,
+      "step": 7038
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.431616492873057,
+      "learning_rate": 4.160470559870869e-06,
+      "loss": 0.3932,
+      "step": 7039
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.680864716856526,
+      "learning_rate": 4.159184355578554e-06,
+      "loss": 0.5945,
+      "step": 7040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.65778336636889,
+      "learning_rate": 4.157898208542064e-06,
+      "loss": 0.4975,
+      "step": 7041
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.960977457101721,
+      "learning_rate": 4.156612118848984e-06,
+      "loss": 0.4808,
+      "step": 7042
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.203618710804229,
+      "learning_rate": 4.155326086586889e-06,
+      "loss": 0.4738,
+      "step": 7043
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.76227303693267,
+      "learning_rate": 4.1540401118433525e-06,
+      "loss": 0.4102,
+      "step": 7044
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 24.85159816499811,
+      "learning_rate": 4.152754194705945e-06,
+      "loss": 0.4905,
+      "step": 7045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.713448757462733,
+      "learning_rate": 4.1514683352622295e-06,
+      "loss": 0.5028,
+      "step": 7046
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.331704968279215,
+      "learning_rate": 4.150182533599771e-06,
+      "loss": 0.4259,
+      "step": 7047
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.147591275865196,
+      "learning_rate": 4.1488967898061216e-06,
+      "loss": 0.5756,
+      "step": 7048
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.642549696510138,
+      "learning_rate": 4.14761110396884e-06,
+      "loss": 0.4913,
+      "step": 7049
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.778073326755476,
+      "learning_rate": 4.1463254761754714e-06,
+      "loss": 0.4624,
+      "step": 7050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.959177998613059,
+      "learning_rate": 4.145039906513565e-06,
+      "loss": 0.4905,
+      "step": 7051
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.812901741187998,
+      "learning_rate": 4.14375439507066e-06,
+      "loss": 0.7097,
+      "step": 7052
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.120291901270267,
+      "learning_rate": 4.142468941934295e-06,
+      "loss": 0.5304,
+      "step": 7053
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.73381177027679,
+      "learning_rate": 4.1411835471920046e-06,
+      "loss": 0.4951,
+      "step": 7054
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9754805703526561,
+      "learning_rate": 4.1398982109313144e-06,
+      "loss": 0.7132,
+      "step": 7055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.552237497506032,
+      "learning_rate": 4.138612933239758e-06,
+      "loss": 0.4221,
+      "step": 7056
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.590603928588253,
+      "learning_rate": 4.137327714204847e-06,
+      "loss": 0.5058,
+      "step": 7057
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.968236051302772,
+      "learning_rate": 4.136042553914108e-06,
+      "loss": 0.5273,
+      "step": 7058
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.415581353638529,
+      "learning_rate": 4.134757452455049e-06,
+      "loss": 0.5697,
+      "step": 7059
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.247866029261974,
+      "learning_rate": 4.133472409915184e-06,
+      "loss": 0.3934,
+      "step": 7060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.225155045877238,
+      "learning_rate": 4.132187426382015e-06,
+      "loss": 0.5769,
+      "step": 7061
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.647703268210841,
+      "learning_rate": 4.130902501943047e-06,
+      "loss": 0.4473,
+      "step": 7062
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.818992063560842,
+      "learning_rate": 4.129617636685773e-06,
+      "loss": 0.4921,
+      "step": 7063
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.883020020098699,
+      "learning_rate": 4.128332830697691e-06,
+      "loss": 0.4094,
+      "step": 7064
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.928948211954985,
+      "learning_rate": 4.1270480840662905e-06,
+      "loss": 0.5178,
+      "step": 7065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.376672661727598,
+      "learning_rate": 4.1257633968790535e-06,
+      "loss": 0.4537,
+      "step": 7066
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0308338673187296,
+      "learning_rate": 4.1244787692234655e-06,
+      "loss": 0.7241,
+      "step": 7067
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.185871313162403,
+      "learning_rate": 4.123194201187001e-06,
+      "loss": 0.5447,
+      "step": 7068
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.90810015320395,
+      "learning_rate": 4.121909692857136e-06,
+      "loss": 0.3824,
+      "step": 7069
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.0792278320099475,
+      "learning_rate": 4.120625244321338e-06,
+      "loss": 0.4636,
+      "step": 7070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.328289810481618,
+      "learning_rate": 4.1193408556670725e-06,
+      "loss": 0.465,
+      "step": 7071
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.798989399521063,
+      "learning_rate": 4.118056526981801e-06,
+      "loss": 0.5183,
+      "step": 7072
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.473558595471034,
+      "learning_rate": 4.116772258352979e-06,
+      "loss": 0.4482,
+      "step": 7073
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.065268852668105,
+      "learning_rate": 4.115488049868063e-06,
+      "loss": 0.4853,
+      "step": 7074
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.006294243067305,
+      "learning_rate": 4.1142039016144995e-06,
+      "loss": 0.4013,
+      "step": 7075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.9002423603214265,
+      "learning_rate": 4.1129198136797344e-06,
+      "loss": 0.4587,
+      "step": 7076
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.819081362474182,
+      "learning_rate": 4.111635786151207e-06,
+      "loss": 0.5719,
+      "step": 7077
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.267753540544896,
+      "learning_rate": 4.1103518191163555e-06,
+      "loss": 0.5781,
+      "step": 7078
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.054089876363212,
+      "learning_rate": 4.10906791266261e-06,
+      "loss": 0.3861,
+      "step": 7079
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.260066768019177,
+      "learning_rate": 4.107784066877403e-06,
+      "loss": 0.4912,
+      "step": 7080
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.3263703166114125,
+      "learning_rate": 4.106500281848154e-06,
+      "loss": 0.5811,
+      "step": 7081
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.81863736854151,
+      "learning_rate": 4.105216557662286e-06,
+      "loss": 0.5726,
+      "step": 7082
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.849123748477755,
+      "learning_rate": 4.103932894407213e-06,
+      "loss": 0.4951,
+      "step": 7083
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.490710074510793,
+      "learning_rate": 4.1026492921703485e-06,
+      "loss": 0.549,
+      "step": 7084
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.480334943205382,
+      "learning_rate": 4.1013657510391e-06,
+      "loss": 0.4884,
+      "step": 7085
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.728879236144009,
+      "learning_rate": 4.100082271100871e-06,
+      "loss": 0.5546,
+      "step": 7086
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.433437935207855,
+      "learning_rate": 4.09879885244306e-06,
+      "loss": 0.4636,
+      "step": 7087
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.024192873822014,
+      "learning_rate": 4.097515495153062e-06,
+      "loss": 0.4411,
+      "step": 7088
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.800891976526664,
+      "learning_rate": 4.09623219931827e-06,
+      "loss": 0.4833,
+      "step": 7089
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.871561654612185,
+      "learning_rate": 4.094948965026068e-06,
+      "loss": 0.5302,
+      "step": 7090
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7484248158910822,
+      "learning_rate": 4.093665792363841e-06,
+      "loss": 0.7224,
+      "step": 7091
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.729700208842774,
+      "learning_rate": 4.092382681418964e-06,
+      "loss": 0.6872,
+      "step": 7092
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.050095024719603,
+      "learning_rate": 4.0910996322788145e-06,
+      "loss": 0.4718,
+      "step": 7093
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.351928349494852,
+      "learning_rate": 4.089816645030762e-06,
+      "loss": 0.4517,
+      "step": 7094
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.880063783905975,
+      "learning_rate": 4.088533719762171e-06,
+      "loss": 0.612,
+      "step": 7095
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.308156109101544,
+      "learning_rate": 4.087250856560405e-06,
+      "loss": 0.5106,
+      "step": 7096
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.057864902792917,
+      "learning_rate": 4.0859680555128185e-06,
+      "loss": 0.494,
+      "step": 7097
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.234993052388797,
+      "learning_rate": 4.0846853167067665e-06,
+      "loss": 0.4591,
+      "step": 7098
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.303134874718797,
+      "learning_rate": 4.083402640229598e-06,
+      "loss": 0.49,
+      "step": 7099
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.767857303331185,
+      "learning_rate": 4.082120026168657e-06,
+      "loss": 0.4581,
+      "step": 7100
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.577467054433335,
+      "learning_rate": 4.080837474611281e-06,
+      "loss": 0.5108,
+      "step": 7101
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.2403453533396105,
+      "learning_rate": 4.0795549856448135e-06,
+      "loss": 0.479,
+      "step": 7102
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.557811370090488,
+      "learning_rate": 4.078272559356578e-06,
+      "loss": 0.5576,
+      "step": 7103
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.6618852977674,
+      "learning_rate": 4.076990195833907e-06,
+      "loss": 0.5457,
+      "step": 7104
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.803423616213671,
+      "learning_rate": 4.075707895164123e-06,
+      "loss": 0.4672,
+      "step": 7105
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.119566682699203,
+      "learning_rate": 4.0744256574345445e-06,
+      "loss": 0.474,
+      "step": 7106
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.33602910621778,
+      "learning_rate": 4.073143482732487e-06,
+      "loss": 0.5028,
+      "step": 7107
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.907395817810793,
+      "learning_rate": 4.071861371145258e-06,
+      "loss": 0.4679,
+      "step": 7108
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.625143846442649,
+      "learning_rate": 4.0705793227601695e-06,
+      "loss": 0.3768,
+      "step": 7109
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.639340201155623,
+      "learning_rate": 4.069297337664516e-06,
+      "loss": 0.4583,
+      "step": 7110
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 11.222726538588173,
+      "learning_rate": 4.0680154159456026e-06,
+      "loss": 0.4299,
+      "step": 7111
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.327509353497567,
+      "learning_rate": 4.066733557690716e-06,
+      "loss": 0.4889,
+      "step": 7112
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.863496869905387,
+      "learning_rate": 4.065451762987147e-06,
+      "loss": 0.3822,
+      "step": 7113
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.833095444348626,
+      "learning_rate": 4.064170031922183e-06,
+      "loss": 0.4582,
+      "step": 7114
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.031894216841307,
+      "learning_rate": 4.062888364583101e-06,
+      "loss": 0.5262,
+      "step": 7115
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 14.985548064224828,
+      "learning_rate": 4.061606761057179e-06,
+      "loss": 0.5479,
+      "step": 7116
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.709166313829414,
+      "learning_rate": 4.060325221431684e-06,
+      "loss": 0.4714,
+      "step": 7117
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.092031030164824,
+      "learning_rate": 4.059043745793893e-06,
+      "loss": 0.4572,
+      "step": 7118
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.635442776463992,
+      "learning_rate": 4.0577623342310565e-06,
+      "loss": 0.4986,
+      "step": 7119
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 19.08489750979561,
+      "learning_rate": 4.0564809868304424e-06,
+      "loss": 0.543,
+      "step": 7120
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.135920387277109,
+      "learning_rate": 4.0551997036793e-06,
+      "loss": 0.4793,
+      "step": 7121
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.600407574236555,
+      "learning_rate": 4.053918484864882e-06,
+      "loss": 0.517,
+      "step": 7122
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.214152850936904,
+      "learning_rate": 4.0526373304744296e-06,
+      "loss": 0.3621,
+      "step": 7123
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.3104121555373,
+      "learning_rate": 4.0513562405951876e-06,
+      "loss": 0.4644,
+      "step": 7124
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.634183128792987,
+      "learning_rate": 4.0500752153143915e-06,
+      "loss": 0.4546,
+      "step": 7125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 9.2189563701757,
+      "learning_rate": 4.048794254719271e-06,
+      "loss": 0.5048,
+      "step": 7126
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.477945374355567,
+      "learning_rate": 4.047513358897061e-06,
+      "loss": 0.451,
+      "step": 7127
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.893886809104686,
+      "learning_rate": 4.046232527934975e-06,
+      "loss": 0.4493,
+      "step": 7128
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.717867267628112,
+      "learning_rate": 4.044951761920239e-06,
+      "loss": 0.5639,
+      "step": 7129
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.947620299872876,
+      "learning_rate": 4.043671060940064e-06,
+      "loss": 0.4544,
+      "step": 7130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.016655336153206,
+      "learning_rate": 4.042390425081664e-06,
+      "loss": 0.4772,
+      "step": 7131
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.4958103300807775,
+      "learning_rate": 4.041109854432239e-06,
+      "loss": 0.4618,
+      "step": 7132
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.584207579564313,
+      "learning_rate": 4.039829349078994e-06,
+      "loss": 0.5795,
+      "step": 7133
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.288015571574118,
+      "learning_rate": 4.0385489091091265e-06,
+      "loss": 0.5167,
+      "step": 7134
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.287276186775196,
+      "learning_rate": 4.037268534609825e-06,
+      "loss": 0.4807,
+      "step": 7135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.112189488356322,
+      "learning_rate": 4.035988225668283e-06,
+      "loss": 0.4699,
+      "step": 7136
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.776753869961253,
+      "learning_rate": 4.034707982371676e-06,
+      "loss": 0.7945,
+      "step": 7137
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.489800558916505,
+      "learning_rate": 4.033427804807191e-06,
+      "loss": 0.5106,
+      "step": 7138
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.391128625612843,
+      "learning_rate": 4.032147693061998e-06,
+      "loss": 0.4121,
+      "step": 7139
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.948454724305481,
+      "learning_rate": 4.030867647223268e-06,
+      "loss": 0.4974,
+      "step": 7140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 66.60067862655514,
+      "learning_rate": 4.029587667378164e-06,
+      "loss": 0.4318,
+      "step": 7141
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.790672062821979,
+      "learning_rate": 4.028307753613853e-06,
+      "loss": 0.3814,
+      "step": 7142
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.477957066498885,
+      "learning_rate": 4.027027906017484e-06,
+      "loss": 0.6005,
+      "step": 7143
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.852667444028915,
+      "learning_rate": 4.0257481246762134e-06,
+      "loss": 0.4666,
+      "step": 7144
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.45546490956839,
+      "learning_rate": 4.024468409677189e-06,
+      "loss": 0.4546,
+      "step": 7145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.173932878732995,
+      "learning_rate": 4.023188761107552e-06,
+      "loss": 0.4894,
+      "step": 7146
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.303441789962552,
+      "learning_rate": 4.021909179054443e-06,
+      "loss": 0.6044,
+      "step": 7147
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.321820308626652,
+      "learning_rate": 4.020629663604992e-06,
+      "loss": 0.4618,
+      "step": 7148
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.517292297476481,
+      "learning_rate": 4.0193502148463325e-06,
+      "loss": 0.5122,
+      "step": 7149
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.667326133490992,
+      "learning_rate": 4.018070832865586e-06,
+      "loss": 0.3978,
+      "step": 7150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.12681358061266,
+      "learning_rate": 4.0167915177498764e-06,
+      "loss": 0.4632,
+      "step": 7151
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.973031744486269,
+      "learning_rate": 4.015512269586316e-06,
+      "loss": 0.5027,
+      "step": 7152
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 34.40867843825787,
+      "learning_rate": 4.014233088462018e-06,
+      "loss": 0.3782,
+      "step": 7153
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.67746060028523,
+      "learning_rate": 4.012953974464085e-06,
+      "loss": 0.4334,
+      "step": 7154
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.073594789907493,
+      "learning_rate": 4.011674927679624e-06,
+      "loss": 0.4293,
+      "step": 7155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.053343807720172,
+      "learning_rate": 4.010395948195731e-06,
+      "loss": 0.5783,
+      "step": 7156
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.38188485956452,
+      "learning_rate": 4.009117036099498e-06,
+      "loss": 0.4799,
+      "step": 7157
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.065548528441402,
+      "learning_rate": 4.007838191478015e-06,
+      "loss": 0.4333,
+      "step": 7158
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.02060239575554,
+      "learning_rate": 4.0065594144183625e-06,
+      "loss": 0.3653,
+      "step": 7159
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.912987168512952,
+      "learning_rate": 4.0052807050076235e-06,
+      "loss": 0.3261,
+      "step": 7160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.942290291699349,
+      "learning_rate": 4.004002063332868e-06,
+      "loss": 0.5151,
+      "step": 7161
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.193415126100362,
+      "learning_rate": 4.00272348948117e-06,
+      "loss": 0.466,
+      "step": 7162
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.493659100547098,
+      "learning_rate": 4.001444983539591e-06,
+      "loss": 0.4893,
+      "step": 7163
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 28.489671308840727,
+      "learning_rate": 4.000166545595194e-06,
+      "loss": 0.4763,
+      "step": 7164
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.864777725407883,
+      "learning_rate": 3.9988881757350365e-06,
+      "loss": 0.4972,
+      "step": 7165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.037011498183503,
+      "learning_rate": 3.997609874046166e-06,
+      "loss": 0.4979,
+      "step": 7166
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 28.322772597107384,
+      "learning_rate": 3.9963316406156335e-06,
+      "loss": 0.443,
+      "step": 7167
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.529316010416663,
+      "learning_rate": 3.9950534755304755e-06,
+      "loss": 0.4286,
+      "step": 7168
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.462739253205685,
+      "learning_rate": 3.993775378877735e-06,
+      "loss": 0.4625,
+      "step": 7169
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.530772333123653,
+      "learning_rate": 3.9924973507444405e-06,
+      "loss": 0.3705,
+      "step": 7170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.435451845004642,
+      "learning_rate": 3.991219391217624e-06,
+      "loss": 0.4045,
+      "step": 7171
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.815308555573564,
+      "learning_rate": 3.989941500384303e-06,
+      "loss": 0.4614,
+      "step": 7172
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.395118854158564,
+      "learning_rate": 3.9886636783315035e-06,
+      "loss": 0.4474,
+      "step": 7173
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.62308251991219,
+      "learning_rate": 3.987385925146233e-06,
+      "loss": 0.5123,
+      "step": 7174
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.818045579652073,
+      "learning_rate": 3.986108240915505e-06,
+      "loss": 0.525,
+      "step": 7175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.461940165814274,
+      "learning_rate": 3.984830625726324e-06,
+      "loss": 0.4628,
+      "step": 7176
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.444799241054962,
+      "learning_rate": 3.9835530796656874e-06,
+      "loss": 0.4759,
+      "step": 7177
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 37.348664213916706,
+      "learning_rate": 3.982275602820593e-06,
+      "loss": 0.4478,
+      "step": 7178
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.0632881688050695,
+      "learning_rate": 3.980998195278028e-06,
+      "loss": 0.4225,
+      "step": 7179
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.988064082607138,
+      "learning_rate": 3.979720857124985e-06,
+      "loss": 0.4503,
+      "step": 7180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.5023995744623955,
+      "learning_rate": 3.978443588448435e-06,
+      "loss": 0.5554,
+      "step": 7181
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 14.880077255034731,
+      "learning_rate": 3.977166389335364e-06,
+      "loss": 0.5162,
+      "step": 7182
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.821054394480333,
+      "learning_rate": 3.9758892598727375e-06,
+      "loss": 0.4383,
+      "step": 7183
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.408536086120154,
+      "learning_rate": 3.974612200147524e-06,
+      "loss": 0.3554,
+      "step": 7184
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.318091194555503,
+      "learning_rate": 3.9733352102466875e-06,
+      "loss": 0.4793,
+      "step": 7185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.413499710971076,
+      "learning_rate": 3.972058290257183e-06,
+      "loss": 0.4723,
+      "step": 7186
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.654467054011796,
+      "learning_rate": 3.970781440265964e-06,
+      "loss": 0.3985,
+      "step": 7187
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.28057725739517,
+      "learning_rate": 3.969504660359977e-06,
+      "loss": 0.5235,
+      "step": 7188
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.608494544151079,
+      "learning_rate": 3.9682279506261695e-06,
+      "loss": 0.4067,
+      "step": 7189
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.059106501105328,
+      "learning_rate": 3.966951311151474e-06,
+      "loss": 0.3274,
+      "step": 7190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.2303669891954945,
+      "learning_rate": 3.965674742022827e-06,
+      "loss": 0.5648,
+      "step": 7191
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.276448737739464,
+      "learning_rate": 3.964398243327157e-06,
+      "loss": 0.4022,
+      "step": 7192
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.602023800862072,
+      "learning_rate": 3.963121815151389e-06,
+      "loss": 0.4391,
+      "step": 7193
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.584851457045087,
+      "learning_rate": 3.961845457582439e-06,
+      "loss": 0.4365,
+      "step": 7194
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.34231398329217,
+      "learning_rate": 3.960569170707224e-06,
+      "loss": 0.3849,
+      "step": 7195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.969879605234271,
+      "learning_rate": 3.959292954612654e-06,
+      "loss": 0.711,
+      "step": 7196
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.77614132572477,
+      "learning_rate": 3.958016809385628e-06,
+      "loss": 0.4436,
+      "step": 7197
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.171516985180822,
+      "learning_rate": 3.956740735113056e-06,
+      "loss": 0.3589,
+      "step": 7198
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.76498601107341,
+      "learning_rate": 3.955464731881823e-06,
+      "loss": 0.5679,
+      "step": 7199
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.478893943904831,
+      "learning_rate": 3.9541887997788255e-06,
+      "loss": 0.4145,
+      "step": 7200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.567486633434086,
+      "learning_rate": 3.952912938890946e-06,
+      "loss": 0.4477,
+      "step": 7201
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.427893659767265,
+      "learning_rate": 3.951637149305066e-06,
+      "loss": 0.3974,
+      "step": 7202
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 24.292349058970995,
+      "learning_rate": 3.95036143110806e-06,
+      "loss": 0.5242,
+      "step": 7203
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.89615937348217,
+      "learning_rate": 3.949085784386799e-06,
+      "loss": 0.5373,
+      "step": 7204
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.6137509868032645,
+      "learning_rate": 3.947810209228152e-06,
+      "loss": 0.4313,
+      "step": 7205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.310598793204418,
+      "learning_rate": 3.946534705718975e-06,
+      "loss": 0.4972,
+      "step": 7206
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.564725383073237,
+      "learning_rate": 3.94525927394613e-06,
+      "loss": 0.5059,
+      "step": 7207
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.735316938857615,
+      "learning_rate": 3.943983913996462e-06,
+      "loss": 0.4791,
+      "step": 7208
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.240524944334,
+      "learning_rate": 3.942708625956822e-06,
+      "loss": 0.4596,
+      "step": 7209
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8747779405993483,
+      "learning_rate": 3.941433409914048e-06,
+      "loss": 0.724,
+      "step": 7210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 21.97255018847919,
+      "learning_rate": 3.94015826595498e-06,
+      "loss": 0.4509,
+      "step": 7211
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.963603497078044,
+      "learning_rate": 3.938883194166447e-06,
+      "loss": 0.4261,
+      "step": 7212
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.818494404482227,
+      "learning_rate": 3.937608194635278e-06,
+      "loss": 0.4061,
+      "step": 7213
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.535519191638733,
+      "learning_rate": 3.936333267448293e-06,
+      "loss": 0.4479,
+      "step": 7214
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.763801401868658,
+      "learning_rate": 3.935058412692308e-06,
+      "loss": 0.4336,
+      "step": 7215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.100804017772587,
+      "learning_rate": 3.933783630454139e-06,
+      "loss": 0.82,
+      "step": 7216
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.494569068564395,
+      "learning_rate": 3.932508920820589e-06,
+      "loss": 0.451,
+      "step": 7217
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.71872070987044,
+      "learning_rate": 3.931234283878463e-06,
+      "loss": 0.5038,
+      "step": 7218
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.068890382174926,
+      "learning_rate": 3.929959719714555e-06,
+      "loss": 0.4314,
+      "step": 7219
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.573732187325161,
+      "learning_rate": 3.928685228415661e-06,
+      "loss": 0.4631,
+      "step": 7220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.599733396312866,
+      "learning_rate": 3.927410810068564e-06,
+      "loss": 0.4844,
+      "step": 7221
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.112968842050028,
+      "learning_rate": 3.92613646476005e-06,
+      "loss": 0.4571,
+      "step": 7222
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.159781608083085,
+      "learning_rate": 3.924862192576893e-06,
+      "loss": 0.4842,
+      "step": 7223
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.952691535903017,
+      "learning_rate": 3.923587993605867e-06,
+      "loss": 0.4765,
+      "step": 7224
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.859315607073104,
+      "learning_rate": 3.922313867933742e-06,
+      "loss": 0.5035,
+      "step": 7225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.273601489687604,
+      "learning_rate": 3.921039815647276e-06,
+      "loss": 0.5352,
+      "step": 7226
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.242320544667261,
+      "learning_rate": 3.919765836833228e-06,
+      "loss": 0.4926,
+      "step": 7227
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.791679879800512,
+      "learning_rate": 3.91849193157835e-06,
+      "loss": 0.503,
+      "step": 7228
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.7295575028767685,
+      "learning_rate": 3.917218099969391e-06,
+      "loss": 0.372,
+      "step": 7229
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.19524786715882,
+      "learning_rate": 3.91594434209309e-06,
+      "loss": 0.4796,
+      "step": 7230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.527948966493679,
+      "learning_rate": 3.914670658036188e-06,
+      "loss": 0.5108,
+      "step": 7231
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 15.977061109537445,
+      "learning_rate": 3.913397047885414e-06,
+      "loss": 0.4931,
+      "step": 7232
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.911948079889376,
+      "learning_rate": 3.9121235117274984e-06,
+      "loss": 0.5461,
+      "step": 7233
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7704395138470437,
+      "learning_rate": 3.910850049649159e-06,
+      "loss": 0.7365,
+      "step": 7234
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.05647280525651,
+      "learning_rate": 3.909576661737117e-06,
+      "loss": 0.5907,
+      "step": 7235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 19.50937788243528,
+      "learning_rate": 3.908303348078084e-06,
+      "loss": 0.46,
+      "step": 7236
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.562244209326464,
+      "learning_rate": 3.907030108758765e-06,
+      "loss": 0.4882,
+      "step": 7237
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.961320576903104,
+      "learning_rate": 3.9057569438658635e-06,
+      "loss": 0.5624,
+      "step": 7238
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.269404379601754,
+      "learning_rate": 3.904483853486076e-06,
+      "loss": 0.3259,
+      "step": 7239
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.062614731833063,
+      "learning_rate": 3.903210837706095e-06,
+      "loss": 0.3739,
+      "step": 7240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.379237650235793,
+      "learning_rate": 3.901937896612604e-06,
+      "loss": 0.7836,
+      "step": 7241
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.07014408934973,
+      "learning_rate": 3.90066503029229e-06,
+      "loss": 0.4163,
+      "step": 7242
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.524612872271065,
+      "learning_rate": 3.899392238831825e-06,
+      "loss": 0.5051,
+      "step": 7243
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.204467806819112,
+      "learning_rate": 3.898119522317882e-06,
+      "loss": 0.5009,
+      "step": 7244
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.634657801418289,
+      "learning_rate": 3.896846880837129e-06,
+      "loss": 0.4581,
+      "step": 7245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.330911571880394,
+      "learning_rate": 3.895574314476225e-06,
+      "loss": 0.621,
+      "step": 7246
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.177936151480273,
+      "learning_rate": 3.894301823321828e-06,
+      "loss": 0.5153,
+      "step": 7247
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.692319794358824,
+      "learning_rate": 3.893029407460587e-06,
+      "loss": 0.4762,
+      "step": 7248
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.498582107280188,
+      "learning_rate": 3.89175706697915e-06,
+      "loss": 0.6574,
+      "step": 7249
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.841386500849422,
+      "learning_rate": 3.890484801964155e-06,
+      "loss": 0.5273,
+      "step": 7250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.653469063652561,
+      "learning_rate": 3.889212612502243e-06,
+      "loss": 0.4641,
+      "step": 7251
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.6458221544998075,
+      "learning_rate": 3.887940498680036e-06,
+      "loss": 0.4628,
+      "step": 7252
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.937051242073952,
+      "learning_rate": 3.886668460584166e-06,
+      "loss": 0.3326,
+      "step": 7253
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.670982507348349,
+      "learning_rate": 3.885396498301251e-06,
+      "loss": 0.574,
+      "step": 7254
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8064962035472172,
+      "learning_rate": 3.884124611917907e-06,
+      "loss": 0.7007,
+      "step": 7255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.253247202983252,
+      "learning_rate": 3.882852801520744e-06,
+      "loss": 0.5107,
+      "step": 7256
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.524699381737873,
+      "learning_rate": 3.8815810671963645e-06,
+      "loss": 0.5044,
+      "step": 7257
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.379800145150051,
+      "learning_rate": 3.8803094090313695e-06,
+      "loss": 0.4805,
+      "step": 7258
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.414470393390935,
+      "learning_rate": 3.879037827112353e-06,
+      "loss": 0.4342,
+      "step": 7259
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 20.458490732675358,
+      "learning_rate": 3.8777663215259055e-06,
+      "loss": 0.4276,
+      "step": 7260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.445664695671052,
+      "learning_rate": 3.876494892358606e-06,
+      "loss": 0.5342,
+      "step": 7261
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2867134000446345,
+      "learning_rate": 3.875223539697041e-06,
+      "loss": 0.7739,
+      "step": 7262
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.978312532559526,
+      "learning_rate": 3.8739522636277775e-06,
+      "loss": 0.431,
+      "step": 7263
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.7562198429275515,
+      "learning_rate": 3.872681064237385e-06,
+      "loss": 0.5064,
+      "step": 7264
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.404791500533959,
+      "learning_rate": 3.8714099416124305e-06,
+      "loss": 0.5592,
+      "step": 7265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.413248150560444,
+      "learning_rate": 3.870138895839468e-06,
+      "loss": 0.3998,
+      "step": 7266
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.917128830241344,
+      "learning_rate": 3.8688679270050514e-06,
+      "loss": 0.4171,
+      "step": 7267
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.624817229722678,
+      "learning_rate": 3.867597035195725e-06,
+      "loss": 0.4,
+      "step": 7268
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7473345257050865,
+      "learning_rate": 3.866326220498038e-06,
+      "loss": 0.7196,
+      "step": 7269
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.748647627475663,
+      "learning_rate": 3.8650554829985194e-06,
+      "loss": 0.523,
+      "step": 7270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.546894254445933,
+      "learning_rate": 3.863784822783706e-06,
+      "loss": 0.4909,
+      "step": 7271
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.033161553320825,
+      "learning_rate": 3.862514239940121e-06,
+      "loss": 0.3674,
+      "step": 7272
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.369917276068326,
+      "learning_rate": 3.861243734554289e-06,
+      "loss": 0.4416,
+      "step": 7273
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.75950748007049,
+      "learning_rate": 3.859973306712722e-06,
+      "loss": 0.6061,
+      "step": 7274
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.485796770054294,
+      "learning_rate": 3.8587029565019315e-06,
+      "loss": 0.4241,
+      "step": 7275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.061121682767835,
+      "learning_rate": 3.857432684008426e-06,
+      "loss": 0.3826,
+      "step": 7276
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.624671751173935,
+      "learning_rate": 3.8561624893187015e-06,
+      "loss": 0.4694,
+      "step": 7277
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.049756357779996,
+      "learning_rate": 3.854892372519256e-06,
+      "loss": 0.4849,
+      "step": 7278
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.908435293587693,
+      "learning_rate": 3.853622333696573e-06,
+      "loss": 0.4693,
+      "step": 7279
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.701663466391404,
+      "learning_rate": 3.852352372937143e-06,
+      "loss": 0.494,
+      "step": 7280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.901439317482915,
+      "learning_rate": 3.851082490327441e-06,
+      "loss": 0.4248,
+      "step": 7281
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9251585971985938,
+      "learning_rate": 3.849812685953943e-06,
+      "loss": 0.7418,
+      "step": 7282
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.69547154646504,
+      "learning_rate": 3.848542959903115e-06,
+      "loss": 0.5573,
+      "step": 7283
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.66514580130179,
+      "learning_rate": 3.84727331226142e-06,
+      "loss": 0.3711,
+      "step": 7284
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.796305231402611,
+      "learning_rate": 3.846003743115316e-06,
+      "loss": 0.4274,
+      "step": 7285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1875934829512698,
+      "learning_rate": 3.844734252551253e-06,
+      "loss": 0.8072,
+      "step": 7286
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.203658769368174,
+      "learning_rate": 3.8434648406556815e-06,
+      "loss": 0.5392,
+      "step": 7287
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 13.15313239106406,
+      "learning_rate": 3.842195507515039e-06,
+      "loss": 0.503,
+      "step": 7288
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.697840751208897,
+      "learning_rate": 3.840926253215766e-06,
+      "loss": 0.3989,
+      "step": 7289
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.441308661206985,
+      "learning_rate": 3.8396570778442885e-06,
+      "loss": 0.5503,
+      "step": 7290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.547973780537745,
+      "learning_rate": 3.838387981487035e-06,
+      "loss": 0.5033,
+      "step": 7291
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.78513247781961,
+      "learning_rate": 3.8371189642304224e-06,
+      "loss": 0.4225,
+      "step": 7292
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 22.765911379053964,
+      "learning_rate": 3.835850026160867e-06,
+      "loss": 0.3631,
+      "step": 7293
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.99209119509691,
+      "learning_rate": 3.834581167364778e-06,
+      "loss": 0.462,
+      "step": 7294
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.512960469784559,
+      "learning_rate": 3.833312387928557e-06,
+      "loss": 0.5078,
+      "step": 7295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.195587578109856,
+      "learning_rate": 3.832043687938606e-06,
+      "loss": 0.5088,
+      "step": 7296
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.639881717510344,
+      "learning_rate": 3.830775067481315e-06,
+      "loss": 0.5112,
+      "step": 7297
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.961366167807653,
+      "learning_rate": 3.829506526643073e-06,
+      "loss": 0.3941,
+      "step": 7298
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.906856102721312,
+      "learning_rate": 3.828238065510261e-06,
+      "loss": 0.5475,
+      "step": 7299
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.488442120651985,
+      "learning_rate": 3.8269696841692555e-06,
+      "loss": 0.5206,
+      "step": 7300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.973074454500591,
+      "learning_rate": 3.825701382706428e-06,
+      "loss": 0.3911,
+      "step": 7301
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.323422908567531,
+      "learning_rate": 3.8244331612081445e-06,
+      "loss": 0.4512,
+      "step": 7302
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.226002491015207,
+      "learning_rate": 3.823165019760765e-06,
+      "loss": 0.3429,
+      "step": 7303
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.564552703625982,
+      "learning_rate": 3.8218969584506435e-06,
+      "loss": 0.4661,
+      "step": 7304
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.9918396026751894,
+      "learning_rate": 3.82062897736413e-06,
+      "loss": 0.3842,
+      "step": 7305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.1999485038195665,
+      "learning_rate": 3.819361076587569e-06,
+      "loss": 0.5619,
+      "step": 7306
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.158235035603813,
+      "learning_rate": 3.8180932562072995e-06,
+      "loss": 0.4431,
+      "step": 7307
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.22225898793764,
+      "learning_rate": 3.816825516309653e-06,
+      "loss": 0.4635,
+      "step": 7308
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.342081766205611,
+      "learning_rate": 3.815557856980957e-06,
+      "loss": 0.4724,
+      "step": 7309
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.814870426141693,
+      "learning_rate": 3.814290278307534e-06,
+      "loss": 0.4103,
+      "step": 7310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.136115618779213,
+      "learning_rate": 3.813022780375702e-06,
+      "loss": 0.4978,
+      "step": 7311
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8217927127115519,
+      "learning_rate": 3.8117553632717676e-06,
+      "loss": 0.7537,
+      "step": 7312
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.906978146074952,
+      "learning_rate": 3.810488027082043e-06,
+      "loss": 0.512,
+      "step": 7313
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.290097713125281,
+      "learning_rate": 3.8092207718928205e-06,
+      "loss": 0.3091,
+      "step": 7314
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.4966943465474,
+      "learning_rate": 3.8079535977904e-06,
+      "loss": 0.4236,
+      "step": 7315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.576367097534572,
+      "learning_rate": 3.8066865048610696e-06,
+      "loss": 0.7177,
+      "step": 7316
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.301064801244509,
+      "learning_rate": 3.8054194931911113e-06,
+      "loss": 0.4559,
+      "step": 7317
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.206475473769652,
+      "learning_rate": 3.804152562866805e-06,
+      "loss": 0.5802,
+      "step": 7318
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.944904596482387,
+      "learning_rate": 3.80288571397442e-06,
+      "loss": 0.5012,
+      "step": 7319
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.32530338519761,
+      "learning_rate": 3.8016189466002262e-06,
+      "loss": 0.4419,
+      "step": 7320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.823351429624577,
+      "learning_rate": 3.800352260830482e-06,
+      "loss": 0.5033,
+      "step": 7321
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.235084601542364,
+      "learning_rate": 3.7990856567514472e-06,
+      "loss": 0.4143,
+      "step": 7322
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.816178858012112,
+      "learning_rate": 3.7978191344493663e-06,
+      "loss": 0.4691,
+      "step": 7323
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.508045979337489,
+      "learning_rate": 3.796552694010489e-06,
+      "loss": 0.4554,
+      "step": 7324
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.139591877288948,
+      "learning_rate": 3.795286335521052e-06,
+      "loss": 0.4367,
+      "step": 7325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.416396185621599,
+      "learning_rate": 3.7940200590672883e-06,
+      "loss": 0.4534,
+      "step": 7326
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.6126631689913244,
+      "learning_rate": 3.7927538647354273e-06,
+      "loss": 0.5003,
+      "step": 7327
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.85471014256461,
+      "learning_rate": 3.7914877526116896e-06,
+      "loss": 0.5897,
+      "step": 7328
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.095064563083409,
+      "learning_rate": 3.7902217227822933e-06,
+      "loss": 0.3921,
+      "step": 7329
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.172282145294008,
+      "learning_rate": 3.7889557753334466e-06,
+      "loss": 0.491,
+      "step": 7330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5312251932121133,
+      "learning_rate": 3.7876899103513605e-06,
+      "loss": 0.7161,
+      "step": 7331
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.479271906883591,
+      "learning_rate": 3.786424127922228e-06,
+      "loss": 0.5145,
+      "step": 7332
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7925231041809362,
+      "learning_rate": 3.7851584281322483e-06,
+      "loss": 0.7489,
+      "step": 7333
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.662276733371492,
+      "learning_rate": 3.7838928110676082e-06,
+      "loss": 0.5249,
+      "step": 7334
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.5857206516687565,
+      "learning_rate": 3.78262727681449e-06,
+      "loss": 0.466,
+      "step": 7335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.642733370019709,
+      "learning_rate": 3.7813618254590726e-06,
+      "loss": 0.3939,
+      "step": 7336
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.524942190469019,
+      "learning_rate": 3.7800964570875267e-06,
+      "loss": 0.5816,
+      "step": 7337
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.06315816620972,
+      "learning_rate": 3.778831171786018e-06,
+      "loss": 0.4491,
+      "step": 7338
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.365991960609869,
+      "learning_rate": 3.7775659696407056e-06,
+      "loss": 0.4541,
+      "step": 7339
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.281378702184767,
+      "learning_rate": 3.7763008507377493e-06,
+      "loss": 0.5104,
+      "step": 7340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.770216761350527,
+      "learning_rate": 3.775035815163291e-06,
+      "loss": 0.4518,
+      "step": 7341
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7786668870230495,
+      "learning_rate": 3.7737708630034797e-06,
+      "loss": 0.7676,
+      "step": 7342
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.250152008881778,
+      "learning_rate": 3.7725059943444496e-06,
+      "loss": 0.437,
+      "step": 7343
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.512185661508073,
+      "learning_rate": 3.7712412092723354e-06,
+      "loss": 0.4708,
+      "step": 7344
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.350792747695898,
+      "learning_rate": 3.769976507873261e-06,
+      "loss": 0.3447,
+      "step": 7345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 91.98625235062138,
+      "learning_rate": 3.7687118902333463e-06,
+      "loss": 0.4439,
+      "step": 7346
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.279133801070472,
+      "learning_rate": 3.7674473564387114e-06,
+      "loss": 0.4922,
+      "step": 7347
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 27.32269626899785,
+      "learning_rate": 3.7661829065754586e-06,
+      "loss": 0.365,
+      "step": 7348
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.887621632952568,
+      "learning_rate": 3.764918540729697e-06,
+      "loss": 0.4244,
+      "step": 7349
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.566753573296773,
+      "learning_rate": 3.76365425898752e-06,
+      "loss": 0.5162,
+      "step": 7350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.213510582890597,
+      "learning_rate": 3.7623900614350234e-06,
+      "loss": 0.5335,
+      "step": 7351
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.110442495201649,
+      "learning_rate": 3.7611259481582905e-06,
+      "loss": 0.5076,
+      "step": 7352
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.013834286714793,
+      "learning_rate": 3.7598619192434036e-06,
+      "loss": 0.5087,
+      "step": 7353
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 18.625034716080258,
+      "learning_rate": 3.7585979747764357e-06,
+      "loss": 0.4728,
+      "step": 7354
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.507494048697268,
+      "learning_rate": 3.7573341148434562e-06,
+      "loss": 0.5361,
+      "step": 7355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.743388373392637,
+      "learning_rate": 3.756070339530532e-06,
+      "loss": 0.478,
+      "step": 7356
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.306476512187217,
+      "learning_rate": 3.7548066489237144e-06,
+      "loss": 0.4154,
+      "step": 7357
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.337101849368334,
+      "learning_rate": 3.7535430431090607e-06,
+      "loss": 0.5238,
+      "step": 7358
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.389498553085149,
+      "learning_rate": 3.752279522172614e-06,
+      "loss": 0.5122,
+      "step": 7359
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.349327876579288,
+      "learning_rate": 3.751016086200415e-06,
+      "loss": 0.4935,
+      "step": 7360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.808971977125463,
+      "learning_rate": 3.749752735278498e-06,
+      "loss": 0.4685,
+      "step": 7361
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.793104352965912,
+      "learning_rate": 3.748489469492893e-06,
+      "loss": 0.5428,
+      "step": 7362
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.3397522354739975,
+      "learning_rate": 3.7472262889296207e-06,
+      "loss": 0.4263,
+      "step": 7363
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.585253376733599,
+      "learning_rate": 3.7459631936746994e-06,
+      "loss": 0.3882,
+      "step": 7364
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.40527668040794,
+      "learning_rate": 3.7447001838141395e-06,
+      "loss": 0.5891,
+      "step": 7365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.583586390814592,
+      "learning_rate": 3.7434372594339453e-06,
+      "loss": 0.5479,
+      "step": 7366
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.746673275610132,
+      "learning_rate": 3.74217442062012e-06,
+      "loss": 0.5611,
+      "step": 7367
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.757203443798999,
+      "learning_rate": 3.7409116674586545e-06,
+      "loss": 0.4399,
+      "step": 7368
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7056508664618253,
+      "learning_rate": 3.7396490000355377e-06,
+      "loss": 0.4707,
+      "step": 7369
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.705175763372979,
+      "learning_rate": 3.7383864184367515e-06,
+      "loss": 0.4265,
+      "step": 7370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.167671617865256,
+      "learning_rate": 3.7371239227482726e-06,
+      "loss": 0.5037,
+      "step": 7371
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.694487463665423,
+      "learning_rate": 3.7358615130560695e-06,
+      "loss": 0.4267,
+      "step": 7372
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.055037187351804,
+      "learning_rate": 3.73459918944611e-06,
+      "loss": 0.502,
+      "step": 7373
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.536771756156835,
+      "learning_rate": 3.7333369520043484e-06,
+      "loss": 0.5309,
+      "step": 7374
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.013112360650442,
+      "learning_rate": 3.7320748008167395e-06,
+      "loss": 0.5109,
+      "step": 7375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.119190199267433,
+      "learning_rate": 3.7308127359692323e-06,
+      "loss": 0.4199,
+      "step": 7376
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.968578941347398,
+      "learning_rate": 3.7295507575477663e-06,
+      "loss": 0.5669,
+      "step": 7377
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5869067478235435,
+      "learning_rate": 3.7282888656382767e-06,
+      "loss": 0.6999,
+      "step": 7378
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.26400603105241,
+      "learning_rate": 3.7270270603266913e-06,
+      "loss": 0.4227,
+      "step": 7379
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.933086421060262,
+      "learning_rate": 3.725765341698936e-06,
+      "loss": 0.5372,
+      "step": 7380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.262510714274333,
+      "learning_rate": 3.7245037098409253e-06,
+      "loss": 0.5327,
+      "step": 7381
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 20.480973672530613,
+      "learning_rate": 3.7232421648385743e-06,
+      "loss": 0.4034,
+      "step": 7382
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.5973136716844145,
+      "learning_rate": 3.721980706777784e-06,
+      "loss": 0.4054,
+      "step": 7383
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.416468501794649,
+      "learning_rate": 3.7207193357444604e-06,
+      "loss": 0.5346,
+      "step": 7384
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.580781320694183,
+      "learning_rate": 3.7194580518244905e-06,
+      "loss": 0.4167,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.875847739756315,
+      "learning_rate": 3.718196855103766e-06,
+      "loss": 0.5766,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.974247624702395,
+      "learning_rate": 3.71693574566817e-06,
+      "loss": 0.534,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.300944949804316,
+      "learning_rate": 3.7156747236035755e-06,
+      "loss": 0.4997,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.168534295295895,
+      "learning_rate": 3.714413788995855e-06,
+      "loss": 0.4704,
+      "step": 7389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.119245748717199,
+      "learning_rate": 3.71315294193087e-06,
+      "loss": 0.5249,
+      "step": 7390
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.756129702414361,
+      "learning_rate": 3.711892182494481e-06,
+      "loss": 0.5138,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.479711786608414,
+      "learning_rate": 3.7106315107725367e-06,
+      "loss": 0.4536,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.065367657564414,
+      "learning_rate": 3.709370926850889e-06,
+      "loss": 0.4689,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.258548611058635,
+      "learning_rate": 3.7081104308153713e-06,
+      "loss": 0.5445,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.709770345384588,
+      "learning_rate": 3.7068500227518224e-06,
+      "loss": 0.4647,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.013068435767742,
+      "learning_rate": 3.7055897027460707e-06,
+      "loss": 0.4344,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.150857597262302,
+      "learning_rate": 3.7043294708839354e-06,
+      "loss": 0.4606,
+      "step": 7397
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.192107405611473,
+      "learning_rate": 3.7030693272512363e-06,
+      "loss": 0.4233,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.7706808814384525,
+      "learning_rate": 3.70180927193378e-06,
+      "loss": 0.4294,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.370726275001276,
+      "learning_rate": 3.7005493050173735e-06,
+      "loss": 0.4606,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.822065956782146,
+      "learning_rate": 3.699289426587812e-06,
+      "loss": 0.4531,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.716433554369583,
+      "learning_rate": 3.698029636730892e-06,
+      "loss": 0.4261,
+      "step": 7402
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 22.335437958805702,
+      "learning_rate": 3.6967699355323947e-06,
+      "loss": 0.486,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.03942078463436,
+      "learning_rate": 3.695510323078105e-06,
+      "loss": 0.4432,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.921901992832587,
+      "learning_rate": 3.694250799453793e-06,
+      "loss": 0.5109,
+      "step": 7405
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.226797554220039,
+      "learning_rate": 3.692991364745228e-06,
+      "loss": 0.4703,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.900255441694452,
+      "learning_rate": 3.691732019038173e-06,
+      "loss": 0.4915,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.252229095939161,
+      "learning_rate": 3.6904727624183823e-06,
+      "loss": 0.5434,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.367925777604584,
+      "learning_rate": 3.689213594971609e-06,
+      "loss": 0.4928,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.476868135244702,
+      "learning_rate": 3.6879545167835905e-06,
+      "loss": 0.5661,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.165927794554301,
+      "learning_rate": 3.6866955279400717e-06,
+      "loss": 0.4746,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.680186188893391,
+      "learning_rate": 3.685436628526778e-06,
+      "loss": 0.5182,
+      "step": 7412
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.8422085539287805,
+      "learning_rate": 3.6841778186294395e-06,
+      "loss": 0.5041,
+      "step": 7413
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.226495546357743,
+      "learning_rate": 3.6829190983337724e-06,
+      "loss": 0.4937,
+      "step": 7414
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.126989883449955,
+      "learning_rate": 3.6816604677254923e-06,
+      "loss": 0.5123,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.5756591862267735,
+      "learning_rate": 3.6804019268903044e-06,
+      "loss": 0.4224,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.2904820315328145,
+      "learning_rate": 3.67914347591391e-06,
+      "loss": 0.5561,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.00917263955098,
+      "learning_rate": 3.6778851148820078e-06,
+      "loss": 0.4717,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.26632385398176,
+      "learning_rate": 3.6766268438802804e-06,
+      "loss": 0.5213,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.945520613873408,
+      "learning_rate": 3.675368662994416e-06,
+      "loss": 0.5187,
+      "step": 7420
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.094741851687447,
+      "learning_rate": 3.6741105723100872e-06,
+      "loss": 0.482,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.746362895569232,
+      "learning_rate": 3.6728525719129677e-06,
+      "loss": 0.5118,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.739169753601003,
+      "learning_rate": 3.671594661888719e-06,
+      "loss": 0.5186,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.756091089861759,
+      "learning_rate": 3.670336842323001e-06,
+      "loss": 0.5297,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.284269570164794,
+      "learning_rate": 3.669079113301464e-06,
+      "loss": 0.5493,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.51555235061726,
+      "learning_rate": 3.667821474909754e-06,
+      "loss": 0.484,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.512697355928559,
+      "learning_rate": 3.6665639272335146e-06,
+      "loss": 0.5166,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.045277954754571,
+      "learning_rate": 3.665306470358373e-06,
+      "loss": 0.5214,
+      "step": 7428
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.627235319327323,
+      "learning_rate": 3.664049104369961e-06,
+      "loss": 0.5335,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.224300587952531,
+      "learning_rate": 3.6627918293538978e-06,
+      "loss": 0.4464,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.437875170085032,
+      "learning_rate": 3.661534645395799e-06,
+      "loss": 0.4366,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.771701813576536,
+      "learning_rate": 3.6602775525812715e-06,
+      "loss": 0.4879,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.152703421313108,
+      "learning_rate": 3.6590205509959213e-06,
+      "loss": 0.512,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.75362661507737,
+      "learning_rate": 3.6577636407253416e-06,
+      "loss": 0.5087,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.234054306233589,
+      "learning_rate": 3.656506821855123e-06,
+      "loss": 0.4724,
+      "step": 7435
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.276120089708246,
+      "learning_rate": 3.655250094470849e-06,
+      "loss": 0.5218,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.6332576923909485,
+      "learning_rate": 3.653993458658097e-06,
+      "loss": 0.4185,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.086140553203543,
+      "learning_rate": 3.652736914502442e-06,
+      "loss": 0.3971,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.471953557653054,
+      "learning_rate": 3.651480462089444e-06,
+      "loss": 0.496,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.466617979433876,
+      "learning_rate": 3.650224101504666e-06,
+      "loss": 0.4019,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 39.676507758449524,
+      "learning_rate": 3.6489678328336574e-06,
+      "loss": 0.4344,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.867281859983907,
+      "learning_rate": 3.647711656161967e-06,
+      "loss": 0.6091,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.030814682765365,
+      "learning_rate": 3.6464555715751324e-06,
+      "loss": 0.3967,
+      "step": 7443
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.576997485824954,
+      "learning_rate": 3.6451995791586896e-06,
+      "loss": 0.4476,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.7392351753391235,
+      "learning_rate": 3.6439436789981646e-06,
+      "loss": 0.4057,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.466218056197062,
+      "learning_rate": 3.642687871179078e-06,
+      "loss": 0.4192,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.010652653168561,
+      "learning_rate": 3.6414321557869485e-06,
+      "loss": 0.4556,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.305949610953985,
+      "learning_rate": 3.6401765329072814e-06,
+      "loss": 0.4993,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.754064726320609,
+      "learning_rate": 3.6389210026255806e-06,
+      "loss": 0.4972,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.305709314683985,
+      "learning_rate": 3.637665565027341e-06,
+      "loss": 0.4832,
+      "step": 7450
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.514295991678427,
+      "learning_rate": 3.6364102201980535e-06,
+      "loss": 0.5446,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.683470138146784,
+      "learning_rate": 3.6351549682232e-06,
+      "loss": 0.509,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.743710205231424,
+      "learning_rate": 3.63389980918826e-06,
+      "loss": 0.3544,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.238901574843912,
+      "learning_rate": 3.6326447431787006e-06,
+      "loss": 0.4064,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9026759021838966,
+      "learning_rate": 3.6313897702799914e-06,
+      "loss": 0.7288,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.30211483366438,
+      "learning_rate": 3.6301348905775845e-06,
+      "loss": 0.4321,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.692536426812503,
+      "learning_rate": 3.6288801041569357e-06,
+      "loss": 0.5216,
+      "step": 7457
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.046245686931389,
+      "learning_rate": 3.6276254111034907e-06,
+      "loss": 0.4378,
+      "step": 7458
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.731860508933889,
+      "learning_rate": 3.6263708115026863e-06,
+      "loss": 0.5469,
+      "step": 7459
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7695158259213912,
+      "learning_rate": 3.6251163054399574e-06,
+      "loss": 0.7212,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.260902517222522,
+      "learning_rate": 3.6238618930007275e-06,
+      "loss": 0.5998,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.045602654550137,
+      "learning_rate": 3.62260757427042e-06,
+      "loss": 0.5687,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.527310395546447,
+      "learning_rate": 3.621353349334443e-06,
+      "loss": 0.4154,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.828243543425439,
+      "learning_rate": 3.6200992182782124e-06,
+      "loss": 0.3882,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.663108307153756,
+      "learning_rate": 3.6188451811871196e-06,
+      "loss": 0.5425,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.211041957069725,
+      "learning_rate": 3.6175912381465643e-06,
+      "loss": 0.4622,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 10.119618480123837,
+      "learning_rate": 3.616337389241935e-06,
+      "loss": 0.3505,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.283269475842635,
+      "learning_rate": 3.61508363455861e-06,
+      "loss": 0.4619,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.763526444985668,
+      "learning_rate": 3.613829974181967e-06,
+      "loss": 0.4525,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.5244301800185305,
+      "learning_rate": 3.6125764081973734e-06,
+      "loss": 0.4583,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.1307501317039215,
+      "learning_rate": 3.6113229366901936e-06,
+      "loss": 0.4529,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.067588943255073,
+      "learning_rate": 3.6100695597457792e-06,
+      "loss": 0.499,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.666391937530504,
+      "learning_rate": 3.608816277449485e-06,
+      "loss": 0.4926,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.227850539667299,
+      "learning_rate": 3.607563089886648e-06,
+      "loss": 0.4906,
+      "step": 7474
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.244567368165772,
+      "learning_rate": 3.606309997142611e-06,
+      "loss": 0.5078,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.376392603081888,
+      "learning_rate": 3.605056999302699e-06,
+      "loss": 0.4227,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.965277541083013,
+      "learning_rate": 3.6038040964522382e-06,
+      "loss": 0.4088,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.153343962536004,
+      "learning_rate": 3.6025512886765455e-06,
+      "loss": 0.5334,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.365342621383148,
+      "learning_rate": 3.6012985760609305e-06,
+      "loss": 0.4022,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.384286549804342,
+      "learning_rate": 3.6000459586907006e-06,
+      "loss": 0.4664,
+      "step": 7480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.089118720409741,
+      "learning_rate": 3.5987934366511478e-06,
+      "loss": 0.4367,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.308668264865686,
+      "learning_rate": 3.59754101002757e-06,
+      "loss": 0.6005,
+      "step": 7482
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.8556552306638086,
+      "learning_rate": 3.5962886789052453e-06,
+      "loss": 0.5203,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.961918894766878,
+      "learning_rate": 3.595036443369457e-06,
+      "loss": 0.3864,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.716813372523942,
+      "learning_rate": 3.5937843035054742e-06,
+      "loss": 0.3809,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.550844553375053,
+      "learning_rate": 3.592532259398563e-06,
+      "loss": 0.3453,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.260334806710094,
+      "learning_rate": 3.591280311133982e-06,
+      "loss": 0.5903,
+      "step": 7487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.206276060486557,
+      "learning_rate": 3.5900284587969827e-06,
+      "loss": 0.4521,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.943047975567826,
+      "learning_rate": 3.5887767024728128e-06,
+      "loss": 0.4923,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.666004050400849,
+      "learning_rate": 3.587525042246707e-06,
+      "loss": 0.5494,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.750711798679176,
+      "learning_rate": 3.5862734782039033e-06,
+      "loss": 0.7329,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.624623883235363,
+      "learning_rate": 3.5850220104296236e-06,
+      "loss": 0.3827,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.336035589006358,
+      "learning_rate": 3.5837706390090897e-06,
+      "loss": 0.4534,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.671404253067525,
+      "learning_rate": 3.5825193640275124e-06,
+      "loss": 0.4538,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.541923589426026,
+      "learning_rate": 3.5812681855701e-06,
+      "loss": 0.5072,
+      "step": 7495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.8920192536082965,
+      "learning_rate": 3.580017103722051e-06,
+      "loss": 0.4536,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.337546927751287,
+      "learning_rate": 3.5787661185685574e-06,
+      "loss": 0.5411,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.86080209240804,
+      "learning_rate": 3.57751523019481e-06,
+      "loss": 0.3845,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.067382999799646,
+      "learning_rate": 3.5762644386859835e-06,
+      "loss": 0.4513,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 11.860155739952575,
+      "learning_rate": 3.5750137441272552e-06,
+      "loss": 0.4568,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.446538002659607,
+      "learning_rate": 3.5737631466037896e-06,
+      "loss": 0.4994,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.985411255790396,
+      "learning_rate": 3.5725126462007485e-06,
+      "loss": 0.442,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.683635342471126,
+      "learning_rate": 3.571262243003284e-06,
+      "loss": 0.4244,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.103526979156652,
+      "learning_rate": 3.5700119370965445e-06,
+      "loss": 0.4404,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6135531212931775,
+      "learning_rate": 3.568761728565668e-06,
+      "loss": 0.7256,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.605511680120488,
+      "learning_rate": 3.567511617495789e-06,
+      "loss": 0.5343,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.55623807378973,
+      "learning_rate": 3.5662616039720387e-06,
+      "loss": 0.4326,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.978668109436803,
+      "learning_rate": 3.5650116880795303e-06,
+      "loss": 0.5485,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.877400265453843,
+      "learning_rate": 3.5637618699033837e-06,
+      "loss": 0.5653,
+      "step": 7509
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.446712562134566,
+      "learning_rate": 3.562512149528702e-06,
+      "loss": 0.492,
+      "step": 7510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 44.565794438794875,
+      "learning_rate": 3.5612625270405884e-06,
+      "loss": 0.4917,
+      "step": 7511
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.689082719999748,
+      "learning_rate": 3.560013002524134e-06,
+      "loss": 0.5514,
+      "step": 7512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.736983639408752,
+      "learning_rate": 3.5587635760644284e-06,
+      "loss": 0.5228,
+      "step": 7513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.101618142941587,
+      "learning_rate": 3.5575142477465502e-06,
+      "loss": 0.5065,
+      "step": 7514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.992958786068263,
+      "learning_rate": 3.5562650176555745e-06,
+      "loss": 0.5159,
+      "step": 7515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.3881525910745625,
+      "learning_rate": 3.5550158858765658e-06,
+      "loss": 0.3737,
+      "step": 7516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.706251713015661,
+      "learning_rate": 3.553766852494588e-06,
+      "loss": 0.5373,
+      "step": 7517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.901931793445404,
+      "learning_rate": 3.5525179175946933e-06,
+      "loss": 0.5359,
+      "step": 7518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.8501631544192385,
+      "learning_rate": 3.5512690812619277e-06,
+      "loss": 0.5046,
+      "step": 7519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.82277969514147,
+      "learning_rate": 3.5500203435813328e-06,
+      "loss": 0.4488,
+      "step": 7520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.204429513145314,
+      "learning_rate": 3.5487717046379412e-06,
+      "loss": 0.5017,
+      "step": 7521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.35397344807425,
+      "learning_rate": 3.54752316451678e-06,
+      "loss": 0.4785,
+      "step": 7522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.700443922535468,
+      "learning_rate": 3.5462747233028695e-06,
+      "loss": 0.4165,
+      "step": 7523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.341857728662021,
+      "learning_rate": 3.5450263810812222e-06,
+      "loss": 0.5439,
+      "step": 7524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.4930132300235845,
+      "learning_rate": 3.543778137936844e-06,
+      "loss": 0.4715,
+      "step": 7525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.550424773172649,
+      "learning_rate": 3.5425299939547377e-06,
+      "loss": 0.4659,
+      "step": 7526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.787766636117513,
+      "learning_rate": 3.5412819492198948e-06,
+      "loss": 0.5602,
+      "step": 7527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.665203275825805,
+      "learning_rate": 3.540034003817301e-06,
+      "loss": 0.5561,
+      "step": 7528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.571489796464472,
+      "learning_rate": 3.538786157831937e-06,
+      "loss": 0.3924,
+      "step": 7529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.166335354539529,
+      "learning_rate": 3.5375384113487744e-06,
+      "loss": 0.5296,
+      "step": 7530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.6272955679576455,
+      "learning_rate": 3.53629076445278e-06,
+      "loss": 0.4302,
+      "step": 7531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.471264855734193,
+      "learning_rate": 3.5350432172289123e-06,
+      "loss": 0.5063,
+      "step": 7532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.870887997023521,
+      "learning_rate": 3.533795769762125e-06,
+      "loss": 0.5844,
+      "step": 7533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.464805581196544,
+      "learning_rate": 3.5325484221373614e-06,
+      "loss": 0.4266,
+      "step": 7534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.231251354154396,
+      "learning_rate": 3.5313011744395646e-06,
+      "loss": 0.6446,
+      "step": 7535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.830480894279239,
+      "learning_rate": 3.5300540267536613e-06,
+      "loss": 0.4326,
+      "step": 7536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 24.94770264214556,
+      "learning_rate": 3.5288069791645806e-06,
+      "loss": 0.5073,
+      "step": 7537
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.653672373604034,
+      "learning_rate": 3.5275600317572405e-06,
+      "loss": 0.48,
+      "step": 7538
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.28773191957685,
+      "learning_rate": 3.526313184616551e-06,
+      "loss": 0.4386,
+      "step": 7539
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.3136861471389265,
+      "learning_rate": 3.525066437827419e-06,
+      "loss": 0.4359,
+      "step": 7540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.5382479426618385,
+      "learning_rate": 3.5238197914747395e-06,
+      "loss": 0.4235,
+      "step": 7541
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.109530421333914,
+      "learning_rate": 3.522573245643407e-06,
+      "loss": 0.519,
+      "step": 7542
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.490834997196043,
+      "learning_rate": 3.521326800418302e-06,
+      "loss": 0.5317,
+      "step": 7543
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.819614544435379,
+      "learning_rate": 3.520080455884307e-06,
+      "loss": 0.4934,
+      "step": 7544
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.891333890728633,
+      "learning_rate": 3.5188342121262864e-06,
+      "loss": 0.5837,
+      "step": 7545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.150433179025753,
+      "learning_rate": 3.51758806922911e-06,
+      "loss": 0.5302,
+      "step": 7546
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.090700565279214,
+      "learning_rate": 3.5163420272776292e-06,
+      "loss": 0.5459,
+      "step": 7547
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.897579769086072,
+      "learning_rate": 3.5150960863566974e-06,
+      "loss": 0.7301,
+      "step": 7548
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.822854856319786,
+      "learning_rate": 3.5138502465511572e-06,
+      "loss": 0.5709,
+      "step": 7549
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.723217523894657,
+      "learning_rate": 3.5126045079458427e-06,
+      "loss": 0.4274,
+      "step": 7550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.013231355990178,
+      "learning_rate": 3.5113588706255872e-06,
+      "loss": 0.4201,
+      "step": 7551
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1893731643236043,
+      "learning_rate": 3.510113334675208e-06,
+      "loss": 0.7843,
+      "step": 7552
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.63661897826078,
+      "learning_rate": 3.5088679001795257e-06,
+      "loss": 0.4598,
+      "step": 7553
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.864930928731228,
+      "learning_rate": 3.5076225672233443e-06,
+      "loss": 0.4507,
+      "step": 7554
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.031799083074731,
+      "learning_rate": 3.5063773358914688e-06,
+      "loss": 0.482,
+      "step": 7555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.048320853447779,
+      "learning_rate": 3.5051322062686922e-06,
+      "loss": 0.6527,
+      "step": 7556
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.0092988464093215,
+      "learning_rate": 3.5038871784398027e-06,
+      "loss": 0.4868,
+      "step": 7557
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.914105625989051,
+      "learning_rate": 3.502642252489582e-06,
+      "loss": 0.5333,
+      "step": 7558
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.14229590883384,
+      "learning_rate": 3.5013974285028018e-06,
+      "loss": 0.5277,
+      "step": 7559
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.930015405413848,
+      "learning_rate": 3.5001527065642338e-06,
+      "loss": 0.4804,
+      "step": 7560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.23104074204308,
+      "learning_rate": 3.4989080867586323e-06,
+      "loss": 0.6083,
+      "step": 7561
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 24.39740122771101,
+      "learning_rate": 3.4976635691707545e-06,
+      "loss": 0.5297,
+      "step": 7562
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.673794535636566,
+      "learning_rate": 3.4964191538853443e-06,
+      "loss": 0.378,
+      "step": 7563
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.617950052596346,
+      "learning_rate": 3.495174840987142e-06,
+      "loss": 0.4945,
+      "step": 7564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.964260917693373,
+      "learning_rate": 3.49393063056088e-06,
+      "loss": 0.4019,
+      "step": 7565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.577863432960622,
+      "learning_rate": 3.492686522691283e-06,
+      "loss": 0.5018,
+      "step": 7566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.59901676458878,
+      "learning_rate": 3.4914425174630684e-06,
+      "loss": 0.4324,
+      "step": 7567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.636432728008058,
+      "learning_rate": 3.4901986149609478e-06,
+      "loss": 0.6197,
+      "step": 7568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.222861141049957,
+      "learning_rate": 3.4889548152696284e-06,
+      "loss": 0.4196,
+      "step": 7569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.700505961987778,
+      "learning_rate": 3.4877111184738022e-06,
+      "loss": 0.4865,
+      "step": 7570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6133785170828885,
+      "learning_rate": 3.486467524658165e-06,
+      "loss": 0.7766,
+      "step": 7571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.727266148318538,
+      "learning_rate": 3.485224033907395e-06,
+      "loss": 0.572,
+      "step": 7572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.189213721053443,
+      "learning_rate": 3.483980646306173e-06,
+      "loss": 0.5137,
+      "step": 7573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.49378231680557,
+      "learning_rate": 3.4827373619391647e-06,
+      "loss": 0.5249,
+      "step": 7574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.456768191714438,
+      "learning_rate": 3.481494180891034e-06,
+      "loss": 0.5345,
+      "step": 7575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.041365382347271,
+      "learning_rate": 3.4802511032464346e-06,
+      "loss": 0.5357,
+      "step": 7576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.069603875969934,
+      "learning_rate": 3.4790081290900153e-06,
+      "loss": 0.5185,
+      "step": 7577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.652341947279292,
+      "learning_rate": 3.47776525850642e-06,
+      "loss": 0.6143,
+      "step": 7578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.478298786154202,
+      "learning_rate": 3.476522491580277e-06,
+      "loss": 0.477,
+      "step": 7579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.693119812803568,
+      "learning_rate": 3.4752798283962184e-06,
+      "loss": 0.4413,
+      "step": 7580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.45037104567122,
+      "learning_rate": 3.474037269038861e-06,
+      "loss": 0.4358,
+      "step": 7581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.228985476170649,
+      "learning_rate": 3.4727948135928195e-06,
+      "loss": 0.4718,
+      "step": 7582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.434567088046845,
+      "learning_rate": 3.4715524621426976e-06,
+      "loss": 0.4665,
+      "step": 7583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.107213015582895,
+      "learning_rate": 3.470310214773097e-06,
+      "loss": 0.5001,
+      "step": 7584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.90468003003,
+      "learning_rate": 3.4690680715686055e-06,
+      "loss": 0.444,
+      "step": 7585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.679277827875397,
+      "learning_rate": 3.46782603261381e-06,
+      "loss": 0.4995,
+      "step": 7586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.909300026504049,
+      "learning_rate": 3.4665840979932865e-06,
+      "loss": 0.581,
+      "step": 7587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.212599884045568,
+      "learning_rate": 3.465342267791606e-06,
+      "loss": 0.4271,
+      "step": 7588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.469107379983786,
+      "learning_rate": 3.4641005420933325e-06,
+      "loss": 0.5043,
+      "step": 7589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.522957269918692,
+      "learning_rate": 3.462858920983021e-06,
+      "loss": 0.4956,
+      "step": 7590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.7005806108140105,
+      "learning_rate": 3.461617404545221e-06,
+      "loss": 0.3643,
+      "step": 7591
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.124451003395508,
+      "learning_rate": 3.460375992864473e-06,
+      "loss": 0.4772,
+      "step": 7592
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.72608289133537,
+      "learning_rate": 3.459134686025314e-06,
+      "loss": 0.3923,
+      "step": 7593
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.43764095777569,
+      "learning_rate": 3.457893484112268e-06,
+      "loss": 0.5003,
+      "step": 7594
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4880495043261177,
+      "learning_rate": 3.456652387209859e-06,
+      "loss": 0.6831,
+      "step": 7595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.36540334185213,
+      "learning_rate": 3.4554113954025964e-06,
+      "loss": 0.5054,
+      "step": 7596
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.856578242443108,
+      "learning_rate": 3.4541705087749895e-06,
+      "loss": 0.3975,
+      "step": 7597
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.3405419837075865,
+      "learning_rate": 3.452929727411537e-06,
+      "loss": 0.5065,
+      "step": 7598
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 17.00734924566977,
+      "learning_rate": 3.4516890513967288e-06,
+      "loss": 0.4925,
+      "step": 7599
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.520831057625048,
+      "learning_rate": 3.4504484808150517e-06,
+      "loss": 0.3866,
+      "step": 7600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.15169220256573,
+      "learning_rate": 3.4492080157509806e-06,
+      "loss": 0.442,
+      "step": 7601
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 18.82265722098901,
+      "learning_rate": 3.447967656288988e-06,
+      "loss": 0.5429,
+      "step": 7602
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.496973939791738,
+      "learning_rate": 3.4467274025135345e-06,
+      "loss": 0.4691,
+      "step": 7603
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 13.28361773974485,
+      "learning_rate": 3.4454872545090783e-06,
+      "loss": 0.4235,
+      "step": 7604
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.832572484720778,
+      "learning_rate": 3.4442472123600645e-06,
+      "loss": 0.465,
+      "step": 7605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.681704312235322,
+      "learning_rate": 3.4430072761509405e-06,
+      "loss": 0.5311,
+      "step": 7606
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.696922416059661,
+      "learning_rate": 3.441767445966133e-06,
+      "loss": 0.599,
+      "step": 7607
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.531213163850436,
+      "learning_rate": 3.4405277218900745e-06,
+      "loss": 0.4059,
+      "step": 7608
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.368773694472418,
+      "learning_rate": 3.439288104007184e-06,
+      "loss": 0.3491,
+      "step": 7609
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.860089427356618,
+      "learning_rate": 3.4380485924018714e-06,
+      "loss": 0.3344,
+      "step": 7610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.275122347123743,
+      "learning_rate": 3.4368091871585453e-06,
+      "loss": 0.4346,
+      "step": 7611
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.610867145675736,
+      "learning_rate": 3.4355698883616007e-06,
+      "loss": 0.4688,
+      "step": 7612
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.977480472565812,
+      "learning_rate": 3.4343306960954304e-06,
+      "loss": 0.5598,
+      "step": 7613
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.664396565149554,
+      "learning_rate": 3.433091610444416e-06,
+      "loss": 0.5345,
+      "step": 7614
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.447502113080331,
+      "learning_rate": 3.4318526314929367e-06,
+      "loss": 0.4728,
+      "step": 7615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.1594787559647,
+      "learning_rate": 3.430613759325357e-06,
+      "loss": 0.3615,
+      "step": 7616
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.638812060894448,
+      "learning_rate": 3.4293749940260423e-06,
+      "loss": 0.554,
+      "step": 7617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.486272659783456,
+      "learning_rate": 3.428136335679347e-06,
+      "loss": 0.5065,
+      "step": 7618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.3603125864406955,
+      "learning_rate": 3.426897784369617e-06,
+      "loss": 0.5459,
+      "step": 7619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.524102684014854,
+      "learning_rate": 3.4256593401811934e-06,
+      "loss": 0.4752,
+      "step": 7620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.1133525385152,
+      "learning_rate": 3.424421003198405e-06,
+      "loss": 0.4776,
+      "step": 7621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.947065916685764,
+      "learning_rate": 3.423182773505584e-06,
+      "loss": 0.3909,
+      "step": 7622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.671112402435932,
+      "learning_rate": 3.4219446511870402e-06,
+      "loss": 0.5532,
+      "step": 7623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.3675922243697505,
+      "learning_rate": 3.420706636327091e-06,
+      "loss": 0.4442,
+      "step": 7624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.9686447398842155,
+      "learning_rate": 3.419468729010036e-06,
+      "loss": 0.4867,
+      "step": 7625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6022269167082153,
+      "learning_rate": 3.4182309293201734e-06,
+      "loss": 0.6901,
+      "step": 7626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.296752200483715,
+      "learning_rate": 3.4169932373417896e-06,
+      "loss": 0.4359,
+      "step": 7627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.67008396621832,
+      "learning_rate": 3.415755653159167e-06,
+      "loss": 0.3733,
+      "step": 7628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.186590210409507,
+      "learning_rate": 3.4145181768565815e-06,
+      "loss": 0.3607,
+      "step": 7629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5974079357033313,
+      "learning_rate": 3.4132808085182955e-06,
+      "loss": 0.7071,
+      "step": 7630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.74360335577476,
+      "learning_rate": 3.412043548228574e-06,
+      "loss": 0.4424,
+      "step": 7631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.800432509648394,
+      "learning_rate": 3.410806396071662e-06,
+      "loss": 0.3664,
+      "step": 7632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.05613712763061,
+      "learning_rate": 3.4095693521318108e-06,
+      "loss": 0.5882,
+      "step": 7633
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.334053691986734,
+      "learning_rate": 3.408332416493253e-06,
+      "loss": 0.5048,
+      "step": 7634
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.304182007459312,
+      "learning_rate": 3.407095589240222e-06,
+      "loss": 0.4011,
+      "step": 7635
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.288588053476838,
+      "learning_rate": 3.4058588704569365e-06,
+      "loss": 0.5699,
+      "step": 7636
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.345336727067094,
+      "learning_rate": 3.404622260227614e-06,
+      "loss": 0.5818,
+      "step": 7637
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.415759278018301,
+      "learning_rate": 3.4033857586364628e-06,
+      "loss": 0.4254,
+      "step": 7638
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.837838216504222,
+      "learning_rate": 3.4021493657676798e-06,
+      "loss": 0.4373,
+      "step": 7639
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.552332054316338,
+      "learning_rate": 3.4009130817054635e-06,
+      "loss": 0.5274,
+      "step": 7640
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.299937456625819,
+      "learning_rate": 3.399676906533993e-06,
+      "loss": 0.4917,
+      "step": 7641
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 19.431675588084932,
+      "learning_rate": 3.398440840337451e-06,
+      "loss": 0.4257,
+      "step": 7642
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3667542984296126,
+      "learning_rate": 3.397204883200006e-06,
+      "loss": 0.6558,
+      "step": 7643
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.3234496479729385,
+      "learning_rate": 3.3959690352058224e-06,
+      "loss": 0.4797,
+      "step": 7644
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.9135082648296295,
+      "learning_rate": 3.3947332964390544e-06,
+      "loss": 0.3785,
+      "step": 7645
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.11832948705307,
+      "learning_rate": 3.393497666983852e-06,
+      "loss": 0.3991,
+      "step": 7646
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.215963821954308,
+      "learning_rate": 3.392262146924355e-06,
+      "loss": 0.4545,
+      "step": 7647
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.046052281594997,
+      "learning_rate": 3.3910267363446956e-06,
+      "loss": 0.4095,
+      "step": 7648
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.3056767270848395,
+      "learning_rate": 3.389791435329005e-06,
+      "loss": 0.6004,
+      "step": 7649
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.7228295452683,
+      "learning_rate": 3.388556243961395e-06,
+      "loss": 0.3928,
+      "step": 7650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9326039117451452,
+      "learning_rate": 3.3873211623259824e-06,
+      "loss": 0.7184,
+      "step": 7651
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.444416175462983,
+      "learning_rate": 3.3860861905068664e-06,
+      "loss": 0.573,
+      "step": 7652
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.379255279802031,
+      "learning_rate": 3.3848513285881467e-06,
+      "loss": 0.5244,
+      "step": 7653
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.948072361319973,
+      "learning_rate": 3.3836165766539087e-06,
+      "loss": 0.549,
+      "step": 7654
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.43081706600722,
+      "learning_rate": 3.3823819347882356e-06,
+      "loss": 0.4535,
+      "step": 7655
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.751952626286303,
+      "learning_rate": 3.3811474030752007e-06,
+      "loss": 0.4402,
+      "step": 7656
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.031786334764353,
+      "learning_rate": 3.3799129815988684e-06,
+      "loss": 0.5595,
+      "step": 7657
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8514846721451637,
+      "learning_rate": 3.3786786704433005e-06,
+      "loss": 0.4426,
+      "step": 7658
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.74969731652426,
+      "learning_rate": 3.3774444696925463e-06,
+      "loss": 0.5449,
+      "step": 7659
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.196733655691576,
+      "learning_rate": 3.3762103794306498e-06,
+      "loss": 0.4321,
+      "step": 7660
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 25.512959886412425,
+      "learning_rate": 3.3749763997416467e-06,
+      "loss": 0.4358,
+      "step": 7661
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.228502357333849,
+      "learning_rate": 3.3737425307095666e-06,
+      "loss": 0.575,
+      "step": 7662
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.9152854371562436,
+      "learning_rate": 3.3725087724184287e-06,
+      "loss": 0.4884,
+      "step": 7663
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.935325972061329,
+      "learning_rate": 3.3712751249522486e-06,
+      "loss": 0.4468,
+      "step": 7664
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.970134060250011,
+      "learning_rate": 3.37004158839503e-06,
+      "loss": 0.4711,
+      "step": 7665
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.052124579259806,
+      "learning_rate": 3.3688081628307733e-06,
+      "loss": 0.5644,
+      "step": 7666
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.512346131914251,
+      "learning_rate": 3.3675748483434668e-06,
+      "loss": 0.4383,
+      "step": 7667
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.288989685654814,
+      "learning_rate": 3.366341645017096e-06,
+      "loss": 0.5193,
+      "step": 7668
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.115494094655428,
+      "learning_rate": 3.3651085529356375e-06,
+      "loss": 0.4538,
+      "step": 7669
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.927486204602384,
+      "learning_rate": 3.363875572183056e-06,
+      "loss": 0.4372,
+      "step": 7670
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.36239492277303,
+      "learning_rate": 3.362642702843316e-06,
+      "loss": 0.5556,
+      "step": 7671
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.979496291016821,
+      "learning_rate": 3.3614099450003677e-06,
+      "loss": 0.4433,
+      "step": 7672
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.618851392793584,
+      "learning_rate": 3.3601772987381575e-06,
+      "loss": 0.4373,
+      "step": 7673
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.982927081377211,
+      "learning_rate": 3.358944764140622e-06,
+      "loss": 0.3918,
+      "step": 7674
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.431940296308849,
+      "learning_rate": 3.3577123412916934e-06,
+      "loss": 0.452,
+      "step": 7675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5716840756400785,
+      "learning_rate": 3.356480030275291e-06,
+      "loss": 0.7431,
+      "step": 7676
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.548941505409785,
+      "learning_rate": 3.355247831175332e-06,
+      "loss": 0.4625,
+      "step": 7677
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.147673302910272,
+      "learning_rate": 3.3540157440757255e-06,
+      "loss": 0.4532,
+      "step": 7678
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.17804906045863,
+      "learning_rate": 3.352783769060368e-06,
+      "loss": 0.4251,
+      "step": 7679
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.74838837608986,
+      "learning_rate": 3.351551906213153e-06,
+      "loss": 0.3733,
+      "step": 7680
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.600179996244236,
+      "learning_rate": 3.350320155617964e-06,
+      "loss": 0.4426,
+      "step": 7681
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.896197220900836,
+      "learning_rate": 3.34908851735868e-06,
+      "loss": 0.5239,
+      "step": 7682
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.7746743014089015,
+      "learning_rate": 3.347856991519166e-06,
+      "loss": 0.4203,
+      "step": 7683
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.441430908487329,
+      "learning_rate": 3.346625578183289e-06,
+      "loss": 0.3638,
+      "step": 7684
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.408026341240622,
+      "learning_rate": 3.345394277434897e-06,
+      "loss": 0.4219,
+      "step": 7685
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.855162644683744,
+      "learning_rate": 3.3441630893578426e-06,
+      "loss": 0.5344,
+      "step": 7686
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.126494374943352,
+      "learning_rate": 3.3429320140359566e-06,
+      "loss": 0.4578,
+      "step": 7687
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.925139156680336,
+      "learning_rate": 3.3417010515530746e-06,
+      "loss": 0.5373,
+      "step": 7688
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 17.804194461530606,
+      "learning_rate": 3.3404702019930206e-06,
+      "loss": 0.5278,
+      "step": 7689
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.990866793418542,
+      "learning_rate": 3.339239465439607e-06,
+      "loss": 0.5286,
+      "step": 7690
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.184769816218745,
+      "learning_rate": 3.3380088419766445e-06,
+      "loss": 0.4606,
+      "step": 7691
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.338849315219319,
+      "learning_rate": 3.3367783316879282e-06,
+      "loss": 0.4188,
+      "step": 7692
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.591162466399526,
+      "learning_rate": 3.3355479346572574e-06,
+      "loss": 0.4221,
+      "step": 7693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.830662314389164,
+      "learning_rate": 3.3343176509684092e-06,
+      "loss": 0.7633,
+      "step": 7694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.255711517472408,
+      "learning_rate": 3.333087480705167e-06,
+      "loss": 0.6044,
+      "step": 7695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.4554762689981215,
+      "learning_rate": 3.3318574239512964e-06,
+      "loss": 0.4435,
+      "step": 7696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.994822046553915,
+      "learning_rate": 3.3306274807905605e-06,
+      "loss": 0.4567,
+      "step": 7697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.926420548674851,
+      "learning_rate": 3.3293976513067105e-06,
+      "loss": 0.5413,
+      "step": 7698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.862593424551983,
+      "learning_rate": 3.3281679355834943e-06,
+      "loss": 0.4611,
+      "step": 7699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.050023036005137,
+      "learning_rate": 3.3269383337046513e-06,
+      "loss": 0.4557,
+      "step": 7700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.975249836956794,
+      "learning_rate": 3.3257088457539076e-06,
+      "loss": 0.5238,
+      "step": 7701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.207922321156639,
+      "learning_rate": 3.324479471814992e-06,
+      "loss": 0.5245,
+      "step": 7702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.778948830229256,
+      "learning_rate": 3.3232502119716124e-06,
+      "loss": 0.533,
+      "step": 7703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.38629995472189,
+      "learning_rate": 3.3220210663074824e-06,
+      "loss": 0.5508,
+      "step": 7704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.49473341095292,
+      "learning_rate": 3.3207920349062983e-06,
+      "loss": 0.4585,
+      "step": 7705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.905065783688508,
+      "learning_rate": 3.3195631178517516e-06,
+      "loss": 0.4826,
+      "step": 7706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.1879549043157205,
+      "learning_rate": 3.318334315227527e-06,
+      "loss": 0.5328,
+      "step": 7707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.007537024092107,
+      "learning_rate": 3.3171056271173e-06,
+      "loss": 0.5252,
+      "step": 7708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.985956557155703,
+      "learning_rate": 3.315877053604739e-06,
+      "loss": 0.3964,
+      "step": 7709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.761487045983302,
+      "learning_rate": 3.314648594773503e-06,
+      "loss": 0.5346,
+      "step": 7710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6605345876032152,
+      "learning_rate": 3.3134202507072487e-06,
+      "loss": 0.71,
+      "step": 7711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.757656605480588,
+      "learning_rate": 3.312192021489615e-06,
+      "loss": 0.5164,
+      "step": 7712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.138670882821769,
+      "learning_rate": 3.3109639072042444e-06,
+      "loss": 0.453,
+      "step": 7713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.644258935918099,
+      "learning_rate": 3.3097359079347624e-06,
+      "loss": 0.4855,
+      "step": 7714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.99165908645555,
+      "learning_rate": 3.308508023764794e-06,
+      "loss": 0.3836,
+      "step": 7715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.449530348631567,
+      "learning_rate": 3.3072802547779475e-06,
+      "loss": 0.471,
+      "step": 7716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.004776617619125,
+      "learning_rate": 3.306052601057833e-06,
+      "loss": 0.4475,
+      "step": 7717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.849740918275965,
+      "learning_rate": 3.3048250626880463e-06,
+      "loss": 0.5303,
+      "step": 7718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.981258448719044,
+      "learning_rate": 3.303597639752176e-06,
+      "loss": 0.4698,
+      "step": 7719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.488800011392672,
+      "learning_rate": 3.3023703323338084e-06,
+      "loss": 0.4509,
+      "step": 7720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.096631794583174,
+      "learning_rate": 3.301143140516513e-06,
+      "loss": 0.3779,
+      "step": 7721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.105467911290484,
+      "learning_rate": 3.2999160643838603e-06,
+      "loss": 0.4266,
+      "step": 7722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.165195017930071,
+      "learning_rate": 3.2986891040194057e-06,
+      "loss": 0.4358,
+      "step": 7723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.873789071579597,
+      "learning_rate": 3.2974622595067017e-06,
+      "loss": 0.4859,
+      "step": 7724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 11.270158467082306,
+      "learning_rate": 3.2962355309292894e-06,
+      "loss": 0.6238,
+      "step": 7725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.878515138578374,
+      "learning_rate": 3.2950089183707046e-06,
+      "loss": 0.5088,
+      "step": 7726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.944295836983655,
+      "learning_rate": 3.293782421914474e-06,
+      "loss": 0.4657,
+      "step": 7727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.411926986695751,
+      "learning_rate": 3.2925560416441145e-06,
+      "loss": 0.4353,
+      "step": 7728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.547256679880388,
+      "learning_rate": 3.291329777643142e-06,
+      "loss": 0.401,
+      "step": 7729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.3271348357476995,
+      "learning_rate": 3.290103629995055e-06,
+      "loss": 0.4842,
+      "step": 7730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.881068982689246,
+      "learning_rate": 3.288877598783352e-06,
+      "loss": 0.5319,
+      "step": 7731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.927258433958593,
+      "learning_rate": 3.287651684091517e-06,
+      "loss": 0.3663,
+      "step": 7732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.602621224144362,
+      "learning_rate": 3.286425886003034e-06,
+      "loss": 0.5049,
+      "step": 7733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.899030307931882,
+      "learning_rate": 3.28520020460137e-06,
+      "loss": 0.5726,
+      "step": 7734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.9693669166488,
+      "learning_rate": 3.28397463996999e-06,
+      "loss": 0.4642,
+      "step": 7735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.3420620582138385,
+      "learning_rate": 3.28274919219235e-06,
+      "loss": 0.4214,
+      "step": 7736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.080745841625166,
+      "learning_rate": 3.281523861351897e-06,
+      "loss": 0.4608,
+      "step": 7737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.0593976645957595,
+      "learning_rate": 3.280298647532071e-06,
+      "loss": 0.4049,
+      "step": 7738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 18.6062755029406,
+      "learning_rate": 3.279073550816303e-06,
+      "loss": 0.5145,
+      "step": 7739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.543251987949612,
+      "learning_rate": 3.277848571288018e-06,
+      "loss": 0.4371,
+      "step": 7740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.812723249123974,
+      "learning_rate": 3.276623709030631e-06,
+      "loss": 0.4075,
+      "step": 7741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.869888409440074,
+      "learning_rate": 3.2753989641275497e-06,
+      "loss": 0.4606,
+      "step": 7742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.504024991483567,
+      "learning_rate": 3.2741743366621727e-06,
+      "loss": 0.5543,
+      "step": 7743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.13556308281986,
+      "learning_rate": 3.272949826717894e-06,
+      "loss": 0.4571,
+      "step": 7744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.617826131489274,
+      "learning_rate": 3.271725434378095e-06,
+      "loss": 0.3947,
+      "step": 7745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.0915810079517625,
+      "learning_rate": 3.270501159726153e-06,
+      "loss": 0.5765,
+      "step": 7746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.719352719327806,
+      "learning_rate": 3.2692770028454333e-06,
+      "loss": 0.4913,
+      "step": 7747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.808180386300823,
+      "learning_rate": 3.268052963819298e-06,
+      "loss": 0.4841,
+      "step": 7748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.843862668495189,
+      "learning_rate": 3.2668290427311e-06,
+      "loss": 0.6131,
+      "step": 7749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.441733890954794,
+      "learning_rate": 3.2656052396641795e-06,
+      "loss": 0.5038,
+      "step": 7750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 13.567262050862778,
+      "learning_rate": 3.2643815547018755e-06,
+      "loss": 0.5146,
+      "step": 7751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.190501505490385,
+      "learning_rate": 3.2631579879275118e-06,
+      "loss": 0.5661,
+      "step": 7752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 29.700400115366776,
+      "learning_rate": 3.2619345394244106e-06,
+      "loss": 0.4284,
+      "step": 7753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.576730303285071,
+      "learning_rate": 3.2607112092758806e-06,
+      "loss": 0.5474,
+      "step": 7754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.702320651367252,
+      "learning_rate": 3.2594879975652304e-06,
+      "loss": 0.5347,
+      "step": 7755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.079145050689527,
+      "learning_rate": 3.2582649043757493e-06,
+      "loss": 0.5283,
+      "step": 7756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.128480552166279,
+      "learning_rate": 3.2570419297907297e-06,
+      "loss": 0.5239,
+      "step": 7757
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.621647483763743,
+      "learning_rate": 3.255819073893445e-06,
+      "loss": 0.5902,
+      "step": 7758
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.065644243423375,
+      "learning_rate": 3.254596336767171e-06,
+      "loss": 0.4554,
+      "step": 7759
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.383382266351018,
+      "learning_rate": 3.25337371849517e-06,
+      "loss": 0.5597,
+      "step": 7760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.439908162259382,
+      "learning_rate": 3.252151219160695e-06,
+      "loss": 0.5262,
+      "step": 7761
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.233641121428627,
+      "learning_rate": 3.2509288388469955e-06,
+      "loss": 0.45,
+      "step": 7762
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.171491054722026,
+      "learning_rate": 3.2497065776373068e-06,
+      "loss": 0.3994,
+      "step": 7763
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.964299149325261,
+      "learning_rate": 3.2484844356148636e-06,
+      "loss": 0.4719,
+      "step": 7764
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.771449265243349,
+      "learning_rate": 3.2472624128628833e-06,
+      "loss": 0.484,
+      "step": 7765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.871414085249174,
+      "learning_rate": 3.246040509464585e-06,
+      "loss": 0.5277,
+      "step": 7766
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.4495352442535445,
+      "learning_rate": 3.2448187255031718e-06,
+      "loss": 0.458,
+      "step": 7767
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.6092579837305845,
+      "learning_rate": 3.2435970610618436e-06,
+      "loss": 0.5199,
+      "step": 7768
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 12.675857545952944,
+      "learning_rate": 3.24237551622379e-06,
+      "loss": 0.4482,
+      "step": 7769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.487635268119066,
+      "learning_rate": 3.2411540910721918e-06,
+      "loss": 0.4108,
+      "step": 7770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.623053213788892,
+      "learning_rate": 3.239932785690224e-06,
+      "loss": 0.4379,
+      "step": 7771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.056608100369023,
+      "learning_rate": 3.2387116001610495e-06,
+      "loss": 0.5495,
+      "step": 7772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.921197117028073,
+      "learning_rate": 3.237490534567831e-06,
+      "loss": 0.4934,
+      "step": 7773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.788144989437217,
+      "learning_rate": 3.236269588993711e-06,
+      "loss": 0.3811,
+      "step": 7774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.817409866902729,
+      "learning_rate": 3.2350487635218353e-06,
+      "loss": 0.3431,
+      "step": 7775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.494267242997826,
+      "learning_rate": 3.2338280582353348e-06,
+      "loss": 0.4247,
+      "step": 7776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.351611357900573,
+      "learning_rate": 3.232607473217335e-06,
+      "loss": 0.4347,
+      "step": 7777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.3160278142468735,
+      "learning_rate": 3.2313870085509513e-06,
+      "loss": 0.496,
+      "step": 7778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.7199187353208885,
+      "learning_rate": 3.230166664319292e-06,
+      "loss": 0.5116,
+      "step": 7779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.483922147719134,
+      "learning_rate": 3.228946440605459e-06,
+      "loss": 0.3956,
+      "step": 7780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.363445424168996,
+      "learning_rate": 3.2277263374925405e-06,
+      "loss": 0.5608,
+      "step": 7781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.530946234179869,
+      "learning_rate": 3.2265063550636255e-06,
+      "loss": 0.4356,
+      "step": 7782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.697925627821314,
+      "learning_rate": 3.2252864934017836e-06,
+      "loss": 0.5195,
+      "step": 7783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.640593772635028,
+      "learning_rate": 3.2240667525900858e-06,
+      "loss": 0.518,
+      "step": 7784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7852584213490477,
+      "learning_rate": 3.2228471327115896e-06,
+      "loss": 0.7642,
+      "step": 7785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.997441675660333,
+      "learning_rate": 3.2216276338493474e-06,
+      "loss": 0.5067,
+      "step": 7786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.967147523889293,
+      "learning_rate": 3.220408256086399e-06,
+      "loss": 0.4166,
+      "step": 7787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.860895313468932,
+      "learning_rate": 3.219188999505779e-06,
+      "loss": 0.4138,
+      "step": 7788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.667837543450829,
+      "learning_rate": 3.2179698641905167e-06,
+      "loss": 0.4435,
+      "step": 7789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.96880126607447,
+      "learning_rate": 3.2167508502236257e-06,
+      "loss": 0.4359,
+      "step": 7790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.028810923118485,
+      "learning_rate": 3.2155319576881183e-06,
+      "loss": 0.4427,
+      "step": 7791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.282096001156605,
+      "learning_rate": 3.2143131866669936e-06,
+      "loss": 0.5887,
+      "step": 7792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.672193002640967,
+      "learning_rate": 3.2130945372432464e-06,
+      "loss": 0.4904,
+      "step": 7793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.669776945373156,
+      "learning_rate": 3.21187600949986e-06,
+      "loss": 0.4726,
+      "step": 7794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.107813098852066,
+      "learning_rate": 3.2106576035198123e-06,
+      "loss": 0.5275,
+      "step": 7795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.385738459997461,
+      "learning_rate": 3.2094393193860697e-06,
+      "loss": 0.4778,
+      "step": 7796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.966663628455039,
+      "learning_rate": 3.2082211571815925e-06,
+      "loss": 0.4966,
+      "step": 7797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.9787058572641,
+      "learning_rate": 3.207003116989332e-06,
+      "loss": 0.4478,
+      "step": 7798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7034670801828136,
+      "learning_rate": 3.20578519889223e-06,
+      "loss": 0.8119,
+      "step": 7799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.228369630249462,
+      "learning_rate": 3.2045674029732255e-06,
+      "loss": 0.4665,
+      "step": 7800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.303196355905041,
+      "learning_rate": 3.203349729315241e-06,
+      "loss": 0.522,
+      "step": 7801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6397479196774853,
+      "learning_rate": 3.202132178001197e-06,
+      "loss": 0.7383,
+      "step": 7802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.842140571883079,
+      "learning_rate": 3.2009147491140023e-06,
+      "loss": 0.5697,
+      "step": 7803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.429327465105057,
+      "learning_rate": 3.199697442736559e-06,
+      "loss": 0.4787,
+      "step": 7804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.531187942937047,
+      "learning_rate": 3.198480258951759e-06,
+      "loss": 0.3583,
+      "step": 7805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.593689213038509,
+      "learning_rate": 3.197263197842489e-06,
+      "loss": 0.4635,
+      "step": 7806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.494505575573879,
+      "learning_rate": 3.196046259491623e-06,
+      "loss": 0.4257,
+      "step": 7807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.744069445902709,
+      "learning_rate": 3.1948294439820306e-06,
+      "loss": 0.3499,
+      "step": 7808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.8738128255874935,
+      "learning_rate": 3.193612751396573e-06,
+      "loss": 0.5095,
+      "step": 7809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.220236268294367,
+      "learning_rate": 3.1923961818180993e-06,
+      "loss": 0.4799,
+      "step": 7810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.268952214363767,
+      "learning_rate": 3.191179735329454e-06,
+      "loss": 0.4363,
+      "step": 7811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.5310972189846,
+      "learning_rate": 3.1899634120134706e-06,
+      "loss": 0.5778,
+      "step": 7812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.573149832372946,
+      "learning_rate": 3.1887472119529768e-06,
+      "loss": 0.5353,
+      "step": 7813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.5773804733806465,
+      "learning_rate": 3.187531135230788e-06,
+      "loss": 0.448,
+      "step": 7814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.438400922994738,
+      "learning_rate": 3.1863151819297167e-06,
+      "loss": 0.5357,
+      "step": 7815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.950692488676776,
+      "learning_rate": 3.185099352132561e-06,
+      "loss": 0.4404,
+      "step": 7816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.58427612039255,
+      "learning_rate": 3.183883645922116e-06,
+      "loss": 0.686,
+      "step": 7817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.891281342587414,
+      "learning_rate": 3.1826680633811635e-06,
+      "loss": 0.5257,
+      "step": 7818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.849560064157923,
+      "learning_rate": 3.181452604592481e-06,
+      "loss": 0.4917,
+      "step": 7819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.674153492711044,
+      "learning_rate": 3.180237269638836e-06,
+      "loss": 0.4654,
+      "step": 7820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9030348692540304,
+      "learning_rate": 3.179022058602987e-06,
+      "loss": 0.4226,
+      "step": 7821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.826209231959457,
+      "learning_rate": 3.177806971567685e-06,
+      "loss": 0.5103,
+      "step": 7822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.963770145603646,
+      "learning_rate": 3.1765920086156697e-06,
+      "loss": 0.4689,
+      "step": 7823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.438262763900532,
+      "learning_rate": 3.1753771698296787e-06,
+      "loss": 0.5367,
+      "step": 7824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.078970455959525,
+      "learning_rate": 3.174162455292432e-06,
+      "loss": 0.5117,
+      "step": 7825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.134466034962138,
+      "learning_rate": 3.1729478650866528e-06,
+      "loss": 0.589,
+      "step": 7826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7968831514813983,
+      "learning_rate": 3.171733399295043e-06,
+      "loss": 0.7081,
+      "step": 7827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.233522062698718,
+      "learning_rate": 3.1705190580003065e-06,
+      "loss": 0.4952,
+      "step": 7828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.933647718486774,
+      "learning_rate": 3.1693048412851324e-06,
+      "loss": 0.3448,
+      "step": 7829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 25.80225154906166,
+      "learning_rate": 3.1680907492322043e-06,
+      "loss": 0.4682,
+      "step": 7830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.282651352629001,
+      "learning_rate": 3.1668767819241973e-06,
+      "loss": 0.6352,
+      "step": 7831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.942947189754542,
+      "learning_rate": 3.1656629394437753e-06,
+      "loss": 0.5197,
+      "step": 7832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.083172489430758,
+      "learning_rate": 3.1644492218735983e-06,
+      "loss": 0.5444,
+      "step": 7833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.5578264763383896,
+      "learning_rate": 3.163235629296311e-06,
+      "loss": 0.586,
+      "step": 7834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.5669087031070985,
+      "learning_rate": 3.1620221617945595e-06,
+      "loss": 0.4914,
+      "step": 7835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.211967521619379,
+      "learning_rate": 3.160808819450969e-06,
+      "loss": 0.4434,
+      "step": 7836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.7113826921045385,
+      "learning_rate": 3.159595602348168e-06,
+      "loss": 0.6103,
+      "step": 7837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.565831971754346,
+      "learning_rate": 3.1583825105687682e-06,
+      "loss": 0.4376,
+      "step": 7838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.737822804990227,
+      "learning_rate": 3.1571695441953763e-06,
+      "loss": 0.4044,
+      "step": 7839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.6656495484146685,
+      "learning_rate": 3.1559567033105927e-06,
+      "loss": 0.484,
+      "step": 7840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 11.439877790937285,
+      "learning_rate": 3.1547439879970023e-06,
+      "loss": 0.4111,
+      "step": 7841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.148430091401379,
+      "learning_rate": 3.1535313983371883e-06,
+      "loss": 0.4925,
+      "step": 7842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.582966005632599,
+      "learning_rate": 3.1523189344137207e-06,
+      "loss": 0.3646,
+      "step": 7843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.258335223873104,
+      "learning_rate": 3.151106596309167e-06,
+      "loss": 0.4851,
+      "step": 7844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.596207115475137,
+      "learning_rate": 3.149894384106077e-06,
+      "loss": 0.3841,
+      "step": 7845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.820937169247346,
+      "learning_rate": 3.1486822978870003e-06,
+      "loss": 0.376,
+      "step": 7846
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0362630738548964,
+      "learning_rate": 3.1474703377344727e-06,
+      "loss": 0.7324,
+      "step": 7847
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.0376312658500435,
+      "learning_rate": 3.1462585037310256e-06,
+      "loss": 0.4239,
+      "step": 7848
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.294110127208884,
+      "learning_rate": 3.1450467959591767e-06,
+      "loss": 0.4898,
+      "step": 7849
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.938750276491764,
+      "learning_rate": 3.1438352145014383e-06,
+      "loss": 0.4864,
+      "step": 7850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.247254247812136,
+      "learning_rate": 3.1426237594403187e-06,
+      "loss": 0.5429,
+      "step": 7851
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.729329705029615,
+      "learning_rate": 3.141412430858305e-06,
+      "loss": 0.4281,
+      "step": 7852
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.384026252204056,
+      "learning_rate": 3.1402012288378903e-06,
+      "loss": 0.4509,
+      "step": 7853
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.035884487761301,
+      "learning_rate": 3.1389901534615462e-06,
+      "loss": 0.4107,
+      "step": 7854
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.7555397429867785,
+      "learning_rate": 3.137779204811746e-06,
+      "loss": 0.4732,
+      "step": 7855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.934065008157668,
+      "learning_rate": 3.136568382970947e-06,
+      "loss": 0.5624,
+      "step": 7856
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.121739898382451,
+      "learning_rate": 3.1353576880216035e-06,
+      "loss": 0.4391,
+      "step": 7857
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.067428036766291,
+      "learning_rate": 3.134147120046156e-06,
+      "loss": 0.569,
+      "step": 7858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.894807768145003,
+      "learning_rate": 3.1329366791270387e-06,
+      "loss": 0.4506,
+      "step": 7859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.543545040114983,
+      "learning_rate": 3.131726365346681e-06,
+      "loss": 0.3726,
+      "step": 7860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1826316767432132,
+      "learning_rate": 3.130516178787495e-06,
+      "loss": 0.7342,
+      "step": 7861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.227152722805717,
+      "learning_rate": 3.129306119531893e-06,
+      "loss": 0.7875,
+      "step": 7862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.888329225144552,
+      "learning_rate": 3.1280961876622718e-06,
+      "loss": 0.4718,
+      "step": 7863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.157400342135505,
+      "learning_rate": 3.1268863832610254e-06,
+      "loss": 0.4714,
+      "step": 7864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.711995574212622,
+      "learning_rate": 3.125676706410533e-06,
+      "loss": 0.4688,
+      "step": 7865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.9813012323594945,
+      "learning_rate": 3.1244671571931706e-06,
+      "loss": 0.4519,
+      "step": 7866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 17.754745301510262,
+      "learning_rate": 3.1232577356913008e-06,
+      "loss": 0.5356,
+      "step": 7867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.524949898153998,
+      "learning_rate": 3.122048441987283e-06,
+      "loss": 0.4407,
+      "step": 7868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.817935014461415,
+      "learning_rate": 3.120839276163462e-06,
+      "loss": 0.3509,
+      "step": 7869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.2257982593773855,
+      "learning_rate": 3.119630238302176e-06,
+      "loss": 0.4863,
+      "step": 7870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.991434507673486,
+      "learning_rate": 3.1184213284857594e-06,
+      "loss": 0.5083,
+      "step": 7871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.702265691889896,
+      "learning_rate": 3.11721254679653e-06,
+      "loss": 0.3665,
+      "step": 7872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.139962462666562,
+      "learning_rate": 3.1160038933168025e-06,
+      "loss": 0.4939,
+      "step": 7873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.644160909865659,
+      "learning_rate": 3.1147953681288785e-06,
+      "loss": 0.4499,
+      "step": 7874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7587902945942582,
+      "learning_rate": 3.1135869713150556e-06,
+      "loss": 0.7261,
+      "step": 7875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.004440194478613,
+      "learning_rate": 3.1123787029576187e-06,
+      "loss": 0.5786,
+      "step": 7876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.755281760086899,
+      "learning_rate": 3.1111705631388465e-06,
+      "loss": 0.5216,
+      "step": 7877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.11752350631519,
+      "learning_rate": 3.109962551941007e-06,
+      "loss": 0.3926,
+      "step": 7878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.314520121601372,
+      "learning_rate": 3.10875466944636e-06,
+      "loss": 0.5226,
+      "step": 7879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.934764489900959,
+      "learning_rate": 3.1075469157371596e-06,
+      "loss": 0.457,
+      "step": 7880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.815403372877019,
+      "learning_rate": 3.1063392908956457e-06,
+      "loss": 0.5265,
+      "step": 7881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.516237048029661,
+      "learning_rate": 3.105131795004055e-06,
+      "loss": 0.3702,
+      "step": 7882
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.098084648536739,
+      "learning_rate": 3.1039244281446103e-06,
+      "loss": 0.5201,
+      "step": 7883
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.515108611463793,
+      "learning_rate": 3.102717190399529e-06,
+      "loss": 0.4788,
+      "step": 7884
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.83289264618579,
+      "learning_rate": 3.101510081851017e-06,
+      "loss": 0.4177,
+      "step": 7885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.109127564245457,
+      "learning_rate": 3.1003031025812757e-06,
+      "loss": 0.6211,
+      "step": 7886
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.524046638107826,
+      "learning_rate": 3.099096252672492e-06,
+      "loss": 0.479,
+      "step": 7887
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.780639790871173,
+      "learning_rate": 3.0978895322068507e-06,
+      "loss": 0.5865,
+      "step": 7888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8755564591682963,
+      "learning_rate": 3.09668294126652e-06,
+      "loss": 0.4114,
+      "step": 7889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.38453625835701,
+      "learning_rate": 3.095476479933667e-06,
+      "loss": 0.453,
+      "step": 7890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.624680371413164,
+      "learning_rate": 3.0942701482904445e-06,
+      "loss": 0.4156,
+      "step": 7891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.954839876103237,
+      "learning_rate": 3.0930639464189988e-06,
+      "loss": 0.5176,
+      "step": 7892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.062949711942215,
+      "learning_rate": 3.0918578744014676e-06,
+      "loss": 0.5612,
+      "step": 7893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.4187844869248805,
+      "learning_rate": 3.090651932319978e-06,
+      "loss": 0.4932,
+      "step": 7894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.912878466025352,
+      "learning_rate": 3.08944612025665e-06,
+      "loss": 0.469,
+      "step": 7895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.849122244953353,
+      "learning_rate": 3.0882404382935917e-06,
+      "loss": 0.4522,
+      "step": 7896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.543569600938349,
+      "learning_rate": 3.0870348865129106e-06,
+      "loss": 0.6247,
+      "step": 7897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.745537655026965,
+      "learning_rate": 3.0858294649966917e-06,
+      "loss": 0.5258,
+      "step": 7898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.967130503153097,
+      "learning_rate": 3.0846241738270243e-06,
+      "loss": 0.509,
+      "step": 7899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.009829937516217,
+      "learning_rate": 3.083419013085983e-06,
+      "loss": 0.4931,
+      "step": 7900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.017951284919972,
+      "learning_rate": 3.082213982855632e-06,
+      "loss": 0.517,
+      "step": 7901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.426939693491139,
+      "learning_rate": 3.0810090832180295e-06,
+      "loss": 0.4528,
+      "step": 7902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.6445686352371816,
+      "learning_rate": 3.0798043142552236e-06,
+      "loss": 0.5113,
+      "step": 7903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.24818996316184,
+      "learning_rate": 3.0785996760492543e-06,
+      "loss": 0.4191,
+      "step": 7904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.505659973488591,
+      "learning_rate": 3.07739516868215e-06,
+      "loss": 0.4793,
+      "step": 7905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.068702292910178,
+      "learning_rate": 3.0761907922359367e-06,
+      "loss": 0.5346,
+      "step": 7906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.839024498606558,
+      "learning_rate": 3.0749865467926216e-06,
+      "loss": 0.4293,
+      "step": 7907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.798661831666047,
+      "learning_rate": 3.0737824324342136e-06,
+      "loss": 0.4363,
+      "step": 7908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.656119019161931,
+      "learning_rate": 3.0725784492427034e-06,
+      "loss": 0.3736,
+      "step": 7909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.066011931316474,
+      "learning_rate": 3.0713745973000796e-06,
+      "loss": 0.5413,
+      "step": 7910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.024719071699814,
+      "learning_rate": 3.0701708766883185e-06,
+      "loss": 0.4846,
+      "step": 7911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 22.10966089847261,
+      "learning_rate": 3.0689672874893883e-06,
+      "loss": 0.5218,
+      "step": 7912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.450105562441478,
+      "learning_rate": 3.067763829785248e-06,
+      "loss": 0.3802,
+      "step": 7913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.877098894723246,
+      "learning_rate": 3.066560503657846e-06,
+      "loss": 0.4681,
+      "step": 7914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.32350256676803,
+      "learning_rate": 3.0653573091891276e-06,
+      "loss": 0.5234,
+      "step": 7915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0473824610551445,
+      "learning_rate": 3.0641542464610198e-06,
+      "loss": 0.4042,
+      "step": 7916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.403094708537746,
+      "learning_rate": 3.06295131555545e-06,
+      "loss": 0.3926,
+      "step": 7917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.718375126968409,
+      "learning_rate": 3.0617485165543315e-06,
+      "loss": 0.4884,
+      "step": 7918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.65670413966091,
+      "learning_rate": 3.060545849539568e-06,
+      "loss": 0.4773,
+      "step": 7919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.261391253999646,
+      "learning_rate": 3.059343314593058e-06,
+      "loss": 0.5218,
+      "step": 7920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.287961448187395,
+      "learning_rate": 3.0581409117966856e-06,
+      "loss": 0.447,
+      "step": 7921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.284654048695235,
+      "learning_rate": 3.056938641232334e-06,
+      "loss": 0.545,
+      "step": 7922
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.099887033076484,
+      "learning_rate": 3.0557365029818673e-06,
+      "loss": 0.4131,
+      "step": 7923
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.644278399862228,
+      "learning_rate": 3.0545344971271517e-06,
+      "loss": 0.5196,
+      "step": 7924
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.212635738221454,
+      "learning_rate": 3.0533326237500315e-06,
+      "loss": 0.4678,
+      "step": 7925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.391468599984539,
+      "learning_rate": 3.052130882932355e-06,
+      "loss": 0.3977,
+      "step": 7926
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.0443647151852495,
+      "learning_rate": 3.0509292747559515e-06,
+      "loss": 0.4747,
+      "step": 7927
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.360456967608878,
+      "learning_rate": 3.0497277993026473e-06,
+      "loss": 0.4894,
+      "step": 7928
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.8351402047853576,
+      "learning_rate": 3.048526456654257e-06,
+      "loss": 0.4744,
+      "step": 7929
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.264050079008128,
+      "learning_rate": 3.047325246892585e-06,
+      "loss": 0.4845,
+      "step": 7930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.9495110289269375,
+      "learning_rate": 3.0461241700994325e-06,
+      "loss": 0.5045,
+      "step": 7931
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8636248706144605,
+      "learning_rate": 3.044923226356583e-06,
+      "loss": 0.7419,
+      "step": 7932
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.305123355429813,
+      "learning_rate": 3.043722415745819e-06,
+      "loss": 0.4943,
+      "step": 7933
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.547146560783414,
+      "learning_rate": 3.042521738348908e-06,
+      "loss": 0.4809,
+      "step": 7934
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.204952685111305,
+      "learning_rate": 3.0413211942476127e-06,
+      "loss": 0.5175,
+      "step": 7935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.897766619138961,
+      "learning_rate": 3.040120783523683e-06,
+      "loss": 0.4265,
+      "step": 7936
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.7130802086228964,
+      "learning_rate": 3.0389205062588633e-06,
+      "loss": 0.4285,
+      "step": 7937
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.537434005226681,
+      "learning_rate": 3.037720362534885e-06,
+      "loss": 0.6546,
+      "step": 7938
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.613028976554931,
+      "learning_rate": 3.036520352433473e-06,
+      "loss": 0.4114,
+      "step": 7939
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.766717115754601,
+      "learning_rate": 3.0353204760363463e-06,
+      "loss": 0.494,
+      "step": 7940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.68842167714242,
+      "learning_rate": 3.0341207334252063e-06,
+      "loss": 0.5385,
+      "step": 7941
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.083176286268589,
+      "learning_rate": 3.0329211246817537e-06,
+      "loss": 0.4358,
+      "step": 7942
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.337128106274374,
+      "learning_rate": 3.0317216498876746e-06,
+      "loss": 0.4989,
+      "step": 7943
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.009952786638241,
+      "learning_rate": 3.030522309124649e-06,
+      "loss": 0.4737,
+      "step": 7944
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.916812354179321,
+      "learning_rate": 3.0293231024743453e-06,
+      "loss": 0.4239,
+      "step": 7945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.921962464281946,
+      "learning_rate": 3.0281240300184263e-06,
+      "loss": 0.5951,
+      "step": 7946
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.180946482480856,
+      "learning_rate": 3.026925091838541e-06,
+      "loss": 0.51,
+      "step": 7947
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.136813187784307,
+      "learning_rate": 3.025726288016334e-06,
+      "loss": 0.4753,
+      "step": 7948
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.452139158801071,
+      "learning_rate": 3.024527618633437e-06,
+      "loss": 0.4082,
+      "step": 7949
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.979237672589254,
+      "learning_rate": 3.0233290837714734e-06,
+      "loss": 0.4384,
+      "step": 7950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5841906366184397,
+      "learning_rate": 3.0221306835120613e-06,
+      "loss": 0.7046,
+      "step": 7951
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.930152252017077,
+      "learning_rate": 3.020932417936804e-06,
+      "loss": 0.4238,
+      "step": 7952
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.2103895619074585,
+      "learning_rate": 3.0197342871272994e-06,
+      "loss": 0.576,
+      "step": 7953
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.431902411067709,
+      "learning_rate": 3.0185362911651333e-06,
+      "loss": 0.4708,
+      "step": 7954
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.227598631953434,
+      "learning_rate": 3.0173384301318857e-06,
+      "loss": 0.4543,
+      "step": 7955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.1491204789038125,
+      "learning_rate": 3.0161407041091238e-06,
+      "loss": 0.5025,
+      "step": 7956
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.020330757658506,
+      "learning_rate": 3.0149431131784094e-06,
+      "loss": 0.4227,
+      "step": 7957
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 17.848025396565195,
+      "learning_rate": 3.0137456574212904e-06,
+      "loss": 0.4289,
+      "step": 7958
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.422666319402721,
+      "learning_rate": 3.012548336919313e-06,
+      "loss": 0.4422,
+      "step": 7959
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.267264492452022,
+      "learning_rate": 3.011351151754003e-06,
+      "loss": 0.4257,
+      "step": 7960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5971725169939945,
+      "learning_rate": 3.0101541020068882e-06,
+      "loss": 0.7091,
+      "step": 7961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 19.246194781448985,
+      "learning_rate": 3.0089571877594812e-06,
+      "loss": 0.5308,
+      "step": 7962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.346159465530541,
+      "learning_rate": 3.007760409093287e-06,
+      "loss": 0.4557,
+      "step": 7963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.43900861600441,
+      "learning_rate": 3.0065637660898e-06,
+      "loss": 0.5632,
+      "step": 7964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8579833627628117,
+      "learning_rate": 3.005367258830506e-06,
+      "loss": 0.3493,
+      "step": 7965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.281175511260044,
+      "learning_rate": 3.004170887396883e-06,
+      "loss": 0.3938,
+      "step": 7966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.903074936731628,
+      "learning_rate": 3.0029746518703974e-06,
+      "loss": 0.4356,
+      "step": 7967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.1486048175059675,
+      "learning_rate": 3.0017785523325103e-06,
+      "loss": 0.4054,
+      "step": 7968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.964643632096027,
+      "learning_rate": 3.000582588864666e-06,
+      "loss": 0.4213,
+      "step": 7969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.9023253941225775,
+      "learning_rate": 2.999386761548308e-06,
+      "loss": 0.4552,
+      "step": 7970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.780448829202564,
+      "learning_rate": 2.998191070464867e-06,
+      "loss": 0.4081,
+      "step": 7971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.003120885302311,
+      "learning_rate": 2.996995515695762e-06,
+      "loss": 0.4971,
+      "step": 7972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.2137138859369,
+      "learning_rate": 2.995800097322408e-06,
+      "loss": 0.5234,
+      "step": 7973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.57177399080408,
+      "learning_rate": 2.9946048154262036e-06,
+      "loss": 0.4882,
+      "step": 7974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.195937130251886,
+      "learning_rate": 2.9934096700885464e-06,
+      "loss": 0.468,
+      "step": 7975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.082329343475688,
+      "learning_rate": 2.992214661390816e-06,
+      "loss": 0.3862,
+      "step": 7976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.4895868749033,
+      "learning_rate": 2.9910197894143934e-06,
+      "loss": 0.5571,
+      "step": 7977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.054462743629645,
+      "learning_rate": 2.989825054240637e-06,
+      "loss": 0.7283,
+      "step": 7978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.722088472107079,
+      "learning_rate": 2.988630455950908e-06,
+      "loss": 0.5361,
+      "step": 7979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 11.136200028372567,
+      "learning_rate": 2.987435994626552e-06,
+      "loss": 0.4384,
+      "step": 7980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.549499106837161,
+      "learning_rate": 2.986241670348906e-06,
+      "loss": 0.4763,
+      "step": 7981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.373319161786858,
+      "learning_rate": 2.9850474831992994e-06,
+      "loss": 0.4349,
+      "step": 7982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.289302248346509,
+      "learning_rate": 2.9838534332590484e-06,
+      "loss": 0.5171,
+      "step": 7983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.447190371191353,
+      "learning_rate": 2.9826595206094664e-06,
+      "loss": 0.3716,
+      "step": 7984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.326126542959448,
+      "learning_rate": 2.981465745331849e-06,
+      "loss": 0.4102,
+      "step": 7985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.418771823710832,
+      "learning_rate": 2.9802721075074924e-06,
+      "loss": 0.4463,
+      "step": 7986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 41.1600651061606,
+      "learning_rate": 2.9790786072176726e-06,
+      "loss": 0.5688,
+      "step": 7987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7810712667119106,
+      "learning_rate": 2.9778852445436657e-06,
+      "loss": 0.353,
+      "step": 7988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.725365140147561,
+      "learning_rate": 2.976692019566732e-06,
+      "loss": 0.5033,
+      "step": 7989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.248505926364286,
+      "learning_rate": 2.975498932368126e-06,
+      "loss": 0.4797,
+      "step": 7990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.074137620008399,
+      "learning_rate": 2.974305983029092e-06,
+      "loss": 0.4534,
+      "step": 7991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.078839212644382,
+      "learning_rate": 2.9731131716308625e-06,
+      "loss": 0.4327,
+      "step": 7992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.846873429698939,
+      "learning_rate": 2.971920498254668e-06,
+      "loss": 0.4376,
+      "step": 7993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.354135675044292,
+      "learning_rate": 2.9707279629817165e-06,
+      "loss": 0.4186,
+      "step": 7994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9740748658725913,
+      "learning_rate": 2.9695355658932202e-06,
+      "loss": 0.7512,
+      "step": 7995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.927791654144954,
+      "learning_rate": 2.9683433070703737e-06,
+      "loss": 0.459,
+      "step": 7996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.232154033230583,
+      "learning_rate": 2.967151186594366e-06,
+      "loss": 0.4151,
+      "step": 7997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.005814041936709,
+      "learning_rate": 2.965959204546373e-06,
+      "loss": 0.5103,
+      "step": 7998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.534391301773193,
+      "learning_rate": 2.9647673610075666e-06,
+      "loss": 0.4792,
+      "step": 7999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.666587904161644,
+      "learning_rate": 2.9635756560591022e-06,
+      "loss": 0.3964,
+      "step": 8000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 9.022445426540314,
+      "learning_rate": 2.962384089782131e-06,
+      "loss": 0.4051,
+      "step": 8001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.878864452133147,
+      "learning_rate": 2.9611926622577974e-06,
+      "loss": 0.4289,
+      "step": 8002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 12.635005874310021,
+      "learning_rate": 2.960001373567226e-06,
+      "loss": 0.5125,
+      "step": 8003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.41429320936359,
+      "learning_rate": 2.958810223791543e-06,
+      "loss": 0.489,
+      "step": 8004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 13.26181503002196,
+      "learning_rate": 2.9576192130118576e-06,
+      "loss": 0.6339,
+      "step": 8005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.491736239297388,
+      "learning_rate": 2.9564283413092754e-06,
+      "loss": 0.4185,
+      "step": 8006
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.161344309795341,
+      "learning_rate": 2.9552376087648853e-06,
+      "loss": 0.4767,
+      "step": 8007
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.51830227693821,
+      "learning_rate": 2.954047015459775e-06,
+      "loss": 0.4097,
+      "step": 8008
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 35.142191039227505,
+      "learning_rate": 2.952856561475016e-06,
+      "loss": 0.6185,
+      "step": 8009
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.354435841229137,
+      "learning_rate": 2.9516662468916722e-06,
+      "loss": 0.4653,
+      "step": 8010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.623338840268087,
+      "learning_rate": 2.9504760717908036e-06,
+      "loss": 0.3838,
+      "step": 8011
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.54204195971582,
+      "learning_rate": 2.9492860362534497e-06,
+      "loss": 0.5204,
+      "step": 8012
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.1550499195841475,
+      "learning_rate": 2.9480961403606516e-06,
+      "loss": 0.4747,
+      "step": 8013
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.835600350110829,
+      "learning_rate": 2.946906384193433e-06,
+      "loss": 0.4598,
+      "step": 8014
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.386797051154908,
+      "learning_rate": 2.9457167678328125e-06,
+      "loss": 0.5639,
+      "step": 8015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.750805497471573,
+      "learning_rate": 2.9445272913597967e-06,
+      "loss": 0.4682,
+      "step": 8016
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.969771861434855,
+      "learning_rate": 2.943337954855385e-06,
+      "loss": 0.4135,
+      "step": 8017
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8159113542755265,
+      "learning_rate": 2.9421487584005646e-06,
+      "loss": 0.4545,
+      "step": 8018
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.53308428038664,
+      "learning_rate": 2.940959702076315e-06,
+      "loss": 0.4382,
+      "step": 8019
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.347600954946329,
+      "learning_rate": 2.939770785963606e-06,
+      "loss": 0.5121,
+      "step": 8020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.397539256407681,
+      "learning_rate": 2.9385820101433954e-06,
+      "loss": 0.593,
+      "step": 8021
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.306195422631803,
+      "learning_rate": 2.9373933746966375e-06,
+      "loss": 0.3773,
+      "step": 8022
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.850366789264117,
+      "learning_rate": 2.93620487970427e-06,
+      "loss": 0.4071,
+      "step": 8023
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.832059838490331,
+      "learning_rate": 2.935016525247227e-06,
+      "loss": 0.5068,
+      "step": 8024
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.819193182838556,
+      "learning_rate": 2.933828311406427e-06,
+      "loss": 0.3629,
+      "step": 8025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.474906460176857,
+      "learning_rate": 2.932640238262784e-06,
+      "loss": 0.5478,
+      "step": 8026
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.469406115920984,
+      "learning_rate": 2.9314523058971992e-06,
+      "loss": 0.3745,
+      "step": 8027
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.402691199591203,
+      "learning_rate": 2.9302645143905673e-06,
+      "loss": 0.4706,
+      "step": 8028
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.252769822929585,
+      "learning_rate": 2.9290768638237693e-06,
+      "loss": 0.4175,
+      "step": 8029
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.482665555568938,
+      "learning_rate": 2.9278893542776808e-06,
+      "loss": 0.4235,
+      "step": 8030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 32.34757722925349,
+      "learning_rate": 2.9267019858331668e-06,
+      "loss": 0.3358,
+      "step": 8031
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.323362402098178,
+      "learning_rate": 2.9255147585710795e-06,
+      "loss": 0.4323,
+      "step": 8032
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.108206973486425,
+      "learning_rate": 2.9243276725722657e-06,
+      "loss": 0.4896,
+      "step": 8033
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.738515202961432,
+      "learning_rate": 2.9231407279175595e-06,
+      "loss": 0.5492,
+      "step": 8034
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.835726946433434,
+      "learning_rate": 2.9219539246877877e-06,
+      "loss": 0.4633,
+      "step": 8035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.330553950011484,
+      "learning_rate": 2.9207672629637648e-06,
+      "loss": 0.4759,
+      "step": 8036
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.273872440962074,
+      "learning_rate": 2.9195807428262986e-06,
+      "loss": 0.6175,
+      "step": 8037
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.850136550585975,
+      "learning_rate": 2.9183943643561835e-06,
+      "loss": 0.5052,
+      "step": 8038
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.898407487630829,
+      "learning_rate": 2.91720812763421e-06,
+      "loss": 0.4428,
+      "step": 8039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.239557687321696,
+      "learning_rate": 2.9160220327411547e-06,
+      "loss": 0.4294,
+      "step": 8040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.020450046749989,
+      "learning_rate": 2.9148360797577825e-06,
+      "loss": 0.4639,
+      "step": 8041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.178799954431261,
+      "learning_rate": 2.913650268764856e-06,
+      "loss": 0.4146,
+      "step": 8042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7591806598250899,
+      "learning_rate": 2.9124645998431193e-06,
+      "loss": 0.7097,
+      "step": 8043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.645773781655424,
+      "learning_rate": 2.9112790730733156e-06,
+      "loss": 0.4074,
+      "step": 8044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.48523200852474,
+      "learning_rate": 2.9100936885361717e-06,
+      "loss": 0.4173,
+      "step": 8045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.992126428028748,
+      "learning_rate": 2.9089084463124075e-06,
+      "loss": 0.4476,
+      "step": 8046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.12979697546695,
+      "learning_rate": 2.90772334648273e-06,
+      "loss": 0.44,
+      "step": 8047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.392426713743033,
+      "learning_rate": 2.906538389127844e-06,
+      "loss": 0.5133,
+      "step": 8048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.453512804306602,
+      "learning_rate": 2.9053535743284367e-06,
+      "loss": 0.4031,
+      "step": 8049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.459204034397595,
+      "learning_rate": 2.9041689021651885e-06,
+      "loss": 0.5429,
+      "step": 8050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.823717173668465,
+      "learning_rate": 2.9029843727187736e-06,
+      "loss": 0.3716,
+      "step": 8051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.632083831116315,
+      "learning_rate": 2.901799986069848e-06,
+      "loss": 0.4888,
+      "step": 8052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.783128204332973,
+      "learning_rate": 2.9006157422990686e-06,
+      "loss": 0.4828,
+      "step": 8053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 26.72539051739062,
+      "learning_rate": 2.899431641487076e-06,
+      "loss": 0.539,
+      "step": 8054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.212500205183668,
+      "learning_rate": 2.8982476837145e-06,
+      "loss": 0.511,
+      "step": 8055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.247887121462852,
+      "learning_rate": 2.8970638690619616e-06,
+      "loss": 0.5153,
+      "step": 8056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.458468105457097,
+      "learning_rate": 2.895880197610078e-06,
+      "loss": 0.4459,
+      "step": 8057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.984298672065954,
+      "learning_rate": 2.89469666943945e-06,
+      "loss": 0.5445,
+      "step": 8058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.529023652840126,
+      "learning_rate": 2.89351328463067e-06,
+      "loss": 0.4766,
+      "step": 8059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.242209567899385,
+      "learning_rate": 2.8923300432643197e-06,
+      "loss": 0.5317,
+      "step": 8060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.215583722087825,
+      "learning_rate": 2.891146945420975e-06,
+      "loss": 0.5093,
+      "step": 8061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.5814134000239015,
+      "learning_rate": 2.8899639911812005e-06,
+      "loss": 0.4545,
+      "step": 8062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.697450428998509,
+      "learning_rate": 2.8887811806255488e-06,
+      "loss": 0.4491,
+      "step": 8063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.540183090108597,
+      "learning_rate": 2.8875985138345643e-06,
+      "loss": 0.449,
+      "step": 8064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.031590240386429,
+      "learning_rate": 2.886415990888779e-06,
+      "loss": 0.5061,
+      "step": 8065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.096840287860323,
+      "learning_rate": 2.885233611868721e-06,
+      "loss": 0.3943,
+      "step": 8066
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.003290909491606,
+      "learning_rate": 2.8840513768549043e-06,
+      "loss": 0.4718,
+      "step": 8067
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.057154573083675,
+      "learning_rate": 2.8828692859278334e-06,
+      "loss": 0.583,
+      "step": 8068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.868716322182097,
+      "learning_rate": 2.881687339168001e-06,
+      "loss": 0.4773,
+      "step": 8069
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.679676190676929,
+      "learning_rate": 2.880505536655895e-06,
+      "loss": 0.5405,
+      "step": 8070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.679915362823444,
+      "learning_rate": 2.8793238784719923e-06,
+      "loss": 0.5264,
+      "step": 8071
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.258071303883763,
+      "learning_rate": 2.878142364696758e-06,
+      "loss": 0.4535,
+      "step": 8072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.420523058609621,
+      "learning_rate": 2.876960995410646e-06,
+      "loss": 0.3508,
+      "step": 8073
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.486286704869663,
+      "learning_rate": 2.875779770694101e-06,
+      "loss": 0.4349,
+      "step": 8074
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.607694408461971,
+      "learning_rate": 2.8745986906275643e-06,
+      "loss": 0.422,
+      "step": 8075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.1846607118784735,
+      "learning_rate": 2.873417755291459e-06,
+      "loss": 0.3741,
+      "step": 8076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.249839524940445,
+      "learning_rate": 2.872236964766202e-06,
+      "loss": 0.5776,
+      "step": 8077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.270523379546728,
+      "learning_rate": 2.8710563191321977e-06,
+      "loss": 0.5173,
+      "step": 8078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.882648165194908,
+      "learning_rate": 2.869875818469847e-06,
+      "loss": 0.4112,
+      "step": 8079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.901177612113115,
+      "learning_rate": 2.868695462859533e-06,
+      "loss": 0.4942,
+      "step": 8080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.608337678721426,
+      "learning_rate": 2.8675152523816374e-06,
+      "loss": 0.4043,
+      "step": 8081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.870241146187632,
+      "learning_rate": 2.8663351871165236e-06,
+      "loss": 0.5181,
+      "step": 8082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.325187209642687,
+      "learning_rate": 2.865155267144547e-06,
+      "loss": 0.4337,
+      "step": 8083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.195725198462789,
+      "learning_rate": 2.8639754925460606e-06,
+      "loss": 0.469,
+      "step": 8084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.572111489183715,
+      "learning_rate": 2.8627958634013986e-06,
+      "loss": 0.4424,
+      "step": 8085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.250047788561733,
+      "learning_rate": 2.861616379790888e-06,
+      "loss": 0.3435,
+      "step": 8086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.508184288589616,
+      "learning_rate": 2.8604370417948457e-06,
+      "loss": 0.4695,
+      "step": 8087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 10.929048141042884,
+      "learning_rate": 2.859257849493583e-06,
+      "loss": 0.501,
+      "step": 8088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.889001258421886,
+      "learning_rate": 2.858078802967394e-06,
+      "loss": 0.4953,
+      "step": 8089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.142268118326588,
+      "learning_rate": 2.8568999022965695e-06,
+      "loss": 0.4153,
+      "step": 8090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.164796880888574,
+      "learning_rate": 2.855721147561386e-06,
+      "loss": 0.4637,
+      "step": 8091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.929830772824639,
+      "learning_rate": 2.85454253884211e-06,
+      "loss": 0.4219,
+      "step": 8092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 14.843825491575336,
+      "learning_rate": 2.8533640762190023e-06,
+      "loss": 0.5092,
+      "step": 8093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.396130630055303,
+      "learning_rate": 2.852185759772308e-06,
+      "loss": 0.4021,
+      "step": 8094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.508885358449051,
+      "learning_rate": 2.8510075895822714e-06,
+      "loss": 0.4142,
+      "step": 8095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.880643592852964,
+      "learning_rate": 2.8498295657291118e-06,
+      "loss": 0.4441,
+      "step": 8096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.605263718316202,
+      "learning_rate": 2.8486516882930544e-06,
+      "loss": 0.4651,
+      "step": 8097
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.55069317757184,
+      "learning_rate": 2.8474739573543026e-06,
+      "loss": 0.4552,
+      "step": 8098
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.538335550696352,
+      "learning_rate": 2.8462963729930595e-06,
+      "loss": 0.5914,
+      "step": 8099
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.8432738039257766,
+      "learning_rate": 2.8451189352895098e-06,
+      "loss": 0.4406,
+      "step": 8100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 15.804471169131913,
+      "learning_rate": 2.8439416443238316e-06,
+      "loss": 0.5093,
+      "step": 8101
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.630205127208717,
+      "learning_rate": 2.842764500176196e-06,
+      "loss": 0.574,
+      "step": 8102
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.127262402182448,
+      "learning_rate": 2.841587502926759e-06,
+      "loss": 0.5527,
+      "step": 8103
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.303523917318596,
+      "learning_rate": 2.8404106526556726e-06,
+      "loss": 0.4508,
+      "step": 8104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.196709540296194,
+      "learning_rate": 2.8392339494430686e-06,
+      "loss": 0.52,
+      "step": 8105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.920632560859806,
+      "learning_rate": 2.838057393369081e-06,
+      "loss": 0.4906,
+      "step": 8106
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.745811310972262,
+      "learning_rate": 2.8368809845138233e-06,
+      "loss": 0.5065,
+      "step": 8107
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.44572434442239,
+      "learning_rate": 2.835704722957409e-06,
+      "loss": 0.509,
+      "step": 8108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.731669217566366,
+      "learning_rate": 2.834528608779934e-06,
+      "loss": 0.4995,
+      "step": 8109
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.716385931451243,
+      "learning_rate": 2.833352642061486e-06,
+      "loss": 0.4505,
+      "step": 8110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.1339000218299145,
+      "learning_rate": 2.832176822882141e-06,
+      "loss": 0.3696,
+      "step": 8111
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.916526679803942,
+      "learning_rate": 2.8310011513219696e-06,
+      "loss": 0.5647,
+      "step": 8112
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.198398787188617,
+      "learning_rate": 2.8298256274610336e-06,
+      "loss": 0.465,
+      "step": 8113
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 23.473264180032416,
+      "learning_rate": 2.8286502513793735e-06,
+      "loss": 0.4691,
+      "step": 8114
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.76765907422539,
+      "learning_rate": 2.8274750231570325e-06,
+      "loss": 0.5622,
+      "step": 8115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.47315301954708,
+      "learning_rate": 2.8262999428740346e-06,
+      "loss": 0.5642,
+      "step": 8116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.183179536956224,
+      "learning_rate": 2.8251250106104017e-06,
+      "loss": 0.4436,
+      "step": 8117
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.487409726736391,
+      "learning_rate": 2.8239502264461395e-06,
+      "loss": 0.5257,
+      "step": 8118
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.559875981696258,
+      "learning_rate": 2.822775590461245e-06,
+      "loss": 0.4939,
+      "step": 8119
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.834556325008122,
+      "learning_rate": 2.821601102735705e-06,
+      "loss": 0.5224,
+      "step": 8120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.1048663969652,
+      "learning_rate": 2.820426763349498e-06,
+      "loss": 0.4294,
+      "step": 8121
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.940955082354335,
+      "learning_rate": 2.819252572382596e-06,
+      "loss": 0.5217,
+      "step": 8122
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.882172334217616,
+      "learning_rate": 2.8180785299149473e-06,
+      "loss": 0.4308,
+      "step": 8123
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.917666384047322,
+      "learning_rate": 2.816904636026505e-06,
+      "loss": 0.5827,
+      "step": 8124
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.735755086687161,
+      "learning_rate": 2.815730890797203e-06,
+      "loss": 0.557,
+      "step": 8125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.6453292251896,
+      "learning_rate": 2.8145572943069715e-06,
+      "loss": 0.576,
+      "step": 8126
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 13.668326976637351,
+      "learning_rate": 2.8133838466357254e-06,
+      "loss": 0.6127,
+      "step": 8127
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.311659051684305,
+      "learning_rate": 2.812210547863372e-06,
+      "loss": 0.4967,
+      "step": 8128
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.2450852317560654,
+      "learning_rate": 2.811037398069804e-06,
+      "loss": 0.6223,
+      "step": 8129
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 16.96936447829512,
+      "learning_rate": 2.8098643973349138e-06,
+      "loss": 0.4734,
+      "step": 8130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.493712777474483,
+      "learning_rate": 2.808691545738572e-06,
+      "loss": 0.5544,
+      "step": 8131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.51366217912579,
+      "learning_rate": 2.8075188433606493e-06,
+      "loss": 0.4614,
+      "step": 8132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.639775510681988,
+      "learning_rate": 2.806346290280999e-06,
+      "loss": 0.711,
+      "step": 8133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.299245720463407,
+      "learning_rate": 2.805173886579465e-06,
+      "loss": 0.4358,
+      "step": 8134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6494076197914787,
+      "learning_rate": 2.804001632335888e-06,
+      "loss": 0.7297,
+      "step": 8135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.861988079517873,
+      "learning_rate": 2.8028295276300897e-06,
+      "loss": 0.485,
+      "step": 8136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.27961185489901,
+      "learning_rate": 2.801657572541886e-06,
+      "loss": 0.5668,
+      "step": 8137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.5780027288028595,
+      "learning_rate": 2.80048576715108e-06,
+      "loss": 0.4822,
+      "step": 8138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.082235742182224,
+      "learning_rate": 2.79931411153747e-06,
+      "loss": 0.3556,
+      "step": 8139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.426467290635284,
+      "learning_rate": 2.798142605780837e-06,
+      "loss": 0.5375,
+      "step": 8140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.0496436035544,
+      "learning_rate": 2.796971249960958e-06,
+      "loss": 0.5233,
+      "step": 8141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.969481741163592,
+      "learning_rate": 2.7958000441575974e-06,
+      "loss": 0.4853,
+      "step": 8142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.160791353120875,
+      "learning_rate": 2.7946289884505063e-06,
+      "loss": 0.4663,
+      "step": 8143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4034707090718077,
+      "learning_rate": 2.793458082919431e-06,
+      "loss": 0.6977,
+      "step": 8144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.726713310092778,
+      "learning_rate": 2.792287327644105e-06,
+      "loss": 0.5125,
+      "step": 8145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.539244123809587,
+      "learning_rate": 2.79111672270425e-06,
+      "loss": 0.5525,
+      "step": 8146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8064218558399467,
+      "learning_rate": 2.7899462681795785e-06,
+      "loss": 0.7457,
+      "step": 8147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.08948261689914,
+      "learning_rate": 2.7887759641497956e-06,
+      "loss": 0.4607,
+      "step": 8148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.196406110668082,
+      "learning_rate": 2.7876058106945915e-06,
+      "loss": 0.3958,
+      "step": 8149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.142328884613204,
+      "learning_rate": 2.7864358078936533e-06,
+      "loss": 0.5745,
+      "step": 8150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.879039359594604,
+      "learning_rate": 2.7852659558266454e-06,
+      "loss": 0.4558,
+      "step": 8151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.647730337186785,
+      "learning_rate": 2.7840962545732337e-06,
+      "loss": 0.368,
+      "step": 8152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.589085524748862,
+      "learning_rate": 2.7829267042130715e-06,
+      "loss": 0.4186,
+      "step": 8153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.001486345116784,
+      "learning_rate": 2.7817573048257977e-06,
+      "loss": 0.4877,
+      "step": 8154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.182620935128873,
+      "learning_rate": 2.7805880564910436e-06,
+      "loss": 0.4561,
+      "step": 8155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.92608388430776,
+      "learning_rate": 2.7794189592884286e-06,
+      "loss": 0.5257,
+      "step": 8156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.904712048039859,
+      "learning_rate": 2.778250013297565e-06,
+      "loss": 0.4142,
+      "step": 8157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.066062775636645,
+      "learning_rate": 2.7770812185980517e-06,
+      "loss": 0.5104,
+      "step": 8158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.298349098672446,
+      "learning_rate": 2.7759125752694814e-06,
+      "loss": 0.4756,
+      "step": 8159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.70130771318051,
+      "learning_rate": 2.7747440833914286e-06,
+      "loss": 0.4559,
+      "step": 8160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.902933348303383,
+      "learning_rate": 2.7735757430434636e-06,
+      "loss": 0.5818,
+      "step": 8161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.936259425391344,
+      "learning_rate": 2.7724075543051487e-06,
+      "loss": 0.4542,
+      "step": 8162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.379010718357974,
+      "learning_rate": 2.7712395172560305e-06,
+      "loss": 0.4093,
+      "step": 8163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.5699541934442,
+      "learning_rate": 2.770071631975646e-06,
+      "loss": 0.4119,
+      "step": 8164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7429942256312203,
+      "learning_rate": 2.7689038985435222e-06,
+      "loss": 0.7061,
+      "step": 8165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.744606173854709,
+      "learning_rate": 2.767736317039181e-06,
+      "loss": 0.4782,
+      "step": 8166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.84268555589088,
+      "learning_rate": 2.7665688875421238e-06,
+      "loss": 0.4846,
+      "step": 8167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.917374360290491,
+      "learning_rate": 2.7654016101318525e-06,
+      "loss": 0.4194,
+      "step": 8168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5092953047198137,
+      "learning_rate": 2.764234484887852e-06,
+      "loss": 0.7024,
+      "step": 8169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.2063161622155585,
+      "learning_rate": 2.763067511889597e-06,
+      "loss": 0.4289,
+      "step": 8170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.303280978408974,
+      "learning_rate": 2.7619006912165523e-06,
+      "loss": 0.4973,
+      "step": 8171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.612620330584065,
+      "learning_rate": 2.760734022948177e-06,
+      "loss": 0.5108,
+      "step": 8172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.957320900091573,
+      "learning_rate": 2.7595675071639138e-06,
+      "loss": 0.4576,
+      "step": 8173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.997539781490162,
+      "learning_rate": 2.758401143943195e-06,
+      "loss": 0.4587,
+      "step": 8174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.911230264930635,
+      "learning_rate": 2.757234933365449e-06,
+      "loss": 0.5614,
+      "step": 8175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.320465053778202,
+      "learning_rate": 2.7560688755100867e-06,
+      "loss": 0.4443,
+      "step": 8176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.438875280615071,
+      "learning_rate": 2.754902970456514e-06,
+      "loss": 0.3942,
+      "step": 8177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.419682457914847,
+      "learning_rate": 2.7537372182841225e-06,
+      "loss": 0.4861,
+      "step": 8178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.922671282231356,
+      "learning_rate": 2.7525716190722945e-06,
+      "loss": 0.4657,
+      "step": 8179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.02735806002203,
+      "learning_rate": 2.7514061729004007e-06,
+      "loss": 0.3945,
+      "step": 8180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 14.33802819256552,
+      "learning_rate": 2.7502408798478064e-06,
+      "loss": 0.4779,
+      "step": 8181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 38.62382852223885,
+      "learning_rate": 2.7490757399938604e-06,
+      "loss": 0.3707,
+      "step": 8182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.684186405244393,
+      "learning_rate": 2.7479107534179027e-06,
+      "loss": 0.4155,
+      "step": 8183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.799723121721658,
+      "learning_rate": 2.746745920199267e-06,
+      "loss": 0.5393,
+      "step": 8184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.185753767030173,
+      "learning_rate": 2.7455812404172693e-06,
+      "loss": 0.4889,
+      "step": 8185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.240948615390031,
+      "learning_rate": 2.7444167141512235e-06,
+      "loss": 0.5702,
+      "step": 8186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.863888413137227,
+      "learning_rate": 2.743252341480427e-06,
+      "loss": 0.4158,
+      "step": 8187
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.625558370615211,
+      "learning_rate": 2.742088122484167e-06,
+      "loss": 0.4957,
+      "step": 8188
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.282738471463883,
+      "learning_rate": 2.740924057241721e-06,
+      "loss": 0.4833,
+      "step": 8189
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.50003656257503,
+      "learning_rate": 2.7397601458323604e-06,
+      "loss": 0.3769,
+      "step": 8190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.168832233401129,
+      "learning_rate": 2.7385963883353407e-06,
+      "loss": 0.4934,
+      "step": 8191
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.5779406319603515,
+      "learning_rate": 2.7374327848299053e-06,
+      "loss": 0.4917,
+      "step": 8192
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.549126547533253,
+      "learning_rate": 2.736269335395296e-06,
+      "loss": 0.4747,
+      "step": 8193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.0218552110926415,
+      "learning_rate": 2.7351060401107336e-06,
+      "loss": 0.489,
+      "step": 8194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.643115887897262,
+      "learning_rate": 2.7339428990554384e-06,
+      "loss": 0.5135,
+      "step": 8195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.948324471672945,
+      "learning_rate": 2.7327799123086117e-06,
+      "loss": 0.3868,
+      "step": 8196
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.577719645890001,
+      "learning_rate": 2.731617079949449e-06,
+      "loss": 0.567,
+      "step": 8197
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.874046241214467,
+      "learning_rate": 2.7304544020571314e-06,
+      "loss": 0.516,
+      "step": 8198
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.070066054333426,
+      "learning_rate": 2.729291878710837e-06,
+      "loss": 0.5112,
+      "step": 8199
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.891909152293295,
+      "learning_rate": 2.7281295099897254e-06,
+      "loss": 0.4009,
+      "step": 8200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.432441434656052,
+      "learning_rate": 2.7269672959729465e-06,
+      "loss": 0.426,
+      "step": 8201
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.04701614923966,
+      "learning_rate": 2.7258052367396475e-06,
+      "loss": 0.3648,
+      "step": 8202
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.852514061213694,
+      "learning_rate": 2.7246433323689546e-06,
+      "loss": 0.5342,
+      "step": 8203
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.523157836814381,
+      "learning_rate": 2.723481582939992e-06,
+      "loss": 0.5344,
+      "step": 8204
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.554588199524301,
+      "learning_rate": 2.7223199885318683e-06,
+      "loss": 0.5085,
+      "step": 8205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.712498925711639,
+      "learning_rate": 2.7211585492236825e-06,
+      "loss": 0.4501,
+      "step": 8206
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.360247595987301,
+      "learning_rate": 2.7199972650945216e-06,
+      "loss": 0.4484,
+      "step": 8207
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.063102541540948,
+      "learning_rate": 2.7188361362234684e-06,
+      "loss": 0.3169,
+      "step": 8208
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.36570069012155,
+      "learning_rate": 2.717675162689588e-06,
+      "loss": 0.4361,
+      "step": 8209
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.918567877779663,
+      "learning_rate": 2.716514344571938e-06,
+      "loss": 0.4955,
+      "step": 8210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.332071978529439,
+      "learning_rate": 2.7153536819495628e-06,
+      "loss": 0.4844,
+      "step": 8211
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.863726017979163,
+      "learning_rate": 2.7141931749015003e-06,
+      "loss": 0.4114,
+      "step": 8212
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.014958471106203,
+      "learning_rate": 2.7130328235067784e-06,
+      "loss": 0.5191,
+      "step": 8213
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.259203060010227,
+      "learning_rate": 2.7118726278444086e-06,
+      "loss": 0.5155,
+      "step": 8214
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.151118657749373,
+      "learning_rate": 2.7107125879933966e-06,
+      "loss": 0.4353,
+      "step": 8215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.292243020052599,
+      "learning_rate": 2.709552704032733e-06,
+      "loss": 0.4318,
+      "step": 8216
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.334915267546146,
+      "learning_rate": 2.708392976041405e-06,
+      "loss": 0.4586,
+      "step": 8217
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.365169714868412,
+      "learning_rate": 2.707233404098384e-06,
+      "loss": 0.518,
+      "step": 8218
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.797411642920956,
+      "learning_rate": 2.7060739882826302e-06,
+      "loss": 0.5157,
+      "step": 8219
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.778825435907153,
+      "learning_rate": 2.7049147286730937e-06,
+      "loss": 0.4782,
+      "step": 8220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.073241499878927,
+      "learning_rate": 2.703755625348718e-06,
+      "loss": 0.5056,
+      "step": 8221
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.571666910212874,
+      "learning_rate": 2.7025966783884296e-06,
+      "loss": 0.3325,
+      "step": 8222
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.657153734190884,
+      "learning_rate": 2.7014378878711518e-06,
+      "loss": 0.4673,
+      "step": 8223
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.447567692189202,
+      "learning_rate": 2.7002792538757904e-06,
+      "loss": 0.3899,
+      "step": 8224
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.935764714514181,
+      "learning_rate": 2.6991207764812423e-06,
+      "loss": 0.5017,
+      "step": 8225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.987816856100806,
+      "learning_rate": 2.6979624557663976e-06,
+      "loss": 0.4853,
+      "step": 8226
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.201813211914182,
+      "learning_rate": 2.6968042918101296e-06,
+      "loss": 0.571,
+      "step": 8227
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.643271852155626,
+      "learning_rate": 2.6956462846913096e-06,
+      "loss": 0.4164,
+      "step": 8228
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.9441073104751725,
+      "learning_rate": 2.694488434488785e-06,
+      "loss": 0.5416,
+      "step": 8229
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.095410734314656,
+      "learning_rate": 2.6933307412814065e-06,
+      "loss": 0.5175,
+      "step": 8230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.256576520176371,
+      "learning_rate": 2.6921732051480037e-06,
+      "loss": 0.3803,
+      "step": 8231
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7257589028393907,
+      "learning_rate": 2.6910158261674035e-06,
+      "loss": 0.7408,
+      "step": 8232
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.528744225055997,
+      "learning_rate": 2.6898586044184165e-06,
+      "loss": 0.5013,
+      "step": 8233
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.084646229046796,
+      "learning_rate": 2.688701539979842e-06,
+      "loss": 0.4698,
+      "step": 8234
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.415231651894723,
+      "learning_rate": 2.687544632930476e-06,
+      "loss": 0.5433,
+      "step": 8235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.499536900840715,
+      "learning_rate": 2.686387883349094e-06,
+      "loss": 0.3998,
+      "step": 8236
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.207101272927306,
+      "learning_rate": 2.6852312913144714e-06,
+      "loss": 0.4889,
+      "step": 8237
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.267240366667913,
+      "learning_rate": 2.684074856905359e-06,
+      "loss": 0.5461,
+      "step": 8238
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.854986261660944,
+      "learning_rate": 2.6829185802005107e-06,
+      "loss": 0.502,
+      "step": 8239
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.184553967548853,
+      "learning_rate": 2.6817624612786615e-06,
+      "loss": 0.3997,
+      "step": 8240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.098859167153723,
+      "learning_rate": 2.68060650021854e-06,
+      "loss": 0.5848,
+      "step": 8241
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.294303434907552,
+      "learning_rate": 2.679450697098861e-06,
+      "loss": 0.4553,
+      "step": 8242
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8567227293160387,
+      "learning_rate": 2.6782950519983264e-06,
+      "loss": 0.774,
+      "step": 8243
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.007614855425416,
+      "learning_rate": 2.6771395649956365e-06,
+      "loss": 0.4867,
+      "step": 8244
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.723398685086085,
+      "learning_rate": 2.675984236169469e-06,
+      "loss": 0.4961,
+      "step": 8245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.157021663444145,
+      "learning_rate": 2.6748290655985042e-06,
+      "loss": 0.4298,
+      "step": 8246
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.895458929769905,
+      "learning_rate": 2.6736740533613948e-06,
+      "loss": 0.5715,
+      "step": 8247
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.725516000249533,
+      "learning_rate": 2.6725191995367988e-06,
+      "loss": 0.3041,
+      "step": 8248
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7015887388242865,
+      "learning_rate": 2.6713645042033526e-06,
+      "loss": 0.6489,
+      "step": 8249
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.025964927236668,
+      "learning_rate": 2.67020996743969e-06,
+      "loss": 0.4811,
+      "step": 8250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.101978479057563,
+      "learning_rate": 2.669055589324427e-06,
+      "loss": 0.4702,
+      "step": 8251
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.528213906900626,
+      "learning_rate": 2.66790136993617e-06,
+      "loss": 0.5346,
+      "step": 8252
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.136858860468303,
+      "learning_rate": 2.6667473093535213e-06,
+      "loss": 0.5017,
+      "step": 8253
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.030912797211254,
+      "learning_rate": 2.6655934076550617e-06,
+      "loss": 0.5394,
+      "step": 8254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.631417626452874,
+      "learning_rate": 2.664439664919374e-06,
+      "loss": 0.4648,
+      "step": 8255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.746315074079219,
+      "learning_rate": 2.6632860812250137e-06,
+      "loss": 0.4409,
+      "step": 8256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.673907404428569,
+      "learning_rate": 2.6621326566505422e-06,
+      "loss": 0.5703,
+      "step": 8257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.80390262988036,
+      "learning_rate": 2.6609793912744987e-06,
+      "loss": 0.6001,
+      "step": 8258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.839211112969233,
+      "learning_rate": 2.6598262851754183e-06,
+      "loss": 0.5295,
+      "step": 8259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.79676789720939,
+      "learning_rate": 2.6586733384318205e-06,
+      "loss": 0.4935,
+      "step": 8260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.918609375345259,
+      "learning_rate": 2.6575205511222163e-06,
+      "loss": 0.3802,
+      "step": 8261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.550809979001924,
+      "learning_rate": 2.656367923325104e-06,
+      "loss": 0.4805,
+      "step": 8262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.622791020722167,
+      "learning_rate": 2.6552154551189737e-06,
+      "loss": 0.4808,
+      "step": 8263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.986056732898103,
+      "learning_rate": 2.654063146582307e-06,
+      "loss": 0.376,
+      "step": 8264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.272798461641571,
+      "learning_rate": 2.6529109977935636e-06,
+      "loss": 0.4221,
+      "step": 8265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.878200712693891,
+      "learning_rate": 2.6517590088312063e-06,
+      "loss": 0.4617,
+      "step": 8266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.77803081368587,
+      "learning_rate": 2.6506071797736756e-06,
+      "loss": 0.4732,
+      "step": 8267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.112931454600654,
+      "learning_rate": 2.6494555106994095e-06,
+      "loss": 0.3517,
+      "step": 8268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.9696224569504475,
+      "learning_rate": 2.648304001686831e-06,
+      "loss": 0.4716,
+      "step": 8269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.427337770516878,
+      "learning_rate": 2.647152652814352e-06,
+      "loss": 0.4425,
+      "step": 8270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.475093454987243,
+      "learning_rate": 2.6460014641603725e-06,
+      "loss": 0.46,
+      "step": 8271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.419304994566888,
+      "learning_rate": 2.644850435803285e-06,
+      "loss": 0.6381,
+      "step": 8272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.948214842868103,
+      "learning_rate": 2.6436995678214716e-06,
+      "loss": 0.3538,
+      "step": 8273
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.175554103518477,
+      "learning_rate": 2.642548860293299e-06,
+      "loss": 0.4457,
+      "step": 8274
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.658753203274776,
+      "learning_rate": 2.6413983132971264e-06,
+      "loss": 0.4117,
+      "step": 8275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.16159903591962,
+      "learning_rate": 2.640247926911298e-06,
+      "loss": 0.4375,
+      "step": 8276
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.191237374825896,
+      "learning_rate": 2.639097701214154e-06,
+      "loss": 0.4136,
+      "step": 8277
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.788776303204404,
+      "learning_rate": 2.637947636284018e-06,
+      "loss": 0.4296,
+      "step": 8278
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.847735766381252,
+      "learning_rate": 2.636797732199204e-06,
+      "loss": 0.4224,
+      "step": 8279
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.934376102805288,
+      "learning_rate": 2.6356479890380137e-06,
+      "loss": 0.3988,
+      "step": 8280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 15.57163510215992,
+      "learning_rate": 2.6344984068787436e-06,
+      "loss": 0.5425,
+      "step": 8281
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.990595911186375,
+      "learning_rate": 2.6333489857996715e-06,
+      "loss": 0.4633,
+      "step": 8282
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.59997296422035,
+      "learning_rate": 2.632199725879071e-06,
+      "loss": 0.4206,
+      "step": 8283
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.320607962123052,
+      "learning_rate": 2.6310506271952e-06,
+      "loss": 0.4842,
+      "step": 8284
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.684841733487877,
+      "learning_rate": 2.6299016898263053e-06,
+      "loss": 0.4524,
+      "step": 8285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.317248769771084,
+      "learning_rate": 2.6287529138506272e-06,
+      "loss": 0.4879,
+      "step": 8286
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.597143812537761,
+      "learning_rate": 2.6276042993463925e-06,
+      "loss": 0.4013,
+      "step": 8287
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.402701532106427,
+      "learning_rate": 2.6264558463918153e-06,
+      "loss": 0.447,
+      "step": 8288
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.07390953680953,
+      "learning_rate": 2.625307555065099e-06,
+      "loss": 0.5071,
+      "step": 8289
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.954570806566237,
+      "learning_rate": 2.6241594254444393e-06,
+      "loss": 0.4373,
+      "step": 8290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.7949434090147,
+      "learning_rate": 2.623011457608017e-06,
+      "loss": 0.4431,
+      "step": 8291
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 28.775536405475908,
+      "learning_rate": 2.621863651634006e-06,
+      "loss": 0.4712,
+      "step": 8292
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.667517261004006,
+      "learning_rate": 2.620716007600566e-06,
+      "loss": 0.4564,
+      "step": 8293
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.47903170213634,
+      "learning_rate": 2.619568525585844e-06,
+      "loss": 0.3917,
+      "step": 8294
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.599998604965875,
+      "learning_rate": 2.6184212056679824e-06,
+      "loss": 0.433,
+      "step": 8295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.352104411496515,
+      "learning_rate": 2.6172740479251064e-06,
+      "loss": 0.6204,
+      "step": 8296
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.557343322173781,
+      "learning_rate": 2.6161270524353327e-06,
+      "loss": 0.4261,
+      "step": 8297
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.78565000107161,
+      "learning_rate": 2.614980219276765e-06,
+      "loss": 0.4455,
+      "step": 8298
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.976648151912242,
+      "learning_rate": 2.6138335485275004e-06,
+      "loss": 0.7937,
+      "step": 8299
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.733001828036288,
+      "learning_rate": 2.612687040265619e-06,
+      "loss": 0.5266,
+      "step": 8300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.674982336302374,
+      "learning_rate": 2.611540694569199e-06,
+      "loss": 0.3656,
+      "step": 8301
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.365553096905968,
+      "learning_rate": 2.6103945115162934e-06,
+      "loss": 0.3575,
+      "step": 8302
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.374798915495283,
+      "learning_rate": 2.6092484911849553e-06,
+      "loss": 0.4938,
+      "step": 8303
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.589427156980829,
+      "learning_rate": 2.6081026336532265e-06,
+      "loss": 0.5277,
+      "step": 8304
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.446026228436789,
+      "learning_rate": 2.6069569389991334e-06,
+      "loss": 0.4673,
+      "step": 8305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7012047302137874,
+      "learning_rate": 2.6058114073006913e-06,
+      "loss": 0.7398,
+      "step": 8306
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.226155775958305,
+      "learning_rate": 2.6046660386359045e-06,
+      "loss": 0.4842,
+      "step": 8307
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.18545381572386,
+      "learning_rate": 2.6035208330827723e-06,
+      "loss": 0.476,
+      "step": 8308
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.302587567333427,
+      "learning_rate": 2.602375790719273e-06,
+      "loss": 0.5228,
+      "step": 8309
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.377993553581671,
+      "learning_rate": 2.6012309116233834e-06,
+      "loss": 0.4257,
+      "step": 8310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.018584087230683,
+      "learning_rate": 2.6000861958730626e-06,
+      "loss": 0.5669,
+      "step": 8311
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8855727554401194,
+      "learning_rate": 2.598941643546259e-06,
+      "loss": 0.7196,
+      "step": 8312
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.551372025177286,
+      "learning_rate": 2.597797254720915e-06,
+      "loss": 0.5705,
+      "step": 8313
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.712588676399018,
+      "learning_rate": 2.596653029474957e-06,
+      "loss": 0.4642,
+      "step": 8314
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 40.40490253463443,
+      "learning_rate": 2.595508967886301e-06,
+      "loss": 0.4927,
+      "step": 8315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.619797471456868,
+      "learning_rate": 2.5943650700328516e-06,
+      "loss": 0.3943,
+      "step": 8316
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.498748089433379,
+      "learning_rate": 2.593221335992506e-06,
+      "loss": 0.4684,
+      "step": 8317
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.366205631383712,
+      "learning_rate": 2.592077765843144e-06,
+      "loss": 0.5267,
+      "step": 8318
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 18.113009775791376,
+      "learning_rate": 2.5909343596626418e-06,
+      "loss": 0.4641,
+      "step": 8319
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 26.763016788614014,
+      "learning_rate": 2.589791117528858e-06,
+      "loss": 0.5316,
+      "step": 8320
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6548965613712059,
+      "learning_rate": 2.5886480395196424e-06,
+      "loss": 0.7551,
+      "step": 8321
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 13.622171223379913,
+      "learning_rate": 2.5875051257128312e-06,
+      "loss": 0.4387,
+      "step": 8322
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.340144750152889,
+      "learning_rate": 2.5863623761862573e-06,
+      "loss": 0.5057,
+      "step": 8323
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.241413606313731,
+      "learning_rate": 2.5852197910177328e-06,
+      "loss": 0.5116,
+      "step": 8324
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.860150188859468,
+      "learning_rate": 2.584077370285062e-06,
+      "loss": 0.4263,
+      "step": 8325
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8902840040143323,
+      "learning_rate": 2.582935114066043e-06,
+      "loss": 0.3208,
+      "step": 8326
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.042999002689903,
+      "learning_rate": 2.581793022438453e-06,
+      "loss": 0.471,
+      "step": 8327
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.891811946992925,
+      "learning_rate": 2.5806510954800685e-06,
+      "loss": 0.3998,
+      "step": 8328
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 12.460624179390054,
+      "learning_rate": 2.579509333268647e-06,
+      "loss": 0.46,
+      "step": 8329
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.7312711084079515,
+      "learning_rate": 2.578367735881938e-06,
+      "loss": 0.4377,
+      "step": 8330
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.572496591429333,
+      "learning_rate": 2.5772263033976776e-06,
+      "loss": 0.4952,
+      "step": 8331
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.501675007337172,
+      "learning_rate": 2.576085035893593e-06,
+      "loss": 0.4707,
+      "step": 8332
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.602590589709619,
+      "learning_rate": 2.574943933447404e-06,
+      "loss": 0.4991,
+      "step": 8333
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.374113233501803,
+      "learning_rate": 2.5738029961368078e-06,
+      "loss": 0.4213,
+      "step": 8334
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.459666501870305,
+      "learning_rate": 2.5726622240395015e-06,
+      "loss": 0.512,
+      "step": 8335
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.342022986389368,
+      "learning_rate": 2.5715216172331634e-06,
+      "loss": 0.4749,
+      "step": 8336
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.807929057368973,
+      "learning_rate": 2.570381175795468e-06,
+      "loss": 0.398,
+      "step": 8337
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.271423045304391,
+      "learning_rate": 2.5692408998040724e-06,
+      "loss": 0.5001,
+      "step": 8338
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.8555386392367295,
+      "learning_rate": 2.568100789336624e-06,
+      "loss": 0.4144,
+      "step": 8339
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.831363889660354,
+      "learning_rate": 2.5669608444707568e-06,
+      "loss": 0.4733,
+      "step": 8340
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.168074770037845,
+      "learning_rate": 2.5658210652841005e-06,
+      "loss": 0.4871,
+      "step": 8341
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.599987168906643,
+      "learning_rate": 2.5646814518542674e-06,
+      "loss": 0.3562,
+      "step": 8342
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.479232051157693,
+      "learning_rate": 2.5635420042588575e-06,
+      "loss": 0.4643,
+      "step": 8343
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.877624481117031,
+      "learning_rate": 2.5624027225754667e-06,
+      "loss": 0.3578,
+      "step": 8344
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.509546079236276,
+      "learning_rate": 2.5612636068816705e-06,
+      "loss": 0.3868,
+      "step": 8345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.7371368817586035,
+      "learning_rate": 2.560124657255042e-06,
+      "loss": 0.5958,
+      "step": 8346
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.702453757029314,
+      "learning_rate": 2.5589858737731364e-06,
+      "loss": 0.528,
+      "step": 8347
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.758883926303038,
+      "learning_rate": 2.5578472565135004e-06,
+      "loss": 0.5087,
+      "step": 8348
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.272027782984969,
+      "learning_rate": 2.556708805553666e-06,
+      "loss": 0.4737,
+      "step": 8349
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9792518082240633,
+      "learning_rate": 2.5555705209711605e-06,
+      "loss": 0.7917,
+      "step": 8350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.53819755598914,
+      "learning_rate": 2.5544324028434954e-06,
+      "loss": 0.5621,
+      "step": 8351
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.786837979868887,
+      "learning_rate": 2.5532944512481684e-06,
+      "loss": 0.3977,
+      "step": 8352
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.060099829589111,
+      "learning_rate": 2.5521566662626734e-06,
+      "loss": 0.5127,
+      "step": 8353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.279443240158875,
+      "learning_rate": 2.5510190479644846e-06,
+      "loss": 0.4319,
+      "step": 8354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.865391481442387,
+      "learning_rate": 2.549881596431072e-06,
+      "loss": 0.3571,
+      "step": 8355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.95907170493632,
+      "learning_rate": 2.5487443117398902e-06,
+      "loss": 0.6249,
+      "step": 8356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.1407254203205195,
+      "learning_rate": 2.5476071939683824e-06,
+      "loss": 0.4719,
+      "step": 8357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5082969327844546,
+      "learning_rate": 2.5464702431939793e-06,
+      "loss": 0.7216,
+      "step": 8358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.35866599718736,
+      "learning_rate": 2.5453334594941064e-06,
+      "loss": 0.5771,
+      "step": 8359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8374134984220254,
+      "learning_rate": 2.5441968429461717e-06,
+      "loss": 0.8486,
+      "step": 8360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.804689559732302,
+      "learning_rate": 2.5430603936275744e-06,
+      "loss": 0.4096,
+      "step": 8361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.130553835666047,
+      "learning_rate": 2.5419241116156986e-06,
+      "loss": 0.4144,
+      "step": 8362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.276773800027273,
+      "learning_rate": 2.5407879969879224e-06,
+      "loss": 0.4696,
+      "step": 8363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.541831850592086,
+      "learning_rate": 2.5396520498216127e-06,
+      "loss": 0.3942,
+      "step": 8364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.340849822961636,
+      "learning_rate": 2.5385162701941195e-06,
+      "loss": 0.4396,
+      "step": 8365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8106733267006176,
+      "learning_rate": 2.5373806581827855e-06,
+      "loss": 0.7523,
+      "step": 8366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.051525813989691,
+      "learning_rate": 2.536245213864938e-06,
+      "loss": 0.4057,
+      "step": 8367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.737384142735209,
+      "learning_rate": 2.5351099373179e-06,
+      "loss": 0.5852,
+      "step": 8368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 14.324210656240762,
+      "learning_rate": 2.5339748286189752e-06,
+      "loss": 0.4469,
+      "step": 8369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.042514364811788,
+      "learning_rate": 2.532839887845465e-06,
+      "loss": 0.511,
+      "step": 8370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.640462366778494,
+      "learning_rate": 2.5317051150746463e-06,
+      "loss": 0.4971,
+      "step": 8371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.837947532426396,
+      "learning_rate": 2.5305705103837973e-06,
+      "loss": 0.4813,
+      "step": 8372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.457150713232199,
+      "learning_rate": 2.529436073850176e-06,
+      "loss": 0.4341,
+      "step": 8373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.568050105019926,
+      "learning_rate": 2.5283018055510365e-06,
+      "loss": 0.4735,
+      "step": 8374
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.979726485802134,
+      "learning_rate": 2.5271677055636163e-06,
+      "loss": 0.444,
+      "step": 8375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8358287906354898,
+      "learning_rate": 2.52603377396514e-06,
+      "loss": 0.7165,
+      "step": 8376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.586783570231614,
+      "learning_rate": 2.524900010832826e-06,
+      "loss": 0.4528,
+      "step": 8377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.24667297141799,
+      "learning_rate": 2.5237664162438765e-06,
+      "loss": 0.3827,
+      "step": 8378
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.953639576249937,
+      "learning_rate": 2.522632990275489e-06,
+      "loss": 0.4513,
+      "step": 8379
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.513698878395498,
+      "learning_rate": 2.5214997330048375e-06,
+      "loss": 0.4287,
+      "step": 8380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.682750581907545,
+      "learning_rate": 2.520366644509098e-06,
+      "loss": 0.4869,
+      "step": 8381
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.420638020998362,
+      "learning_rate": 2.5192337248654236e-06,
+      "loss": 0.6232,
+      "step": 8382
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.69758472225401,
+      "learning_rate": 2.5181009741509667e-06,
+      "loss": 0.5311,
+      "step": 8383
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.715974656090575,
+      "learning_rate": 2.516968392442859e-06,
+      "loss": 0.4491,
+      "step": 8384
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.965495678470104,
+      "learning_rate": 2.515835979818224e-06,
+      "loss": 0.4739,
+      "step": 8385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.351148419600344,
+      "learning_rate": 2.514703736354177e-06,
+      "loss": 0.6391,
+      "step": 8386
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.248711770093001,
+      "learning_rate": 2.513571662127815e-06,
+      "loss": 0.5225,
+      "step": 8387
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.60850561904693,
+      "learning_rate": 2.512439757216233e-06,
+      "loss": 0.4567,
+      "step": 8388
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.473043301069546,
+      "learning_rate": 2.5113080216965e-06,
+      "loss": 0.3858,
+      "step": 8389
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.099217097456793,
+      "learning_rate": 2.51017645564569e-06,
+      "loss": 0.4974,
+      "step": 8390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.879447171938422,
+      "learning_rate": 2.5090450591408532e-06,
+      "loss": 0.5624,
+      "step": 8391
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.996563595513791,
+      "learning_rate": 2.5079138322590357e-06,
+      "loss": 0.5208,
+      "step": 8392
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.788312720896016,
+      "learning_rate": 2.5067827750772674e-06,
+      "loss": 0.4301,
+      "step": 8393
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.739503871272222,
+      "learning_rate": 2.5056518876725667e-06,
+      "loss": 0.5318,
+      "step": 8394
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.577462061047846,
+      "learning_rate": 2.504521170121945e-06,
+      "loss": 0.4083,
+      "step": 8395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.417932630065632,
+      "learning_rate": 2.5033906225023974e-06,
+      "loss": 0.482,
+      "step": 8396
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.997483140009356,
+      "learning_rate": 2.5022602448909126e-06,
+      "loss": 0.4433,
+      "step": 8397
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.112668230528898,
+      "learning_rate": 2.501130037364458e-06,
+      "loss": 0.4936,
+      "step": 8398
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.852395171881739,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.481,
+      "step": 8399
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.0891596782925905,
+      "learning_rate": 2.4988701328744887e-06,
+      "loss": 0.4322,
+      "step": 8400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.126896592898223,
+      "learning_rate": 2.497740436064863e-06,
+      "loss": 0.4343,
+      "step": 8401
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.575835890744914,
+      "learning_rate": 2.4966109096480505e-06,
+      "loss": 0.5688,
+      "step": 8402
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.8200873240701005,
+      "learning_rate": 2.4954815537009646e-06,
+      "loss": 0.4682,
+      "step": 8403
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.626915463729746,
+      "learning_rate": 2.494352368300513e-06,
+      "loss": 0.4398,
+      "step": 8404
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.847219848326296,
+      "learning_rate": 2.493223353523585e-06,
+      "loss": 0.5217,
+      "step": 8405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.753200259273822,
+      "learning_rate": 2.4920945094470645e-06,
+      "loss": 0.4909,
+      "step": 8406
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.375622901958085,
+      "learning_rate": 2.490965836147819e-06,
+      "loss": 0.5064,
+      "step": 8407
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.694255705715502,
+      "learning_rate": 2.489837333702707e-06,
+      "loss": 0.5011,
+      "step": 8408
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.836271028243706,
+      "learning_rate": 2.488709002188572e-06,
+      "loss": 0.4002,
+      "step": 8409
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.952311532001265,
+      "learning_rate": 2.4875808416822524e-06,
+      "loss": 0.7892,
+      "step": 8410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.169585523531593,
+      "learning_rate": 2.4864528522605684e-06,
+      "loss": 0.4835,
+      "step": 8411
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.973324008459716,
+      "learning_rate": 2.485325034000332e-06,
+      "loss": 0.5154,
+      "step": 8412
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.788430468348612,
+      "learning_rate": 2.4841973869783403e-06,
+      "loss": 0.5379,
+      "step": 8413
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.09897055743218,
+      "learning_rate": 2.483069911271383e-06,
+      "loss": 0.473,
+      "step": 8414
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.4362768895826425,
+      "learning_rate": 2.481942606956238e-06,
+      "loss": 0.436,
+      "step": 8415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.146746092784682,
+      "learning_rate": 2.4808154741096684e-06,
+      "loss": 0.4558,
+      "step": 8416
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.160604642385182,
+      "learning_rate": 2.4796885128084264e-06,
+      "loss": 0.4473,
+      "step": 8417
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.276864125208138,
+      "learning_rate": 2.4785617231292513e-06,
+      "loss": 0.4457,
+      "step": 8418
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.421095110323542,
+      "learning_rate": 2.477435105148876e-06,
+      "loss": 0.5611,
+      "step": 8419
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.238934033836905,
+      "learning_rate": 2.476308658944017e-06,
+      "loss": 0.4036,
+      "step": 8420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.980940130593611,
+      "learning_rate": 2.47518238459138e-06,
+      "loss": 0.4604,
+      "step": 8421
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.921744372598869,
+      "learning_rate": 2.4740562821676572e-06,
+      "loss": 0.4567,
+      "step": 8422
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8727948108736165,
+      "learning_rate": 2.472930351749533e-06,
+      "loss": 0.7977,
+      "step": 8423
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.485467459067358,
+      "learning_rate": 2.471804593413681e-06,
+      "loss": 0.5209,
+      "step": 8424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.5709103225027405,
+      "learning_rate": 2.470679007236757e-06,
+      "loss": 0.4539,
+      "step": 8425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.325095753947501,
+      "learning_rate": 2.46955359329541e-06,
+      "loss": 0.399,
+      "step": 8426
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.782939567603281,
+      "learning_rate": 2.4684283516662727e-06,
+      "loss": 0.4986,
+      "step": 8427
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.406244847427008,
+      "learning_rate": 2.4673032824259736e-06,
+      "loss": 0.5081,
+      "step": 8428
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.402373148224667,
+      "learning_rate": 2.4661783856511227e-06,
+      "loss": 0.5548,
+      "step": 8429
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.648327380471803,
+      "learning_rate": 2.4650536614183206e-06,
+      "loss": 0.5355,
+      "step": 8430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.28112470011452,
+      "learning_rate": 2.463929109804154e-06,
+      "loss": 0.6464,
+      "step": 8431
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.261847298046554,
+      "learning_rate": 2.4628047308852044e-06,
+      "loss": 0.5008,
+      "step": 8432
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6424622939290832,
+      "learning_rate": 2.4616805247380314e-06,
+      "loss": 0.7382,
+      "step": 8433
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.916179399205672,
+      "learning_rate": 2.4605564914391942e-06,
+      "loss": 0.4848,
+      "step": 8434
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.508977681671639,
+      "learning_rate": 2.459432631065232e-06,
+      "loss": 0.4268,
+      "step": 8435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.508569541797374,
+      "learning_rate": 2.458308943692673e-06,
+      "loss": 0.555,
+      "step": 8436
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 22.858569457231354,
+      "learning_rate": 2.4571854293980385e-06,
+      "loss": 0.4882,
+      "step": 8437
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.138972944374872,
+      "learning_rate": 2.456062088257833e-06,
+      "loss": 0.4803,
+      "step": 8438
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.952446765110848,
+      "learning_rate": 2.4549389203485517e-06,
+      "loss": 0.5007,
+      "step": 8439
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.269740962129305,
+      "learning_rate": 2.4538159257466754e-06,
+      "loss": 0.5587,
+      "step": 8440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.713473558726356,
+      "learning_rate": 2.4526931045286784e-06,
+      "loss": 0.4653,
+      "step": 8441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.220689639580762,
+      "learning_rate": 2.451570456771016e-06,
+      "loss": 0.4027,
+      "step": 8442
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.8471141216852125,
+      "learning_rate": 2.45044798255014e-06,
+      "loss": 0.4882,
+      "step": 8443
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.5832685239244455,
+      "learning_rate": 2.4493256819424833e-06,
+      "loss": 0.4077,
+      "step": 8444
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5990932139187397,
+      "learning_rate": 2.4482035550244682e-06,
+      "loss": 0.7052,
+      "step": 8445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.317798095490597,
+      "learning_rate": 2.4470816018725103e-06,
+      "loss": 0.4913,
+      "step": 8446
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.116564537913475,
+      "learning_rate": 2.4459598225630077e-06,
+      "loss": 0.4391,
+      "step": 8447
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.326008899257364,
+      "learning_rate": 2.4448382171723482e-06,
+      "loss": 0.5498,
+      "step": 8448
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.309985087278461,
+      "learning_rate": 2.4437167857769073e-06,
+      "loss": 0.556,
+      "step": 8449
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7184051658097983,
+      "learning_rate": 2.442595528453052e-06,
+      "loss": 0.6666,
+      "step": 8450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.9599415976036445,
+      "learning_rate": 2.441474445277133e-06,
+      "loss": 0.5091,
+      "step": 8451
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 20.130387704046232,
+      "learning_rate": 2.4403535363254927e-06,
+      "loss": 0.4862,
+      "step": 8452
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.099813711354454,
+      "learning_rate": 2.4392328016744595e-06,
+      "loss": 0.4896,
+      "step": 8453
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.502781322484505,
+      "learning_rate": 2.4381122414003483e-06,
+      "loss": 0.4521,
+      "step": 8454
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.91894556997297,
+      "learning_rate": 2.436991855579468e-06,
+      "loss": 0.5616,
+      "step": 8455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.012672266390863,
+      "learning_rate": 2.43587164428811e-06,
+      "loss": 0.4192,
+      "step": 8456
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.481782379321548,
+      "learning_rate": 2.4347516076025567e-06,
+      "loss": 0.4492,
+      "step": 8457
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 23.474724737792506,
+      "learning_rate": 2.4336317455990744e-06,
+      "loss": 0.5304,
+      "step": 8458
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.740374875567073,
+      "learning_rate": 2.4325120583539246e-06,
+      "loss": 0.4808,
+      "step": 8459
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.763503836669693,
+      "learning_rate": 2.43139254594335e-06,
+      "loss": 0.4771,
+      "step": 8460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.04015205254057,
+      "learning_rate": 2.4302732084435882e-06,
+      "loss": 0.3761,
+      "step": 8461
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9598236935965931,
+      "learning_rate": 2.4291540459308583e-06,
+      "loss": 0.7138,
+      "step": 8462
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.778482131127479,
+      "learning_rate": 2.4280350584813695e-06,
+      "loss": 0.4383,
+      "step": 8463
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.503703313166499,
+      "learning_rate": 2.4269162461713237e-06,
+      "loss": 0.6204,
+      "step": 8464
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.771107193162395,
+      "learning_rate": 2.425797609076902e-06,
+      "loss": 0.4518,
+      "step": 8465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.171462748259504,
+      "learning_rate": 2.424679147274286e-06,
+      "loss": 0.4413,
+      "step": 8466
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.384315428385744,
+      "learning_rate": 2.423560860839629e-06,
+      "loss": 0.4135,
+      "step": 8467
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 19.80600769120747,
+      "learning_rate": 2.4224427498490877e-06,
+      "loss": 0.3977,
+      "step": 8468
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.227948143534077,
+      "learning_rate": 2.421324814378797e-06,
+      "loss": 0.3778,
+      "step": 8469
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.981360177692674,
+      "learning_rate": 2.420207054504887e-06,
+      "loss": 0.5162,
+      "step": 8470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.193359033962748,
+      "learning_rate": 2.4190894703034694e-06,
+      "loss": 0.4411,
+      "step": 8471
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.410374347272899,
+      "learning_rate": 2.417972061850648e-06,
+      "loss": 0.4144,
+      "step": 8472
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.246227729492414,
+      "learning_rate": 2.416854829222511e-06,
+      "loss": 0.4686,
+      "step": 8473
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 19.479793531344214,
+      "learning_rate": 2.4157377724951384e-06,
+      "loss": 0.4228,
+      "step": 8474
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.288730476428201,
+      "learning_rate": 2.414620891744602e-06,
+      "loss": 0.5093,
+      "step": 8475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.1057655378199645,
+      "learning_rate": 2.413504187046947e-06,
+      "loss": 0.4322,
+      "step": 8476
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.4198038061250084,
+      "learning_rate": 2.412387658478223e-06,
+      "loss": 0.4601,
+      "step": 8477
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.176828107954057,
+      "learning_rate": 2.4112713061144566e-06,
+      "loss": 0.4977,
+      "step": 8478
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.465343263417026,
+      "learning_rate": 2.4101551300316704e-06,
+      "loss": 0.4507,
+      "step": 8479
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.246947013552541,
+      "learning_rate": 2.4090391303058687e-06,
+      "loss": 0.4448,
+      "step": 8480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.707117606236512,
+      "learning_rate": 2.4079233070130466e-06,
+      "loss": 0.5522,
+      "step": 8481
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.068903509460999,
+      "learning_rate": 2.4068076602291845e-06,
+      "loss": 0.558,
+      "step": 8482
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 15.254690397839655,
+      "learning_rate": 2.4056921900302547e-06,
+      "loss": 0.458,
+      "step": 8483
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.154299450629704,
+      "learning_rate": 2.4045768964922204e-06,
+      "loss": 0.4062,
+      "step": 8484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 25.912415888467223,
+      "learning_rate": 2.4034617796910204e-06,
+      "loss": 0.467,
+      "step": 8485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.878865090251958,
+      "learning_rate": 2.402346839702594e-06,
+      "loss": 0.4835,
+      "step": 8486
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.248380714642936,
+      "learning_rate": 2.4012320766028607e-06,
+      "loss": 0.5365,
+      "step": 8487
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.335635195087593,
+      "learning_rate": 2.4001174904677344e-06,
+      "loss": 0.5384,
+      "step": 8488
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.76848168498275,
+      "learning_rate": 2.399003081373112e-06,
+      "loss": 0.4256,
+      "step": 8489
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.886631651824431,
+      "learning_rate": 2.397888849394879e-06,
+      "loss": 0.5005,
+      "step": 8490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.56200178529093,
+      "learning_rate": 2.3967747946089094e-06,
+      "loss": 0.4226,
+      "step": 8491
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.041828230360391,
+      "learning_rate": 2.3956609170910673e-06,
+      "loss": 0.3918,
+      "step": 8492
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.655118943331605,
+      "learning_rate": 2.394547216917202e-06,
+      "loss": 0.4788,
+      "step": 8493
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 12.473450920871652,
+      "learning_rate": 2.393433694163149e-06,
+      "loss": 0.4023,
+      "step": 8494
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.3489211207864304,
+      "learning_rate": 2.3923203489047393e-06,
+      "loss": 0.4758,
+      "step": 8495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.367899498807331,
+      "learning_rate": 2.3912071812177817e-06,
+      "loss": 0.4587,
+      "step": 8496
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.894695136822928,
+      "learning_rate": 2.3900941911780827e-06,
+      "loss": 0.3856,
+      "step": 8497
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.467310062382953,
+      "learning_rate": 2.388981378861429e-06,
+      "loss": 0.4798,
+      "step": 8498
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.471572291607465,
+      "learning_rate": 2.3878687443435995e-06,
+      "loss": 0.4015,
+      "step": 8499
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.716738496361494,
+      "learning_rate": 2.386756287700357e-06,
+      "loss": 0.447,
+      "step": 8500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.94789005130783,
+      "learning_rate": 2.385644009007459e-06,
+      "loss": 0.4305,
+      "step": 8501
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.741764187729974,
+      "learning_rate": 2.3845319083406427e-06,
+      "loss": 0.4441,
+      "step": 8502
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.082583902046768,
+      "learning_rate": 2.3834199857756428e-06,
+      "loss": 0.5,
+      "step": 8503
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7311186900384068,
+      "learning_rate": 2.382308241388169e-06,
+      "loss": 0.7542,
+      "step": 8504
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.926981601966539,
+      "learning_rate": 2.381196675253931e-06,
+      "loss": 0.4457,
+      "step": 8505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.716193290453301,
+      "learning_rate": 2.3800852874486212e-06,
+      "loss": 0.364,
+      "step": 8506
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.445378048901853,
+      "learning_rate": 2.37897407804792e-06,
+      "loss": 0.5527,
+      "step": 8507
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.261484602998139,
+      "learning_rate": 2.377863047127495e-06,
+      "loss": 0.4096,
+      "step": 8508
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.651427771742239,
+      "learning_rate": 2.376752194763001e-06,
+      "loss": 0.5331,
+      "step": 8509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.523034290153022,
+      "learning_rate": 2.375641521030086e-06,
+      "loss": 0.4365,
+      "step": 8510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.457792523035693,
+      "learning_rate": 2.374531026004378e-06,
+      "loss": 0.3733,
+      "step": 8511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.467876263617779,
+      "learning_rate": 2.3734207097615026e-06,
+      "loss": 0.4483,
+      "step": 8512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 27.718292144163094,
+      "learning_rate": 2.3723105723770596e-06,
+      "loss": 0.4925,
+      "step": 8513
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.011164429850615,
+      "learning_rate": 2.3712006139266484e-06,
+      "loss": 0.45,
+      "step": 8514
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.229802116785182,
+      "learning_rate": 2.3700908344858543e-06,
+      "loss": 0.4452,
+      "step": 8515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.370581577400184,
+      "learning_rate": 2.368981234130246e-06,
+      "loss": 0.5217,
+      "step": 8516
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.469805233930124,
+      "learning_rate": 2.367871812935383e-06,
+      "loss": 0.5208,
+      "step": 8517
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.88837982578067,
+      "learning_rate": 2.36676257097681e-06,
+      "loss": 0.4432,
+      "step": 8518
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7867688286389705,
+      "learning_rate": 2.3656535083300646e-06,
+      "loss": 0.7502,
+      "step": 8519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.792582137917243,
+      "learning_rate": 2.3645446250706662e-06,
+      "loss": 0.4324,
+      "step": 8520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.441727860577986,
+      "learning_rate": 2.36343592127413e-06,
+      "loss": 0.5104,
+      "step": 8521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.940568943298557,
+      "learning_rate": 2.3623273970159462e-06,
+      "loss": 0.4396,
+      "step": 8522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.208669357529233,
+      "learning_rate": 2.361219052371606e-06,
+      "loss": 0.4834,
+      "step": 8523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 19.33311646797293,
+      "learning_rate": 2.3601108874165796e-06,
+      "loss": 0.5239,
+      "step": 8524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8422416362611507,
+      "learning_rate": 2.359002902226331e-06,
+      "loss": 0.7622,
+      "step": 8525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.9227642023565705,
+      "learning_rate": 2.357895096876308e-06,
+      "loss": 0.3724,
+      "step": 8526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.160636815236785,
+      "learning_rate": 2.3567874714419454e-06,
+      "loss": 0.447,
+      "step": 8527
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.737783689949892,
+      "learning_rate": 2.3556800259986705e-06,
+      "loss": 0.6156,
+      "step": 8528
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.899725659782887,
+      "learning_rate": 2.3545727606218926e-06,
+      "loss": 0.4786,
+      "step": 8529
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.503903144678186,
+      "learning_rate": 2.3534656753870167e-06,
+      "loss": 0.4213,
+      "step": 8530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.161633370561175,
+      "learning_rate": 2.3523587703694227e-06,
+      "loss": 0.5117,
+      "step": 8531
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8875620741751487,
+      "learning_rate": 2.351252045644492e-06,
+      "loss": 0.657,
+      "step": 8532
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.447127746536065,
+      "learning_rate": 2.350145501287584e-06,
+      "loss": 0.5374,
+      "step": 8533
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.903815334729887,
+      "learning_rate": 2.349039137374053e-06,
+      "loss": 0.3849,
+      "step": 8534
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.225928085249856,
+      "learning_rate": 2.347932953979235e-06,
+      "loss": 0.5172,
+      "step": 8535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.26471207090271,
+      "learning_rate": 2.346826951178456e-06,
+      "loss": 0.5062,
+      "step": 8536
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.959738117803344,
+      "learning_rate": 2.3457211290470316e-06,
+      "loss": 0.538,
+      "step": 8537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 36.99693671325953,
+      "learning_rate": 2.3446154876602617e-06,
+      "loss": 0.4459,
+      "step": 8538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.290249653742539,
+      "learning_rate": 2.3435100270934397e-06,
+      "loss": 0.4876,
+      "step": 8539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.085039203026689,
+      "learning_rate": 2.3424047474218353e-06,
+      "loss": 0.4637,
+      "step": 8540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.194650164965142,
+      "learning_rate": 2.341299648720719e-06,
+      "loss": 0.4831,
+      "step": 8541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.472654267445689,
+      "learning_rate": 2.3401947310653394e-06,
+      "loss": 0.4633,
+      "step": 8542
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 35.99250147919997,
+      "learning_rate": 2.3390899945309407e-06,
+      "loss": 0.5066,
+      "step": 8543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.723321120221548,
+      "learning_rate": 2.337985439192748e-06,
+      "loss": 0.4592,
+      "step": 8544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.015290077053302,
+      "learning_rate": 2.3368810651259757e-06,
+      "loss": 0.4368,
+      "step": 8545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 56.102419282534456,
+      "learning_rate": 2.335776872405829e-06,
+      "loss": 0.5083,
+      "step": 8546
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.054397722248105,
+      "learning_rate": 2.334672861107496e-06,
+      "loss": 0.4389,
+      "step": 8547
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.399291255626862,
+      "learning_rate": 2.3335690313061583e-06,
+      "loss": 0.5248,
+      "step": 8548
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 24.25070600486648,
+      "learning_rate": 2.33246538307698e-06,
+      "loss": 0.5061,
+      "step": 8549
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.10605775518798,
+      "learning_rate": 2.331361916495115e-06,
+      "loss": 0.563,
+      "step": 8550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.418340342595648,
+      "learning_rate": 2.3302586316357024e-06,
+      "loss": 0.4983,
+      "step": 8551
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.6493870638053005,
+      "learning_rate": 2.3291555285738743e-06,
+      "loss": 0.4956,
+      "step": 8552
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.421464412507502,
+      "learning_rate": 2.328052607384746e-06,
+      "loss": 0.4317,
+      "step": 8553
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.59350278668601,
+      "learning_rate": 2.326949868143419e-06,
+      "loss": 0.4276,
+      "step": 8554
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.587823304463882,
+      "learning_rate": 2.325847310924989e-06,
+      "loss": 0.4556,
+      "step": 8555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.675905364393403,
+      "learning_rate": 2.3247449358045322e-06,
+      "loss": 0.4501,
+      "step": 8556
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.678241538676846,
+      "learning_rate": 2.3236427428571185e-06,
+      "loss": 0.468,
+      "step": 8557
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.912687631552218,
+      "learning_rate": 2.3225407321578003e-06,
+      "loss": 0.4826,
+      "step": 8558
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.032258135498734,
+      "learning_rate": 2.3214389037816197e-06,
+      "loss": 0.4964,
+      "step": 8559
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.99883633975859,
+      "learning_rate": 2.3203372578036056e-06,
+      "loss": 0.3471,
+      "step": 8560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.047032195476314,
+      "learning_rate": 2.3192357942987775e-06,
+      "loss": 0.4376,
+      "step": 8561
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.639739674309688,
+      "learning_rate": 2.318134513342139e-06,
+      "loss": 0.4743,
+      "step": 8562
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.921433438841955,
+      "learning_rate": 2.3170334150086827e-06,
+      "loss": 0.4102,
+      "step": 8563
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.889251323434111,
+      "learning_rate": 2.315932499373386e-06,
+      "loss": 0.4264,
+      "step": 8564
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.52231155459897,
+      "learning_rate": 2.3148317665112184e-06,
+      "loss": 0.5852,
+      "step": 8565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.802567502914844,
+      "learning_rate": 2.3137312164971377e-06,
+      "loss": 0.5102,
+      "step": 8566
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.144389271619197,
+      "learning_rate": 2.3126308494060834e-06,
+      "loss": 0.4508,
+      "step": 8567
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.68887016728004,
+      "learning_rate": 2.3115306653129867e-06,
+      "loss": 0.4654,
+      "step": 8568
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.75858690214661,
+      "learning_rate": 2.310430664292762e-06,
+      "loss": 0.4328,
+      "step": 8569
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.880354846178724,
+      "learning_rate": 2.30933084642032e-06,
+      "loss": 0.4622,
+      "step": 8570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.89843465424081,
+      "learning_rate": 2.30823121177055e-06,
+      "loss": 0.5043,
+      "step": 8571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.750987386847673,
+      "learning_rate": 2.3071317604183334e-06,
+      "loss": 0.444,
+      "step": 8572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.825797247372465,
+      "learning_rate": 2.306032492438536e-06,
+      "loss": 0.4773,
+      "step": 8573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.134220837480662,
+      "learning_rate": 2.304933407906014e-06,
+      "loss": 0.4913,
+      "step": 8574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.070424285183325,
+      "learning_rate": 2.3038345068956123e-06,
+      "loss": 0.4694,
+      "step": 8575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.214650691708197,
+      "learning_rate": 2.3027357894821605e-06,
+      "loss": 0.4904,
+      "step": 8576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.798496227989345,
+      "learning_rate": 2.3016372557404748e-06,
+      "loss": 0.4697,
+      "step": 8577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.404816897699964,
+      "learning_rate": 2.3005389057453604e-06,
+      "loss": 0.4717,
+      "step": 8578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.385730755552666,
+      "learning_rate": 2.299440739571612e-06,
+      "loss": 0.4988,
+      "step": 8579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.633215482843183,
+      "learning_rate": 2.2983427572940094e-06,
+      "loss": 0.7366,
+      "step": 8580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.663480446541049,
+      "learning_rate": 2.297244958987319e-06,
+      "loss": 0.387,
+      "step": 8581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.412637098048178,
+      "learning_rate": 2.2961473447262953e-06,
+      "loss": 0.397,
+      "step": 8582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.788604427888426,
+      "learning_rate": 2.295049914585685e-06,
+      "loss": 0.5326,
+      "step": 8583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.334770762323364,
+      "learning_rate": 2.293952668640213e-06,
+      "loss": 0.4305,
+      "step": 8584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.333335044653351,
+      "learning_rate": 2.2928556069646018e-06,
+      "loss": 0.4798,
+      "step": 8585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.145997288540878,
+      "learning_rate": 2.2917587296335543e-06,
+      "loss": 0.4422,
+      "step": 8586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.733445025558634,
+      "learning_rate": 2.290662036721761e-06,
+      "loss": 0.5264,
+      "step": 8587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.1937658362534584,
+      "learning_rate": 2.2895655283039052e-06,
+      "loss": 0.5137,
+      "step": 8588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 18.453173827034778,
+      "learning_rate": 2.2884692044546537e-06,
+      "loss": 0.4174,
+      "step": 8589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.009967662403863,
+      "learning_rate": 2.2873730652486596e-06,
+      "loss": 0.5242,
+      "step": 8590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.796702284005117,
+      "learning_rate": 2.2862771107605647e-06,
+      "loss": 0.4948,
+      "step": 8591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.888918389820799,
+      "learning_rate": 2.285181341065002e-06,
+      "loss": 0.4398,
+      "step": 8592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.436738102077553,
+      "learning_rate": 2.2840857562365842e-06,
+      "loss": 0.4854,
+      "step": 8593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.422296829776924,
+      "learning_rate": 2.28299035634992e-06,
+      "loss": 0.4249,
+      "step": 8594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.100158732789358,
+      "learning_rate": 2.281895141479599e-06,
+      "loss": 0.4859,
+      "step": 8595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.445506408702879,
+      "learning_rate": 2.2808001117001997e-06,
+      "loss": 0.5104,
+      "step": 8596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 30.94417371686238,
+      "learning_rate": 2.279705267086292e-06,
+      "loss": 0.4419,
+      "step": 8597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.331052969991267,
+      "learning_rate": 2.2786106077124277e-06,
+      "loss": 0.4194,
+      "step": 8598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.570851512391831,
+      "learning_rate": 2.277516133653148e-06,
+      "loss": 0.5175,
+      "step": 8599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.962704261809225,
+      "learning_rate": 2.2764218449829804e-06,
+      "loss": 0.4061,
+      "step": 8600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.17817354444738,
+      "learning_rate": 2.2753277417764457e-06,
+      "loss": 0.4784,
+      "step": 8601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.016005691882443,
+      "learning_rate": 2.2742338241080425e-06,
+      "loss": 0.5325,
+      "step": 8602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7545357276025082,
+      "learning_rate": 2.273140092052266e-06,
+      "loss": 0.6736,
+      "step": 8603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 13.328963744570464,
+      "learning_rate": 2.2720465456835918e-06,
+      "loss": 0.4106,
+      "step": 8604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.149274770445103,
+      "learning_rate": 2.270953185076485e-06,
+      "loss": 0.4977,
+      "step": 8605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 15.921987890613453,
+      "learning_rate": 2.269860010305402e-06,
+      "loss": 0.4542,
+      "step": 8606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.684407339340719,
+      "learning_rate": 2.2687670214447786e-06,
+      "loss": 0.395,
+      "step": 8607
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.644074542589994,
+      "learning_rate": 2.267674218569049e-06,
+      "loss": 0.4262,
+      "step": 8608
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.9035076253663075,
+      "learning_rate": 2.266581601752621e-06,
+      "loss": 0.5721,
+      "step": 8609
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.0170484165354265,
+      "learning_rate": 2.2654891710699018e-06,
+      "loss": 0.4763,
+      "step": 8610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.044225385159121,
+      "learning_rate": 2.2643969265952778e-06,
+      "loss": 0.4668,
+      "step": 8611
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 27.58113779644849,
+      "learning_rate": 2.263304868403129e-06,
+      "loss": 0.453,
+      "step": 8612
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.76702469005677,
+      "learning_rate": 2.262212996567819e-06,
+      "loss": 0.3868,
+      "step": 8613
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.908625177899079,
+      "learning_rate": 2.261121311163697e-06,
+      "loss": 0.4239,
+      "step": 8614
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.153826076932114,
+      "learning_rate": 2.260029812265106e-06,
+      "loss": 0.4558,
+      "step": 8615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.441156253732774,
+      "learning_rate": 2.2589384999463676e-06,
+      "loss": 0.5321,
+      "step": 8616
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.239412415420867,
+      "learning_rate": 2.257847374281802e-06,
+      "loss": 0.5759,
+      "step": 8617
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.0862880930274335,
+      "learning_rate": 2.2567564353457023e-06,
+      "loss": 0.4865,
+      "step": 8618
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.047379228264797,
+      "learning_rate": 2.2556656832123617e-06,
+      "loss": 0.5128,
+      "step": 8619
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.8356614694680236,
+      "learning_rate": 2.2545751179560527e-06,
+      "loss": 0.4888,
+      "step": 8620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.631576875944434,
+      "learning_rate": 2.2534847396510413e-06,
+      "loss": 0.5169,
+      "step": 8621
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.04008197137332,
+      "learning_rate": 2.2523945483715747e-06,
+      "loss": 0.4994,
+      "step": 8622
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.897629218069863,
+      "learning_rate": 2.251304544191892e-06,
+      "loss": 0.5135,
+      "step": 8623
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.907883108513862,
+      "learning_rate": 2.250214727186214e-06,
+      "loss": 0.3706,
+      "step": 8624
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.48065358113969,
+      "learning_rate": 2.2491250974287555e-06,
+      "loss": 0.5131,
+      "step": 8625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.135863310275216,
+      "learning_rate": 2.248035654993718e-06,
+      "loss": 0.4575,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4852973569366874,
+      "learning_rate": 2.2469463999552815e-06,
+      "loss": 0.8165,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.50847219337655,
+      "learning_rate": 2.245857332387625e-06,
+      "loss": 0.5689,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.993691424066995,
+      "learning_rate": 2.244768452364904e-06,
+      "loss": 0.4136,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.63494825481348,
+      "learning_rate": 2.2436797599612716e-06,
+      "loss": 0.425,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.052936010510939,
+      "learning_rate": 2.2425912552508603e-06,
+      "loss": 0.5106,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.421532879923172,
+      "learning_rate": 2.2415029383077935e-06,
+      "loss": 0.3486,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.54917935545021,
+      "learning_rate": 2.240414809206177e-06,
+      "loss": 0.5328,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.8158000064923066,
+      "learning_rate": 2.2393268680201136e-06,
+      "loss": 0.4629,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.811240622089398,
+      "learning_rate": 2.238239114823684e-06,
+      "loss": 0.5006,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.186405626646764,
+      "learning_rate": 2.2371515496909572e-06,
+      "loss": 0.5232,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.9380494065522695,
+      "learning_rate": 2.2360641726959965e-06,
+      "loss": 0.5004,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.4131002606136525,
+      "learning_rate": 2.2349769839128434e-06,
+      "loss": 0.4751,
+      "step": 8638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.311865906518162,
+      "learning_rate": 2.2338899834155337e-06,
+      "loss": 0.5014,
+      "step": 8639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.149205586973366,
+      "learning_rate": 2.2328031712780858e-06,
+      "loss": 0.4956,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.466316364210822,
+      "learning_rate": 2.2317165475745066e-06,
+      "loss": 0.4649,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.054811091071187,
+      "learning_rate": 2.230630112378789e-06,
+      "loss": 0.3949,
+      "step": 8642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9230937970520272,
+      "learning_rate": 2.229543865764919e-06,
+      "loss": 0.6621,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.313103383080335,
+      "learning_rate": 2.2284578078068593e-06,
+      "loss": 0.4773,
+      "step": 8644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.194143754903795,
+      "learning_rate": 2.2273719385785706e-06,
+      "loss": 0.5039,
+      "step": 8645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.4291160714676,
+      "learning_rate": 2.2262862581539945e-06,
+      "loss": 0.4509,
+      "step": 8646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.641859646006467,
+      "learning_rate": 2.225200766607058e-06,
+      "loss": 0.4154,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.640540788329445,
+      "learning_rate": 2.2241154640116825e-06,
+      "loss": 0.7621,
+      "step": 8648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.830110274786195,
+      "learning_rate": 2.2230303504417704e-06,
+      "loss": 0.5491,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.813715149125619,
+      "learning_rate": 2.2219454259712133e-06,
+      "loss": 0.4402,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.864181607469838,
+      "learning_rate": 2.2208606906738877e-06,
+      "loss": 0.5373,
+      "step": 8651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.204262110573334,
+      "learning_rate": 2.2197761446236633e-06,
+      "loss": 0.4722,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.817902064128082,
+      "learning_rate": 2.218691787894389e-06,
+      "loss": 0.522,
+      "step": 8653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.222471115945547,
+      "learning_rate": 2.2176076205599094e-06,
+      "loss": 0.4478,
+      "step": 8654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.756636121665563,
+      "learning_rate": 2.2165236426940455e-06,
+      "loss": 0.5119,
+      "step": 8655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 15.177033038373162,
+      "learning_rate": 2.2154398543706145e-06,
+      "loss": 0.4967,
+      "step": 8656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.15419662306264,
+      "learning_rate": 2.2143562556634197e-06,
+      "loss": 0.5024,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.8314280787447155,
+      "learning_rate": 2.213272846646247e-06,
+      "loss": 0.5577,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.435047568222542,
+      "learning_rate": 2.2121896273928724e-06,
+      "loss": 0.4981,
+      "step": 8659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.985623848074779,
+      "learning_rate": 2.211106597977056e-06,
+      "loss": 0.4322,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.664170645542459,
+      "learning_rate": 2.210023758472551e-06,
+      "loss": 0.5601,
+      "step": 8661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.692026872635217,
+      "learning_rate": 2.2089411089530897e-06,
+      "loss": 0.7515,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.328350869064101,
+      "learning_rate": 2.207858649492403e-06,
+      "loss": 0.4837,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.740153377780432,
+      "learning_rate": 2.206776380164193e-06,
+      "loss": 0.4556,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.642575385959243,
+      "learning_rate": 2.205694301042162e-06,
+      "loss": 0.4363,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.596764961022647,
+      "learning_rate": 2.2046124121999952e-06,
+      "loss": 0.5319,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.176020709155823,
+      "learning_rate": 2.2035307137113638e-06,
+      "loss": 0.44,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.363419645710461,
+      "learning_rate": 2.202449205649926e-06,
+      "loss": 0.4185,
+      "step": 8668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.168177485201426,
+      "learning_rate": 2.201367888089326e-06,
+      "loss": 0.4795,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.595452435971986,
+      "learning_rate": 2.2002867611032013e-06,
+      "loss": 0.5055,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.943052484500221,
+      "learning_rate": 2.1992058247651676e-06,
+      "loss": 0.4985,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.629938765411543,
+      "learning_rate": 2.1981250791488366e-06,
+      "loss": 0.488,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.709833830196618,
+      "learning_rate": 2.1970445243277967e-06,
+      "loss": 0.3711,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.101898927664848,
+      "learning_rate": 2.1959641603756323e-06,
+      "loss": 0.5466,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.119786314977136,
+      "learning_rate": 2.1948839873659093e-06,
+      "loss": 0.3854,
+      "step": 8675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.639602117123383,
+      "learning_rate": 2.193804005372186e-06,
+      "loss": 0.5799,
+      "step": 8676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.768096477954297,
+      "learning_rate": 2.1927242144680024e-06,
+      "loss": 0.4286,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.274823641349625,
+      "learning_rate": 2.1916446147268853e-06,
+      "loss": 0.4578,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.385356852573535,
+      "learning_rate": 2.1905652062223555e-06,
+      "loss": 0.4345,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.726524530526605,
+      "learning_rate": 2.189485989027911e-06,
+      "loss": 0.5244,
+      "step": 8680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.969999628874794,
+      "learning_rate": 2.188406963217046e-06,
+      "loss": 0.4408,
+      "step": 8681
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.018305240049651,
+      "learning_rate": 2.187328128863235e-06,
+      "loss": 0.4156,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.150607522437795,
+      "learning_rate": 2.186249486039943e-06,
+      "loss": 0.5038,
+      "step": 8683
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.277083016447537,
+      "learning_rate": 2.1851710348206182e-06,
+      "loss": 0.4813,
+      "step": 8684
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.000157330652725,
+      "learning_rate": 2.1840927752787023e-06,
+      "loss": 0.4003,
+      "step": 8685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.954826194406623,
+      "learning_rate": 2.183014707487617e-06,
+      "loss": 0.4982,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.3991358383704515,
+      "learning_rate": 2.1819368315207738e-06,
+      "loss": 0.4477,
+      "step": 8687
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 21.35757855408503,
+      "learning_rate": 2.1808591474515743e-06,
+      "loss": 0.4517,
+      "step": 8688
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.033229219654291,
+      "learning_rate": 2.1797816553534007e-06,
+      "loss": 0.4494,
+      "step": 8689
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.117977265407111,
+      "learning_rate": 2.1787043552996284e-06,
+      "loss": 0.4792,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.385877468116847,
+      "learning_rate": 2.177627247363615e-06,
+      "loss": 0.4222,
+      "step": 8691
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.819347983818907,
+      "learning_rate": 2.176550331618707e-06,
+      "loss": 0.4102,
+      "step": 8692
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.66611380398822,
+      "learning_rate": 2.175473608138236e-06,
+      "loss": 0.5164,
+      "step": 8693
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.998286309418944,
+      "learning_rate": 2.1743970769955257e-06,
+      "loss": 0.4441,
+      "step": 8694
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.406951844538684,
+      "learning_rate": 2.1733207382638817e-06,
+      "loss": 0.5542,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.834101695719847,
+      "learning_rate": 2.1722445920165948e-06,
+      "loss": 0.5176,
+      "step": 8696
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.70802201644078,
+      "learning_rate": 2.1711686383269505e-06,
+      "loss": 0.4326,
+      "step": 8697
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.6607325417291365,
+      "learning_rate": 2.1700928772682122e-06,
+      "loss": 0.5008,
+      "step": 8698
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.658487194909283,
+      "learning_rate": 2.1690173089136385e-06,
+      "loss": 0.4387,
+      "step": 8699
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.272692157213619,
+      "learning_rate": 2.1679419333364694e-06,
+      "loss": 0.3951,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.465268946748954,
+      "learning_rate": 2.1668667506099326e-06,
+      "loss": 0.503,
+      "step": 8701
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.694095112122439,
+      "learning_rate": 2.1657917608072417e-06,
+      "loss": 0.5437,
+      "step": 8702
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.88994222805743,
+      "learning_rate": 2.1647169640016025e-06,
+      "loss": 0.4852,
+      "step": 8703
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4620814561866788,
+      "learning_rate": 2.1636423602662015e-06,
+      "loss": 0.6518,
+      "step": 8704
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.816301510296299,
+      "learning_rate": 2.1625679496742136e-06,
+      "loss": 0.5275,
+      "step": 8705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.739332333089108,
+      "learning_rate": 2.1614937322988045e-06,
+      "loss": 0.4596,
+      "step": 8706
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.190967597643145,
+      "learning_rate": 2.1604197082131203e-06,
+      "loss": 0.358,
+      "step": 8707
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.822134519673913,
+      "learning_rate": 2.1593458774903005e-06,
+      "loss": 0.4471,
+      "step": 8708
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.020451995824693,
+      "learning_rate": 2.158272240203466e-06,
+      "loss": 0.4385,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.37913124685045,
+      "learning_rate": 2.157198796425728e-06,
+      "loss": 0.3951,
+      "step": 8710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.601058959964349,
+      "learning_rate": 2.156125546230181e-06,
+      "loss": 0.7221,
+      "step": 8711
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.4258755008224915,
+      "learning_rate": 2.1550524896899116e-06,
+      "loss": 0.562,
+      "step": 8712
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.533741764859348,
+      "learning_rate": 2.153979626877989e-06,
+      "loss": 0.4414,
+      "step": 8713
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.905989689388768,
+      "learning_rate": 2.1529069578674705e-06,
+      "loss": 0.3566,
+      "step": 8714
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 39.484340620288386,
+      "learning_rate": 2.1518344827313975e-06,
+      "loss": 0.4437,
+      "step": 8715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.9235034358155865,
+      "learning_rate": 2.1507622015428035e-06,
+      "loss": 0.4763,
+      "step": 8716
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.275472625800579,
+      "learning_rate": 2.1496901143747078e-06,
+      "loss": 0.5629,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.401984100278845,
+      "learning_rate": 2.1486182213001127e-06,
+      "loss": 0.4788,
+      "step": 8718
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.231397478825444,
+      "learning_rate": 2.1475465223920096e-06,
+      "loss": 0.4838,
+      "step": 8719
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.696476983179565,
+      "learning_rate": 2.1464750177233738e-06,
+      "loss": 0.4304,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.966238759267172,
+      "learning_rate": 2.1454037073671746e-06,
+      "loss": 0.4429,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 48.912176679823155,
+      "learning_rate": 2.144332591396362e-06,
+      "loss": 0.4115,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6620078534212572,
+      "learning_rate": 2.143261669883873e-06,
+      "loss": 0.5021,
+      "step": 8723
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5425275190262235,
+      "learning_rate": 2.142190942902632e-06,
+      "loss": 0.7211,
+      "step": 8724
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.683599861814631,
+      "learning_rate": 2.1411204105255524e-06,
+      "loss": 0.4453,
+      "step": 8725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.435223799981985,
+      "learning_rate": 2.140050072825534e-06,
+      "loss": 0.5725,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.901852456386388,
+      "learning_rate": 2.1389799298754603e-06,
+      "loss": 0.337,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.928294738819841,
+      "learning_rate": 2.1379099817482037e-06,
+      "loss": 0.7195,
+      "step": 8728
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.714327766013639,
+      "learning_rate": 2.1368402285166205e-06,
+      "loss": 0.3677,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.321778256008231,
+      "learning_rate": 2.1357706702535607e-06,
+      "loss": 0.5859,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.046896792484198,
+      "learning_rate": 2.1347013070318538e-06,
+      "loss": 0.5056,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.735960592570185,
+      "learning_rate": 2.1336321389243193e-06,
+      "loss": 0.5314,
+      "step": 8732
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.554537130637814,
+      "learning_rate": 2.132563166003761e-06,
+      "loss": 0.3988,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.344814478627503,
+      "learning_rate": 2.1314943883429745e-06,
+      "loss": 0.457,
+      "step": 8734
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.83263490050865,
+      "learning_rate": 2.130425806014735e-06,
+      "loss": 0.4718,
+      "step": 8735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.627672578564743,
+      "learning_rate": 2.1293574190918126e-06,
+      "loss": 0.5098,
+      "step": 8736
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.3099061126036595,
+      "learning_rate": 2.1282892276469574e-06,
+      "loss": 0.4115,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.7535577326222525,
+      "learning_rate": 2.1272212317529074e-06,
+      "loss": 0.4948,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.17130969990903,
+      "learning_rate": 2.1261534314823906e-06,
+      "loss": 0.3529,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.269317374947544,
+      "learning_rate": 2.125085826908117e-06,
+      "loss": 0.5123,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 23.30493965698663,
+      "learning_rate": 2.1240184181027913e-06,
+      "loss": 0.3813,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.256161115171935,
+      "learning_rate": 2.1229512051390915e-06,
+      "loss": 0.4168,
+      "step": 8742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.458924712579665,
+      "learning_rate": 2.1218841880896955e-06,
+      "loss": 0.5681,
+      "step": 8743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.0319374427725485,
+      "learning_rate": 2.1208173670272593e-06,
+      "loss": 0.4178,
+      "step": 8744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.632203831724398,
+      "learning_rate": 2.1197507420244317e-06,
+      "loss": 0.5036,
+      "step": 8745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.331180272110838,
+      "learning_rate": 2.1186843131538435e-06,
+      "loss": 0.3286,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.092008066852502,
+      "learning_rate": 2.1176180804881123e-06,
+      "loss": 0.4295,
+      "step": 8747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.26679244767734,
+      "learning_rate": 2.1165520440998477e-06,
+      "loss": 0.4505,
+      "step": 8748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.676687897658877,
+      "learning_rate": 2.115486204061638e-06,
+      "loss": 0.4553,
+      "step": 8749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.560306663541412,
+      "learning_rate": 2.114420560446067e-06,
+      "loss": 0.4995,
+      "step": 8750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.2995577356165,
+      "learning_rate": 2.113355113325695e-06,
+      "loss": 0.5395,
+      "step": 8751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.801886315396308,
+      "learning_rate": 2.112289862773077e-06,
+      "loss": 0.4466,
+      "step": 8752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.112130794807796,
+      "learning_rate": 2.1112248088607506e-06,
+      "loss": 0.4531,
+      "step": 8753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.260210069290397,
+      "learning_rate": 2.1101599516612437e-06,
+      "loss": 0.4216,
+      "step": 8754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.34765838220826,
+      "learning_rate": 2.109095291247066e-06,
+      "loss": 0.4088,
+      "step": 8755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 28.63258185075136,
+      "learning_rate": 2.1080308276907165e-06,
+      "loss": 0.5276,
+      "step": 8756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.10438721874686,
+      "learning_rate": 2.1069665610646816e-06,
+      "loss": 0.4948,
+      "step": 8757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.070665491870807,
+      "learning_rate": 2.105902491441431e-06,
+      "loss": 0.4876,
+      "step": 8758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.85660321848395,
+      "learning_rate": 2.104838618893429e-06,
+      "loss": 0.4446,
+      "step": 8759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.550164011060895,
+      "learning_rate": 2.1037749434931116e-06,
+      "loss": 0.3834,
+      "step": 8760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.82245905876571,
+      "learning_rate": 2.102711465312917e-06,
+      "loss": 0.3724,
+      "step": 8761
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.5540333293429685,
+      "learning_rate": 2.10164818442526e-06,
+      "loss": 0.3281,
+      "step": 8762
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.711843152838805,
+      "learning_rate": 2.100585100902548e-06,
+      "loss": 0.3323,
+      "step": 8763
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.210627500862625,
+      "learning_rate": 2.099522214817171e-06,
+      "loss": 0.3699,
+      "step": 8764
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.303835496098473,
+      "learning_rate": 2.0984595262415076e-06,
+      "loss": 0.4575,
+      "step": 8765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5589314895802016,
+      "learning_rate": 2.097397035247919e-06,
+      "loss": 0.6817,
+      "step": 8766
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.487771405085343,
+      "learning_rate": 2.0963347419087586e-06,
+      "loss": 0.5816,
+      "step": 8767
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.9412396056172865,
+      "learning_rate": 2.0952726462963673e-06,
+      "loss": 0.4299,
+      "step": 8768
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 30.421566675766755,
+      "learning_rate": 2.0942107484830627e-06,
+      "loss": 0.4135,
+      "step": 8769
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.65361059392014,
+      "learning_rate": 2.093149048541159e-06,
+      "loss": 0.5765,
+      "step": 8770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.451846263872759,
+      "learning_rate": 2.0920875465429517e-06,
+      "loss": 0.4777,
+      "step": 8771
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.627995961414963,
+      "learning_rate": 2.091026242560727e-06,
+      "loss": 0.457,
+      "step": 8772
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.991611884244574,
+      "learning_rate": 2.0899651366667535e-06,
+      "loss": 0.472,
+      "step": 8773
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.163183779080112,
+      "learning_rate": 2.088904228933287e-06,
+      "loss": 0.3683,
+      "step": 8774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.646407161529433,
+      "learning_rate": 2.0878435194325697e-06,
+      "loss": 0.7325,
+      "step": 8775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.024217999543797,
+      "learning_rate": 2.0867830082368324e-06,
+      "loss": 0.4848,
+      "step": 8776
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.840955036919461,
+      "learning_rate": 2.085722695418294e-06,
+      "loss": 0.5151,
+      "step": 8777
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.868964970065498,
+      "learning_rate": 2.0846625810491544e-06,
+      "loss": 0.4429,
+      "step": 8778
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.865781865588861,
+      "learning_rate": 2.083602665201603e-06,
+      "loss": 0.4466,
+      "step": 8779
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.191263809861065,
+      "learning_rate": 2.0825429479478134e-06,
+      "loss": 0.5174,
+      "step": 8780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.724437106321801,
+      "learning_rate": 2.081483429359951e-06,
+      "loss": 0.4898,
+      "step": 8781
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 17.936868186691008,
+      "learning_rate": 2.0804241095101643e-06,
+      "loss": 0.333,
+      "step": 8782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.533800562188461,
+      "learning_rate": 2.0793649884705857e-06,
+      "loss": 0.4894,
+      "step": 8783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.412023588755224,
+      "learning_rate": 2.0783060663133364e-06,
+      "loss": 0.4834,
+      "step": 8784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.546066127544378,
+      "learning_rate": 2.077247343110528e-06,
+      "loss": 0.5204,
+      "step": 8785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8026764103453243,
+      "learning_rate": 2.076188818934251e-06,
+      "loss": 0.75,
+      "step": 8786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.228542972547276,
+      "learning_rate": 2.075130493856589e-06,
+      "loss": 0.4034,
+      "step": 8787
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.252146276523305,
+      "learning_rate": 2.074072367949609e-06,
+      "loss": 0.5336,
+      "step": 8788
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 46.017961592745536,
+      "learning_rate": 2.0730144412853612e-06,
+      "loss": 0.5692,
+      "step": 8789
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0259557575271323,
+      "learning_rate": 2.0719567139358916e-06,
+      "loss": 0.7742,
+      "step": 8790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.378338639925984,
+      "learning_rate": 2.0708991859732225e-06,
+      "loss": 0.4105,
+      "step": 8791
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.902131868208017,
+      "learning_rate": 2.069841857469368e-06,
+      "loss": 0.4683,
+      "step": 8792
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.925087037719733,
+      "learning_rate": 2.068784728496326e-06,
+      "loss": 0.4193,
+      "step": 8793
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.367562920391357,
+      "learning_rate": 2.0677277991260854e-06,
+      "loss": 0.4732,
+      "step": 8794
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.136745601795462,
+      "learning_rate": 2.0666710694306146e-06,
+      "loss": 0.4214,
+      "step": 8795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8707006569604827,
+      "learning_rate": 2.0656145394818767e-06,
+      "loss": 0.7442,
+      "step": 8796
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.9440710854198855,
+      "learning_rate": 2.064558209351814e-06,
+      "loss": 0.4596,
+      "step": 8797
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.954820978263291,
+      "learning_rate": 2.0635020791123566e-06,
+      "loss": 0.489,
+      "step": 8798
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.350895774939309,
+      "learning_rate": 2.0624461488354257e-06,
+      "loss": 0.5618,
+      "step": 8799
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.52212389514741,
+      "learning_rate": 2.061390418592924e-06,
+      "loss": 0.4535,
+      "step": 8800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.97959145644919,
+      "learning_rate": 2.060334888456742e-06,
+      "loss": 0.4157,
+      "step": 8801
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.46354732587755,
+      "learning_rate": 2.0592795584987547e-06,
+      "loss": 0.4018,
+      "step": 8802
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.121060327306538,
+      "learning_rate": 2.0582244287908286e-06,
+      "loss": 0.4958,
+      "step": 8803
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.9707220343499,
+      "learning_rate": 2.057169499404811e-06,
+      "loss": 0.433,
+      "step": 8804
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.228598333785644,
+      "learning_rate": 2.0561147704125416e-06,
+      "loss": 0.4502,
+      "step": 8805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.357996050638889,
+      "learning_rate": 2.055060241885837e-06,
+      "loss": 0.5443,
+      "step": 8806
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.419666385362556,
+      "learning_rate": 2.0540059138965085e-06,
+      "loss": 0.4839,
+      "step": 8807
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.661443559822804,
+      "learning_rate": 2.052951786516354e-06,
+      "loss": 0.5155,
+      "step": 8808
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.951476537375176,
+      "learning_rate": 2.0518978598171523e-06,
+      "loss": 0.5592,
+      "step": 8809
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.739676256529399,
+      "learning_rate": 2.0508441338706713e-06,
+      "loss": 0.38,
+      "step": 8810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 12.767645610062132,
+      "learning_rate": 2.0497906087486637e-06,
+      "loss": 0.4558,
+      "step": 8811
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.677742947107992,
+      "learning_rate": 2.048737284522873e-06,
+      "loss": 0.5917,
+      "step": 8812
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.209563501440961,
+      "learning_rate": 2.047684161265023e-06,
+      "loss": 0.4765,
+      "step": 8813
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.489109511678827,
+      "learning_rate": 2.0466312390468294e-06,
+      "loss": 0.5193,
+      "step": 8814
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.787471988067459,
+      "learning_rate": 2.04557851793999e-06,
+      "loss": 0.4565,
+      "step": 8815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.452858764124095,
+      "learning_rate": 2.0445259980161895e-06,
+      "loss": 0.412,
+      "step": 8816
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.380028940975517,
+      "learning_rate": 2.0434736793471026e-06,
+      "loss": 0.499,
+      "step": 8817
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.81622142390869,
+      "learning_rate": 2.0424215620043857e-06,
+      "loss": 0.4549,
+      "step": 8818
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.987266761958851,
+      "learning_rate": 2.0413696460596843e-06,
+      "loss": 0.5614,
+      "step": 8819
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.099604498663268,
+      "learning_rate": 2.040317931584626e-06,
+      "loss": 0.3906,
+      "step": 8820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.458788626572742,
+      "learning_rate": 2.039266418650833e-06,
+      "loss": 0.434,
+      "step": 8821
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.161547968334357,
+      "learning_rate": 2.0382151073299038e-06,
+      "loss": 0.4291,
+      "step": 8822
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.320485462036287,
+      "learning_rate": 2.0371639976934327e-06,
+      "loss": 0.3882,
+      "step": 8823
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.478178142679451,
+      "learning_rate": 2.0361130898129926e-06,
+      "loss": 0.3804,
+      "step": 8824
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.479869657557463,
+      "learning_rate": 2.0350623837601467e-06,
+      "loss": 0.3754,
+      "step": 8825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.045041388920711,
+      "learning_rate": 2.034011879606441e-06,
+      "loss": 0.4003,
+      "step": 8826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.915613344868492,
+      "learning_rate": 2.0329615774234145e-06,
+      "loss": 0.4744,
+      "step": 8827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.434106856861309,
+      "learning_rate": 2.031911477282586e-06,
+      "loss": 0.5067,
+      "step": 8828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.650694248449775,
+      "learning_rate": 2.030861579255461e-06,
+      "loss": 0.5052,
+      "step": 8829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.590804189110345,
+      "learning_rate": 2.029811883413536e-06,
+      "loss": 0.5402,
+      "step": 8830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.570896958860576,
+      "learning_rate": 2.028762389828287e-06,
+      "loss": 0.5286,
+      "step": 8831
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.825713347319787,
+      "learning_rate": 2.0277130985711848e-06,
+      "loss": 0.5157,
+      "step": 8832
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.689830256207932,
+      "learning_rate": 2.026664009713678e-06,
+      "loss": 0.5246,
+      "step": 8833
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.03067629112632,
+      "learning_rate": 2.025615123327206e-06,
+      "loss": 0.5042,
+      "step": 8834
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.6595778891476085,
+      "learning_rate": 2.024566439483191e-06,
+      "loss": 0.4653,
+      "step": 8835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 15.227541694900543,
+      "learning_rate": 2.0235179582530455e-06,
+      "loss": 0.5372,
+      "step": 8836
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.477838923028485,
+      "learning_rate": 2.022469679708171e-06,
+      "loss": 0.4787,
+      "step": 8837
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.106835713154479,
+      "learning_rate": 2.021421603919942e-06,
+      "loss": 0.4089,
+      "step": 8838
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.589372875348684,
+      "learning_rate": 2.0203737309597344e-06,
+      "loss": 0.48,
+      "step": 8839
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.7082537983505475,
+      "learning_rate": 2.019326060898899e-06,
+      "loss": 0.5054,
+      "step": 8840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.129024089572718,
+      "learning_rate": 2.018278593808782e-06,
+      "loss": 0.4573,
+      "step": 8841
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.304842468947788,
+      "learning_rate": 2.01723132976071e-06,
+      "loss": 0.5061,
+      "step": 8842
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.896243214975588,
+      "learning_rate": 2.016184268825995e-06,
+      "loss": 0.4701,
+      "step": 8843
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 22.191974515918968,
+      "learning_rate": 2.0151374110759376e-06,
+      "loss": 0.4682,
+      "step": 8844
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.159201399577228,
+      "learning_rate": 2.0140907565818263e-06,
+      "loss": 0.4437,
+      "step": 8845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.398638205943563,
+      "learning_rate": 2.0130443054149328e-06,
+      "loss": 0.4322,
+      "step": 8846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.158844630025835,
+      "learning_rate": 2.0119980576465136e-06,
+      "loss": 0.4737,
+      "step": 8847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.052625502853391,
+      "learning_rate": 2.0109520133478165e-06,
+      "loss": 0.3916,
+      "step": 8848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.4574565882349955,
+      "learning_rate": 2.0099061725900702e-06,
+      "loss": 0.339,
+      "step": 8849
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.07269603214835,
+      "learning_rate": 2.008860535444495e-06,
+      "loss": 0.4138,
+      "step": 8850
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.950591230616255,
+      "learning_rate": 2.0078151019822915e-06,
+      "loss": 0.4472,
+      "step": 8851
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.835599985545487,
+      "learning_rate": 2.00676987227465e-06,
+      "loss": 0.4222,
+      "step": 8852
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.662101448361124,
+      "learning_rate": 2.005724846392744e-06,
+      "loss": 0.4985,
+      "step": 8853
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.967348371121625,
+      "learning_rate": 2.0046800244077384e-06,
+      "loss": 0.5278,
+      "step": 8854
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.776841136699124,
+      "learning_rate": 2.00363540639078e-06,
+      "loss": 0.562,
+      "step": 8855
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.883372011280812,
+      "learning_rate": 2.0025909924129998e-06,
+      "loss": 0.4267,
+      "step": 8856
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.917604073831822,
+      "learning_rate": 2.0015467825455226e-06,
+      "loss": 0.3587,
+      "step": 8857
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 38.0960474302947,
+      "learning_rate": 2.0005027768594492e-06,
+      "loss": 0.4599,
+      "step": 8858
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.607204317905186,
+      "learning_rate": 1.999458975425877e-06,
+      "loss": 0.4541,
+      "step": 8859
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.795891368142088,
+      "learning_rate": 1.9984153783158818e-06,
+      "loss": 0.4996,
+      "step": 8860
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.0339401598289175,
+      "learning_rate": 1.997371985600528e-06,
+      "loss": 0.4451,
+      "step": 8861
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.917106433588623,
+      "learning_rate": 1.996328797350864e-06,
+      "loss": 0.4929,
+      "step": 8862
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7473558514127812,
+      "learning_rate": 1.99528581363793e-06,
+      "loss": 0.7087,
+      "step": 8863
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.270983370158149,
+      "learning_rate": 1.994243034532747e-06,
+      "loss": 0.52,
+      "step": 8864
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.621468755002804,
+      "learning_rate": 1.9932004601063236e-06,
+      "loss": 0.4682,
+      "step": 8865
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9821589576492615,
+      "learning_rate": 1.9921580904296528e-06,
+      "loss": 0.7476,
+      "step": 8866
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6247875291179226,
+      "learning_rate": 1.9911159255737165e-06,
+      "loss": 0.7225,
+      "step": 8867
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.48183662798863,
+      "learning_rate": 1.9900739656094844e-06,
+      "loss": 0.4521,
+      "step": 8868
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.352481734624755,
+      "learning_rate": 1.9890322106079063e-06,
+      "loss": 0.3385,
+      "step": 8869
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.233760485079216,
+      "learning_rate": 1.9879906606399222e-06,
+      "loss": 0.5147,
+      "step": 8870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.805629989727698,
+      "learning_rate": 1.9869493157764546e-06,
+      "loss": 0.4923,
+      "step": 8871
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.3966893716198,
+      "learning_rate": 1.9859081760884187e-06,
+      "loss": 0.4442,
+      "step": 8872
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.519104942649744,
+      "learning_rate": 1.984867241646709e-06,
+      "loss": 0.5151,
+      "step": 8873
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 11.968823860124964,
+      "learning_rate": 1.983826512522209e-06,
+      "loss": 0.4242,
+      "step": 8874
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.2253553186179955,
+      "learning_rate": 1.982785988785786e-06,
+      "loss": 0.4669,
+      "step": 8875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.106649379965141,
+      "learning_rate": 1.9817456705082967e-06,
+      "loss": 0.4794,
+      "step": 8876
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6524828692110132,
+      "learning_rate": 1.980705557760584e-06,
+      "loss": 0.7154,
+      "step": 8877
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.685187177542853,
+      "learning_rate": 1.9796656506134725e-06,
+      "loss": 0.388,
+      "step": 8878
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.163915030620245,
+      "learning_rate": 1.9786259491377763e-06,
+      "loss": 0.5737,
+      "step": 8879
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 15.089545238576996,
+      "learning_rate": 1.977586453404292e-06,
+      "loss": 0.4769,
+      "step": 8880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.908268738697892,
+      "learning_rate": 1.976547163483809e-06,
+      "loss": 0.5723,
+      "step": 8881
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.82385168052258,
+      "learning_rate": 1.9755080794470942e-06,
+      "loss": 0.4147,
+      "step": 8882
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.3775734109998625,
+      "learning_rate": 1.97446920136491e-06,
+      "loss": 0.5328,
+      "step": 8883
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.638529049821173,
+      "learning_rate": 1.9734305293079923e-06,
+      "loss": 0.4194,
+      "step": 8884
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.143048086621099,
+      "learning_rate": 1.972392063347076e-06,
+      "loss": 0.4348,
+      "step": 8885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.505249200525774,
+      "learning_rate": 1.9713538035528712e-06,
+      "loss": 0.3318,
+      "step": 8886
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.816437697793296,
+      "learning_rate": 1.970315749996084e-06,
+      "loss": 0.4846,
+      "step": 8887
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.507671793185436,
+      "learning_rate": 1.9692779027473978e-06,
+      "loss": 0.5341,
+      "step": 8888
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.7972544487021,
+      "learning_rate": 1.968240261877485e-06,
+      "loss": 0.4449,
+      "step": 8889
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.566511779110961,
+      "learning_rate": 1.967202827457007e-06,
+      "loss": 0.4737,
+      "step": 8890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.318995559303981,
+      "learning_rate": 1.9661655995566054e-06,
+      "loss": 0.4876,
+      "step": 8891
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.257252765683422,
+      "learning_rate": 1.9651285782469158e-06,
+      "loss": 0.4388,
+      "step": 8892
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.117004033214545,
+      "learning_rate": 1.9640917635985485e-06,
+      "loss": 0.4011,
+      "step": 8893
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.2386986350071,
+      "learning_rate": 1.9630551556821105e-06,
+      "loss": 0.5401,
+      "step": 8894
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.484707899209008,
+      "learning_rate": 1.962018754568187e-06,
+      "loss": 0.4937,
+      "step": 8895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.940916276495458,
+      "learning_rate": 1.960982560327355e-06,
+      "loss": 0.4026,
+      "step": 8896
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.356926677621773,
+      "learning_rate": 1.9599465730301746e-06,
+      "loss": 0.5411,
+      "step": 8897
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.112883854826345,
+      "learning_rate": 1.958910792747189e-06,
+      "loss": 0.5529,
+      "step": 8898
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 34.14435927692349,
+      "learning_rate": 1.957875219548934e-06,
+      "loss": 0.4526,
+      "step": 8899
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.336576127650033,
+      "learning_rate": 1.956839853505925e-06,
+      "loss": 0.4534,
+      "step": 8900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.866250720342892,
+      "learning_rate": 1.9558046946886694e-06,
+      "loss": 0.4328,
+      "step": 8901
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.046038925817068,
+      "learning_rate": 1.954769743167651e-06,
+      "loss": 0.4496,
+      "step": 8902
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.358743074808958,
+      "learning_rate": 1.95373499901335e-06,
+      "loss": 0.4884,
+      "step": 8903
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.486799288834162,
+      "learning_rate": 1.9527004622962258e-06,
+      "loss": 0.7263,
+      "step": 8904
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.4691322877080015,
+      "learning_rate": 1.9516661330867283e-06,
+      "loss": 0.4149,
+      "step": 8905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.029258363218366,
+      "learning_rate": 1.9506320114552884e-06,
+      "loss": 0.4616,
+      "step": 8906
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.402256975045645,
+      "learning_rate": 1.949598097472325e-06,
+      "loss": 0.4153,
+      "step": 8907
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.77474305277218,
+      "learning_rate": 1.9485643912082447e-06,
+      "loss": 0.4745,
+      "step": 8908
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 18.03949137836611,
+      "learning_rate": 1.9475308927334363e-06,
+      "loss": 0.4026,
+      "step": 8909
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.930914249950985,
+      "learning_rate": 1.9464976021182813e-06,
+      "loss": 0.4639,
+      "step": 8910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.650473322323725,
+      "learning_rate": 1.945464519433135e-06,
+      "loss": 0.399,
+      "step": 8911
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.277404458906204,
+      "learning_rate": 1.944431644748352e-06,
+      "loss": 0.4996,
+      "step": 8912
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.945744507408695,
+      "learning_rate": 1.943398978134261e-06,
+      "loss": 0.4645,
+      "step": 8913
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.36118299177403,
+      "learning_rate": 1.9423665196611876e-06,
+      "loss": 0.5018,
+      "step": 8914
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.761900478481481,
+      "learning_rate": 1.941334269399434e-06,
+      "loss": 0.4396,
+      "step": 8915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.897151053197249,
+      "learning_rate": 1.940302227419294e-06,
+      "loss": 0.444,
+      "step": 8916
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.163137262976447,
+      "learning_rate": 1.939270393791042e-06,
+      "loss": 0.5002,
+      "step": 8917
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.5257541077358825,
+      "learning_rate": 1.9382387685849423e-06,
+      "loss": 0.5298,
+      "step": 8918
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.62451250365885,
+      "learning_rate": 1.9372073518712474e-06,
+      "loss": 0.4476,
+      "step": 8919
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.582650856402758,
+      "learning_rate": 1.9361761437201903e-06,
+      "loss": 0.5789,
+      "step": 8920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.04505749772403,
+      "learning_rate": 1.935145144201991e-06,
+      "loss": 0.4476,
+      "step": 8921
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.913148609390449,
+      "learning_rate": 1.934114353386854e-06,
+      "loss": 0.4269,
+      "step": 8922
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.862908193117456,
+      "learning_rate": 1.9330837713449765e-06,
+      "loss": 0.5528,
+      "step": 8923
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.603560458383871,
+      "learning_rate": 1.9320533981465336e-06,
+      "loss": 0.4025,
+      "step": 8924
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.080608633229456,
+      "learning_rate": 1.9310232338616884e-06,
+      "loss": 0.3589,
+      "step": 8925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.224870590923406,
+      "learning_rate": 1.929993278560591e-06,
+      "loss": 0.5051,
+      "step": 8926
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.608486726603578,
+      "learning_rate": 1.9289635323133776e-06,
+      "loss": 0.4537,
+      "step": 8927
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.240236071117971,
+      "learning_rate": 1.9279339951901703e-06,
+      "loss": 0.3901,
+      "step": 8928
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.833551188411776,
+      "learning_rate": 1.9269046672610753e-06,
+      "loss": 0.4325,
+      "step": 8929
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.055270334724876,
+      "learning_rate": 1.925875548596185e-06,
+      "loss": 0.4888,
+      "step": 8930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.111029209987065,
+      "learning_rate": 1.924846639265576e-06,
+      "loss": 0.52,
+      "step": 8931
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.079484698803487,
+      "learning_rate": 1.923817939339315e-06,
+      "loss": 0.4881,
+      "step": 8932
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.429691282375729,
+      "learning_rate": 1.922789448887452e-06,
+      "loss": 0.4743,
+      "step": 8933
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.875672112975075,
+      "learning_rate": 1.921761167980022e-06,
+      "loss": 0.5007,
+      "step": 8934
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.580539895735415,
+      "learning_rate": 1.9207330966870435e-06,
+      "loss": 0.4682,
+      "step": 8935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.163126457744808,
+      "learning_rate": 1.919705235078529e-06,
+      "loss": 0.4479,
+      "step": 8936
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.794706988416724,
+      "learning_rate": 1.9186775832244663e-06,
+      "loss": 0.4886,
+      "step": 8937
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.131818207891895,
+      "learning_rate": 1.9176501411948376e-06,
+      "loss": 0.376,
+      "step": 8938
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.580421099920281,
+      "learning_rate": 1.9166229090596065e-06,
+      "loss": 0.399,
+      "step": 8939
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.821524085180783,
+      "learning_rate": 1.91559588688872e-06,
+      "loss": 0.4523,
+      "step": 8940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.635840567131084,
+      "learning_rate": 1.9145690747521184e-06,
+      "loss": 0.459,
+      "step": 8941
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.094795276881082,
+      "learning_rate": 1.9135424727197204e-06,
+      "loss": 0.4319,
+      "step": 8942
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.3997281548874145,
+      "learning_rate": 1.9125160808614336e-06,
+      "loss": 0.5396,
+      "step": 8943
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.0909325286240295,
+      "learning_rate": 1.9114898992471487e-06,
+      "loss": 0.3682,
+      "step": 8944
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0295466434933225,
+      "learning_rate": 1.9104639279467472e-06,
+      "loss": 0.7046,
+      "step": 8945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.681890433748291,
+      "learning_rate": 1.909438167030091e-06,
+      "loss": 0.3781,
+      "step": 8946
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.698054561523357,
+      "learning_rate": 1.908412616567032e-06,
+      "loss": 0.5474,
+      "step": 8947
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.471433993641075,
+      "learning_rate": 1.9073872766274042e-06,
+      "loss": 0.4006,
+      "step": 8948
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7640792890537556,
+      "learning_rate": 1.9063621472810278e-06,
+      "loss": 0.7504,
+      "step": 8949
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.87619605860848,
+      "learning_rate": 1.9053372285977128e-06,
+      "loss": 0.4553,
+      "step": 8950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.073807509136249,
+      "learning_rate": 1.904312520647249e-06,
+      "loss": 0.426,
+      "step": 8951
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.626178019057471,
+      "learning_rate": 1.9032880234994156e-06,
+      "loss": 0.4648,
+      "step": 8952
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.763135625061332,
+      "learning_rate": 1.9022637372239738e-06,
+      "loss": 0.4406,
+      "step": 8953
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.886088416976914,
+      "learning_rate": 1.9012396618906764e-06,
+      "loss": 0.4196,
+      "step": 8954
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.782475588698819,
+      "learning_rate": 1.900215797569256e-06,
+      "loss": 0.415,
+      "step": 8955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.49285447722343,
+      "learning_rate": 1.8991921443294353e-06,
+      "loss": 0.4558,
+      "step": 8956
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.9917689153276354,
+      "learning_rate": 1.8981687022409202e-06,
+      "loss": 0.4775,
+      "step": 8957
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.713972310503083,
+      "learning_rate": 1.8971454713733993e-06,
+      "loss": 0.4536,
+      "step": 8958
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.363715654038064,
+      "learning_rate": 1.8961224517965548e-06,
+      "loss": 0.5068,
+      "step": 8959
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.691878827607632,
+      "learning_rate": 1.895099643580048e-06,
+      "loss": 0.4982,
+      "step": 8960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.07682330782135,
+      "learning_rate": 1.8940770467935265e-06,
+      "loss": 0.5075,
+      "step": 8961
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.584382996624707,
+      "learning_rate": 1.8930546615066243e-06,
+      "loss": 0.3832,
+      "step": 8962
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.453168628575387,
+      "learning_rate": 1.892032487788964e-06,
+      "loss": 0.4147,
+      "step": 8963
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.564772979642031,
+      "learning_rate": 1.8910105257101475e-06,
+      "loss": 0.5235,
+      "step": 8964
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.645021647035638,
+      "learning_rate": 1.88998877533977e-06,
+      "loss": 0.4561,
+      "step": 8965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.83678242528266,
+      "learning_rate": 1.888967236747406e-06,
+      "loss": 0.4522,
+      "step": 8966
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.116542575280371,
+      "learning_rate": 1.8879459100026165e-06,
+      "loss": 0.7402,
+      "step": 8967
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.353521064044411,
+      "learning_rate": 1.886924795174952e-06,
+      "loss": 0.4401,
+      "step": 8968
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.458361778757079,
+      "learning_rate": 1.8859038923339446e-06,
+      "loss": 0.3565,
+      "step": 8969
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.981162222468127,
+      "learning_rate": 1.8848832015491136e-06,
+      "loss": 0.44,
+      "step": 8970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.343870394447352,
+      "learning_rate": 1.8838627228899614e-06,
+      "loss": 0.3403,
+      "step": 8971
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.250058289373222,
+      "learning_rate": 1.8828424564259818e-06,
+      "loss": 0.4864,
+      "step": 8972
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 24.943761482687307,
+      "learning_rate": 1.8818224022266462e-06,
+      "loss": 0.4253,
+      "step": 8973
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.542389145415691,
+      "learning_rate": 1.8808025603614205e-06,
+      "loss": 0.4734,
+      "step": 8974
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.640074949137933,
+      "learning_rate": 1.8797829308997484e-06,
+      "loss": 0.4191,
+      "step": 8975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.34362465718164,
+      "learning_rate": 1.8787635139110634e-06,
+      "loss": 0.5164,
+      "step": 8976
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.201520589620806,
+      "learning_rate": 1.8777443094647807e-06,
+      "loss": 0.4039,
+      "step": 8977
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.896102301916154,
+      "learning_rate": 1.8767253176303058e-06,
+      "loss": 0.5446,
+      "step": 8978
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.948104431200906,
+      "learning_rate": 1.8757065384770306e-06,
+      "loss": 0.5887,
+      "step": 8979
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 20.168311468491993,
+      "learning_rate": 1.8746879720743227e-06,
+      "loss": 0.4221,
+      "step": 8980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.167129904666227,
+      "learning_rate": 1.873669618491547e-06,
+      "loss": 0.4653,
+      "step": 8981
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 44.557519464868626,
+      "learning_rate": 1.8726514777980464e-06,
+      "loss": 0.4571,
+      "step": 8982
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.632680747034867,
+      "learning_rate": 1.8716335500631539e-06,
+      "loss": 0.3999,
+      "step": 8983
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.2216060062001945,
+      "learning_rate": 1.8706158353561849e-06,
+      "loss": 0.4551,
+      "step": 8984
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.851269363567186,
+      "learning_rate": 1.8695983337464406e-06,
+      "loss": 0.5445,
+      "step": 8985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.447710065129156,
+      "learning_rate": 1.8685810453032077e-06,
+      "loss": 0.5169,
+      "step": 8986
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.376878530526714,
+      "learning_rate": 1.86756397009576e-06,
+      "loss": 0.4873,
+      "step": 8987
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.650841863559374,
+      "learning_rate": 1.866547108193359e-06,
+      "loss": 0.4992,
+      "step": 8988
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.717690323805085,
+      "learning_rate": 1.8655304596652423e-06,
+      "loss": 0.5009,
+      "step": 8989
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 14.100932424981899,
+      "learning_rate": 1.8645140245806448e-06,
+      "loss": 0.4622,
+      "step": 8990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.688904053727264,
+      "learning_rate": 1.8634978030087763e-06,
+      "loss": 0.5803,
+      "step": 8991
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.922157508551335,
+      "learning_rate": 1.8624817950188413e-06,
+      "loss": 0.4462,
+      "step": 8992
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.8605578208401825,
+      "learning_rate": 1.8614660006800239e-06,
+      "loss": 0.439,
+      "step": 8993
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.047014153714704,
+      "learning_rate": 1.8604504200614948e-06,
+      "loss": 0.5737,
+      "step": 8994
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.8960475244891715,
+      "learning_rate": 1.859435053232409e-06,
+      "loss": 0.4592,
+      "step": 8995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.062686775204564,
+      "learning_rate": 1.858419900261912e-06,
+      "loss": 0.3607,
+      "step": 8996
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.216281619969886,
+      "learning_rate": 1.8574049612191296e-06,
+      "loss": 0.4173,
+      "step": 8997
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.342196632145372,
+      "learning_rate": 1.8563902361731728e-06,
+      "loss": 0.4382,
+      "step": 8998
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.250726567975269,
+      "learning_rate": 1.8553757251931426e-06,
+      "loss": 0.4406,
+      "step": 8999
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.901862145678687,
+      "learning_rate": 1.8543614283481208e-06,
+      "loss": 0.5445,
+      "step": 9000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.308790824148833,
+      "learning_rate": 1.8533473457071787e-06,
+      "loss": 0.4944,
+      "step": 9001
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.047357709168803,
+      "learning_rate": 1.8523334773393697e-06,
+      "loss": 0.4389,
+      "step": 9002
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.838406818214714,
+      "learning_rate": 1.8513198233137336e-06,
+      "loss": 0.5053,
+      "step": 9003
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.757428205088448,
+      "learning_rate": 1.850306383699294e-06,
+      "loss": 0.4617,
+      "step": 9004
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.7518306910912385,
+      "learning_rate": 1.8492931585650654e-06,
+      "loss": 0.4698,
+      "step": 9005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.939159368083446,
+      "learning_rate": 1.8482801479800417e-06,
+      "loss": 0.4656,
+      "step": 9006
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.428716019892416,
+      "learning_rate": 1.847267352013203e-06,
+      "loss": 0.5591,
+      "step": 9007
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.868521118261123,
+      "learning_rate": 1.8462547707335198e-06,
+      "loss": 0.4961,
+      "step": 9008
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.437439162165304,
+      "learning_rate": 1.845242404209941e-06,
+      "loss": 0.444,
+      "step": 9009
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.161929272318783,
+      "learning_rate": 1.844230252511407e-06,
+      "loss": 0.4203,
+      "step": 9010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.021673448606824,
+      "learning_rate": 1.8432183157068401e-06,
+      "loss": 0.4803,
+      "step": 9011
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4902972882781615,
+      "learning_rate": 1.8422065938651484e-06,
+      "loss": 0.7208,
+      "step": 9012
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.165819127992922,
+      "learning_rate": 1.8411950870552241e-06,
+      "loss": 0.3814,
+      "step": 9013
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9411829413846768,
+      "learning_rate": 1.840183795345949e-06,
+      "loss": 0.6966,
+      "step": 9014
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.419427849526632,
+      "learning_rate": 1.8391727188061852e-06,
+      "loss": 0.4487,
+      "step": 9015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.661852161041486,
+      "learning_rate": 1.8381618575047872e-06,
+      "loss": 0.3934,
+      "step": 9016
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.0432834472663695,
+      "learning_rate": 1.8371512115105827e-06,
+      "loss": 0.3918,
+      "step": 9017
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.42745394516785,
+      "learning_rate": 1.836140780892397e-06,
+      "loss": 0.4607,
+      "step": 9018
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.19804711045645,
+      "learning_rate": 1.8351305657190366e-06,
+      "loss": 0.56,
+      "step": 9019
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.242862088514801,
+      "learning_rate": 1.8341205660592915e-06,
+      "loss": 0.4806,
+      "step": 9020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.931419842563761,
+      "learning_rate": 1.8331107819819383e-06,
+      "loss": 0.4664,
+      "step": 9021
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.316977368278463,
+      "learning_rate": 1.8321012135557365e-06,
+      "loss": 0.4683,
+      "step": 9022
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.156541834102017,
+      "learning_rate": 1.831091860849437e-06,
+      "loss": 0.3858,
+      "step": 9023
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.866173496438203,
+      "learning_rate": 1.8300827239317692e-06,
+      "loss": 0.4735,
+      "step": 9024
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.118070226268124,
+      "learning_rate": 1.8290738028714555e-06,
+      "loss": 0.4231,
+      "step": 9025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.051955022546204,
+      "learning_rate": 1.8280650977371923e-06,
+      "loss": 0.5289,
+      "step": 9026
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.432806718105663,
+      "learning_rate": 1.8270566085976709e-06,
+      "loss": 0.429,
+      "step": 9027
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.164323092816092,
+      "learning_rate": 1.826048335521568e-06,
+      "loss": 0.8231,
+      "step": 9028
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.67704725692609,
+      "learning_rate": 1.8250402785775396e-06,
+      "loss": 0.5288,
+      "step": 9029
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.561902203268932,
+      "learning_rate": 1.82403243783423e-06,
+      "loss": 0.4446,
+      "step": 9030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.160418611031752,
+      "learning_rate": 1.8230248133602673e-06,
+      "loss": 0.4381,
+      "step": 9031
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.52280905379834,
+      "learning_rate": 1.82201740522427e-06,
+      "loss": 0.5172,
+      "step": 9032
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.395367261986311,
+      "learning_rate": 1.8210102134948342e-06,
+      "loss": 0.3825,
+      "step": 9033
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.451729574299913,
+      "learning_rate": 1.8200032382405508e-06,
+      "loss": 0.4407,
+      "step": 9034
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.536425345014642,
+      "learning_rate": 1.8189964795299831e-06,
+      "loss": 0.7251,
+      "step": 9035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.606690102827661,
+      "learning_rate": 1.8179899374316916e-06,
+      "loss": 0.4453,
+      "step": 9036
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.704378281677464,
+      "learning_rate": 1.8169836120142153e-06,
+      "loss": 0.3807,
+      "step": 9037
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.298681491454375,
+      "learning_rate": 1.8159775033460825e-06,
+      "loss": 0.5091,
+      "step": 9038
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.034465788741745,
+      "learning_rate": 1.8149716114958044e-06,
+      "loss": 0.4756,
+      "step": 9039
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.257941297002628,
+      "learning_rate": 1.813965936531875e-06,
+      "loss": 0.5731,
+      "step": 9040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.149997069064929,
+      "learning_rate": 1.81296047852278e-06,
+      "loss": 0.4693,
+      "step": 9041
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.174874487912931,
+      "learning_rate": 1.8119552375369836e-06,
+      "loss": 0.4238,
+      "step": 9042
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.1780993082967,
+      "learning_rate": 1.810950213642943e-06,
+      "loss": 0.5138,
+      "step": 9043
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.027437684013894,
+      "learning_rate": 1.8099454069090893e-06,
+      "loss": 0.4397,
+      "step": 9044
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.487744501548201,
+      "learning_rate": 1.8089408174038504e-06,
+      "loss": 0.5286,
+      "step": 9045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.290284247264932,
+      "learning_rate": 1.807936445195631e-06,
+      "loss": 0.504,
+      "step": 9046
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.328606851567976,
+      "learning_rate": 1.8069322903528274e-06,
+      "loss": 0.4983,
+      "step": 9047
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.425753763954716,
+      "learning_rate": 1.8059283529438171e-06,
+      "loss": 0.5127,
+      "step": 9048
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8179186904303242,
+      "learning_rate": 1.804924633036962e-06,
+      "loss": 0.7381,
+      "step": 9049
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.323907895134152,
+      "learning_rate": 1.8039211307006144e-06,
+      "loss": 0.3818,
+      "step": 9050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.893484608104433,
+      "learning_rate": 1.8029178460031045e-06,
+      "loss": 0.4677,
+      "step": 9051
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.974111546754266,
+      "learning_rate": 1.801914779012755e-06,
+      "loss": 0.4158,
+      "step": 9052
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.183986981674218,
+      "learning_rate": 1.8009119297978688e-06,
+      "loss": 0.4944,
+      "step": 9053
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 16.19721208516461,
+      "learning_rate": 1.7999092984267358e-06,
+      "loss": 0.5103,
+      "step": 9054
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.062893043049471,
+      "learning_rate": 1.7989068849676289e-06,
+      "loss": 0.4115,
+      "step": 9055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.9065053827549745,
+      "learning_rate": 1.7979046894888114e-06,
+      "loss": 0.4821,
+      "step": 9056
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.957648021940427,
+      "learning_rate": 1.7969027120585264e-06,
+      "loss": 0.4989,
+      "step": 9057
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.11978888970482,
+      "learning_rate": 1.7959009527450028e-06,
+      "loss": 0.4618,
+      "step": 9058
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.5298085776776205,
+      "learning_rate": 1.7948994116164592e-06,
+      "loss": 0.4046,
+      "step": 9059
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.5679727505717675,
+      "learning_rate": 1.7938980887410924e-06,
+      "loss": 0.4149,
+      "step": 9060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.841526171112369,
+      "learning_rate": 1.7928969841870919e-06,
+      "loss": 0.3884,
+      "step": 9061
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.427841996113155,
+      "learning_rate": 1.7918960980226268e-06,
+      "loss": 0.498,
+      "step": 9062
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.5283560009273325,
+      "learning_rate": 1.7908954303158533e-06,
+      "loss": 0.5587,
+      "step": 9063
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.816760884849613,
+      "learning_rate": 1.7898949811349102e-06,
+      "loss": 0.475,
+      "step": 9064
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.70787461196738,
+      "learning_rate": 1.7888947505479271e-06,
+      "loss": 0.4194,
+      "step": 9065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.564305087129842,
+      "learning_rate": 1.787894738623014e-06,
+      "loss": 0.3626,
+      "step": 9066
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.095406243349979,
+      "learning_rate": 1.7868949454282668e-06,
+      "loss": 0.3135,
+      "step": 9067
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.7485415046848365,
+      "learning_rate": 1.7858953710317656e-06,
+      "loss": 0.5005,
+      "step": 9068
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.498347235839075,
+      "learning_rate": 1.7848960155015783e-06,
+      "loss": 0.4053,
+      "step": 9069
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.8010376064381255,
+      "learning_rate": 1.7838968789057587e-06,
+      "loss": 0.4541,
+      "step": 9070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.271649648048054,
+      "learning_rate": 1.7828979613123422e-06,
+      "loss": 0.3527,
+      "step": 9071
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.057863988143898,
+      "learning_rate": 1.78189926278935e-06,
+      "loss": 0.4566,
+      "step": 9072
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.262072965570786,
+      "learning_rate": 1.7809007834047875e-06,
+      "loss": 0.5233,
+      "step": 9073
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.225191730425506,
+      "learning_rate": 1.7799025232266504e-06,
+      "loss": 0.3796,
+      "step": 9074
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.92448756818129,
+      "learning_rate": 1.7789044823229135e-06,
+      "loss": 0.4491,
+      "step": 9075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.562138523863875,
+      "learning_rate": 1.7779066607615398e-06,
+      "loss": 0.5325,
+      "step": 9076
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.296259148173362,
+      "learning_rate": 1.7769090586104748e-06,
+      "loss": 0.4351,
+      "step": 9077
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 23.8397931974722,
+      "learning_rate": 1.7759116759376516e-06,
+      "loss": 0.4673,
+      "step": 9078
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 17.9821557393871,
+      "learning_rate": 1.77491451281099e-06,
+      "loss": 0.4927,
+      "step": 9079
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.306900447863779,
+      "learning_rate": 1.7739175692983906e-06,
+      "loss": 0.5293,
+      "step": 9080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.157667422041948,
+      "learning_rate": 1.7729208454677404e-06,
+      "loss": 0.5216,
+      "step": 9081
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.267101790469941,
+      "learning_rate": 1.771924341386911e-06,
+      "loss": 0.5143,
+      "step": 9082
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.26906796360511,
+      "learning_rate": 1.7709280571237624e-06,
+      "loss": 0.482,
+      "step": 9083
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.040305725140371,
+      "learning_rate": 1.769931992746136e-06,
+      "loss": 0.495,
+      "step": 9084
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.761055038427074,
+      "learning_rate": 1.7689361483218598e-06,
+      "loss": 0.4729,
+      "step": 9085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.9662319051670885,
+      "learning_rate": 1.7679405239187436e-06,
+      "loss": 0.4723,
+      "step": 9086
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.474181283147697,
+      "learning_rate": 1.7669451196045895e-06,
+      "loss": 0.516,
+      "step": 9087
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.728303916988624,
+      "learning_rate": 1.7659499354471755e-06,
+      "loss": 0.393,
+      "step": 9088
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8836505074731884,
+      "learning_rate": 1.7649549715142738e-06,
+      "loss": 0.733,
+      "step": 9089
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.742383189061871,
+      "learning_rate": 1.763960227873635e-06,
+      "loss": 0.7336,
+      "step": 9090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.249560971136589,
+      "learning_rate": 1.7629657045929948e-06,
+      "loss": 0.4886,
+      "step": 9091
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.621394796040356,
+      "learning_rate": 1.7619714017400797e-06,
+      "loss": 0.4458,
+      "step": 9092
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.734449660554445,
+      "learning_rate": 1.7609773193825952e-06,
+      "loss": 0.7379,
+      "step": 9093
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5850821990773012,
+      "learning_rate": 1.759983457588234e-06,
+      "loss": 0.6847,
+      "step": 9094
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.592381868251106,
+      "learning_rate": 1.7589898164246722e-06,
+      "loss": 0.3453,
+      "step": 9095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.641098116135386,
+      "learning_rate": 1.7579963959595753e-06,
+      "loss": 0.4659,
+      "step": 9096
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.9023842637548,
+      "learning_rate": 1.7570031962605882e-06,
+      "loss": 0.4529,
+      "step": 9097
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.541567548437864,
+      "learning_rate": 1.7560102173953458e-06,
+      "loss": 0.4591,
+      "step": 9098
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.692184742353421,
+      "learning_rate": 1.755017459431465e-06,
+      "loss": 0.4823,
+      "step": 9099
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.01316382155734,
+      "learning_rate": 1.7540249224365458e-06,
+      "loss": 0.4755,
+      "step": 9100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.825736124827947,
+      "learning_rate": 1.7530326064781788e-06,
+      "loss": 0.3612,
+      "step": 9101
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 42.66695174199032,
+      "learning_rate": 1.7520405116239353e-06,
+      "loss": 0.4861,
+      "step": 9102
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.857261980229537,
+      "learning_rate": 1.751048637941372e-06,
+      "loss": 0.4648,
+      "step": 9103
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.856806376329087,
+      "learning_rate": 1.7500569854980303e-06,
+      "loss": 0.5263,
+      "step": 9104
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.847400591895665,
+      "learning_rate": 1.7490655543614392e-06,
+      "loss": 0.4835,
+      "step": 9105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 12.74415654851217,
+      "learning_rate": 1.7480743445991089e-06,
+      "loss": 0.497,
+      "step": 9106
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.188386056201312,
+      "learning_rate": 1.7470833562785389e-06,
+      "loss": 0.4136,
+      "step": 9107
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.269975584578635,
+      "learning_rate": 1.7460925894672098e-06,
+      "loss": 0.4308,
+      "step": 9108
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.629528722356391,
+      "learning_rate": 1.7451020442325866e-06,
+      "loss": 0.4196,
+      "step": 9109
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.742102846890894,
+      "learning_rate": 1.7441117206421248e-06,
+      "loss": 0.4496,
+      "step": 9110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.033705291452167,
+      "learning_rate": 1.743121618763257e-06,
+      "loss": 0.4539,
+      "step": 9111
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.746058098809605,
+      "learning_rate": 1.7421317386634101e-06,
+      "loss": 0.4164,
+      "step": 9112
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.907686983289239,
+      "learning_rate": 1.7411420804099844e-06,
+      "loss": 0.4876,
+      "step": 9113
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.988539925044043,
+      "learning_rate": 1.7401526440703753e-06,
+      "loss": 0.4647,
+      "step": 9114
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.922650805495328,
+      "learning_rate": 1.7391634297119564e-06,
+      "loss": 0.4242,
+      "step": 9115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.805140693281307,
+      "learning_rate": 1.7381744374020921e-06,
+      "loss": 0.4115,
+      "step": 9116
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.508151859100443,
+      "learning_rate": 1.7371856672081266e-06,
+      "loss": 0.6834,
+      "step": 9117
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.816391643986165,
+      "learning_rate": 1.7361971191973886e-06,
+      "loss": 0.4253,
+      "step": 9118
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 29.733938118289764,
+      "learning_rate": 1.7352087934371981e-06,
+      "loss": 0.4935,
+      "step": 9119
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.471492750429394,
+      "learning_rate": 1.734220689994852e-06,
+      "loss": 0.5125,
+      "step": 9120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.196201925658418,
+      "learning_rate": 1.7332328089376405e-06,
+      "loss": 0.4964,
+      "step": 9121
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.713769965021835,
+      "learning_rate": 1.7322451503328274e-06,
+      "loss": 0.7525,
+      "step": 9122
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.623946948668396,
+      "learning_rate": 1.7312577142476733e-06,
+      "loss": 0.4907,
+      "step": 9123
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.312598814827226,
+      "learning_rate": 1.7302705007494146e-06,
+      "loss": 0.4418,
+      "step": 9124
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.133956835496447,
+      "learning_rate": 1.7292835099052797e-06,
+      "loss": 0.5065,
+      "step": 9125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.618429794083283,
+      "learning_rate": 1.7282967417824759e-06,
+      "loss": 0.4543,
+      "step": 9126
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.440402512776516,
+      "learning_rate": 1.7273101964481987e-06,
+      "loss": 0.5488,
+      "step": 9127
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4684960578883017,
+      "learning_rate": 1.7263238739696253e-06,
+      "loss": 0.399,
+      "step": 9128
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.476524799041712,
+      "learning_rate": 1.7253377744139215e-06,
+      "loss": 0.6358,
+      "step": 9129
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.5356723626946085,
+      "learning_rate": 1.7243518978482398e-06,
+      "loss": 0.4525,
+      "step": 9130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 96.17188333645741,
+      "learning_rate": 1.7233662443397075e-06,
+      "loss": 0.5131,
+      "step": 9131
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.18231640332025,
+      "learning_rate": 1.7223808139554483e-06,
+      "loss": 0.4066,
+      "step": 9132
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.571214600088348,
+      "learning_rate": 1.721395606762562e-06,
+      "loss": 0.5913,
+      "step": 9133
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.248685615361051,
+      "learning_rate": 1.7204106228281404e-06,
+      "loss": 0.4824,
+      "step": 9134
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.891141188822021,
+      "learning_rate": 1.719425862219255e-06,
+      "loss": 0.524,
+      "step": 9135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 22.532103341242276,
+      "learning_rate": 1.718441325002963e-06,
+      "loss": 0.4382,
+      "step": 9136
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.165401204582089,
+      "learning_rate": 1.7174570112463063e-06,
+      "loss": 0.5296,
+      "step": 9137
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.137539757524278,
+      "learning_rate": 1.7164729210163127e-06,
+      "loss": 0.4796,
+      "step": 9138
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.644561210381383,
+      "learning_rate": 1.7154890543799995e-06,
+      "loss": 0.475,
+      "step": 9139
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.972873609536837,
+      "learning_rate": 1.7145054114043558e-06,
+      "loss": 0.5715,
+      "step": 9140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.51822959781618,
+      "learning_rate": 1.7135219921563684e-06,
+      "loss": 0.5408,
+      "step": 9141
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.727266874135474,
+      "learning_rate": 1.7125387967030005e-06,
+      "loss": 0.4152,
+      "step": 9142
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.527898251542155,
+      "learning_rate": 1.711555825111207e-06,
+      "loss": 0.4803,
+      "step": 9143
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.336154844135925,
+      "learning_rate": 1.7105730774479223e-06,
+      "loss": 0.5019,
+      "step": 9144
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.610462255768721,
+      "learning_rate": 1.7095905537800666e-06,
+      "loss": 0.4389,
+      "step": 9145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4888695258361178,
+      "learning_rate": 1.7086082541745442e-06,
+      "loss": 0.6776,
+      "step": 9146
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.035260248292465,
+      "learning_rate": 1.707626178698249e-06,
+      "loss": 0.4712,
+      "step": 9147
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.340612100420514,
+      "learning_rate": 1.7066443274180517e-06,
+      "loss": 0.5821,
+      "step": 9148
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.428846429013289,
+      "learning_rate": 1.7056627004008165e-06,
+      "loss": 0.3406,
+      "step": 9149
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.6828907211908,
+      "learning_rate": 1.7046812977133853e-06,
+      "loss": 0.503,
+      "step": 9150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 20.84604538354915,
+      "learning_rate": 1.703700119422586e-06,
+      "loss": 0.349,
+      "step": 9151
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.40273646329424,
+      "learning_rate": 1.7027191655952357e-06,
+      "loss": 0.4475,
+      "step": 9152
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.383840080501646,
+      "learning_rate": 1.7017384362981315e-06,
+      "loss": 0.5493,
+      "step": 9153
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.304358483669469,
+      "learning_rate": 1.7007579315980566e-06,
+      "loss": 0.4097,
+      "step": 9154
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.0418203840435085,
+      "learning_rate": 1.6997776515617776e-06,
+      "loss": 0.3588,
+      "step": 9155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.394222658745727,
+      "learning_rate": 1.6987975962560505e-06,
+      "loss": 0.5384,
+      "step": 9156
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.081772806482439,
+      "learning_rate": 1.697817765747609e-06,
+      "loss": 0.5122,
+      "step": 9157
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.423380282780968,
+      "learning_rate": 1.6968381601031791e-06,
+      "loss": 0.4145,
+      "step": 9158
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.222173241140555,
+      "learning_rate": 1.6958587793894653e-06,
+      "loss": 0.426,
+      "step": 9159
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.388287190359698,
+      "learning_rate": 1.694879623673158e-06,
+      "loss": 0.4231,
+      "step": 9160
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8646998288180994,
+      "learning_rate": 1.6939006930209368e-06,
+      "loss": 0.5215,
+      "step": 9161
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.414413438181507,
+      "learning_rate": 1.6929219874994608e-06,
+      "loss": 0.451,
+      "step": 9162
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.540723490005302,
+      "learning_rate": 1.6919435071753748e-06,
+      "loss": 0.4068,
+      "step": 9163
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.787005944577965,
+      "learning_rate": 1.690965252115308e-06,
+      "loss": 0.4326,
+      "step": 9164
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.425607277549345,
+      "learning_rate": 1.6899872223858787e-06,
+      "loss": 0.5037,
+      "step": 9165
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.645267486324951,
+      "learning_rate": 1.6890094180536825e-06,
+      "loss": 0.4639,
+      "step": 9166
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.341227180056252,
+      "learning_rate": 1.6880318391853096e-06,
+      "loss": 0.5489,
+      "step": 9167
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.68388597583724,
+      "learning_rate": 1.6870544858473204e-06,
+      "loss": 0.529,
+      "step": 9168
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.186935401672587,
+      "learning_rate": 1.6860773581062733e-06,
+      "loss": 0.4718,
+      "step": 9169
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.903374234526819,
+      "learning_rate": 1.6851004560287071e-06,
+      "loss": 0.3787,
+      "step": 9170
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.625118310237632,
+      "learning_rate": 1.6841237796811433e-06,
+      "loss": 0.3857,
+      "step": 9171
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.518444480049813,
+      "learning_rate": 1.6831473291300893e-06,
+      "loss": 0.5272,
+      "step": 9172
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.032261986803552,
+      "learning_rate": 1.6821711044420347e-06,
+      "loss": 0.4469,
+      "step": 9173
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.488170286928793,
+      "learning_rate": 1.6811951056834603e-06,
+      "loss": 0.7095,
+      "step": 9174
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.030292370509693,
+      "learning_rate": 1.6802193329208233e-06,
+      "loss": 0.3955,
+      "step": 9175
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.628288384371166,
+      "learning_rate": 1.6792437862205752e-06,
+      "loss": 0.4249,
+      "step": 9176
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5898409433335081,
+      "learning_rate": 1.6782684656491389e-06,
+      "loss": 0.665,
+      "step": 9177
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.370489532701671,
+      "learning_rate": 1.677293371272935e-06,
+      "loss": 0.3673,
+      "step": 9178
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.612730408769744,
+      "learning_rate": 1.67631850315836e-06,
+      "loss": 0.4607,
+      "step": 9179
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.4590226835392555,
+      "learning_rate": 1.675343861371801e-06,
+      "loss": 0.4145,
+      "step": 9180
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.7595212564253275,
+      "learning_rate": 1.6743694459796245e-06,
+      "loss": 0.3976,
+      "step": 9181
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.159555565892328,
+      "learning_rate": 1.6733952570481837e-06,
+      "loss": 0.4799,
+      "step": 9182
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.396242215994909,
+      "learning_rate": 1.672421294643819e-06,
+      "loss": 0.4188,
+      "step": 9183
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.0634741337818,
+      "learning_rate": 1.6714475588328493e-06,
+      "loss": 0.4115,
+      "step": 9184
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.885170585674469,
+      "learning_rate": 1.6704740496815874e-06,
+      "loss": 0.4792,
+      "step": 9185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.923533647614091,
+      "learning_rate": 1.6695007672563179e-06,
+      "loss": 0.5293,
+      "step": 9186
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.070313377038205,
+      "learning_rate": 1.668527711623323e-06,
+      "loss": 0.514,
+      "step": 9187
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.618334473580328,
+      "learning_rate": 1.667554882848858e-06,
+      "loss": 0.4177,
+      "step": 9188
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.157831374975704,
+      "learning_rate": 1.6665822809991738e-06,
+      "loss": 0.4424,
+      "step": 9189
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.487455500246155,
+      "learning_rate": 1.6656099061404973e-06,
+      "loss": 0.4555,
+      "step": 9190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.380775875246066,
+      "learning_rate": 1.6646377583390415e-06,
+      "loss": 0.4864,
+      "step": 9191
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.599722286738395,
+      "learning_rate": 1.6636658376610094e-06,
+      "loss": 0.432,
+      "step": 9192
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.197421987373542,
+      "learning_rate": 1.6626941441725803e-06,
+      "loss": 0.4163,
+      "step": 9193
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.228089622575509,
+      "learning_rate": 1.6617226779399259e-06,
+      "loss": 0.4209,
+      "step": 9194
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.229544173507005,
+      "learning_rate": 1.6607514390291974e-06,
+      "loss": 0.4677,
+      "step": 9195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.087396103436387,
+      "learning_rate": 1.659780427506531e-06,
+      "loss": 0.4255,
+      "step": 9196
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.464567820919152,
+      "learning_rate": 1.6588096434380474e-06,
+      "loss": 0.4452,
+      "step": 9197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.566961169396738,
+      "learning_rate": 1.657839086889856e-06,
+      "loss": 0.5156,
+      "step": 9198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.9723736556897515,
+      "learning_rate": 1.6568687579280457e-06,
+      "loss": 0.5323,
+      "step": 9199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.58706693840563,
+      "learning_rate": 1.6558986566186897e-06,
+      "loss": 0.506,
+      "step": 9200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.504883595329638,
+      "learning_rate": 1.6549287830278504e-06,
+      "loss": 0.4516,
+      "step": 9201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.459428759595458,
+      "learning_rate": 1.6539591372215701e-06,
+      "loss": 0.3532,
+      "step": 9202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.298938204012212,
+      "learning_rate": 1.6529897192658794e-06,
+      "loss": 0.4964,
+      "step": 9203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.360312764583399,
+      "learning_rate": 1.65202052922679e-06,
+      "loss": 0.4536,
+      "step": 9204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.242504739233025,
+      "learning_rate": 1.6510515671702997e-06,
+      "loss": 0.5205,
+      "step": 9205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.580382079598913,
+      "learning_rate": 1.650082833162389e-06,
+      "loss": 0.3863,
+      "step": 9206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.730478067040884,
+      "learning_rate": 1.6491143272690274e-06,
+      "loss": 0.4489,
+      "step": 9207
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.5923597793452595,
+      "learning_rate": 1.6481460495561646e-06,
+      "loss": 0.3944,
+      "step": 9208
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.985966544233741,
+      "learning_rate": 1.6471780000897335e-06,
+      "loss": 0.31,
+      "step": 9209
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.3953498044803405,
+      "learning_rate": 1.6462101789356587e-06,
+      "loss": 0.4299,
+      "step": 9210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.141064251590272,
+      "learning_rate": 1.6452425861598398e-06,
+      "loss": 0.4193,
+      "step": 9211
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.280138722269025,
+      "learning_rate": 1.6442752218281698e-06,
+      "loss": 0.4851,
+      "step": 9212
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.840520105123765,
+      "learning_rate": 1.6433080860065197e-06,
+      "loss": 0.4458,
+      "step": 9213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.378448065747301,
+      "learning_rate": 1.642341178760748e-06,
+      "loss": 0.6207,
+      "step": 9214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.209392824039832,
+      "learning_rate": 1.6413745001566938e-06,
+      "loss": 0.5499,
+      "step": 9215
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.287033375700262,
+      "learning_rate": 1.6404080502601882e-06,
+      "loss": 0.449,
+      "step": 9216
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.729831383243684,
+      "learning_rate": 1.6394418291370405e-06,
+      "loss": 0.7167,
+      "step": 9217
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.571906783944251,
+      "learning_rate": 1.638475836853045e-06,
+      "loss": 0.5126,
+      "step": 9218
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.688839779524193,
+      "learning_rate": 1.6375100734739812e-06,
+      "loss": 0.4523,
+      "step": 9219
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.439355648571806,
+      "learning_rate": 1.6365445390656137e-06,
+      "loss": 0.4072,
+      "step": 9220
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.86209158597294,
+      "learning_rate": 1.6355792336936937e-06,
+      "loss": 0.5552,
+      "step": 9221
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.326978620751007,
+      "learning_rate": 1.634614157423952e-06,
+      "loss": 0.4862,
+      "step": 9222
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7960255971801244,
+      "learning_rate": 1.6336493103221064e-06,
+      "loss": 0.7026,
+      "step": 9223
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 16.9648233214494,
+      "learning_rate": 1.632684692453857e-06,
+      "loss": 0.3639,
+      "step": 9224
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.827423768488429,
+      "learning_rate": 1.6317203038848932e-06,
+      "loss": 0.4684,
+      "step": 9225
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.813619195042758,
+      "learning_rate": 1.630756144680884e-06,
+      "loss": 0.4287,
+      "step": 9226
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.710796119247545,
+      "learning_rate": 1.6297922149074846e-06,
+      "loss": 0.4003,
+      "step": 9227
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.04724455313512,
+      "learning_rate": 1.6288285146303323e-06,
+      "loss": 0.4869,
+      "step": 9228
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.255141847777493,
+      "learning_rate": 1.6278650439150523e-06,
+      "loss": 0.3285,
+      "step": 9229
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.123487494613439,
+      "learning_rate": 1.6269018028272554e-06,
+      "loss": 0.5558,
+      "step": 9230
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.286663844126089,
+      "learning_rate": 1.6259387914325314e-06,
+      "loss": 0.4194,
+      "step": 9231
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.969380641920462,
+      "learning_rate": 1.6249760097964573e-06,
+      "loss": 0.4382,
+      "step": 9232
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.766418225204345,
+      "learning_rate": 1.6240134579845928e-06,
+      "loss": 0.5039,
+      "step": 9233
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.258666418526158,
+      "learning_rate": 1.6230511360624868e-06,
+      "loss": 0.462,
+      "step": 9234
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.449034935380948,
+      "learning_rate": 1.622089044095667e-06,
+      "loss": 0.3844,
+      "step": 9235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.13131472131507,
+      "learning_rate": 1.6211271821496489e-06,
+      "loss": 0.5888,
+      "step": 9236
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 15.930466881950707,
+      "learning_rate": 1.6201655502899277e-06,
+      "loss": 0.455,
+      "step": 9237
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.785170288731015,
+      "learning_rate": 1.6192041485819903e-06,
+      "loss": 0.4791,
+      "step": 9238
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.7231648042965935,
+      "learning_rate": 1.618242977091301e-06,
+      "loss": 0.4843,
+      "step": 9239
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.374737543852143,
+      "learning_rate": 1.6172820358833147e-06,
+      "loss": 0.4063,
+      "step": 9240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 19.739179256624574,
+      "learning_rate": 1.616321325023465e-06,
+      "loss": 0.4793,
+      "step": 9241
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.269278666495595,
+      "learning_rate": 1.6153608445771702e-06,
+      "loss": 0.5397,
+      "step": 9242
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.194636802119901,
+      "learning_rate": 1.614400594609839e-06,
+      "loss": 0.5205,
+      "step": 9243
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.677103732494173,
+      "learning_rate": 1.6134405751868576e-06,
+      "loss": 0.5693,
+      "step": 9244
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.001300895389132,
+      "learning_rate": 1.6124807863735998e-06,
+      "loss": 0.6969,
+      "step": 9245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.074006357368152,
+      "learning_rate": 1.6115212282354208e-06,
+      "loss": 0.4541,
+      "step": 9246
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.334133145134759,
+      "learning_rate": 1.610561900837666e-06,
+      "loss": 0.3527,
+      "step": 9247
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.981294366577287,
+      "learning_rate": 1.609602804245658e-06,
+      "loss": 0.5281,
+      "step": 9248
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.109042039881173,
+      "learning_rate": 1.6086439385247099e-06,
+      "loss": 0.4997,
+      "step": 9249
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.568803065906587,
+      "learning_rate": 1.6076853037401148e-06,
+      "loss": 0.4856,
+      "step": 9250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.214705655317854,
+      "learning_rate": 1.60672689995715e-06,
+      "loss": 0.417,
+      "step": 9251
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.043322751806079,
+      "learning_rate": 1.6057687272410816e-06,
+      "loss": 0.4504,
+      "step": 9252
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.814275876543829,
+      "learning_rate": 1.6048107856571543e-06,
+      "loss": 0.4338,
+      "step": 9253
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.163072659228886,
+      "learning_rate": 1.6038530752706032e-06,
+      "loss": 0.448,
+      "step": 9254
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.46226850372341,
+      "learning_rate": 1.6028955961466392e-06,
+      "loss": 0.4301,
+      "step": 9255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.104832036665454,
+      "learning_rate": 1.6019383483504663e-06,
+      "loss": 0.5132,
+      "step": 9256
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6599442979539782,
+      "learning_rate": 1.600981331947266e-06,
+      "loss": 0.4203,
+      "step": 9257
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.730145536399041,
+      "learning_rate": 1.6000245470022103e-06,
+      "loss": 0.417,
+      "step": 9258
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.741531784437342,
+      "learning_rate": 1.5990679935804498e-06,
+      "loss": 0.4176,
+      "step": 9259
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8052411817219642,
+      "learning_rate": 1.5981116717471202e-06,
+      "loss": 0.4667,
+      "step": 9260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.899167362068836,
+      "learning_rate": 1.5971555815673461e-06,
+      "loss": 0.4499,
+      "step": 9261
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.209512606473791,
+      "learning_rate": 1.5961997231062298e-06,
+      "loss": 0.5117,
+      "step": 9262
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.717771553051222,
+      "learning_rate": 1.5952440964288658e-06,
+      "loss": 0.3922,
+      "step": 9263
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.993901906416523,
+      "learning_rate": 1.5942887016003216e-06,
+      "loss": 0.4258,
+      "step": 9264
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.545334643738985,
+      "learning_rate": 1.5933335386856602e-06,
+      "loss": 0.4207,
+      "step": 9265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.384494716205579,
+      "learning_rate": 1.5923786077499204e-06,
+      "loss": 0.3632,
+      "step": 9266
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.382043874278709,
+      "learning_rate": 1.591423908858133e-06,
+      "loss": 0.5084,
+      "step": 9267
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.7454163971252745,
+      "learning_rate": 1.5904694420753065e-06,
+      "loss": 0.4743,
+      "step": 9268
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.8192220386089035,
+      "learning_rate": 1.589515207466434e-06,
+      "loss": 0.3944,
+      "step": 9269
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 16.25514462107929,
+      "learning_rate": 1.5885612050964983e-06,
+      "loss": 0.562,
+      "step": 9270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.523305847187808,
+      "learning_rate": 1.5876074350304593e-06,
+      "loss": 0.4953,
+      "step": 9271
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4136874003476936,
+      "learning_rate": 1.5866538973332695e-06,
+      "loss": 0.4776,
+      "step": 9272
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.926959017754351,
+      "learning_rate": 1.5857005920698543e-06,
+      "loss": 0.5449,
+      "step": 9273
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7192452522401942,
+      "learning_rate": 1.5847475193051336e-06,
+      "loss": 0.7118,
+      "step": 9274
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.074644102902866,
+      "learning_rate": 1.5837946791040053e-06,
+      "loss": 0.4467,
+      "step": 9275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.583842276083845,
+      "learning_rate": 1.5828420715313564e-06,
+      "loss": 0.4163,
+      "step": 9276
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.951449433319246,
+      "learning_rate": 1.5818896966520535e-06,
+      "loss": 0.4793,
+      "step": 9277
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.413240029925636,
+      "learning_rate": 1.580937554530949e-06,
+      "loss": 0.447,
+      "step": 9278
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.7915270514949135,
+      "learning_rate": 1.5799856452328788e-06,
+      "loss": 0.5474,
+      "step": 9279
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.866211303764798,
+      "learning_rate": 1.579033968822664e-06,
+      "loss": 0.5084,
+      "step": 9280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.200485033257351,
+      "learning_rate": 1.5780825253651138e-06,
+      "loss": 0.4959,
+      "step": 9281
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.599969095885136,
+      "learning_rate": 1.5771313149250112e-06,
+      "loss": 0.4681,
+      "step": 9282
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9859789650006023,
+      "learning_rate": 1.5761803375671326e-06,
+      "loss": 0.47,
+      "step": 9283
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8952526154509406,
+      "learning_rate": 1.5752295933562333e-06,
+      "loss": 0.6754,
+      "step": 9284
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9357598069836384,
+      "learning_rate": 1.5742790823570585e-06,
+      "loss": 0.8048,
+      "step": 9285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.82684970785863,
+      "learning_rate": 1.5733288046343308e-06,
+      "loss": 0.4505,
+      "step": 9286
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.733630212194706,
+      "learning_rate": 1.5723787602527612e-06,
+      "loss": 0.5028,
+      "step": 9287
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.60130437266907,
+      "learning_rate": 1.5714289492770407e-06,
+      "loss": 0.4178,
+      "step": 9288
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.176180670087994,
+      "learning_rate": 1.5704793717718497e-06,
+      "loss": 0.5164,
+      "step": 9289
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.094380859155923,
+      "learning_rate": 1.5695300278018521e-06,
+      "loss": 0.496,
+      "step": 9290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.504886095538684,
+      "learning_rate": 1.5685809174316912e-06,
+      "loss": 0.6053,
+      "step": 9291
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.010356245168092,
+      "learning_rate": 1.567632040725998e-06,
+      "loss": 0.4678,
+      "step": 9292
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.426267161022059,
+      "learning_rate": 1.5666833977493857e-06,
+      "loss": 0.4742,
+      "step": 9293
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.361866241734769,
+      "learning_rate": 1.5657349885664552e-06,
+      "loss": 0.5001,
+      "step": 9294
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.692453867398974,
+      "learning_rate": 1.5647868132417876e-06,
+      "loss": 0.4885,
+      "step": 9295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.716297287696114,
+      "learning_rate": 1.5638388718399488e-06,
+      "loss": 0.4495,
+      "step": 9296
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.320472139142913,
+      "learning_rate": 1.562891164425489e-06,
+      "loss": 0.5904,
+      "step": 9297
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.944674058787178,
+      "learning_rate": 1.5619436910629454e-06,
+      "loss": 0.538,
+      "step": 9298
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.818801698229147,
+      "learning_rate": 1.5609964518168336e-06,
+      "loss": 0.4847,
+      "step": 9299
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.837484088364063,
+      "learning_rate": 1.560049446751659e-06,
+      "loss": 0.5195,
+      "step": 9300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.711201481213373,
+      "learning_rate": 1.5591026759319083e-06,
+      "loss": 0.4676,
+      "step": 9301
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.175679803711659,
+      "learning_rate": 1.5581561394220496e-06,
+      "loss": 0.5699,
+      "step": 9302
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.71041351523031,
+      "learning_rate": 1.5572098372865414e-06,
+      "loss": 0.5011,
+      "step": 9303
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.275328142496613,
+      "learning_rate": 1.5562637695898214e-06,
+      "loss": 0.4408,
+      "step": 9304
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.864174620312938,
+      "learning_rate": 1.5553179363963123e-06,
+      "loss": 0.4977,
+      "step": 9305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 47.42764244931894,
+      "learning_rate": 1.5543723377704195e-06,
+      "loss": 0.5484,
+      "step": 9306
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.864172142321674,
+      "learning_rate": 1.5534269737765373e-06,
+      "loss": 0.4924,
+      "step": 9307
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.40210291873308,
+      "learning_rate": 1.552481844479038e-06,
+      "loss": 0.4383,
+      "step": 9308
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.03554625594599,
+      "learning_rate": 1.551536949942285e-06,
+      "loss": 0.4397,
+      "step": 9309
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.8150024423097415,
+      "learning_rate": 1.5505922902306153e-06,
+      "loss": 0.4772,
+      "step": 9310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9287689930314544,
+      "learning_rate": 1.549647865408359e-06,
+      "loss": 0.7153,
+      "step": 9311
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.5119078461303905,
+      "learning_rate": 1.54870367553983e-06,
+      "loss": 0.5154,
+      "step": 9312
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.331295311922123,
+      "learning_rate": 1.5477597206893197e-06,
+      "loss": 0.3938,
+      "step": 9313
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.377503737101946,
+      "learning_rate": 1.5468160009211091e-06,
+      "loss": 0.4726,
+      "step": 9314
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.439800185711812,
+      "learning_rate": 1.5458725162994593e-06,
+      "loss": 0.4848,
+      "step": 9315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.631140814098113,
+      "learning_rate": 1.54492926688862e-06,
+      "loss": 0.5187,
+      "step": 9316
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.389281895415378,
+      "learning_rate": 1.5439862527528205e-06,
+      "loss": 0.4658,
+      "step": 9317
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.601522129614987,
+      "learning_rate": 1.543043473956279e-06,
+      "loss": 0.4374,
+      "step": 9318
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.386272259626181,
+      "learning_rate": 1.542100930563189e-06,
+      "loss": 0.5387,
+      "step": 9319
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.0931468488326335,
+      "learning_rate": 1.5411586226377368e-06,
+      "loss": 0.3846,
+      "step": 9320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.827584239804985,
+      "learning_rate": 1.5402165502440907e-06,
+      "loss": 0.4124,
+      "step": 9321
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.0056267190696015,
+      "learning_rate": 1.5392747134464004e-06,
+      "loss": 0.5206,
+      "step": 9322
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.1743887292706345,
+      "learning_rate": 1.5383331123088007e-06,
+      "loss": 0.4256,
+      "step": 9323
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.750457717583861,
+      "learning_rate": 1.5373917468954087e-06,
+      "loss": 0.5031,
+      "step": 9324
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.7197284819544185,
+      "learning_rate": 1.536450617270331e-06,
+      "loss": 0.4955,
+      "step": 9325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.584816693053496,
+      "learning_rate": 1.5355097234976502e-06,
+      "loss": 0.4905,
+      "step": 9326
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.549914204699812,
+      "learning_rate": 1.534569065641441e-06,
+      "loss": 0.4967,
+      "step": 9327
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.676166687387654,
+      "learning_rate": 1.533628643765756e-06,
+      "loss": 0.4563,
+      "step": 9328
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.836662222184803,
+      "learning_rate": 1.532688457934634e-06,
+      "loss": 0.3336,
+      "step": 9329
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.273873178805856,
+      "learning_rate": 1.5317485082120959e-06,
+      "loss": 0.512,
+      "step": 9330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.3937252607822765,
+      "learning_rate": 1.5308087946621508e-06,
+      "loss": 0.4899,
+      "step": 9331
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.186645195185389,
+      "learning_rate": 1.5298693173487888e-06,
+      "loss": 0.4866,
+      "step": 9332
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.591390288471547,
+      "learning_rate": 1.528930076335981e-06,
+      "loss": 0.421,
+      "step": 9333
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.179441799597755,
+      "learning_rate": 1.5279910716876895e-06,
+      "loss": 0.5489,
+      "step": 9334
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.494291949190769,
+      "learning_rate": 1.527052303467853e-06,
+      "loss": 0.442,
+      "step": 9335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.231014348530702,
+      "learning_rate": 1.5261137717404012e-06,
+      "loss": 0.4615,
+      "step": 9336
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.463097942177479,
+      "learning_rate": 1.5251754765692417e-06,
+      "loss": 0.4893,
+      "step": 9337
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.517101724124062,
+      "learning_rate": 1.5242374180182683e-06,
+      "loss": 0.551,
+      "step": 9338
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 14.989488843843413,
+      "learning_rate": 1.5232995961513575e-06,
+      "loss": 0.4899,
+      "step": 9339
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.286292023199837,
+      "learning_rate": 1.5223620110323734e-06,
+      "loss": 0.4235,
+      "step": 9340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.061023319161972,
+      "learning_rate": 1.5214246627251606e-06,
+      "loss": 0.481,
+      "step": 9341
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.574157082340874,
+      "learning_rate": 1.5204875512935458e-06,
+      "loss": 0.3878,
+      "step": 9342
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 75.95089472106707,
+      "learning_rate": 1.5195506768013457e-06,
+      "loss": 0.4464,
+      "step": 9343
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.501768435974091,
+      "learning_rate": 1.5186140393123544e-06,
+      "loss": 0.4671,
+      "step": 9344
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.801739431334497,
+      "learning_rate": 1.5176776388903564e-06,
+      "loss": 0.5132,
+      "step": 9345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.252556816653595,
+      "learning_rate": 1.5167414755991138e-06,
+      "loss": 0.3854,
+      "step": 9346
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.366727666290247,
+      "learning_rate": 1.5158055495023755e-06,
+      "loss": 0.5141,
+      "step": 9347
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.056549118390898,
+      "learning_rate": 1.5148698606638723e-06,
+      "loss": 0.4669,
+      "step": 9348
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.451904270605207,
+      "learning_rate": 1.5139344091473247e-06,
+      "loss": 0.5047,
+      "step": 9349
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.12849909690733,
+      "learning_rate": 1.5129991950164296e-06,
+      "loss": 0.5001,
+      "step": 9350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.045936704073063,
+      "learning_rate": 1.5120642183348706e-06,
+      "loss": 0.7616,
+      "step": 9351
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.372084930325746,
+      "learning_rate": 1.5111294791663172e-06,
+      "loss": 0.4748,
+      "step": 9352
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.462082659764787,
+      "learning_rate": 1.5101949775744195e-06,
+      "loss": 0.5334,
+      "step": 9353
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.123081491985098,
+      "learning_rate": 1.5092607136228155e-06,
+      "loss": 0.3344,
+      "step": 9354
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.610841943472097,
+      "learning_rate": 1.5083266873751229e-06,
+      "loss": 0.4903,
+      "step": 9355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.9244600831935355,
+      "learning_rate": 1.5073928988949443e-06,
+      "loss": 0.4247,
+      "step": 9356
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.864364859236207,
+      "learning_rate": 1.506459348245865e-06,
+      "loss": 0.506,
+      "step": 9357
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.6443695009435215,
+      "learning_rate": 1.5055260354914597e-06,
+      "loss": 0.4646,
+      "step": 9358
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7413210446360698,
+      "learning_rate": 1.5045929606952803e-06,
+      "loss": 0.7313,
+      "step": 9359
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.634059923461196,
+      "learning_rate": 1.5036601239208637e-06,
+      "loss": 0.3603,
+      "step": 9360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6187357737052084,
+      "learning_rate": 1.5027275252317358e-06,
+      "loss": 0.465,
+      "step": 9361
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.295706492054143,
+      "learning_rate": 1.5017951646913987e-06,
+      "loss": 0.4291,
+      "step": 9362
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.638239227569301,
+      "learning_rate": 1.5008630423633452e-06,
+      "loss": 0.7209,
+      "step": 9363
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.864982974701132,
+      "learning_rate": 1.4999311583110476e-06,
+      "loss": 0.7387,
+      "step": 9364
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.860629017658727,
+      "learning_rate": 1.4989995125979624e-06,
+      "loss": 0.4256,
+      "step": 9365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.422794521677054,
+      "learning_rate": 1.498068105287529e-06,
+      "loss": 0.4694,
+      "step": 9366
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 12.43097312219174,
+      "learning_rate": 1.4971369364431759e-06,
+      "loss": 0.4542,
+      "step": 9367
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.7531023751307755,
+      "learning_rate": 1.4962060061283095e-06,
+      "loss": 0.4444,
+      "step": 9368
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.103350241990673,
+      "learning_rate": 1.4952753144063225e-06,
+      "loss": 0.4464,
+      "step": 9369
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.673740404290092,
+      "learning_rate": 1.4943448613405892e-06,
+      "loss": 0.7117,
+      "step": 9370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.8867796528264265,
+      "learning_rate": 1.4934146469944704e-06,
+      "loss": 0.4569,
+      "step": 9371
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.345786257795443,
+      "learning_rate": 1.4924846714313112e-06,
+      "loss": 0.5252,
+      "step": 9372
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.812542612515497,
+      "learning_rate": 1.491554934714438e-06,
+      "loss": 0.4312,
+      "step": 9373
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.887144416708999,
+      "learning_rate": 1.490625436907161e-06,
+      "loss": 0.3721,
+      "step": 9374
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.604169119863901,
+      "learning_rate": 1.4896961780727743e-06,
+      "loss": 0.4345,
+      "step": 9375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.0313049971816675,
+      "learning_rate": 1.4887671582745578e-06,
+      "loss": 0.3948,
+      "step": 9376
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.973000457188021,
+      "learning_rate": 1.4878383775757738e-06,
+      "loss": 0.4514,
+      "step": 9377
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.051307252459054,
+      "learning_rate": 1.4869098360396672e-06,
+      "loss": 0.3977,
+      "step": 9378
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.024370841232978,
+      "learning_rate": 1.485981533729466e-06,
+      "loss": 0.3661,
+      "step": 9379
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.208808649950171,
+      "learning_rate": 1.4850534707083853e-06,
+      "loss": 0.409,
+      "step": 9380
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.897965891020932,
+      "learning_rate": 1.4841256470396236e-06,
+      "loss": 0.543,
+      "step": 9381
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.903766976311757,
+      "learning_rate": 1.4831980627863606e-06,
+      "loss": 0.7346,
+      "step": 9382
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.134799291685776,
+      "learning_rate": 1.4822707180117596e-06,
+      "loss": 0.4761,
+      "step": 9383
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.826197844255786,
+      "learning_rate": 1.4813436127789683e-06,
+      "loss": 0.4331,
+      "step": 9384
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.385486268866553,
+      "learning_rate": 1.48041674715112e-06,
+      "loss": 0.4232,
+      "step": 9385
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.013669611805694,
+      "learning_rate": 1.4794901211913287e-06,
+      "loss": 0.5081,
+      "step": 9386
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.245810027919594,
+      "learning_rate": 1.4785637349626968e-06,
+      "loss": 0.4406,
+      "step": 9387
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.181950894517239,
+      "learning_rate": 1.4776375885283023e-06,
+      "loss": 0.4757,
+      "step": 9388
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.832588164862777,
+      "learning_rate": 1.4767116819512151e-06,
+      "loss": 0.4473,
+      "step": 9389
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.562436616832732,
+      "learning_rate": 1.475786015294483e-06,
+      "loss": 0.4734,
+      "step": 9390
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.699877105240892,
+      "learning_rate": 1.4748605886211426e-06,
+      "loss": 0.4081,
+      "step": 9391
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.210690906751231,
+      "learning_rate": 1.4739354019942098e-06,
+      "loss": 0.3229,
+      "step": 9392
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.855077660673735,
+      "learning_rate": 1.4730104554766844e-06,
+      "loss": 0.5462,
+      "step": 9393
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.575940810303795,
+      "learning_rate": 1.4720857491315537e-06,
+      "loss": 0.4366,
+      "step": 9394
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2603667669063965,
+      "learning_rate": 1.4711612830217836e-06,
+      "loss": 0.5404,
+      "step": 9395
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.257779311870572,
+      "learning_rate": 1.4702370572103308e-06,
+      "loss": 0.3778,
+      "step": 9396
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.137346189235281,
+      "learning_rate": 1.4693130717601246e-06,
+      "loss": 0.4555,
+      "step": 9397
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.187734850247025,
+      "learning_rate": 1.4683893267340888e-06,
+      "loss": 0.5803,
+      "step": 9398
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.884883275474808,
+      "learning_rate": 1.4674658221951232e-06,
+      "loss": 0.3895,
+      "step": 9399
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.301444516445843,
+      "learning_rate": 1.4665425582061183e-06,
+      "loss": 0.4167,
+      "step": 9400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.233324940954399,
+      "learning_rate": 1.4656195348299419e-06,
+      "loss": 0.3414,
+      "step": 9401
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.2984081346533305,
+      "learning_rate": 1.4646967521294463e-06,
+      "loss": 0.4454,
+      "step": 9402
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6617220933994403,
+      "learning_rate": 1.4637742101674723e-06,
+      "loss": 0.4002,
+      "step": 9403
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.387161437896506,
+      "learning_rate": 1.4628519090068382e-06,
+      "loss": 0.4826,
+      "step": 9404
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.686692290548038,
+      "learning_rate": 1.4619298487103523e-06,
+      "loss": 0.4486,
+      "step": 9405
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.915966636745442,
+      "learning_rate": 1.4610080293407974e-06,
+      "loss": 0.5479,
+      "step": 9406
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.816588005681993,
+      "learning_rate": 1.4600864509609498e-06,
+      "loss": 0.4816,
+      "step": 9407
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.315367745858888,
+      "learning_rate": 1.4591651136335617e-06,
+      "loss": 0.5247,
+      "step": 9408
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.496808058798536,
+      "learning_rate": 1.4582440174213751e-06,
+      "loss": 0.3687,
+      "step": 9409
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.47661056160978,
+      "learning_rate": 1.457323162387112e-06,
+      "loss": 0.4117,
+      "step": 9410
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.09001754512499,
+      "learning_rate": 1.4564025485934752e-06,
+      "loss": 0.4079,
+      "step": 9411
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.195369755975003,
+      "learning_rate": 1.455482176103159e-06,
+      "loss": 0.4913,
+      "step": 9412
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.724008234271033,
+      "learning_rate": 1.4545620449788327e-06,
+      "loss": 0.5113,
+      "step": 9413
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.169086355291608,
+      "learning_rate": 1.4536421552831592e-06,
+      "loss": 0.5399,
+      "step": 9414
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.4798496315508025,
+      "learning_rate": 1.452722507078771e-06,
+      "loss": 0.5088,
+      "step": 9415
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3139188481940653,
+      "learning_rate": 1.4518031004282974e-06,
+      "loss": 0.3154,
+      "step": 9416
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.798273238800471,
+      "learning_rate": 1.4508839353943432e-06,
+      "loss": 0.3947,
+      "step": 9417
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.6535299391798475,
+      "learning_rate": 1.449965012039502e-06,
+      "loss": 0.4757,
+      "step": 9418
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.757167846307963,
+      "learning_rate": 1.4490463304263474e-06,
+      "loss": 0.5138,
+      "step": 9419
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.129200640445519,
+      "learning_rate": 1.4481278906174357e-06,
+      "loss": 0.4333,
+      "step": 9420
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.538388558372169,
+      "learning_rate": 1.4472096926753122e-06,
+      "loss": 0.4846,
+      "step": 9421
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.092170722278735,
+      "learning_rate": 1.4462917366624978e-06,
+      "loss": 0.5026,
+      "step": 9422
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.437108970547818,
+      "learning_rate": 1.4453740226415052e-06,
+      "loss": 0.4229,
+      "step": 9423
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.739462844869097,
+      "learning_rate": 1.4444565506748255e-06,
+      "loss": 0.4136,
+      "step": 9424
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.52853062624691,
+      "learning_rate": 1.4435393208249338e-06,
+      "loss": 0.4634,
+      "step": 9425
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.948329419226072,
+      "learning_rate": 1.4426223331542883e-06,
+      "loss": 0.4119,
+      "step": 9426
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.721275156530062,
+      "learning_rate": 1.4417055877253344e-06,
+      "loss": 0.5217,
+      "step": 9427
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.096907704176193,
+      "learning_rate": 1.4407890846004975e-06,
+      "loss": 0.3601,
+      "step": 9428
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.845705595580839,
+      "learning_rate": 1.439872823842186e-06,
+      "loss": 0.4522,
+      "step": 9429
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.904927508439084,
+      "learning_rate": 1.438956805512794e-06,
+      "loss": 0.4946,
+      "step": 9430
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.395985504199167,
+      "learning_rate": 1.4380410296746976e-06,
+      "loss": 0.5543,
+      "step": 9431
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8419054785176836,
+      "learning_rate": 1.4371254963902598e-06,
+      "loss": 0.7115,
+      "step": 9432
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.108594801419684,
+      "learning_rate": 1.436210205721823e-06,
+      "loss": 0.4625,
+      "step": 9433
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 19.044753907722185,
+      "learning_rate": 1.4352951577317136e-06,
+      "loss": 0.4619,
+      "step": 9434
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.007321120739263,
+      "learning_rate": 1.4343803524822413e-06,
+      "loss": 0.5141,
+      "step": 9435
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.340109803391117,
+      "learning_rate": 1.4334657900357036e-06,
+      "loss": 0.4834,
+      "step": 9436
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.656397198105975,
+      "learning_rate": 1.4325514704543764e-06,
+      "loss": 0.4898,
+      "step": 9437
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.174681753719594,
+      "learning_rate": 1.43163739380052e-06,
+      "loss": 0.4032,
+      "step": 9438
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.372735888353386,
+      "learning_rate": 1.430723560136379e-06,
+      "loss": 0.4767,
+      "step": 9439
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.929362519766427,
+      "learning_rate": 1.4298099695241829e-06,
+      "loss": 0.4925,
+      "step": 9440
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.404501254958982,
+      "learning_rate": 1.4288966220261418e-06,
+      "loss": 0.5336,
+      "step": 9441
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8766244812493875,
+      "learning_rate": 1.427983517704452e-06,
+      "loss": 0.3762,
+      "step": 9442
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.632331766580607,
+      "learning_rate": 1.4270706566212916e-06,
+      "loss": 0.5156,
+      "step": 9443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.047965417865702,
+      "learning_rate": 1.4261580388388202e-06,
+      "loss": 0.5182,
+      "step": 9444
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.121116689632738,
+      "learning_rate": 1.4252456644191864e-06,
+      "loss": 0.4737,
+      "step": 9445
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.645696327256138,
+      "learning_rate": 1.4243335334245173e-06,
+      "loss": 0.5047,
+      "step": 9446
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.1093849119359085,
+      "learning_rate": 1.4234216459169249e-06,
+      "loss": 0.4505,
+      "step": 9447
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.777105220400049,
+      "learning_rate": 1.422510001958503e-06,
+      "loss": 0.412,
+      "step": 9448
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.8429800689331275,
+      "learning_rate": 1.4215986016113337e-06,
+      "loss": 0.4363,
+      "step": 9449
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 22.11288692320535,
+      "learning_rate": 1.4206874449374764e-06,
+      "loss": 0.3493,
+      "step": 9450
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.7272233333381415,
+      "learning_rate": 1.4197765319989803e-06,
+      "loss": 0.4321,
+      "step": 9451
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.394582620117091,
+      "learning_rate": 1.4188658628578721e-06,
+      "loss": 0.3706,
+      "step": 9452
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.939016197767407,
+      "learning_rate": 1.4179554375761633e-06,
+      "loss": 0.4684,
+      "step": 9453
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.0529415506405035,
+      "learning_rate": 1.4170452562158532e-06,
+      "loss": 0.3301,
+      "step": 9454
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.854114134443096,
+      "learning_rate": 1.4161353188389193e-06,
+      "loss": 0.4825,
+      "step": 9455
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7612483279576348,
+      "learning_rate": 1.4152256255073237e-06,
+      "loss": 0.6955,
+      "step": 9456
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.650108175821752,
+      "learning_rate": 1.414316176283012e-06,
+      "loss": 0.4176,
+      "step": 9457
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.808204947682487,
+      "learning_rate": 1.4134069712279163e-06,
+      "loss": 0.4302,
+      "step": 9458
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.455918061628744,
+      "learning_rate": 1.412498010403946e-06,
+      "loss": 0.4631,
+      "step": 9459
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.511210682435372,
+      "learning_rate": 1.4115892938730025e-06,
+      "loss": 0.3815,
+      "step": 9460
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.491435809514115,
+      "learning_rate": 1.4106808216969591e-06,
+      "loss": 0.4104,
+      "step": 9461
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.339097948824057,
+      "learning_rate": 1.4097725939376816e-06,
+      "loss": 0.5095,
+      "step": 9462
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.28936436116714,
+      "learning_rate": 1.4088646106570182e-06,
+      "loss": 0.2891,
+      "step": 9463
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.388009494487879,
+      "learning_rate": 1.4079568719167963e-06,
+      "loss": 0.4985,
+      "step": 9464
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.532972682928396,
+      "learning_rate": 1.4070493777788298e-06,
+      "loss": 0.4878,
+      "step": 9465
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.50917219227271,
+      "learning_rate": 1.4061421283049131e-06,
+      "loss": 0.3903,
+      "step": 9466
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.131479699093958,
+      "learning_rate": 1.4052351235568284e-06,
+      "loss": 0.4328,
+      "step": 9467
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.902851377615033,
+      "learning_rate": 1.4043283635963367e-06,
+      "loss": 0.4403,
+      "step": 9468
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.673057306516663,
+      "learning_rate": 1.4034218484851864e-06,
+      "loss": 0.4888,
+      "step": 9469
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.089246729342264,
+      "learning_rate": 1.4025155782851063e-06,
+      "loss": 0.4923,
+      "step": 9470
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 13.943699607205131,
+      "learning_rate": 1.4016095530578078e-06,
+      "loss": 0.4079,
+      "step": 9471
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.037717685850033,
+      "learning_rate": 1.4007037728649902e-06,
+      "loss": 0.4541,
+      "step": 9472
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.9390492222371964,
+      "learning_rate": 1.399798237768331e-06,
+      "loss": 0.5233,
+      "step": 9473
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.194885044486799,
+      "learning_rate": 1.3988929478294943e-06,
+      "loss": 0.4501,
+      "step": 9474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.972050897168191,
+      "learning_rate": 1.397987903110124e-06,
+      "loss": 0.5049,
+      "step": 9475
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.019293399588567,
+      "learning_rate": 1.3970831036718525e-06,
+      "loss": 0.449,
+      "step": 9476
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.455699481306302,
+      "learning_rate": 1.3961785495762904e-06,
+      "loss": 0.479,
+      "step": 9477
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.443606788207415,
+      "learning_rate": 1.395274240885036e-06,
+      "loss": 0.4154,
+      "step": 9478
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7947103954797217,
+      "learning_rate": 1.3943701776596685e-06,
+      "loss": 0.3835,
+      "step": 9479
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.763035302417368,
+      "learning_rate": 1.3934663599617493e-06,
+      "loss": 0.484,
+      "step": 9480
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.036053597824214,
+      "learning_rate": 1.392562787852823e-06,
+      "loss": 0.3346,
+      "step": 9481
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.0190662232822,
+      "learning_rate": 1.3916594613944213e-06,
+      "loss": 0.449,
+      "step": 9482
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.2637508593237,
+      "learning_rate": 1.3907563806480584e-06,
+      "loss": 0.3858,
+      "step": 9483
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.5698780220912605,
+      "learning_rate": 1.3898535456752254e-06,
+      "loss": 0.4574,
+      "step": 9484
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.039549475584403,
+      "learning_rate": 1.3889509565374055e-06,
+      "loss": 0.4021,
+      "step": 9485
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.565912846382535,
+      "learning_rate": 1.3880486132960574e-06,
+      "loss": 0.4798,
+      "step": 9486
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.829581039902234,
+      "learning_rate": 1.3871465160126307e-06,
+      "loss": 0.4516,
+      "step": 9487
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.7564295943741355,
+      "learning_rate": 1.3862446647485516e-06,
+      "loss": 0.4486,
+      "step": 9488
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.599182707182837,
+      "learning_rate": 1.3853430595652328e-06,
+      "loss": 0.4719,
+      "step": 9489
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.969400319382041,
+      "learning_rate": 1.3844417005240685e-06,
+      "loss": 0.4781,
+      "step": 9490
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.055104973926132,
+      "learning_rate": 1.3835405876864384e-06,
+      "loss": 0.5473,
+      "step": 9491
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.104087596291008,
+      "learning_rate": 1.382639721113707e-06,
+      "loss": 0.4427,
+      "step": 9492
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.434338166622639,
+      "learning_rate": 1.3817391008672137e-06,
+      "loss": 0.5486,
+      "step": 9493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.880868250224413,
+      "learning_rate": 1.3808387270082907e-06,
+      "loss": 0.5167,
+      "step": 9494
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.16553814873599,
+      "learning_rate": 1.3799385995982474e-06,
+      "loss": 0.3592,
+      "step": 9495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.828156957666391,
+      "learning_rate": 1.379038718698381e-06,
+      "loss": 0.5512,
+      "step": 9496
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.61855298298978,
+      "learning_rate": 1.3781390843699678e-06,
+      "loss": 0.6432,
+      "step": 9497
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.432412440421972,
+      "learning_rate": 1.377239696674269e-06,
+      "loss": 0.3999,
+      "step": 9498
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.482852724630668,
+      "learning_rate": 1.376340555672528e-06,
+      "loss": 0.4415,
+      "step": 9499
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.228423794787605,
+      "learning_rate": 1.3754416614259742e-06,
+      "loss": 0.3955,
+      "step": 9500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.632114414210458,
+      "learning_rate": 1.3745430139958177e-06,
+      "loss": 0.5301,
+      "step": 9501
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.171642710715887,
+      "learning_rate": 1.3736446134432502e-06,
+      "loss": 0.4598,
+      "step": 9502
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.743037045543048,
+      "learning_rate": 1.3727464598294526e-06,
+      "loss": 0.3961,
+      "step": 9503
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.9615897060615195,
+      "learning_rate": 1.3718485532155817e-06,
+      "loss": 0.3786,
+      "step": 9504
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.934166234639333,
+      "learning_rate": 1.3709508936627846e-06,
+      "loss": 0.397,
+      "step": 9505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.651504937712948,
+      "learning_rate": 1.3700534812321852e-06,
+      "loss": 0.7653,
+      "step": 9506
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.107633479279002,
+      "learning_rate": 1.3691563159848947e-06,
+      "loss": 0.4364,
+      "step": 9507
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.57978406506225,
+      "learning_rate": 1.3682593979820035e-06,
+      "loss": 0.5062,
+      "step": 9508
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.350314477815543,
+      "learning_rate": 1.3673627272845919e-06,
+      "loss": 0.4289,
+      "step": 9509
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.424423744465594,
+      "learning_rate": 1.3664663039537163e-06,
+      "loss": 0.6238,
+      "step": 9510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.781730038898662,
+      "learning_rate": 1.3655701280504185e-06,
+      "loss": 0.4588,
+      "step": 9511
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.901440516635677,
+      "learning_rate": 1.3646741996357276e-06,
+      "loss": 0.4847,
+      "step": 9512
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.716886762141675,
+      "learning_rate": 1.3637785187706481e-06,
+      "loss": 0.4626,
+      "step": 9513
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.220364455380273,
+      "learning_rate": 1.3628830855161762e-06,
+      "loss": 0.443,
+      "step": 9514
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.11409369265574,
+      "learning_rate": 1.361987899933284e-06,
+      "loss": 0.4384,
+      "step": 9515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.452468741466397,
+      "learning_rate": 1.3610929620829312e-06,
+      "loss": 0.5052,
+      "step": 9516
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.601857837454479,
+      "learning_rate": 1.3601982720260564e-06,
+      "loss": 0.4084,
+      "step": 9517
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.252005696144822,
+      "learning_rate": 1.3593038298235878e-06,
+      "loss": 0.4369,
+      "step": 9518
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.372223553947322,
+      "learning_rate": 1.3584096355364312e-06,
+      "loss": 0.4715,
+      "step": 9519
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.2731846561311855,
+      "learning_rate": 1.3575156892254776e-06,
+      "loss": 0.4648,
+      "step": 9520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6895126453399334,
+      "learning_rate": 1.3566219909515987e-06,
+      "loss": 0.6404,
+      "step": 9521
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.97730100949056,
+      "learning_rate": 1.3557285407756532e-06,
+      "loss": 0.5471,
+      "step": 9522
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 27.938140879823607,
+      "learning_rate": 1.3548353387584822e-06,
+      "loss": 0.3696,
+      "step": 9523
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.16062906142292,
+      "learning_rate": 1.353942384960908e-06,
+      "loss": 0.4234,
+      "step": 9524
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.879944566965991,
+      "learning_rate": 1.3530496794437365e-06,
+      "loss": 0.5009,
+      "step": 9525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.07186176039751,
+      "learning_rate": 1.3521572222677554e-06,
+      "loss": 0.3918,
+      "step": 9526
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.982203673805953,
+      "learning_rate": 1.3512650134937405e-06,
+      "loss": 0.4483,
+      "step": 9527
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.086479547776732,
+      "learning_rate": 1.350373053182444e-06,
+      "loss": 0.4229,
+      "step": 9528
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.061372978322919,
+      "learning_rate": 1.3494813413946084e-06,
+      "loss": 0.5094,
+      "step": 9529
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.631876862555988,
+      "learning_rate": 1.34858987819095e-06,
+      "loss": 0.5542,
+      "step": 9530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 22.684503620707375,
+      "learning_rate": 1.3476986636321764e-06,
+      "loss": 0.3983,
+      "step": 9531
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.172982529259228,
+      "learning_rate": 1.3468076977789769e-06,
+      "loss": 0.3987,
+      "step": 9532
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.996023574121942,
+      "learning_rate": 1.345916980692021e-06,
+      "loss": 0.4632,
+      "step": 9533
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.269682295794443,
+      "learning_rate": 1.3450265124319617e-06,
+      "loss": 0.477,
+      "step": 9534
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.790943676458244,
+      "learning_rate": 1.3441362930594355e-06,
+      "loss": 0.4774,
+      "step": 9535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.483943244263358,
+      "learning_rate": 1.343246322635065e-06,
+      "loss": 0.4731,
+      "step": 9536
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.568601696721457,
+      "learning_rate": 1.34235660121945e-06,
+      "loss": 0.5447,
+      "step": 9537
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.1227970045924645,
+      "learning_rate": 1.3414671288731817e-06,
+      "loss": 0.4931,
+      "step": 9538
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.394561216357166,
+      "learning_rate": 1.340577905656823e-06,
+      "loss": 0.4066,
+      "step": 9539
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.187866051925514,
+      "learning_rate": 1.3396889316309297e-06,
+      "loss": 0.3938,
+      "step": 9540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.167769238617918,
+      "learning_rate": 1.3388002068560351e-06,
+      "loss": 0.4873,
+      "step": 9541
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.561272167728547,
+      "learning_rate": 1.3379117313926604e-06,
+      "loss": 0.4788,
+      "step": 9542
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.644118826417635,
+      "learning_rate": 1.3370235053013048e-06,
+      "loss": 0.4528,
+      "step": 9543
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.304457904410296,
+      "learning_rate": 1.3361355286424515e-06,
+      "loss": 0.4441,
+      "step": 9544
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 60.76648464518586,
+      "learning_rate": 1.3352478014765702e-06,
+      "loss": 0.5195,
+      "step": 9545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.704126687500302,
+      "learning_rate": 1.334360323864109e-06,
+      "loss": 0.3251,
+      "step": 9546
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.593404102345097,
+      "learning_rate": 1.3334730958655052e-06,
+      "loss": 0.4411,
+      "step": 9547
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.265729688101146,
+      "learning_rate": 1.3325861175411686e-06,
+      "loss": 0.4288,
+      "step": 9548
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.525979641619574,
+      "learning_rate": 1.3316993889515039e-06,
+      "loss": 0.4819,
+      "step": 9549
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9399238703700883,
+      "learning_rate": 1.3308129101568907e-06,
+      "loss": 0.704,
+      "step": 9550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.314022791580014,
+      "learning_rate": 1.3299266812176958e-06,
+      "loss": 0.4884,
+      "step": 9551
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7674422231901923,
+      "learning_rate": 1.3290407021942676e-06,
+      "loss": 0.7092,
+      "step": 9552
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6306027743050153,
+      "learning_rate": 1.3281549731469345e-06,
+      "loss": 0.6973,
+      "step": 9553
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.553669062396029,
+      "learning_rate": 1.3272694941360143e-06,
+      "loss": 0.4569,
+      "step": 9554
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.332889218522519,
+      "learning_rate": 1.3263842652218012e-06,
+      "loss": 0.4528,
+      "step": 9555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.872695143169823,
+      "learning_rate": 1.3254992864645804e-06,
+      "loss": 0.5053,
+      "step": 9556
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.33963748068033,
+      "learning_rate": 1.3246145579246084e-06,
+      "loss": 0.3513,
+      "step": 9557
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.927149558885829,
+      "learning_rate": 1.3237300796621355e-06,
+      "loss": 0.4912,
+      "step": 9558
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.54450581594129,
+      "learning_rate": 1.3228458517373888e-06,
+      "loss": 0.5537,
+      "step": 9559
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.7598458040643985,
+      "learning_rate": 1.3219618742105821e-06,
+      "loss": 0.4762,
+      "step": 9560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.697965538715794,
+      "learning_rate": 1.3210781471419098e-06,
+      "loss": 0.3988,
+      "step": 9561
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.234146138937893,
+      "learning_rate": 1.3201946705915481e-06,
+      "loss": 0.5666,
+      "step": 9562
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.3289249377053425,
+      "learning_rate": 1.319311444619661e-06,
+      "loss": 0.5261,
+      "step": 9563
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.99052668306936,
+      "learning_rate": 1.3184284692863886e-06,
+      "loss": 0.518,
+      "step": 9564
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.851444489375293,
+      "learning_rate": 1.317545744651861e-06,
+      "loss": 0.3988,
+      "step": 9565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.262045487737215,
+      "learning_rate": 1.3166632707761863e-06,
+      "loss": 0.4342,
+      "step": 9566
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.3883696346876775,
+      "learning_rate": 1.3157810477194576e-06,
+      "loss": 0.434,
+      "step": 9567
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 17.561094393607437,
+      "learning_rate": 1.314899075541748e-06,
+      "loss": 0.5007,
+      "step": 9568
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.442406698832486,
+      "learning_rate": 1.3140173543031193e-06,
+      "loss": 0.4762,
+      "step": 9569
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.345228456793325,
+      "learning_rate": 1.3131358840636117e-06,
+      "loss": 0.4885,
+      "step": 9570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.718693617888686,
+      "learning_rate": 1.3122546648832474e-06,
+      "loss": 0.4505,
+      "step": 9571
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.946969430960271,
+      "learning_rate": 1.311373696822037e-06,
+      "loss": 0.4226,
+      "step": 9572
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.728855284452317,
+      "learning_rate": 1.310492979939967e-06,
+      "loss": 0.4264,
+      "step": 9573
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 25.203284564549794,
+      "learning_rate": 1.3096125142970134e-06,
+      "loss": 0.4187,
+      "step": 9574
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.7835075777415454,
+      "learning_rate": 1.308732299953131e-06,
+      "loss": 0.5994,
+      "step": 9575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.023510094336324,
+      "learning_rate": 1.3078523369682577e-06,
+      "loss": 0.4421,
+      "step": 9576
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.053091973283836,
+      "learning_rate": 1.3069726254023147e-06,
+      "loss": 0.4401,
+      "step": 9577
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.106227431801681,
+      "learning_rate": 1.3060931653152082e-06,
+      "loss": 0.4935,
+      "step": 9578
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.574549308194356,
+      "learning_rate": 1.3052139567668253e-06,
+      "loss": 0.3921,
+      "step": 9579
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.883156214461331,
+      "learning_rate": 1.3043349998170352e-06,
+      "loss": 0.3349,
+      "step": 9580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.777913316569882,
+      "learning_rate": 1.3034562945256902e-06,
+      "loss": 0.4721,
+      "step": 9581
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.808987134516254,
+      "learning_rate": 1.3025778409526274e-06,
+      "loss": 0.4872,
+      "step": 9582
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.42301631000535,
+      "learning_rate": 1.301699639157667e-06,
+      "loss": 0.3941,
+      "step": 9583
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.241380127853,
+      "learning_rate": 1.3008216892006093e-06,
+      "loss": 0.5257,
+      "step": 9584
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.552399793656974,
+      "learning_rate": 1.2999439911412388e-06,
+      "loss": 0.5147,
+      "step": 9585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.457182850347924,
+      "learning_rate": 1.2990665450393215e-06,
+      "loss": 0.5076,
+      "step": 9586
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.613793524112904,
+      "learning_rate": 1.29818935095461e-06,
+      "loss": 0.472,
+      "step": 9587
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 13.592663452846194,
+      "learning_rate": 1.297312408946837e-06,
+      "loss": 0.4668,
+      "step": 9588
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.2753446085578695,
+      "learning_rate": 1.296435719075717e-06,
+      "loss": 0.5197,
+      "step": 9589
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.906064323583943,
+      "learning_rate": 1.2955592814009483e-06,
+      "loss": 0.5048,
+      "step": 9590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.017159653907684,
+      "learning_rate": 1.2946830959822148e-06,
+      "loss": 0.4892,
+      "step": 9591
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.587717384568787,
+      "learning_rate": 1.2938071628791782e-06,
+      "loss": 0.4686,
+      "step": 9592
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.928863437752519,
+      "learning_rate": 1.2929314821514889e-06,
+      "loss": 0.4793,
+      "step": 9593
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.410912074175311,
+      "learning_rate": 1.2920560538587745e-06,
+      "loss": 0.4236,
+      "step": 9594
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.503195034936861,
+      "learning_rate": 1.2911808780606466e-06,
+      "loss": 0.4942,
+      "step": 9595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.340876922377755,
+      "learning_rate": 1.2903059548167047e-06,
+      "loss": 0.4894,
+      "step": 9596
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.4163952179269,
+      "learning_rate": 1.2894312841865248e-06,
+      "loss": 0.3779,
+      "step": 9597
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.852648833857511,
+      "learning_rate": 1.2885568662296682e-06,
+      "loss": 0.4094,
+      "step": 9598
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.390132560120112,
+      "learning_rate": 1.2876827010056775e-06,
+      "loss": 0.4003,
+      "step": 9599
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.084522932629677,
+      "learning_rate": 1.2868087885740827e-06,
+      "loss": 0.476,
+      "step": 9600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.6515329887024865,
+      "learning_rate": 1.28593512899439e-06,
+      "loss": 0.5571,
+      "step": 9601
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 19.047922594869302,
+      "learning_rate": 1.2850617223260948e-06,
+      "loss": 0.4162,
+      "step": 9602
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.471940509011786,
+      "learning_rate": 1.284188568628671e-06,
+      "loss": 0.5511,
+      "step": 9603
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.962165264065241,
+      "learning_rate": 1.283315667961575e-06,
+      "loss": 0.4072,
+      "step": 9604
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.1012182125303385,
+      "learning_rate": 1.2824430203842503e-06,
+      "loss": 0.3534,
+      "step": 9605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.325076572737491,
+      "learning_rate": 1.281570625956119e-06,
+      "loss": 0.5469,
+      "step": 9606
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.4321150550335044,
+      "learning_rate": 1.2806984847365871e-06,
+      "loss": 0.3956,
+      "step": 9607
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.269571559332485,
+      "learning_rate": 1.2798265967850421e-06,
+      "loss": 0.4288,
+      "step": 9608
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5984789794276544,
+      "learning_rate": 1.2789549621608588e-06,
+      "loss": 0.6437,
+      "step": 9609
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.670010529985582,
+      "learning_rate": 1.2780835809233888e-06,
+      "loss": 0.5635,
+      "step": 9610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.506291049292636,
+      "learning_rate": 1.2772124531319717e-06,
+      "loss": 0.3472,
+      "step": 9611
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.056712881082144,
+      "learning_rate": 1.2763415788459265e-06,
+      "loss": 0.3859,
+      "step": 9612
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.204652741052241,
+      "learning_rate": 1.2754709581245545e-06,
+      "loss": 0.5273,
+      "step": 9613
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.382615243221466,
+      "learning_rate": 1.2746005910271442e-06,
+      "loss": 0.5247,
+      "step": 9614
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.001503499392404,
+      "learning_rate": 1.2737304776129616e-06,
+      "loss": 0.4253,
+      "step": 9615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.21107599291536,
+      "learning_rate": 1.2728606179412578e-06,
+      "loss": 0.47,
+      "step": 9616
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.140377243960094,
+      "learning_rate": 1.2719910120712654e-06,
+      "loss": 0.6137,
+      "step": 9617
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.616322233370761,
+      "learning_rate": 1.271121660062203e-06,
+      "loss": 0.3885,
+      "step": 9618
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.42432150489394,
+      "learning_rate": 1.270252561973267e-06,
+      "loss": 0.5291,
+      "step": 9619
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.418878554160873,
+      "learning_rate": 1.2693837178636426e-06,
+      "loss": 0.468,
+      "step": 9620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1614596128171493,
+      "learning_rate": 1.2685151277924918e-06,
+      "loss": 0.7455,
+      "step": 9621
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.835069450820099,
+      "learning_rate": 1.2676467918189605e-06,
+      "loss": 0.4332,
+      "step": 9622
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.3904273170477275,
+      "learning_rate": 1.2667787100021823e-06,
+      "loss": 0.455,
+      "step": 9623
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.44106321152613,
+      "learning_rate": 1.2659108824012657e-06,
+      "loss": 0.4834,
+      "step": 9624
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.030336816003572,
+      "learning_rate": 1.2650433090753112e-06,
+      "loss": 0.4208,
+      "step": 9625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.9740733165989495,
+      "learning_rate": 1.2641759900833904e-06,
+      "loss": 0.5589,
+      "step": 9626
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.36020312570067,
+      "learning_rate": 1.2633089254845682e-06,
+      "loss": 0.4055,
+      "step": 9627
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.905862471105787,
+      "learning_rate": 1.2624421153378852e-06,
+      "loss": 0.4956,
+      "step": 9628
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.897409451627987,
+      "learning_rate": 1.26157555970237e-06,
+      "loss": 0.7705,
+      "step": 9629
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.996954332787288,
+      "learning_rate": 1.2607092586370302e-06,
+      "loss": 0.5924,
+      "step": 9630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.176892975700083,
+      "learning_rate": 1.2598432122008563e-06,
+      "loss": 0.4577,
+      "step": 9631
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.444322166985948,
+      "learning_rate": 1.2589774204528217e-06,
+      "loss": 0.4992,
+      "step": 9632
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.240849150157368,
+      "learning_rate": 1.2581118834518841e-06,
+      "loss": 0.4115,
+      "step": 9633
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.221201340584855,
+      "learning_rate": 1.257246601256985e-06,
+      "loss": 0.4721,
+      "step": 9634
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.655673132477501,
+      "learning_rate": 1.2563815739270413e-06,
+      "loss": 0.7132,
+      "step": 9635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.477920089490887,
+      "learning_rate": 1.2555168015209618e-06,
+      "loss": 0.4697,
+      "step": 9636
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7555446028987454,
+      "learning_rate": 1.2546522840976304e-06,
+      "loss": 0.7404,
+      "step": 9637
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.228252171056763,
+      "learning_rate": 1.25378802171592e-06,
+      "loss": 0.4463,
+      "step": 9638
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.508287185188864,
+      "learning_rate": 1.2529240144346816e-06,
+      "loss": 0.4662,
+      "step": 9639
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.239284798800447,
+      "learning_rate": 1.2520602623127503e-06,
+      "loss": 0.3328,
+      "step": 9640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 29.475402611261842,
+      "learning_rate": 1.2511967654089424e-06,
+      "loss": 0.423,
+      "step": 9641
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.785023518532459,
+      "learning_rate": 1.2503335237820597e-06,
+      "loss": 0.4993,
+      "step": 9642
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.605285362440908,
+      "learning_rate": 1.2494705374908878e-06,
+      "loss": 0.5201,
+      "step": 9643
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8241926751108206,
+      "learning_rate": 1.2486078065941864e-06,
+      "loss": 0.4731,
+      "step": 9644
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.789446063145882,
+      "learning_rate": 1.2477453311507083e-06,
+      "loss": 0.4426,
+      "step": 9645
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.597723786417638,
+      "learning_rate": 1.2468831112191815e-06,
+      "loss": 0.5077,
+      "step": 9646
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.249887816855991,
+      "learning_rate": 1.2460211468583216e-06,
+      "loss": 0.461,
+      "step": 9647
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7795639168373412,
+      "learning_rate": 1.245159438126824e-06,
+      "loss": 0.7092,
+      "step": 9648
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.847521203636981,
+      "learning_rate": 1.2442979850833663e-06,
+      "loss": 0.3813,
+      "step": 9649
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.555591132965734,
+      "learning_rate": 1.2434367877866094e-06,
+      "loss": 0.3665,
+      "step": 9650
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.1257315753587545,
+      "learning_rate": 1.2425758462951986e-06,
+      "loss": 0.4747,
+      "step": 9651
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0410003469196396,
+      "learning_rate": 1.2417151606677597e-06,
+      "loss": 0.7547,
+      "step": 9652
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.723787648105459,
+      "learning_rate": 1.2408547309628998e-06,
+      "loss": 0.4458,
+      "step": 9653
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.12957247536028,
+      "learning_rate": 1.239994557239213e-06,
+      "loss": 0.5285,
+      "step": 9654
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.957211050361633,
+      "learning_rate": 1.239134639555271e-06,
+      "loss": 0.4503,
+      "step": 9655
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.700168645404824,
+      "learning_rate": 1.238274977969633e-06,
+      "loss": 0.4483,
+      "step": 9656
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.837361273142933,
+      "learning_rate": 1.2374155725408366e-06,
+      "loss": 0.7473,
+      "step": 9657
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.16243127067313,
+      "learning_rate": 1.2365564233274036e-06,
+      "loss": 0.4755,
+      "step": 9658
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.595502350624964,
+      "learning_rate": 1.2356975303878371e-06,
+      "loss": 0.4405,
+      "step": 9659
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.483019598289132,
+      "learning_rate": 1.234838893780626e-06,
+      "loss": 0.4802,
+      "step": 9660
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.586016536007802,
+      "learning_rate": 1.2339805135642373e-06,
+      "loss": 0.4829,
+      "step": 9661
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.598419918161517,
+      "learning_rate": 1.233122389797126e-06,
+      "loss": 0.4648,
+      "step": 9662
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.446047639838879,
+      "learning_rate": 1.232264522537725e-06,
+      "loss": 0.3418,
+      "step": 9663
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.073903765228717,
+      "learning_rate": 1.2314069118444499e-06,
+      "loss": 0.6417,
+      "step": 9664
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.527676764396996,
+      "learning_rate": 1.230549557775702e-06,
+      "loss": 0.4671,
+      "step": 9665
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.880562579615705,
+      "learning_rate": 1.2296924603898637e-06,
+      "loss": 0.4468,
+      "step": 9666
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.4362268561466625,
+      "learning_rate": 1.2288356197452983e-06,
+      "loss": 0.4396,
+      "step": 9667
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.241392436745322,
+      "learning_rate": 1.2279790359003514e-06,
+      "loss": 0.4664,
+      "step": 9668
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.748710798607389,
+      "learning_rate": 1.227122708913356e-06,
+      "loss": 0.469,
+      "step": 9669
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.072407373188526,
+      "learning_rate": 1.2262666388426209e-06,
+      "loss": 0.4709,
+      "step": 9670
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7626829507536557,
+      "learning_rate": 1.2254108257464454e-06,
+      "loss": 0.7496,
+      "step": 9671
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.409254734263653,
+      "learning_rate": 1.2245552696831002e-06,
+      "loss": 0.4022,
+      "step": 9672
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7748280691847413,
+      "learning_rate": 1.2236999707108488e-06,
+      "loss": 0.7321,
+      "step": 9673
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.411671792939352,
+      "learning_rate": 1.2228449288879341e-06,
+      "loss": 0.501,
+      "step": 9674
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.185069631629457,
+      "learning_rate": 1.2219901442725785e-06,
+      "loss": 0.4821,
+      "step": 9675
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.160873826069727,
+      "learning_rate": 1.2211356169229905e-06,
+      "loss": 0.4186,
+      "step": 9676
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.226213607089933,
+      "learning_rate": 1.2202813468973567e-06,
+      "loss": 0.4102,
+      "step": 9677
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.453197898917776,
+      "learning_rate": 1.2194273342538532e-06,
+      "loss": 0.4844,
+      "step": 9678
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.614931280592476,
+      "learning_rate": 1.2185735790506314e-06,
+      "loss": 0.491,
+      "step": 9679
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.872886841918123,
+      "learning_rate": 1.2177200813458318e-06,
+      "loss": 0.4179,
+      "step": 9680
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5570343474604798,
+      "learning_rate": 1.2168668411975688e-06,
+      "loss": 0.6514,
+      "step": 9681
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.440873559570951,
+      "learning_rate": 1.216013858663947e-06,
+      "loss": 0.5287,
+      "step": 9682
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.627892418475344,
+      "learning_rate": 1.2151611338030522e-06,
+      "loss": 0.6816,
+      "step": 9683
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6401568502970623,
+      "learning_rate": 1.2143086666729497e-06,
+      "loss": 0.6671,
+      "step": 9684
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8169324881483346,
+      "learning_rate": 1.2134564573316888e-06,
+      "loss": 0.7084,
+      "step": 9685
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.76544582703945,
+      "learning_rate": 1.2126045058372992e-06,
+      "loss": 0.4615,
+      "step": 9686
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.719141448050818,
+      "learning_rate": 1.211752812247799e-06,
+      "loss": 0.4706,
+      "step": 9687
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.010815694840511,
+      "learning_rate": 1.2109013766211808e-06,
+      "loss": 0.4473,
+      "step": 9688
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.436702392247351,
+      "learning_rate": 1.2100501990154284e-06,
+      "loss": 0.5402,
+      "step": 9689
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 14.539742380595632,
+      "learning_rate": 1.2091992794884977e-06,
+      "loss": 0.458,
+      "step": 9690
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.935751867073743,
+      "learning_rate": 1.2083486180983366e-06,
+      "loss": 0.5323,
+      "step": 9691
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 16.127943209389464,
+      "learning_rate": 1.2074982149028685e-06,
+      "loss": 0.3922,
+      "step": 9692
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.775749861901072,
+      "learning_rate": 1.2066480699600053e-06,
+      "loss": 0.4106,
+      "step": 9693
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.958532207698214,
+      "learning_rate": 1.2057981833276362e-06,
+      "loss": 0.4367,
+      "step": 9694
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.99540452999758,
+      "learning_rate": 1.204948555063633e-06,
+      "loss": 0.4511,
+      "step": 9695
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2874431369667,
+      "learning_rate": 1.2040991852258554e-06,
+      "loss": 0.6881,
+      "step": 9696
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.093338095992158,
+      "learning_rate": 1.2032500738721386e-06,
+      "loss": 0.4743,
+      "step": 9697
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8190694782381898,
+      "learning_rate": 1.2024012210603058e-06,
+      "loss": 0.693,
+      "step": 9698
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.82643748436257,
+      "learning_rate": 1.2015526268481592e-06,
+      "loss": 0.3931,
+      "step": 9699
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.047513150746792,
+      "learning_rate": 1.200704291293484e-06,
+      "loss": 0.4535,
+      "step": 9700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.405188969716176,
+      "learning_rate": 1.1998562144540466e-06,
+      "loss": 0.4649,
+      "step": 9701
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.099835887355772,
+      "learning_rate": 1.1990083963875998e-06,
+      "loss": 0.464,
+      "step": 9702
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.589706894073316,
+      "learning_rate": 1.1981608371518761e-06,
+      "loss": 0.452,
+      "step": 9703
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.665160637427526,
+      "learning_rate": 1.197313536804588e-06,
+      "loss": 0.5198,
+      "step": 9704
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 11.480306022567511,
+      "learning_rate": 1.1964664954034355e-06,
+      "loss": 0.4185,
+      "step": 9705
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5400257350146525,
+      "learning_rate": 1.1956197130060964e-06,
+      "loss": 0.6381,
+      "step": 9706
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.444299232692264,
+      "learning_rate": 1.1947731896702358e-06,
+      "loss": 0.4996,
+      "step": 9707
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.237254625211817,
+      "learning_rate": 1.1939269254534958e-06,
+      "loss": 0.4316,
+      "step": 9708
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.336784528867742,
+      "learning_rate": 1.193080920413504e-06,
+      "loss": 0.5278,
+      "step": 9709
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.110535575202922,
+      "learning_rate": 1.192235174607868e-06,
+      "loss": 0.4564,
+      "step": 9710
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.676816448729076,
+      "learning_rate": 1.1913896880941822e-06,
+      "loss": 0.5738,
+      "step": 9711
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 15.191224133031023,
+      "learning_rate": 1.1905444609300193e-06,
+      "loss": 0.4692,
+      "step": 9712
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.078603226779375,
+      "learning_rate": 1.1896994931729334e-06,
+      "loss": 0.5632,
+      "step": 9713
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.2272879153082465,
+      "learning_rate": 1.1888547848804667e-06,
+      "loss": 0.46,
+      "step": 9714
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6181294793022567,
+      "learning_rate": 1.1880103361101364e-06,
+      "loss": 0.4303,
+      "step": 9715
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 23.408403332707344,
+      "learning_rate": 1.1871661469194495e-06,
+      "loss": 0.5513,
+      "step": 9716
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.81432819798275,
+      "learning_rate": 1.1863222173658894e-06,
+      "loss": 0.4753,
+      "step": 9717
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.264249328254296,
+      "learning_rate": 1.1854785475069248e-06,
+      "loss": 0.4459,
+      "step": 9718
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.709765945349405,
+      "learning_rate": 1.1846351374000032e-06,
+      "loss": 0.4094,
+      "step": 9719
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.722725907031328,
+      "learning_rate": 1.1837919871025611e-06,
+      "loss": 0.447,
+      "step": 9720
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.097589532871186,
+      "learning_rate": 1.182949096672012e-06,
+      "loss": 0.4504,
+      "step": 9721
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.770892830025188,
+      "learning_rate": 1.1821064661657517e-06,
+      "loss": 0.5994,
+      "step": 9722
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.389866638336848,
+      "learning_rate": 1.1812640956411592e-06,
+      "loss": 0.3811,
+      "step": 9723
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.403997546491213,
+      "learning_rate": 1.1804219851555976e-06,
+      "loss": 0.4398,
+      "step": 9724
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.473896716087147,
+      "learning_rate": 1.1795801347664122e-06,
+      "loss": 0.5266,
+      "step": 9725
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.707631660972172,
+      "learning_rate": 1.1787385445309273e-06,
+      "loss": 0.431,
+      "step": 9726
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.7983381453151885,
+      "learning_rate": 1.177897214506452e-06,
+      "loss": 0.3847,
+      "step": 9727
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.979538937500849,
+      "learning_rate": 1.1770561447502764e-06,
+      "loss": 0.4471,
+      "step": 9728
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.454026320972909,
+      "learning_rate": 1.1762153353196754e-06,
+      "loss": 0.4337,
+      "step": 9729
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.024667806196671,
+      "learning_rate": 1.1753747862719033e-06,
+      "loss": 0.5667,
+      "step": 9730
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.707828201649887,
+      "learning_rate": 1.1745344976641975e-06,
+      "loss": 0.5319,
+      "step": 9731
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.238440365042922,
+      "learning_rate": 1.1736944695537777e-06,
+      "loss": 0.4739,
+      "step": 9732
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.979513300023861,
+      "learning_rate": 1.1728547019978458e-06,
+      "loss": 0.4423,
+      "step": 9733
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.86603193647017,
+      "learning_rate": 1.1720151950535891e-06,
+      "loss": 0.5306,
+      "step": 9734
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.833628109486405,
+      "learning_rate": 1.171175948778172e-06,
+      "loss": 0.4427,
+      "step": 9735
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7799040564995714,
+      "learning_rate": 1.1703369632287442e-06,
+      "loss": 0.7427,
+      "step": 9736
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.849627574782396,
+      "learning_rate": 1.1694982384624348e-06,
+      "loss": 0.4556,
+      "step": 9737
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.086887980964914,
+      "learning_rate": 1.1686597745363609e-06,
+      "loss": 0.4375,
+      "step": 9738
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 46.65856993367782,
+      "learning_rate": 1.1678215715076157e-06,
+      "loss": 0.4976,
+      "step": 9739
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.026001920970876,
+      "learning_rate": 1.1669836294332775e-06,
+      "loss": 0.4462,
+      "step": 9740
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.473246669791515,
+      "learning_rate": 1.1661459483704058e-06,
+      "loss": 0.5003,
+      "step": 9741
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.344719769528781,
+      "learning_rate": 1.165308528376045e-06,
+      "loss": 0.4213,
+      "step": 9742
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.24431098889315,
+      "learning_rate": 1.164471369507217e-06,
+      "loss": 0.3488,
+      "step": 9743
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7841041405505522,
+      "learning_rate": 1.1636344718209314e-06,
+      "loss": 0.7594,
+      "step": 9744
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.160240243138242,
+      "learning_rate": 1.1627978353741764e-06,
+      "loss": 0.5385,
+      "step": 9745
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.10589608315741,
+      "learning_rate": 1.161961460223921e-06,
+      "loss": 0.5057,
+      "step": 9746
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.502209372581922,
+      "learning_rate": 1.1611253464271217e-06,
+      "loss": 0.4885,
+      "step": 9747
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.175664427264435,
+      "learning_rate": 1.1602894940407134e-06,
+      "loss": 0.4425,
+      "step": 9748
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.652363315182823,
+      "learning_rate": 1.1594539031216128e-06,
+      "loss": 0.5567,
+      "step": 9749
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.9029393127088925,
+      "learning_rate": 1.1586185737267197e-06,
+      "loss": 0.3871,
+      "step": 9750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8806915685037844,
+      "learning_rate": 1.1577835059129183e-06,
+      "loss": 0.7095,
+      "step": 9751
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.078923383999804,
+      "learning_rate": 1.1569486997370705e-06,
+      "loss": 0.4376,
+      "step": 9752
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.251857656587057,
+      "learning_rate": 1.1561141552560261e-06,
+      "loss": 0.4715,
+      "step": 9753
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.791326240184102,
+      "learning_rate": 1.1552798725266123e-06,
+      "loss": 0.4681,
+      "step": 9754
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.3715057629610525,
+      "learning_rate": 1.1544458516056378e-06,
+      "loss": 0.4867,
+      "step": 9755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.273650639174018,
+      "learning_rate": 1.1536120925499e-06,
+      "loss": 0.4363,
+      "step": 9756
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.751308128436501,
+      "learning_rate": 1.1527785954161701e-06,
+      "loss": 0.4484,
+      "step": 9757
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.913782157808516,
+      "learning_rate": 1.1519453602612107e-06,
+      "loss": 0.5287,
+      "step": 9758
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.938515176894494,
+      "learning_rate": 1.151112387141755e-06,
+      "loss": 0.4615,
+      "step": 9759
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.301952125846965,
+      "learning_rate": 1.1502796761145303e-06,
+      "loss": 0.4257,
+      "step": 9760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.472884425098053,
+      "learning_rate": 1.149447227236236e-06,
+      "loss": 0.7035,
+      "step": 9761
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.205141836170976,
+      "learning_rate": 1.1486150405635622e-06,
+      "loss": 0.488,
+      "step": 9762
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.270380711518324,
+      "learning_rate": 1.1477831161531755e-06,
+      "loss": 0.4302,
+      "step": 9763
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.532535527262439,
+      "learning_rate": 1.1469514540617248e-06,
+      "loss": 0.5045,
+      "step": 9764
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.14516307461931,
+      "learning_rate": 1.146120054345845e-06,
+      "loss": 0.4251,
+      "step": 9765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.07600477085039,
+      "learning_rate": 1.145288917062149e-06,
+      "loss": 0.3966,
+      "step": 9766
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.238823811842288,
+      "learning_rate": 1.1444580422672364e-06,
+      "loss": 0.4922,
+      "step": 9767
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.5401992607785955,
+      "learning_rate": 1.143627430017682e-06,
+      "loss": 0.4433,
+      "step": 9768
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.877020451758641,
+      "learning_rate": 1.1427970803700495e-06,
+      "loss": 0.4844,
+      "step": 9769
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.890596290142113,
+      "learning_rate": 1.1419669933808803e-06,
+      "loss": 0.5372,
+      "step": 9770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.532994447137327,
+      "learning_rate": 1.1411371691067018e-06,
+      "loss": 0.5322,
+      "step": 9771
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.658134748036034,
+      "learning_rate": 1.140307607604021e-06,
+      "loss": 0.4511,
+      "step": 9772
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.435892656097289,
+      "learning_rate": 1.1394783089293242e-06,
+      "loss": 0.475,
+      "step": 9773
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.740325567627726,
+      "learning_rate": 1.138649273139088e-06,
+      "loss": 0.4432,
+      "step": 9774
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.873242225742145,
+      "learning_rate": 1.1378205002897613e-06,
+      "loss": 0.5271,
+      "step": 9775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.018615904971757,
+      "learning_rate": 1.1369919904377852e-06,
+      "loss": 0.5204,
+      "step": 9776
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.683141194297925,
+      "learning_rate": 1.1361637436395718e-06,
+      "loss": 0.7127,
+      "step": 9777
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.659991697477546,
+      "learning_rate": 1.1353357599515246e-06,
+      "loss": 0.3802,
+      "step": 9778
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.36684449690318,
+      "learning_rate": 1.1345080394300233e-06,
+      "loss": 0.5149,
+      "step": 9779
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6513491530394457,
+      "learning_rate": 1.133680582131435e-06,
+      "loss": 0.6453,
+      "step": 9780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.740903919169664,
+      "learning_rate": 1.1328533881121045e-06,
+      "loss": 0.59,
+      "step": 9781
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.828540268659538,
+      "learning_rate": 1.1320264574283597e-06,
+      "loss": 0.5114,
+      "step": 9782
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3101545052393577,
+      "learning_rate": 1.1311997901365097e-06,
+      "loss": 0.4178,
+      "step": 9783
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.318194043884338,
+      "learning_rate": 1.130373386292848e-06,
+      "loss": 0.4788,
+      "step": 9784
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.478596645841774,
+      "learning_rate": 1.129547245953652e-06,
+      "loss": 0.4352,
+      "step": 9785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.438510768436602,
+      "learning_rate": 1.1287213691751736e-06,
+      "loss": 0.4811,
+      "step": 9786
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.818350892506356,
+      "learning_rate": 1.1278957560136538e-06,
+      "loss": 0.4323,
+      "step": 9787
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.245425833096599,
+      "learning_rate": 1.1270704065253118e-06,
+      "loss": 0.5556,
+      "step": 9788
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.047210650397568,
+      "learning_rate": 1.126245320766352e-06,
+      "loss": 0.4895,
+      "step": 9789
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.976119588283257,
+      "learning_rate": 1.1254204987929585e-06,
+      "loss": 0.5694,
+      "step": 9790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.421808212902823,
+      "learning_rate": 1.1245959406612978e-06,
+      "loss": 0.4754,
+      "step": 9791
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.953700309066504,
+      "learning_rate": 1.1237716464275172e-06,
+      "loss": 0.4503,
+      "step": 9792
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.520353809776631,
+      "learning_rate": 1.1229476161477488e-06,
+      "loss": 0.511,
+      "step": 9793
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.124066695279729,
+      "learning_rate": 1.1221238498781085e-06,
+      "loss": 0.3469,
+      "step": 9794
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.997969154391987,
+      "learning_rate": 1.1213003476746853e-06,
+      "loss": 0.4313,
+      "step": 9795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.721221122099877,
+      "learning_rate": 1.1204771095935602e-06,
+      "loss": 0.411,
+      "step": 9796
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.934358594710091,
+      "learning_rate": 1.1196541356907893e-06,
+      "loss": 0.4057,
+      "step": 9797
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.692141690226565,
+      "learning_rate": 1.1188314260224164e-06,
+      "loss": 0.533,
+      "step": 9798
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.964891200695824,
+      "learning_rate": 1.1180089806444632e-06,
+      "loss": 0.3778,
+      "step": 9799
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.990720716336439,
+      "learning_rate": 1.1171867996129344e-06,
+      "loss": 0.4237,
+      "step": 9800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.252989621730221,
+      "learning_rate": 1.1163648829838153e-06,
+      "loss": 0.5638,
+      "step": 9801
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.737390775846092,
+      "learning_rate": 1.1155432308130776e-06,
+      "loss": 0.3414,
+      "step": 9802
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.040029231626813,
+      "learning_rate": 1.1147218431566698e-06,
+      "loss": 0.7766,
+      "step": 9803
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.753593947413735,
+      "learning_rate": 1.113900720070527e-06,
+      "loss": 0.4837,
+      "step": 9804
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.115877537085814,
+      "learning_rate": 1.113079861610563e-06,
+      "loss": 0.472,
+      "step": 9805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.455330708072131,
+      "learning_rate": 1.1122592678326733e-06,
+      "loss": 0.4089,
+      "step": 9806
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 16.365290505729792,
+      "learning_rate": 1.1114389387927393e-06,
+      "loss": 0.4542,
+      "step": 9807
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.8092252356581575,
+      "learning_rate": 1.1106188745466207e-06,
+      "loss": 0.5048,
+      "step": 9808
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.874019145126758,
+      "learning_rate": 1.1097990751501598e-06,
+      "loss": 0.5182,
+      "step": 9809
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.58079203174489,
+      "learning_rate": 1.1089795406591803e-06,
+      "loss": 0.6728,
+      "step": 9810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.7799911472765135,
+      "learning_rate": 1.1081602711294915e-06,
+      "loss": 0.3893,
+      "step": 9811
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.210913627492282,
+      "learning_rate": 1.1073412666168786e-06,
+      "loss": 0.5949,
+      "step": 9812
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.277070022638215,
+      "learning_rate": 1.1065225271771163e-06,
+      "loss": 0.3623,
+      "step": 9813
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 41.812530879821836,
+      "learning_rate": 1.1057040528659552e-06,
+      "loss": 0.495,
+      "step": 9814
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.408782474893611,
+      "learning_rate": 1.1048858437391275e-06,
+      "loss": 0.3409,
+      "step": 9815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9839968085235653,
+      "learning_rate": 1.1040678998523534e-06,
+      "loss": 0.444,
+      "step": 9816
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.130515050442439,
+      "learning_rate": 1.10325022126133e-06,
+      "loss": 0.4103,
+      "step": 9817
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.392066474553413,
+      "learning_rate": 1.1024328080217372e-06,
+      "loss": 0.3913,
+      "step": 9818
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.31568562691919,
+      "learning_rate": 1.1016156601892352e-06,
+      "loss": 0.5168,
+      "step": 9819
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7319261642574495,
+      "learning_rate": 1.1007987778194722e-06,
+      "loss": 0.4912,
+      "step": 9820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.472923761814414,
+      "learning_rate": 1.0999821609680711e-06,
+      "loss": 0.5608,
+      "step": 9821
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.887982972353759,
+      "learning_rate": 1.0991658096906433e-06,
+      "loss": 0.7482,
+      "step": 9822
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.913439268442911,
+      "learning_rate": 1.0983497240427738e-06,
+      "loss": 0.4189,
+      "step": 9823
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.028555282274816,
+      "learning_rate": 1.0975339040800377e-06,
+      "loss": 0.498,
+      "step": 9824
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.99897461362907,
+      "learning_rate": 1.0967183498579887e-06,
+      "loss": 0.4841,
+      "step": 9825
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.24026673885149,
+      "learning_rate": 1.0959030614321625e-06,
+      "loss": 0.5324,
+      "step": 9826
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.767690069186493,
+      "learning_rate": 1.0950880388580754e-06,
+      "loss": 0.3627,
+      "step": 9827
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.325134215733407,
+      "learning_rate": 1.0942732821912267e-06,
+      "loss": 0.4096,
+      "step": 9828
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.640164668409184,
+      "learning_rate": 1.0934587914870998e-06,
+      "loss": 0.4675,
+      "step": 9829
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.011065546211295,
+      "learning_rate": 1.092644566801155e-06,
+      "loss": 0.5181,
+      "step": 9830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.73687418241901,
+      "learning_rate": 1.0918306081888413e-06,
+      "loss": 0.4594,
+      "step": 9831
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.206655744147253,
+      "learning_rate": 1.091016915705581e-06,
+      "loss": 0.492,
+      "step": 9832
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.365678204385204,
+      "learning_rate": 1.0902034894067859e-06,
+      "loss": 0.5208,
+      "step": 9833
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.9865238012329565,
+      "learning_rate": 1.0893903293478469e-06,
+      "loss": 0.4735,
+      "step": 9834
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.347838010054963,
+      "learning_rate": 1.0885774355841362e-06,
+      "loss": 0.5466,
+      "step": 9835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.098921886212574,
+      "learning_rate": 1.087764808171008e-06,
+      "loss": 0.4649,
+      "step": 9836
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 24.481895577694466,
+      "learning_rate": 1.0869524471637966e-06,
+      "loss": 0.4222,
+      "step": 9837
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.632553074566873,
+      "learning_rate": 1.0861403526178238e-06,
+      "loss": 0.4449,
+      "step": 9838
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 14.452526793102786,
+      "learning_rate": 1.0853285245883871e-06,
+      "loss": 0.414,
+      "step": 9839
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.342668543176886,
+      "learning_rate": 1.0845169631307706e-06,
+      "loss": 0.4644,
+      "step": 9840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.192918695193241,
+      "learning_rate": 1.0837056683002362e-06,
+      "loss": 0.5675,
+      "step": 9841
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.920070708786701,
+      "learning_rate": 1.0828946401520306e-06,
+      "loss": 0.408,
+      "step": 9842
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.117905114150529,
+      "learning_rate": 1.082083878741379e-06,
+      "loss": 0.4478,
+      "step": 9843
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.370661033903687,
+      "learning_rate": 1.081273384123494e-06,
+      "loss": 0.5302,
+      "step": 9844
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.124071994891702,
+      "learning_rate": 1.080463156353565e-06,
+      "loss": 0.422,
+      "step": 9845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.688367820285766,
+      "learning_rate": 1.0796531954867633e-06,
+      "loss": 0.5345,
+      "step": 9846
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.668394981095045,
+      "learning_rate": 1.0788435015782472e-06,
+      "loss": 0.4886,
+      "step": 9847
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.852035913534146,
+      "learning_rate": 1.0780340746831503e-06,
+      "loss": 0.5906,
+      "step": 9848
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.270389526633087,
+      "learning_rate": 1.0772249148565928e-06,
+      "loss": 0.4739,
+      "step": 9849
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.33364387190669,
+      "learning_rate": 1.0764160221536745e-06,
+      "loss": 0.4655,
+      "step": 9850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.193371405661308,
+      "learning_rate": 1.0756073966294766e-06,
+      "loss": 0.3959,
+      "step": 9851
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8746243474594793,
+      "learning_rate": 1.0747990383390628e-06,
+      "loss": 0.4892,
+      "step": 9852
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.193284978166919,
+      "learning_rate": 1.0739909473374805e-06,
+      "loss": 0.4609,
+      "step": 9853
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.138297483224627,
+      "learning_rate": 1.0731831236797563e-06,
+      "loss": 0.4827,
+      "step": 9854
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.0281769427239835,
+      "learning_rate": 1.0723755674208975e-06,
+      "loss": 0.396,
+      "step": 9855
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.025519275590605,
+      "learning_rate": 1.0715682786158977e-06,
+      "loss": 0.468,
+      "step": 9856
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5582614528197305,
+      "learning_rate": 1.070761257319728e-06,
+      "loss": 0.7286,
+      "step": 9857
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.2260324622709025,
+      "learning_rate": 1.0699545035873449e-06,
+      "loss": 0.387,
+      "step": 9858
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.5515699805955645,
+      "learning_rate": 1.0691480174736836e-06,
+      "loss": 0.4252,
+      "step": 9859
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.632793992305576,
+      "learning_rate": 1.068341799033662e-06,
+      "loss": 0.3731,
+      "step": 9860
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.48749906957977,
+      "learning_rate": 1.0675358483221787e-06,
+      "loss": 0.5158,
+      "step": 9861
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.483014107402699,
+      "learning_rate": 1.0667301653941186e-06,
+      "loss": 0.6094,
+      "step": 9862
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.143633131732924,
+      "learning_rate": 1.0659247503043434e-06,
+      "loss": 0.4478,
+      "step": 9863
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.856992997910271,
+      "learning_rate": 1.065119603107697e-06,
+      "loss": 0.4227,
+      "step": 9864
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 17.57383809747451,
+      "learning_rate": 1.064314723859009e-06,
+      "loss": 0.5177,
+      "step": 9865
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 25.895391831258323,
+      "learning_rate": 1.063510112613086e-06,
+      "loss": 0.407,
+      "step": 9866
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.156570263372904,
+      "learning_rate": 1.06270576942472e-06,
+      "loss": 0.5486,
+      "step": 9867
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.205033996972472,
+      "learning_rate": 1.0619016943486825e-06,
+      "loss": 0.4323,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.71393301312438,
+      "learning_rate": 1.0610978874397282e-06,
+      "loss": 0.4081,
+      "step": 9869
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7065975549752976,
+      "learning_rate": 1.0602943487525901e-06,
+      "loss": 0.4343,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.56707048609558,
+      "learning_rate": 1.0594910783419894e-06,
+      "loss": 0.6181,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.180576062914177,
+      "learning_rate": 1.0586880762626229e-06,
+      "loss": 0.4429,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.838991376715647,
+      "learning_rate": 1.0578853425691727e-06,
+      "loss": 0.5457,
+      "step": 9873
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.187885977341015,
+      "learning_rate": 1.0570828773162989e-06,
+      "loss": 0.476,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.011000006182703,
+      "learning_rate": 1.0562806805586478e-06,
+      "loss": 0.4686,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.218095726027101,
+      "learning_rate": 1.0554787523508469e-06,
+      "loss": 0.5085,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.2954465610263055,
+      "learning_rate": 1.0546770927475025e-06,
+      "loss": 0.4735,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.095158822797178,
+      "learning_rate": 1.0538757018032037e-06,
+      "loss": 0.5365,
+      "step": 9878
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9166558859486114,
+      "learning_rate": 1.0530745795725211e-06,
+      "loss": 0.3594,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.230499152588206,
+      "learning_rate": 1.0522737261100097e-06,
+      "loss": 0.5035,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.448397519149125,
+      "learning_rate": 1.0514731414702023e-06,
+      "loss": 0.4667,
+      "step": 9881
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.86040821755989,
+      "learning_rate": 1.050672825707616e-06,
+      "loss": 0.4253,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.266904058175232,
+      "learning_rate": 1.0498727788767477e-06,
+      "loss": 0.4369,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 19.81617907532653,
+      "learning_rate": 1.0490730010320772e-06,
+      "loss": 0.3056,
+      "step": 9884
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.035557561186081,
+      "learning_rate": 1.048273492228068e-06,
+      "loss": 0.4225,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.994936042413801,
+      "learning_rate": 1.0474742525191616e-06,
+      "loss": 0.4876,
+      "step": 9886
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.819968177719112,
+      "learning_rate": 1.0466752819597824e-06,
+      "loss": 0.4746,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5282286354238606,
+      "learning_rate": 1.0458765806043364e-06,
+      "loss": 0.4448,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.254314048302119,
+      "learning_rate": 1.0450781485072132e-06,
+      "loss": 0.342,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.923048666388706,
+      "learning_rate": 1.0442799857227815e-06,
+      "loss": 0.4143,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.567877409278247,
+      "learning_rate": 1.043482092305393e-06,
+      "loss": 0.6313,
+      "step": 9891
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.458998529718431,
+      "learning_rate": 1.042684468309379e-06,
+      "loss": 0.4341,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.1284346872422075,
+      "learning_rate": 1.0418871137890568e-06,
+      "loss": 0.4949,
+      "step": 9893
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.080841856755196,
+      "learning_rate": 1.0410900287987202e-06,
+      "loss": 0.4412,
+      "step": 9894
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.453566658458534,
+      "learning_rate": 1.0402932133926496e-06,
+      "loss": 0.4266,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.00906220926383,
+      "learning_rate": 1.039496667625104e-06,
+      "loss": 0.5154,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.775331430924791,
+      "learning_rate": 1.0387003915503223e-06,
+      "loss": 0.449,
+      "step": 9897
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.699425512843945,
+      "learning_rate": 1.0379043852225312e-06,
+      "loss": 0.3033,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.463142346375131,
+      "learning_rate": 1.0371086486959314e-06,
+      "loss": 0.4815,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.8690261571635025,
+      "learning_rate": 1.0363131820247136e-06,
+      "loss": 0.4361,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.681187315148487,
+      "learning_rate": 1.0355179852630404e-06,
+      "loss": 0.4719,
+      "step": 9901
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.179069202836915,
+      "learning_rate": 1.0347230584650647e-06,
+      "loss": 0.3907,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.2556166035802,
+      "learning_rate": 1.0339284016849148e-06,
+      "loss": 0.5014,
+      "step": 9903
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.659848862869948,
+      "learning_rate": 1.0331340149767065e-06,
+      "loss": 0.4275,
+      "step": 9904
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.953832014140802,
+      "learning_rate": 1.0323398983945326e-06,
+      "loss": 0.4701,
+      "step": 9905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.0408074933805205,
+      "learning_rate": 1.0315460519924675e-06,
+      "loss": 0.4802,
+      "step": 9906
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.618361912934795,
+      "learning_rate": 1.0307524758245713e-06,
+      "loss": 0.407,
+      "step": 9907
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 21.138921538164794,
+      "learning_rate": 1.0299591699448802e-06,
+      "loss": 0.4655,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.028474309546885,
+      "learning_rate": 1.0291661344074188e-06,
+      "loss": 0.4602,
+      "step": 9909
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.757116284524929,
+      "learning_rate": 1.0283733692661846e-06,
+      "loss": 0.507,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.99388046072818,
+      "learning_rate": 1.0275808745751642e-06,
+      "loss": 0.4633,
+      "step": 9911
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.348115270157726,
+      "learning_rate": 1.0267886503883212e-06,
+      "loss": 0.4579,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.641507891973419,
+      "learning_rate": 1.0259966967596053e-06,
+      "loss": 0.5394,
+      "step": 9913
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.558884836802893,
+      "learning_rate": 1.0252050137429436e-06,
+      "loss": 0.4849,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.1585977902568185,
+      "learning_rate": 1.0244136013922446e-06,
+      "loss": 0.3898,
+      "step": 9915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.0964829509124225,
+      "learning_rate": 1.0236224597614024e-06,
+      "loss": 0.4507,
+      "step": 9916
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6671914742710094,
+      "learning_rate": 1.0228315889042883e-06,
+      "loss": 0.688,
+      "step": 9917
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.450091739647564,
+      "learning_rate": 1.0220409888747607e-06,
+      "loss": 0.5532,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7782849735746626,
+      "learning_rate": 1.0212506597266508e-06,
+      "loss": 0.6984,
+      "step": 9919
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.77403999041046,
+      "learning_rate": 1.0204606015137809e-06,
+      "loss": 0.4307,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.174753953964963,
+      "learning_rate": 1.0196708142899463e-06,
+      "loss": 0.4157,
+      "step": 9921
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.188772905861553,
+      "learning_rate": 1.0188812981089325e-06,
+      "loss": 0.4282,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.160967121015079,
+      "learning_rate": 1.0180920530244997e-06,
+      "loss": 0.3762,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.314072663874268,
+      "learning_rate": 1.017303079090391e-06,
+      "loss": 0.486,
+      "step": 9924
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 30.43450404526416,
+      "learning_rate": 1.0165143763603346e-06,
+      "loss": 0.3995,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.230468000915424,
+      "learning_rate": 1.0157259448880352e-06,
+      "loss": 0.4888,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.4253145505003415,
+      "learning_rate": 1.0149377847271853e-06,
+      "loss": 0.4297,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.136128677638098,
+      "learning_rate": 1.0141498959314493e-06,
+      "loss": 0.4349,
+      "step": 9928
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.4486777416040875,
+      "learning_rate": 1.013362278554484e-06,
+      "loss": 0.4051,
+      "step": 9929
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.995650392091797,
+      "learning_rate": 1.012574932649919e-06,
+      "loss": 0.4318,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.693218418621899,
+      "learning_rate": 1.0117878582713724e-06,
+      "loss": 0.5785,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 18.991353663115717,
+      "learning_rate": 1.0110010554724392e-06,
+      "loss": 0.4885,
+      "step": 9932
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.323385198651309,
+      "learning_rate": 1.0102145243066963e-06,
+      "loss": 0.5245,
+      "step": 9933
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.76273467375305,
+      "learning_rate": 1.009428264827703e-06,
+      "loss": 0.5047,
+      "step": 9934
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.59827428179482,
+      "learning_rate": 1.0086422770890004e-06,
+      "loss": 0.5323,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.044754585222374,
+      "learning_rate": 1.0078565611441127e-06,
+      "loss": 0.4847,
+      "step": 9936
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.528746595558108,
+      "learning_rate": 1.007071117046542e-06,
+      "loss": 0.3751,
+      "step": 9937
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.651063012574595,
+      "learning_rate": 1.0062859448497731e-06,
+      "loss": 0.4671,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.307928647817108,
+      "learning_rate": 1.0055010446072728e-06,
+      "loss": 0.4875,
+      "step": 9939
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.431478500670533,
+      "learning_rate": 1.0047164163724905e-06,
+      "loss": 0.4382,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.585100353673137,
+      "learning_rate": 1.0039320601988557e-06,
+      "loss": 0.4675,
+      "step": 9941
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.38368858433458,
+      "learning_rate": 1.0031479761397788e-06,
+      "loss": 0.524,
+      "step": 9942
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.51718608920969,
+      "learning_rate": 1.002364164248652e-06,
+      "loss": 0.486,
+      "step": 9943
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.147872125838308,
+      "learning_rate": 1.0015806245788501e-06,
+      "loss": 0.4737,
+      "step": 9944
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.344975049515822,
+      "learning_rate": 1.0007973571837298e-06,
+      "loss": 0.4465,
+      "step": 9945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.398546749694117,
+      "learning_rate": 1.0000143621166275e-06,
+      "loss": 0.4956,
+      "step": 9946
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.577049160220395,
+      "learning_rate": 9.992316394308615e-07,
+      "loss": 0.4438,
+      "step": 9947
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.599520054802324,
+      "learning_rate": 9.984491891797299e-07,
+      "loss": 0.4472,
+      "step": 9948
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.388467141029817,
+      "learning_rate": 9.97667011416517e-07,
+      "loss": 0.6227,
+      "step": 9949
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.2985908000965445,
+      "learning_rate": 9.96885106194485e-07,
+      "loss": 0.4873,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.889992760373661,
+      "learning_rate": 9.96103473566878e-07,
+      "loss": 0.4762,
+      "step": 9951
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.240084619651356,
+      "learning_rate": 9.95322113586919e-07,
+      "loss": 0.4508,
+      "step": 9952
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.640121186227123,
+      "learning_rate": 9.945410263078197e-07,
+      "loss": 0.375,
+      "step": 9953
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.352140065053361,
+      "learning_rate": 9.937602117827648e-07,
+      "loss": 0.5672,
+      "step": 9954
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.899313072591287,
+      "learning_rate": 9.929796700649274e-07,
+      "loss": 0.4569,
+      "step": 9955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.225273193221054,
+      "learning_rate": 9.921994012074581e-07,
+      "loss": 0.4098,
+      "step": 9956
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.662306848696348,
+      "learning_rate": 9.914194052634873e-07,
+      "loss": 0.5074,
+      "step": 9957
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 12.220862280600517,
+      "learning_rate": 9.90639682286133e-07,
+      "loss": 0.492,
+      "step": 9958
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.8094789917913285,
+      "learning_rate": 9.89860232328489e-07,
+      "loss": 0.4442,
+      "step": 9959
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5616055748885738,
+      "learning_rate": 9.890810554436326e-07,
+      "loss": 0.7891,
+      "step": 9960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.440404293305404,
+      "learning_rate": 9.88302151684621e-07,
+      "loss": 0.424,
+      "step": 9961
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.367340719754102,
+      "learning_rate": 9.875235211044959e-07,
+      "loss": 0.4303,
+      "step": 9962
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.210232042873584,
+      "learning_rate": 9.867451637562774e-07,
+      "loss": 0.4797,
+      "step": 9963
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.02429763883308,
+      "learning_rate": 9.859670796929698e-07,
+      "loss": 0.4951,
+      "step": 9964
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.893241367092685,
+      "learning_rate": 9.851892689675564e-07,
+      "loss": 0.3683,
+      "step": 9965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.231360819860959,
+      "learning_rate": 9.84411731633001e-07,
+      "loss": 0.4269,
+      "step": 9966
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.0187991921417705,
+      "learning_rate": 9.836344677422539e-07,
+      "loss": 0.331,
+      "step": 9967
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.01279981109111,
+      "learning_rate": 9.828574773482402e-07,
+      "loss": 0.4546,
+      "step": 9968
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.158145942938377,
+      "learning_rate": 9.820807605038717e-07,
+      "loss": 0.4658,
+      "step": 9969
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.074347695854923,
+      "learning_rate": 9.813043172620363e-07,
+      "loss": 0.5073,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.359413341495795,
+      "learning_rate": 9.805281476756101e-07,
+      "loss": 0.4674,
+      "step": 9971
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 13.479916187985644,
+      "learning_rate": 9.797522517974439e-07,
+      "loss": 0.5318,
+      "step": 9972
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.44793996157733,
+      "learning_rate": 9.789766296803749e-07,
+      "loss": 0.5194,
+      "step": 9973
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.803310833019045,
+      "learning_rate": 9.78201281377219e-07,
+      "loss": 0.4772,
+      "step": 9974
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.135577191070853,
+      "learning_rate": 9.774262069407725e-07,
+      "loss": 0.4611,
+      "step": 9975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.528100168198068,
+      "learning_rate": 9.766514064238163e-07,
+      "loss": 0.4795,
+      "step": 9976
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.347883022200273,
+      "learning_rate": 9.758768798791106e-07,
+      "loss": 0.4743,
+      "step": 9977
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.495055009102982,
+      "learning_rate": 9.751026273593972e-07,
+      "loss": 0.5176,
+      "step": 9978
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.728854339436317,
+      "learning_rate": 9.743286489173975e-07,
+      "loss": 0.4773,
+      "step": 9979
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.165313982827575,
+      "learning_rate": 9.735549446058184e-07,
+      "loss": 0.4765,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 50.23314900170075,
+      "learning_rate": 9.727815144773438e-07,
+      "loss": 0.5028,
+      "step": 9981
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.059489039964209,
+      "learning_rate": 9.720083585846435e-07,
+      "loss": 0.416,
+      "step": 9982
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.517385244075628,
+      "learning_rate": 9.712354769803638e-07,
+      "loss": 0.461,
+      "step": 9983
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.542676401828732,
+      "learning_rate": 9.704628697171353e-07,
+      "loss": 0.4905,
+      "step": 9984
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.310953332726143,
+      "learning_rate": 9.696905368475678e-07,
+      "loss": 0.4971,
+      "step": 9985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.942155793985625,
+      "learning_rate": 9.689184784242562e-07,
+      "loss": 0.4115,
+      "step": 9986
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.955108299600008,
+      "learning_rate": 9.68146694499773e-07,
+      "loss": 0.4758,
+      "step": 9987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.2181474005559005,
+      "learning_rate": 9.673751851266716e-07,
+      "loss": 0.4611,
+      "step": 9988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.041329530591966,
+      "learning_rate": 9.66603950357492e-07,
+      "loss": 0.423,
+      "step": 9989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.396072983025943,
+      "learning_rate": 9.658329902447478e-07,
+      "loss": 0.4932,
+      "step": 9990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.598141569428424,
+      "learning_rate": 9.650623048409419e-07,
+      "loss": 0.5432,
+      "step": 9991
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.004070707467756,
+      "learning_rate": 9.642918941985524e-07,
+      "loss": 0.5076,
+      "step": 9992
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.746988275349766,
+      "learning_rate": 9.635217583700413e-07,
+      "loss": 0.38,
+      "step": 9993
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.992260522550975,
+      "learning_rate": 9.627518974078504e-07,
+      "loss": 0.4491,
+      "step": 9994
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.008730899911219,
+      "learning_rate": 9.619823113644045e-07,
+      "loss": 0.6265,
+      "step": 9995
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.84987106127966,
+      "learning_rate": 9.612130002921115e-07,
+      "loss": 0.4067,
+      "step": 9996
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.81169599565335,
+      "learning_rate": 9.604439642433538e-07,
+      "loss": 0.4439,
+      "step": 9997
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.500139466266395,
+      "learning_rate": 9.596752032705026e-07,
+      "loss": 0.4516,
+      "step": 9998
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.051423133036314,
+      "learning_rate": 9.589067174259042e-07,
+      "loss": 0.4325,
+      "step": 9999
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7572877004712661,
+      "learning_rate": 9.581385067618925e-07,
+      "loss": 0.6808,
+      "step": 10000
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.554308913411037,
+      "learning_rate": 9.573705713307774e-07,
+      "loss": 0.3914,
+      "step": 10001
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.8053539781401335,
+      "learning_rate": 9.566029111848524e-07,
+      "loss": 0.3679,
+      "step": 10002
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.522622413922228,
+      "learning_rate": 9.558355263763896e-07,
+      "loss": 0.5155,
+      "step": 10003
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.521526432285482,
+      "learning_rate": 9.550684169576484e-07,
+      "loss": 0.3976,
+      "step": 10004
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.756731220037941,
+      "learning_rate": 9.543015829808632e-07,
+      "loss": 0.4582,
+      "step": 10005
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.598359791595431,
+      "learning_rate": 9.535350244982506e-07,
+      "loss": 0.5529,
+      "step": 10006
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.750059225538125,
+      "learning_rate": 9.527687415620135e-07,
+      "loss": 0.3232,
+      "step": 10007
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.71658606969646,
+      "learning_rate": 9.520027342243288e-07,
+      "loss": 0.5305,
+      "step": 10008
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.136275752668137,
+      "learning_rate": 9.512370025373613e-07,
+      "loss": 0.416,
+      "step": 10009
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.546729170770495,
+      "learning_rate": 9.504715465532521e-07,
+      "loss": 0.3639,
+      "step": 10010
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.841694314540368,
+      "learning_rate": 9.497063663241263e-07,
+      "loss": 0.389,
+      "step": 10011
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.511965874450835,
+      "learning_rate": 9.48941461902087e-07,
+      "loss": 0.4889,
+      "step": 10012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.393831020924914,
+      "learning_rate": 9.481768333392238e-07,
+      "loss": 0.5948,
+      "step": 10013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 21.389705120303784,
+      "learning_rate": 9.47412480687604e-07,
+      "loss": 0.4255,
+      "step": 10014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.9916003945168566,
+      "learning_rate": 9.466484039992741e-07,
+      "loss": 0.466,
+      "step": 10015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.892352877571813,
+      "learning_rate": 9.458846033262675e-07,
+      "loss": 0.4411,
+      "step": 10016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 33.51243427064841,
+      "learning_rate": 9.451210787205928e-07,
+      "loss": 0.4133,
+      "step": 10017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.076663509624877,
+      "learning_rate": 9.443578302342459e-07,
+      "loss": 0.4635,
+      "step": 10018
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.796922758423589,
+      "learning_rate": 9.435948579191989e-07,
+      "loss": 0.3607,
+      "step": 10019
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.608443887713639,
+      "learning_rate": 9.428321618274066e-07,
+      "loss": 0.4387,
+      "step": 10020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.330366458345106,
+      "learning_rate": 9.420697420108044e-07,
+      "loss": 0.4038,
+      "step": 10021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.97081038536865,
+      "learning_rate": 9.413075985213122e-07,
+      "loss": 0.4033,
+      "step": 10022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.180788802963225,
+      "learning_rate": 9.405457314108269e-07,
+      "loss": 0.3256,
+      "step": 10023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.00578523934972,
+      "learning_rate": 9.397841407312286e-07,
+      "loss": 0.3433,
+      "step": 10024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.442948277794774,
+      "learning_rate": 9.39022826534377e-07,
+      "loss": 0.3705,
+      "step": 10025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 16.90147514155133,
+      "learning_rate": 9.382617888721151e-07,
+      "loss": 0.5424,
+      "step": 10026
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.313080980987303,
+      "learning_rate": 9.375010277962687e-07,
+      "loss": 0.3615,
+      "step": 10027
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.203850154839431,
+      "learning_rate": 9.367405433586396e-07,
+      "loss": 0.3472,
+      "step": 10028
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.341411719483625,
+      "learning_rate": 9.359803356110142e-07,
+      "loss": 0.5629,
+      "step": 10029
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.515750659741559,
+      "learning_rate": 9.352204046051577e-07,
+      "loss": 0.3464,
+      "step": 10030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.202275964342792,
+      "learning_rate": 9.344607503928205e-07,
+      "loss": 0.493,
+      "step": 10031
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.221582495551978,
+      "learning_rate": 9.337013730257294e-07,
+      "loss": 0.4805,
+      "step": 10032
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.88344063346139,
+      "learning_rate": 9.329422725555987e-07,
+      "loss": 0.5442,
+      "step": 10033
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.200341923463222,
+      "learning_rate": 9.321834490341142e-07,
+      "loss": 0.467,
+      "step": 10034
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 15.793329215114793,
+      "learning_rate": 9.314249025129507e-07,
+      "loss": 0.4664,
+      "step": 10035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.392181342351366,
+      "learning_rate": 9.306666330437641e-07,
+      "loss": 0.4402,
+      "step": 10036
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.234265165816602,
+      "learning_rate": 9.299086406781871e-07,
+      "loss": 0.5284,
+      "step": 10037
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.555665864579173,
+      "learning_rate": 9.29150925467836e-07,
+      "loss": 0.4621,
+      "step": 10038
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5640945298647055,
+      "learning_rate": 9.283934874643069e-07,
+      "loss": 0.6766,
+      "step": 10039
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.937671700445241,
+      "learning_rate": 9.276363267191795e-07,
+      "loss": 0.5236,
+      "step": 10040
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.458764996985592,
+      "learning_rate": 9.268794432840117e-07,
+      "loss": 0.41,
+      "step": 10041
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.925267345520518,
+      "learning_rate": 9.26122837210347e-07,
+      "loss": 0.472,
+      "step": 10042
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.295462552754058,
+      "learning_rate": 9.253665085497021e-07,
+      "loss": 0.4722,
+      "step": 10043
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.671903792194122,
+      "learning_rate": 9.246104573535836e-07,
+      "loss": 0.5051,
+      "step": 10044
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.056867026113015,
+      "learning_rate": 9.238546836734724e-07,
+      "loss": 0.4138,
+      "step": 10045
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.1025357925947965,
+      "learning_rate": 9.230991875608358e-07,
+      "loss": 0.33,
+      "step": 10046
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.331004383485055,
+      "learning_rate": 9.223439690671193e-07,
+      "loss": 0.5478,
+      "step": 10047
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.151434429248363,
+      "learning_rate": 9.215890282437473e-07,
+      "loss": 0.5039,
+      "step": 10048
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 18.509946941561992,
+      "learning_rate": 9.208343651421315e-07,
+      "loss": 0.5072,
+      "step": 10049
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.184223905852008,
+      "learning_rate": 9.200799798136584e-07,
+      "loss": 0.5025,
+      "step": 10050
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.667259825955901,
+      "learning_rate": 9.19325872309702e-07,
+      "loss": 0.4329,
+      "step": 10051
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.525052689281104,
+      "learning_rate": 9.185720426816086e-07,
+      "loss": 0.4589,
+      "step": 10052
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.727289032538636,
+      "learning_rate": 9.178184909807147e-07,
+      "loss": 0.404,
+      "step": 10053
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.2962489404468105,
+      "learning_rate": 9.170652172583311e-07,
+      "loss": 0.4269,
+      "step": 10054
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.868103018265902,
+      "learning_rate": 9.163122215657544e-07,
+      "loss": 0.2901,
+      "step": 10055
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.950426671337524,
+      "learning_rate": 9.155595039542602e-07,
+      "loss": 0.4989,
+      "step": 10056
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7210478090456518,
+      "learning_rate": 9.148070644751033e-07,
+      "loss": 0.7171,
+      "step": 10057
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.836898730520573,
+      "learning_rate": 9.140549031795237e-07,
+      "loss": 0.4591,
+      "step": 10058
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.086730678394488,
+      "learning_rate": 9.133030201187382e-07,
+      "loss": 0.4192,
+      "step": 10059
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.123422563875917,
+      "learning_rate": 9.125514153439508e-07,
+      "loss": 0.3345,
+      "step": 10060
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.805484589459496,
+      "learning_rate": 9.118000889063361e-07,
+      "loss": 0.4622,
+      "step": 10061
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.332995405593328,
+      "learning_rate": 9.110490408570616e-07,
+      "loss": 0.5003,
+      "step": 10062
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.112269295364976,
+      "learning_rate": 9.102982712472669e-07,
+      "loss": 0.4468,
+      "step": 10063
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.136469478577768,
+      "learning_rate": 9.095477801280788e-07,
+      "loss": 0.5011,
+      "step": 10064
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.379725532414513,
+      "learning_rate": 9.087975675506006e-07,
+      "loss": 0.484,
+      "step": 10065
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.336510759717281,
+      "learning_rate": 9.080476335659183e-07,
+      "loss": 0.5095,
+      "step": 10066
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.925383474395844,
+      "learning_rate": 9.072979782251006e-07,
+      "loss": 0.4402,
+      "step": 10067
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.090139350998761,
+      "learning_rate": 9.065486015791941e-07,
+      "loss": 0.4423,
+      "step": 10068
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.805819864093451,
+      "learning_rate": 9.0579950367923e-07,
+      "loss": 0.4804,
+      "step": 10069
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.56726603494845,
+      "learning_rate": 9.050506845762169e-07,
+      "loss": 0.3948,
+      "step": 10070
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.8011085614112945,
+      "learning_rate": 9.043021443211469e-07,
+      "loss": 0.4824,
+      "step": 10071
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.753779989948985,
+      "learning_rate": 9.035538829649903e-07,
+      "loss": 0.5065,
+      "step": 10072
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.180544343493936,
+      "learning_rate": 9.028059005587036e-07,
+      "loss": 0.3784,
+      "step": 10073
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.800244551724872,
+      "learning_rate": 9.020581971532194e-07,
+      "loss": 0.3483,
+      "step": 10074
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.040637727362575,
+      "learning_rate": 9.013107727994519e-07,
+      "loss": 0.4543,
+      "step": 10075
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.097823888250758,
+      "learning_rate": 9.005636275482999e-07,
+      "loss": 0.4525,
+      "step": 10076
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.708753167442674,
+      "learning_rate": 8.998167614506381e-07,
+      "loss": 0.4515,
+      "step": 10077
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.0075361156992315,
+      "learning_rate": 8.990701745573272e-07,
+      "loss": 0.4197,
+      "step": 10078
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.063538037858205,
+      "learning_rate": 8.983238669192057e-07,
+      "loss": 0.5032,
+      "step": 10079
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.52150204006643,
+      "learning_rate": 8.975778385870931e-07,
+      "loss": 0.4946,
+      "step": 10080
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.154960559319015,
+      "learning_rate": 8.9683208961179e-07,
+      "loss": 0.5003,
+      "step": 10081
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.578207853956743,
+      "learning_rate": 8.960866200440804e-07,
+      "loss": 0.3428,
+      "step": 10082
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.528482511306491,
+      "learning_rate": 8.953414299347269e-07,
+      "loss": 0.4138,
+      "step": 10083
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.095487039022025,
+      "learning_rate": 8.945965193344736e-07,
+      "loss": 0.4948,
+      "step": 10084
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.924874230736446,
+      "learning_rate": 8.938518882940444e-07,
+      "loss": 0.5056,
+      "step": 10085
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 26.99625571426257,
+      "learning_rate": 8.931075368641462e-07,
+      "loss": 0.348,
+      "step": 10086
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.0161453996047785,
+      "learning_rate": 8.923634650954677e-07,
+      "loss": 0.5475,
+      "step": 10087
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.169249669399409,
+      "learning_rate": 8.916196730386751e-07,
+      "loss": 0.3732,
+      "step": 10088
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.987847862040848,
+      "learning_rate": 8.908761607444183e-07,
+      "loss": 0.4405,
+      "step": 10089
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.063122516149561,
+      "learning_rate": 8.901329282633247e-07,
+      "loss": 0.529,
+      "step": 10090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 19.279864183844982,
+      "learning_rate": 8.893899756460089e-07,
+      "loss": 0.3766,
+      "step": 10091
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.752232973012822,
+      "learning_rate": 8.886473029430609e-07,
+      "loss": 0.4715,
+      "step": 10092
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.316947888861525,
+      "learning_rate": 8.879049102050529e-07,
+      "loss": 0.4093,
+      "step": 10093
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.870948369894107,
+      "learning_rate": 8.87162797482538e-07,
+      "loss": 0.4608,
+      "step": 10094
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.068802301894884,
+      "learning_rate": 8.864209648260519e-07,
+      "loss": 0.3802,
+      "step": 10095
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.512902538331238,
+      "learning_rate": 8.856794122861118e-07,
+      "loss": 0.432,
+      "step": 10096
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.741896368404952,
+      "learning_rate": 8.849381399132118e-07,
+      "loss": 0.4659,
+      "step": 10097
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.952224147052782,
+      "learning_rate": 8.841971477578304e-07,
+      "loss": 0.4869,
+      "step": 10098
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 20.291425446236094,
+      "learning_rate": 8.834564358704234e-07,
+      "loss": 0.5229,
+      "step": 10099
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.754846252358386,
+      "learning_rate": 8.827160043014338e-07,
+      "loss": 0.4616,
+      "step": 10100
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.242092874898631,
+      "learning_rate": 8.819758531012801e-07,
+      "loss": 0.5059,
+      "step": 10101
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.782367715528662,
+      "learning_rate": 8.812359823203626e-07,
+      "loss": 0.3437,
+      "step": 10102
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.6014943594082105,
+      "learning_rate": 8.804963920090632e-07,
+      "loss": 0.3949,
+      "step": 10103
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.16402932210199,
+      "learning_rate": 8.797570822177459e-07,
+      "loss": 0.5282,
+      "step": 10104
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.016492893362459,
+      "learning_rate": 8.790180529967529e-07,
+      "loss": 0.4506,
+      "step": 10105
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.715351993615023,
+      "learning_rate": 8.782793043964111e-07,
+      "loss": 0.5335,
+      "step": 10106
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.510071021384293,
+      "learning_rate": 8.775408364670241e-07,
+      "loss": 0.4673,
+      "step": 10107
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 42.61915159575654,
+      "learning_rate": 8.768026492588782e-07,
+      "loss": 0.4058,
+      "step": 10108
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.2911927170334,
+      "learning_rate": 8.760647428222424e-07,
+      "loss": 0.4933,
+      "step": 10109
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.283830658910393,
+      "learning_rate": 8.753271172073635e-07,
+      "loss": 0.4123,
+      "step": 10110
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7798740795775678,
+      "learning_rate": 8.745897724644714e-07,
+      "loss": 0.8092,
+      "step": 10111
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.708645662026051,
+      "learning_rate": 8.738527086437737e-07,
+      "loss": 0.4785,
+      "step": 10112
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.57380067976828,
+      "learning_rate": 8.731159257954646e-07,
+      "loss": 0.3475,
+      "step": 10113
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.287389973041275,
+      "learning_rate": 8.723794239697125e-07,
+      "loss": 0.471,
+      "step": 10114
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.36368840350955,
+      "learning_rate": 8.716432032166733e-07,
+      "loss": 0.3775,
+      "step": 10115
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.284065990354922,
+      "learning_rate": 8.709072635864785e-07,
+      "loss": 0.5262,
+      "step": 10116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.622615546250698,
+      "learning_rate": 8.701716051292408e-07,
+      "loss": 0.4614,
+      "step": 10117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.456990558923723,
+      "learning_rate": 8.694362278950586e-07,
+      "loss": 0.5041,
+      "step": 10118
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1406500843866447,
+      "learning_rate": 8.687011319340056e-07,
+      "loss": 0.7264,
+      "step": 10119
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.353107962828701,
+      "learning_rate": 8.679663172961394e-07,
+      "loss": 0.4134,
+      "step": 10120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.072534603919392,
+      "learning_rate": 8.672317840314964e-07,
+      "loss": 0.504,
+      "step": 10121
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 19.261932732610752,
+      "learning_rate": 8.66497532190097e-07,
+      "loss": 0.4945,
+      "step": 10122
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.832765155722948,
+      "learning_rate": 8.657635618219384e-07,
+      "loss": 0.4357,
+      "step": 10123
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.335759524958377,
+      "learning_rate": 8.650298729770035e-07,
+      "loss": 0.4575,
+      "step": 10124
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9674311788608543,
+      "learning_rate": 8.642964657052511e-07,
+      "loss": 0.7347,
+      "step": 10125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7428317410777194,
+      "learning_rate": 8.635633400566229e-07,
+      "loss": 0.683,
+      "step": 10126
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.697624638819015,
+      "learning_rate": 8.62830496081043e-07,
+      "loss": 0.4705,
+      "step": 10127
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.773690375892421,
+      "learning_rate": 8.620979338284141e-07,
+      "loss": 0.3715,
+      "step": 10128
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.086746915803422,
+      "learning_rate": 8.613656533486203e-07,
+      "loss": 0.5511,
+      "step": 10129
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.68218132657136,
+      "learning_rate": 8.606336546915256e-07,
+      "loss": 0.3727,
+      "step": 10130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 32.4289576780104,
+      "learning_rate": 8.599019379069778e-07,
+      "loss": 0.5073,
+      "step": 10131
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.7336788718370055,
+      "learning_rate": 8.591705030448022e-07,
+      "loss": 0.5442,
+      "step": 10132
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.804905651784086,
+      "learning_rate": 8.584393501548077e-07,
+      "loss": 0.5179,
+      "step": 10133
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.014549183896766,
+      "learning_rate": 8.577084792867818e-07,
+      "loss": 0.4691,
+      "step": 10134
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.645748196726168,
+      "learning_rate": 8.569778904904929e-07,
+      "loss": 0.4525,
+      "step": 10135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.438614664180797,
+      "learning_rate": 8.562475838156908e-07,
+      "loss": 0.4747,
+      "step": 10136
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.5840123179857635,
+      "learning_rate": 8.555175593121068e-07,
+      "loss": 0.3902,
+      "step": 10137
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.27343585678152,
+      "learning_rate": 8.54787817029455e-07,
+      "loss": 0.3572,
+      "step": 10138
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.237720755415661,
+      "learning_rate": 8.540583570174221e-07,
+      "loss": 0.4599,
+      "step": 10139
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.456382530588732,
+      "learning_rate": 8.533291793256859e-07,
+      "loss": 0.4608,
+      "step": 10140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 12.901005201686413,
+      "learning_rate": 8.526002840038966e-07,
+      "loss": 0.432,
+      "step": 10141
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.36986533796349,
+      "learning_rate": 8.518716711016917e-07,
+      "loss": 0.4237,
+      "step": 10142
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.2804309490715315,
+      "learning_rate": 8.511433406686847e-07,
+      "loss": 0.4419,
+      "step": 10143
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.939476689379586,
+      "learning_rate": 8.504152927544729e-07,
+      "loss": 0.5106,
+      "step": 10144
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.008089282071587,
+      "learning_rate": 8.496875274086308e-07,
+      "loss": 0.4415,
+      "step": 10145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.103754262885621,
+      "learning_rate": 8.489600446807178e-07,
+      "loss": 0.4888,
+      "step": 10146
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.110985921342472,
+      "learning_rate": 8.482328446202748e-07,
+      "loss": 0.381,
+      "step": 10147
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.7008304398069285,
+      "learning_rate": 8.475059272768155e-07,
+      "loss": 0.4493,
+      "step": 10148
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.746534920654861,
+      "learning_rate": 8.467792926998437e-07,
+      "loss": 0.4905,
+      "step": 10149
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.611485117920906,
+      "learning_rate": 8.46052940938838e-07,
+      "loss": 0.4501,
+      "step": 10150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.966738967076735,
+      "learning_rate": 8.453268720432611e-07,
+      "loss": 0.5082,
+      "step": 10151
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.998598213110106,
+      "learning_rate": 8.446010860625547e-07,
+      "loss": 0.5195,
+      "step": 10152
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.466699163800547,
+      "learning_rate": 8.438755830461415e-07,
+      "loss": 0.4952,
+      "step": 10153
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 20.85855052632193,
+      "learning_rate": 8.431503630434241e-07,
+      "loss": 0.5214,
+      "step": 10154
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.538015128034495,
+      "learning_rate": 8.424254261037884e-07,
+      "loss": 0.4405,
+      "step": 10155
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.709930965665688,
+      "learning_rate": 8.417007722765985e-07,
+      "loss": 0.4218,
+      "step": 10156
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.210738376960126,
+      "learning_rate": 8.409764016111993e-07,
+      "loss": 0.4967,
+      "step": 10157
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.670078287341043,
+      "learning_rate": 8.402523141569197e-07,
+      "loss": 0.4851,
+      "step": 10158
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.6865286108753565,
+      "learning_rate": 8.39528509963064e-07,
+      "loss": 0.4862,
+      "step": 10159
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.5329967936816296,
+      "learning_rate": 8.388049890789224e-07,
+      "loss": 0.4983,
+      "step": 10160
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.5131045916658845,
+      "learning_rate": 8.380817515537626e-07,
+      "loss": 0.5139,
+      "step": 10161
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.890056839512206,
+      "learning_rate": 8.373587974368336e-07,
+      "loss": 0.4342,
+      "step": 10162
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 26.049630869098692,
+      "learning_rate": 8.366361267773648e-07,
+      "loss": 0.3596,
+      "step": 10163
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.589765129105103,
+      "learning_rate": 8.359137396245687e-07,
+      "loss": 0.5286,
+      "step": 10164
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.908231524000042,
+      "learning_rate": 8.351916360276352e-07,
+      "loss": 0.4795,
+      "step": 10165
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.602086856421593,
+      "learning_rate": 8.344698160357356e-07,
+      "loss": 0.46,
+      "step": 10166
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.052763647167044,
+      "learning_rate": 8.337482796980256e-07,
+      "loss": 0.4559,
+      "step": 10167
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.358568796177347,
+      "learning_rate": 8.330270270636354e-07,
+      "loss": 0.5033,
+      "step": 10168
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.983777175654074,
+      "learning_rate": 8.323060581816816e-07,
+      "loss": 0.5314,
+      "step": 10169
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.491395387771526,
+      "learning_rate": 8.31585373101258e-07,
+      "loss": 0.4154,
+      "step": 10170
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.1916138317391916,
+      "learning_rate": 8.308649718714401e-07,
+      "loss": 0.4606,
+      "step": 10171
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.256475641581051,
+      "learning_rate": 8.301448545412827e-07,
+      "loss": 0.2979,
+      "step": 10172
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.347818619750402,
+      "learning_rate": 8.294250211598248e-07,
+      "loss": 0.4363,
+      "step": 10173
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.038629411788658,
+      "learning_rate": 8.287054717760823e-07,
+      "loss": 0.5434,
+      "step": 10174
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.989205350194723,
+      "learning_rate": 8.279862064390559e-07,
+      "loss": 0.4649,
+      "step": 10175
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.106209725919705,
+      "learning_rate": 8.272672251977204e-07,
+      "loss": 0.4282,
+      "step": 10176
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.145939242845098,
+      "learning_rate": 8.265485281010372e-07,
+      "loss": 0.3982,
+      "step": 10177
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.69887955258301,
+      "learning_rate": 8.258301151979475e-07,
+      "loss": 0.5282,
+      "step": 10178
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.089386899312101,
+      "learning_rate": 8.251119865373714e-07,
+      "loss": 0.3982,
+      "step": 10179
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 17.04756735721419,
+      "learning_rate": 8.2439414216821e-07,
+      "loss": 0.4062,
+      "step": 10180
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.639028211366384,
+      "learning_rate": 8.236765821393433e-07,
+      "loss": 0.3856,
+      "step": 10181
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.27167530751515,
+      "learning_rate": 8.229593064996377e-07,
+      "loss": 0.3946,
+      "step": 10182
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.298728872961092,
+      "learning_rate": 8.222423152979336e-07,
+      "loss": 0.4827,
+      "step": 10183
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.991446778755614,
+      "learning_rate": 8.215256085830581e-07,
+      "loss": 0.477,
+      "step": 10184
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.534397614932672,
+      "learning_rate": 8.208091864038114e-07,
+      "loss": 0.4416,
+      "step": 10185
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.911432599041636,
+      "learning_rate": 8.200930488089814e-07,
+      "loss": 0.4512,
+      "step": 10186
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.311065849354394,
+      "learning_rate": 8.193771958473345e-07,
+      "loss": 0.3919,
+      "step": 10187
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 21.443594351518588,
+      "learning_rate": 8.186616275676156e-07,
+      "loss": 0.4292,
+      "step": 10188
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.767429284685199,
+      "learning_rate": 8.179463440185531e-07,
+      "loss": 0.4967,
+      "step": 10189
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 15.329728284033289,
+      "learning_rate": 8.172313452488523e-07,
+      "loss": 0.3899,
+      "step": 10190
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.38460415971196,
+      "learning_rate": 8.165166313072037e-07,
+      "loss": 0.4218,
+      "step": 10191
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.780928832187982,
+      "learning_rate": 8.158022022422745e-07,
+      "loss": 0.4341,
+      "step": 10192
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.1989890367867435,
+      "learning_rate": 8.150880581027176e-07,
+      "loss": 0.3498,
+      "step": 10193
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.358118180019961,
+      "learning_rate": 8.143741989371584e-07,
+      "loss": 0.5382,
+      "step": 10194
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.198176581628715,
+      "learning_rate": 8.136606247942103e-07,
+      "loss": 0.44,
+      "step": 10195
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.770633457556139,
+      "learning_rate": 8.129473357224633e-07,
+      "loss": 0.4329,
+      "step": 10196
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.942837798130381,
+      "learning_rate": 8.122343317704906e-07,
+      "loss": 0.4925,
+      "step": 10197
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.747558673427397,
+      "learning_rate": 8.115216129868436e-07,
+      "loss": 0.4178,
+      "step": 10198
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.910169325560929,
+      "learning_rate": 8.10809179420055e-07,
+      "loss": 0.5062,
+      "step": 10199
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.79889095123164,
+      "learning_rate": 8.100970311186401e-07,
+      "loss": 0.6071,
+      "step": 10200
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.727333228250036,
+      "learning_rate": 8.093851681310904e-07,
+      "loss": 0.4388,
+      "step": 10201
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.141358728192876,
+      "learning_rate": 8.086735905058846e-07,
+      "loss": 0.4803,
+      "step": 10202
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.664627277351387,
+      "learning_rate": 8.079622982914736e-07,
+      "loss": 0.42,
+      "step": 10203
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 10.21854471887122,
+      "learning_rate": 8.072512915362956e-07,
+      "loss": 0.4469,
+      "step": 10204
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.182912837853294,
+      "learning_rate": 8.065405702887657e-07,
+      "loss": 0.4547,
+      "step": 10205
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 27.2368529690893,
+      "learning_rate": 8.058301345972835e-07,
+      "loss": 0.4788,
+      "step": 10206
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.204344841589182,
+      "learning_rate": 8.051199845102243e-07,
+      "loss": 0.5282,
+      "step": 10207
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.109128033089089,
+      "learning_rate": 8.044101200759452e-07,
+      "loss": 0.5486,
+      "step": 10208
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.755191580953285,
+      "learning_rate": 8.037005413427879e-07,
+      "loss": 0.5176,
+      "step": 10209
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.761876031681286,
+      "learning_rate": 8.029912483590685e-07,
+      "loss": 0.4273,
+      "step": 10210
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 59.0534099629245,
+      "learning_rate": 8.022822411730896e-07,
+      "loss": 0.4593,
+      "step": 10211
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.81537693086377,
+      "learning_rate": 8.015735198331303e-07,
+      "loss": 0.4006,
+      "step": 10212
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.918180793180587,
+      "learning_rate": 8.008650843874505e-07,
+      "loss": 0.5406,
+      "step": 10213
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.414264016779281,
+      "learning_rate": 8.001569348842913e-07,
+      "loss": 0.4852,
+      "step": 10214
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.85618793470409,
+      "learning_rate": 7.994490713718766e-07,
+      "loss": 0.4303,
+      "step": 10215
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.5177532367579145,
+      "learning_rate": 7.987414938984079e-07,
+      "loss": 0.4204,
+      "step": 10216
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.7254330227390975,
+      "learning_rate": 7.980342025120664e-07,
+      "loss": 0.4031,
+      "step": 10217
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6011377616744695,
+      "learning_rate": 7.97327197261018e-07,
+      "loss": 0.713,
+      "step": 10218
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.440415063527272,
+      "learning_rate": 7.96620478193404e-07,
+      "loss": 0.4746,
+      "step": 10219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.274476612519814,
+      "learning_rate": 7.959140453573522e-07,
+      "loss": 0.5095,
+      "step": 10220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.555062540433794,
+      "learning_rate": 7.952078988009653e-07,
+      "loss": 0.4319,
+      "step": 10221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.107390850298797,
+      "learning_rate": 7.94502038572329e-07,
+      "loss": 0.4241,
+      "step": 10222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7255476439517494,
+      "learning_rate": 7.93796464719509e-07,
+      "loss": 0.4091,
+      "step": 10223
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.167470336526223,
+      "learning_rate": 7.930911772905525e-07,
+      "loss": 0.3295,
+      "step": 10224
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.549952197460975,
+      "learning_rate": 7.923861763334867e-07,
+      "loss": 0.6856,
+      "step": 10225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7684922170113253,
+      "learning_rate": 7.916814618963175e-07,
+      "loss": 0.6797,
+      "step": 10226
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.856152804704237,
+      "learning_rate": 7.909770340270345e-07,
+      "loss": 0.4075,
+      "step": 10227
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 16.57091151691455,
+      "learning_rate": 7.902728927736047e-07,
+      "loss": 0.489,
+      "step": 10228
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6451915758631048,
+      "learning_rate": 7.895690381839788e-07,
+      "loss": 0.7685,
+      "step": 10229
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.227915698894213,
+      "learning_rate": 7.88865470306085e-07,
+      "loss": 0.4408,
+      "step": 10230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.532884662470714,
+      "learning_rate": 7.881621891878338e-07,
+      "loss": 0.5697,
+      "step": 10231
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.986056699257861,
+      "learning_rate": 7.874591948771132e-07,
+      "loss": 0.4476,
+      "step": 10232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.591042438035359,
+      "learning_rate": 7.867564874217976e-07,
+      "loss": 0.5087,
+      "step": 10233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.782236938695508,
+      "learning_rate": 7.860540668697364e-07,
+      "loss": 0.4107,
+      "step": 10234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.628301162198461,
+      "learning_rate": 7.853519332687615e-07,
+      "loss": 0.4926,
+      "step": 10235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.023006101989973,
+      "learning_rate": 7.846500866666834e-07,
+      "loss": 0.4258,
+      "step": 10236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.309383047459592,
+      "learning_rate": 7.839485271112967e-07,
+      "loss": 0.4749,
+      "step": 10237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.887208590936651,
+      "learning_rate": 7.832472546503756e-07,
+      "loss": 0.4078,
+      "step": 10238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.495569067616138,
+      "learning_rate": 7.825462693316721e-07,
+      "loss": 0.5243,
+      "step": 10239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.054486652621074,
+      "learning_rate": 7.818455712029205e-07,
+      "loss": 0.4597,
+      "step": 10240
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.388462524247033,
+      "learning_rate": 7.811451603118342e-07,
+      "loss": 0.4859,
+      "step": 10241
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.507605719914513,
+      "learning_rate": 7.804450367061095e-07,
+      "loss": 0.4152,
+      "step": 10242
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.206298354554233,
+      "learning_rate": 7.797452004334217e-07,
+      "loss": 0.5345,
+      "step": 10243
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.393543732232727,
+      "learning_rate": 7.790456515414258e-07,
+      "loss": 0.5441,
+      "step": 10244
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.446107893794446,
+      "learning_rate": 7.783463900777572e-07,
+      "loss": 0.4282,
+      "step": 10245
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.527720582083138,
+      "learning_rate": 7.776474160900338e-07,
+      "loss": 0.4894,
+      "step": 10246
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.77179295838648,
+      "learning_rate": 7.769487296258532e-07,
+      "loss": 0.4889,
+      "step": 10247
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.896501106413803,
+      "learning_rate": 7.762503307327924e-07,
+      "loss": 0.4579,
+      "step": 10248
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.218680680037386,
+      "learning_rate": 7.755522194584087e-07,
+      "loss": 0.4314,
+      "step": 10249
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.501057332238288,
+      "learning_rate": 7.748543958502396e-07,
+      "loss": 0.4567,
+      "step": 10250
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.483444561829112,
+      "learning_rate": 7.741568599558058e-07,
+      "loss": 0.4254,
+      "step": 10251
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.667845693561465,
+      "learning_rate": 7.734596118226057e-07,
+      "loss": 0.4328,
+      "step": 10252
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.986626402722206,
+      "learning_rate": 7.727626514981185e-07,
+      "loss": 0.5595,
+      "step": 10253
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.167210265490292,
+      "learning_rate": 7.72065979029803e-07,
+      "loss": 0.4104,
+      "step": 10254
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.779959517595863,
+      "learning_rate": 7.713695944651017e-07,
+      "loss": 0.3582,
+      "step": 10255
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.34850976344241,
+      "learning_rate": 7.706734978514335e-07,
+      "loss": 0.5424,
+      "step": 10256
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.189922133993687,
+      "learning_rate": 7.69977689236201e-07,
+      "loss": 0.3998,
+      "step": 10257
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.186514994629393,
+      "learning_rate": 7.692821686667856e-07,
+      "loss": 0.4184,
+      "step": 10258
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.417545388884876,
+      "learning_rate": 7.685869361905468e-07,
+      "loss": 0.503,
+      "step": 10259
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.005061375542694,
+      "learning_rate": 7.678919918548306e-07,
+      "loss": 0.4586,
+      "step": 10260
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.620087002815506,
+      "learning_rate": 7.671973357069568e-07,
+      "loss": 0.4094,
+      "step": 10261
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.19243538804253,
+      "learning_rate": 7.665029677942298e-07,
+      "loss": 0.3717,
+      "step": 10262
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.23141442243463,
+      "learning_rate": 7.658088881639314e-07,
+      "loss": 0.4812,
+      "step": 10263
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.925134067017576,
+      "learning_rate": 7.65115096863327e-07,
+      "loss": 0.4851,
+      "step": 10264
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.069595301949004,
+      "learning_rate": 7.644215939396593e-07,
+      "loss": 0.4632,
+      "step": 10265
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.169862670348313,
+      "learning_rate": 7.637283794401551e-07,
+      "loss": 0.5524,
+      "step": 10266
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.536707473508523,
+      "learning_rate": 7.630354534120172e-07,
+      "loss": 0.423,
+      "step": 10267
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.720560529812948,
+      "learning_rate": 7.623428159024305e-07,
+      "loss": 0.4524,
+      "step": 10268
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.276074460004801,
+      "learning_rate": 7.616504669585628e-07,
+      "loss": 0.711,
+      "step": 10269
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.148055123998724,
+      "learning_rate": 7.609584066275572e-07,
+      "loss": 0.4788,
+      "step": 10270
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.991562145935618,
+      "learning_rate": 7.602666349565435e-07,
+      "loss": 0.4234,
+      "step": 10271
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.577597228000968,
+      "learning_rate": 7.595751519926242e-07,
+      "loss": 0.4196,
+      "step": 10272
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.490263074972535,
+      "learning_rate": 7.588839577828893e-07,
+      "loss": 0.6517,
+      "step": 10273
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.557945805963383,
+      "learning_rate": 7.581930523744036e-07,
+      "loss": 0.3953,
+      "step": 10274
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.438696799288935,
+      "learning_rate": 7.575024358142174e-07,
+      "loss": 0.4557,
+      "step": 10275
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.787385670808862,
+      "learning_rate": 7.568121081493573e-07,
+      "loss": 0.4533,
+      "step": 10276
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.583623396526351,
+      "learning_rate": 7.561220694268301e-07,
+      "loss": 0.527,
+      "step": 10277
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 21.98473746445431,
+      "learning_rate": 7.554323196936275e-07,
+      "loss": 0.3475,
+      "step": 10278
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.973424868455671,
+      "learning_rate": 7.547428589967154e-07,
+      "loss": 0.3835,
+      "step": 10279
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6889134933832208,
+      "learning_rate": 7.540536873830468e-07,
+      "loss": 0.6687,
+      "step": 10280
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.447383318684343,
+      "learning_rate": 7.533648048995462e-07,
+      "loss": 0.4042,
+      "step": 10281
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.069434906769151,
+      "learning_rate": 7.526762115931274e-07,
+      "loss": 0.4845,
+      "step": 10282
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.031768114335188,
+      "learning_rate": 7.519879075106779e-07,
+      "loss": 0.4646,
+      "step": 10283
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.557760973935841,
+      "learning_rate": 7.512998926990705e-07,
+      "loss": 0.7312,
+      "step": 10284
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.051523049821311,
+      "learning_rate": 7.506121672051559e-07,
+      "loss": 0.4862,
+      "step": 10285
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6921965711043703,
+      "learning_rate": 7.499247310757634e-07,
+      "loss": 0.6966,
+      "step": 10286
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.578678013204452,
+      "learning_rate": 7.492375843577044e-07,
+      "loss": 0.4255,
+      "step": 10287
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.657610804045239,
+      "learning_rate": 7.48550727097771e-07,
+      "loss": 0.3873,
+      "step": 10288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5917523154180873,
+      "learning_rate": 7.478641593427377e-07,
+      "loss": 0.4711,
+      "step": 10289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.5822068044330155,
+      "learning_rate": 7.471778811393526e-07,
+      "loss": 0.4637,
+      "step": 10290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.767657169552659,
+      "learning_rate": 7.464918925343511e-07,
+      "loss": 0.3879,
+      "step": 10291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.643979902870738,
+      "learning_rate": 7.458061935744437e-07,
+      "loss": 0.5003,
+      "step": 10292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.792147794670904,
+      "learning_rate": 7.451207843063263e-07,
+      "loss": 0.472,
+      "step": 10293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 17.76247269136627,
+      "learning_rate": 7.444356647766704e-07,
+      "loss": 0.5232,
+      "step": 10294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.960365257659569,
+      "learning_rate": 7.437508350321304e-07,
+      "loss": 0.4055,
+      "step": 10295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.893697784260606,
+      "learning_rate": 7.430662951193385e-07,
+      "loss": 0.4232,
+      "step": 10296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.797180435003266,
+      "learning_rate": 7.423820450849095e-07,
+      "loss": 0.4523,
+      "step": 10297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.4158920756938125,
+      "learning_rate": 7.416980849754413e-07,
+      "loss": 0.4459,
+      "step": 10298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.184064916815126,
+      "learning_rate": 7.410144148375026e-07,
+      "loss": 0.4383,
+      "step": 10299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.3973286760158645,
+      "learning_rate": 7.403310347176529e-07,
+      "loss": 0.4393,
+      "step": 10300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.19684901449279,
+      "learning_rate": 7.396479446624244e-07,
+      "loss": 0.6005,
+      "step": 10301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.881350129490674,
+      "learning_rate": 7.389651447183355e-07,
+      "loss": 0.463,
+      "step": 10302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8356424269715002,
+      "learning_rate": 7.3828263493188e-07,
+      "loss": 0.7784,
+      "step": 10303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.40780488323278,
+      "learning_rate": 7.376004153495342e-07,
+      "loss": 0.4416,
+      "step": 10304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.122924972573127,
+      "learning_rate": 7.369184860177525e-07,
+      "loss": 0.4129,
+      "step": 10305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.981247682749559,
+      "learning_rate": 7.362368469829745e-07,
+      "loss": 0.5068,
+      "step": 10306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.031190528486965,
+      "learning_rate": 7.355554982916141e-07,
+      "loss": 0.5973,
+      "step": 10307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.07959973825253,
+      "learning_rate": 7.348744399900698e-07,
+      "loss": 0.5174,
+      "step": 10308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 24.941193979196562,
+      "learning_rate": 7.341936721247184e-07,
+      "loss": 0.4086,
+      "step": 10309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.707588442902993,
+      "learning_rate": 7.335131947419161e-07,
+      "loss": 0.5532,
+      "step": 10310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.861077069715557,
+      "learning_rate": 7.328330078880019e-07,
+      "loss": 0.5181,
+      "step": 10311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.020368581159936,
+      "learning_rate": 7.321531116092928e-07,
+      "loss": 0.5644,
+      "step": 10312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.6183073039736735,
+      "learning_rate": 7.314735059520867e-07,
+      "loss": 0.4905,
+      "step": 10313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.0684528487208285,
+      "learning_rate": 7.307941909626603e-07,
+      "loss": 0.432,
+      "step": 10314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.538386773294028,
+      "learning_rate": 7.301151666872747e-07,
+      "loss": 0.471,
+      "step": 10315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.3699224692430905,
+      "learning_rate": 7.29436433172166e-07,
+      "loss": 0.4372,
+      "step": 10316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.610481393000784,
+      "learning_rate": 7.287579904635556e-07,
+      "loss": 0.454,
+      "step": 10317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.117286155135122,
+      "learning_rate": 7.280798386076405e-07,
+      "loss": 0.4644,
+      "step": 10318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.1594445469726,
+      "learning_rate": 7.274019776505992e-07,
+      "loss": 0.4095,
+      "step": 10319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.793946527622873,
+      "learning_rate": 7.267244076385938e-07,
+      "loss": 0.4641,
+      "step": 10320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.207747161363574,
+      "learning_rate": 7.260471286177611e-07,
+      "loss": 0.4524,
+      "step": 10321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.633174687187186,
+      "learning_rate": 7.253701406342229e-07,
+      "loss": 0.4133,
+      "step": 10322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.035010125474021,
+      "learning_rate": 7.246934437340764e-07,
+      "loss": 0.4558,
+      "step": 10323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.769781279943063,
+      "learning_rate": 7.240170379634043e-07,
+      "loss": 0.4798,
+      "step": 10324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.5545188231923905,
+      "learning_rate": 7.233409233682648e-07,
+      "loss": 0.5203,
+      "step": 10325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.054025838929874,
+      "learning_rate": 7.226650999947016e-07,
+      "loss": 0.4788,
+      "step": 10326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.358264501728286,
+      "learning_rate": 7.219895678887307e-07,
+      "loss": 0.4649,
+      "step": 10327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.883400466258075,
+      "learning_rate": 7.213143270963558e-07,
+      "loss": 0.4378,
+      "step": 10328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 37.657558503435894,
+      "learning_rate": 7.206393776635578e-07,
+      "loss": 0.4584,
+      "step": 10329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.78149968577508,
+      "learning_rate": 7.199647196362969e-07,
+      "loss": 0.5503,
+      "step": 10330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.742150217086569,
+      "learning_rate": 7.19290353060515e-07,
+      "loss": 0.5299,
+      "step": 10331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.187614618476711,
+      "learning_rate": 7.186162779821321e-07,
+      "loss": 0.4583,
+      "step": 10332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3989175948702206,
+      "learning_rate": 7.179424944470514e-07,
+      "loss": 0.4116,
+      "step": 10333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.825521031915665,
+      "learning_rate": 7.172690025011531e-07,
+      "loss": 0.3194,
+      "step": 10334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.022018640549585,
+      "learning_rate": 7.165958021903019e-07,
+      "loss": 0.6979,
+      "step": 10335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 14.239379232429181,
+      "learning_rate": 7.159228935603357e-07,
+      "loss": 0.4234,
+      "step": 10336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5489621561371625,
+      "learning_rate": 7.152502766570785e-07,
+      "loss": 0.6613,
+      "step": 10337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.727056096479397,
+      "learning_rate": 7.145779515263334e-07,
+      "loss": 0.4894,
+      "step": 10338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.677408246501608,
+      "learning_rate": 7.139059182138824e-07,
+      "loss": 0.4413,
+      "step": 10339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6171288566313935,
+      "learning_rate": 7.132341767654871e-07,
+      "loss": 0.444,
+      "step": 10340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.8092111113876985,
+      "learning_rate": 7.125627272268898e-07,
+      "loss": 0.4397,
+      "step": 10341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.962761852781529,
+      "learning_rate": 7.118915696438156e-07,
+      "loss": 0.4878,
+      "step": 10342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 25.998246926367777,
+      "learning_rate": 7.11220704061964e-07,
+      "loss": 0.389,
+      "step": 10343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.998661530016385,
+      "learning_rate": 7.105501305270212e-07,
+      "loss": 0.3088,
+      "step": 10344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.198627412517219,
+      "learning_rate": 7.098798490846492e-07,
+      "loss": 0.5053,
+      "step": 10345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8261309473980967,
+      "learning_rate": 7.092098597804903e-07,
+      "loss": 0.6914,
+      "step": 10346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.539851021415837,
+      "learning_rate": 7.085401626601679e-07,
+      "loss": 0.3469,
+      "step": 10347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.266609610168836,
+      "learning_rate": 7.078707577692867e-07,
+      "loss": 0.4568,
+      "step": 10348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 23.709863092381323,
+      "learning_rate": 7.072016451534291e-07,
+      "loss": 0.4265,
+      "step": 10349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.03238072148009,
+      "learning_rate": 7.065328248581582e-07,
+      "loss": 0.4103,
+      "step": 10350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 12.743704511909572,
+      "learning_rate": 7.058642969290197e-07,
+      "loss": 0.4121,
+      "step": 10351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.385691940066793,
+      "learning_rate": 7.051960614115356e-07,
+      "loss": 0.443,
+      "step": 10352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.899613897094804,
+      "learning_rate": 7.045281183512109e-07,
+      "loss": 0.5375,
+      "step": 10353
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.857702836385562,
+      "learning_rate": 7.038604677935296e-07,
+      "loss": 0.371,
+      "step": 10354
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.080631002390868,
+      "learning_rate": 7.031931097839545e-07,
+      "loss": 0.5231,
+      "step": 10355
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.050081048771366,
+      "learning_rate": 7.025260443679299e-07,
+      "loss": 0.4251,
+      "step": 10356
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.626016865996316,
+      "learning_rate": 7.018592715908818e-07,
+      "loss": 0.4758,
+      "step": 10357
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.926604838803196,
+      "learning_rate": 7.011927914982131e-07,
+      "loss": 0.5908,
+      "step": 10358
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.912545480800339,
+      "learning_rate": 7.00526604135307e-07,
+      "loss": 0.4682,
+      "step": 10359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.071025479858539,
+      "learning_rate": 6.998607095475307e-07,
+      "loss": 0.5638,
+      "step": 10360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.997971832651093,
+      "learning_rate": 6.991951077802256e-07,
+      "loss": 0.4255,
+      "step": 10361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.437117756411667,
+      "learning_rate": 6.985297988787193e-07,
+      "loss": 0.4538,
+      "step": 10362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.992437527731263,
+      "learning_rate": 6.978647828883151e-07,
+      "loss": 0.4718,
+      "step": 10363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.988786943990995,
+      "learning_rate": 6.972000598542972e-07,
+      "loss": 0.4423,
+      "step": 10364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.442190530910961,
+      "learning_rate": 6.965356298219294e-07,
+      "loss": 0.472,
+      "step": 10365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.28514540840861,
+      "learning_rate": 6.958714928364585e-07,
+      "loss": 0.5106,
+      "step": 10366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 16.641572767581476,
+      "learning_rate": 6.952076489431092e-07,
+      "loss": 0.4742,
+      "step": 10367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.035979747111728,
+      "learning_rate": 6.945440981870838e-07,
+      "loss": 0.4846,
+      "step": 10368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.9728649722583835,
+      "learning_rate": 6.938808406135705e-07,
+      "loss": 0.4232,
+      "step": 10369
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.531127454456704,
+      "learning_rate": 6.932178762677311e-07,
+      "loss": 0.5209,
+      "step": 10370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.279398089579307,
+      "learning_rate": 6.925552051947132e-07,
+      "loss": 0.3727,
+      "step": 10371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.589364081407366,
+      "learning_rate": 6.918928274396408e-07,
+      "loss": 0.4687,
+      "step": 10372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.578351571969703,
+      "learning_rate": 6.912307430476189e-07,
+      "loss": 0.5177,
+      "step": 10373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.577532358101966,
+      "learning_rate": 6.90568952063731e-07,
+      "loss": 0.5157,
+      "step": 10374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.803117734108216,
+      "learning_rate": 6.899074545330447e-07,
+      "loss": 0.5033,
+      "step": 10375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.805719047108606,
+      "learning_rate": 6.892462505006036e-07,
+      "loss": 0.4032,
+      "step": 10376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.845885675677648,
+      "learning_rate": 6.885853400114323e-07,
+      "loss": 0.4176,
+      "step": 10377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.929217505384955,
+      "learning_rate": 6.879247231105379e-07,
+      "loss": 0.4353,
+      "step": 10378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.117751749768837,
+      "learning_rate": 6.872643998429024e-07,
+      "loss": 0.5103,
+      "step": 10379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.475404845567762,
+      "learning_rate": 6.866043702534947e-07,
+      "loss": 0.4338,
+      "step": 10380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.890205691271916,
+      "learning_rate": 6.859446343872572e-07,
+      "loss": 0.4742,
+      "step": 10381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.566953220560184,
+      "learning_rate": 6.852851922891157e-07,
+      "loss": 0.3632,
+      "step": 10382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.54953099054524,
+      "learning_rate": 6.846260440039743e-07,
+      "loss": 0.5114,
+      "step": 10383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.943801153802767,
+      "learning_rate": 6.839671895767203e-07,
+      "loss": 0.513,
+      "step": 10384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.963805576397885,
+      "learning_rate": 6.833086290522173e-07,
+      "loss": 0.4632,
+      "step": 10385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9361121423502894,
+      "learning_rate": 6.826503624753111e-07,
+      "loss": 0.371,
+      "step": 10386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.904596700545476,
+      "learning_rate": 6.819923898908249e-07,
+      "loss": 0.4206,
+      "step": 10387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8595788743160906,
+      "learning_rate": 6.813347113435648e-07,
+      "loss": 0.6515,
+      "step": 10388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.836690637415546,
+      "learning_rate": 6.806773268783174e-07,
+      "loss": 0.4945,
+      "step": 10389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.06885486067001,
+      "learning_rate": 6.800202365398462e-07,
+      "loss": 0.7055,
+      "step": 10390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.601018204367996,
+      "learning_rate": 6.79363440372896e-07,
+      "loss": 0.448,
+      "step": 10391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.671560121220129,
+      "learning_rate": 6.787069384221911e-07,
+      "loss": 0.4388,
+      "step": 10392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.63573861172682,
+      "learning_rate": 6.780507307324385e-07,
+      "loss": 0.4477,
+      "step": 10393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.741660385453111,
+      "learning_rate": 6.773948173483219e-07,
+      "loss": 0.4387,
+      "step": 10394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.400496982843,
+      "learning_rate": 6.767391983145061e-07,
+      "loss": 0.5343,
+      "step": 10395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.16970292300231,
+      "learning_rate": 6.760838736756348e-07,
+      "loss": 0.4664,
+      "step": 10396
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.145728469134591,
+      "learning_rate": 6.754288434763345e-07,
+      "loss": 0.3927,
+      "step": 10397
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.404368379823699,
+      "learning_rate": 6.74774107761208e-07,
+      "loss": 0.6051,
+      "step": 10398
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.533545349915169,
+      "learning_rate": 6.741196665748418e-07,
+      "loss": 0.5661,
+      "step": 10399
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.09605620001057,
+      "learning_rate": 6.734655199618001e-07,
+      "loss": 0.4615,
+      "step": 10400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7996740899303034,
+      "learning_rate": 6.728116679666258e-07,
+      "loss": 0.7601,
+      "step": 10401
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.738549414482861,
+      "learning_rate": 6.721581106338448e-07,
+      "loss": 0.4375,
+      "step": 10402
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.0773029084228645,
+      "learning_rate": 6.715048480079606e-07,
+      "loss": 0.3939,
+      "step": 10403
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.857756061104507,
+      "learning_rate": 6.708518801334601e-07,
+      "loss": 0.4645,
+      "step": 10404
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.634884462123731,
+      "learning_rate": 6.701992070548036e-07,
+      "loss": 0.4216,
+      "step": 10405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.898033081493702,
+      "learning_rate": 6.695468288164375e-07,
+      "loss": 0.4458,
+      "step": 10406
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.718605948963959,
+      "learning_rate": 6.688947454627848e-07,
+      "loss": 0.4221,
+      "step": 10407
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.455553306111614,
+      "learning_rate": 6.682429570382509e-07,
+      "loss": 0.3789,
+      "step": 10408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.076554799121526,
+      "learning_rate": 6.675914635872194e-07,
+      "loss": 0.468,
+      "step": 10409
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.799466577086512,
+      "learning_rate": 6.669402651540524e-07,
+      "loss": 0.5108,
+      "step": 10410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.960082354720697,
+      "learning_rate": 6.662893617830956e-07,
+      "loss": 0.4637,
+      "step": 10411
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.066840895352672,
+      "learning_rate": 6.656387535186714e-07,
+      "loss": 0.4917,
+      "step": 10412
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.614944958391556,
+      "learning_rate": 6.649884404050854e-07,
+      "loss": 0.4615,
+      "step": 10413
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.961426536974871,
+      "learning_rate": 6.643384224866179e-07,
+      "loss": 0.4567,
+      "step": 10414
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.961723668420323,
+      "learning_rate": 6.636886998075348e-07,
+      "loss": 0.3917,
+      "step": 10415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.54925858376471,
+      "learning_rate": 6.630392724120777e-07,
+      "loss": 0.4837,
+      "step": 10416
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.903104640371842,
+      "learning_rate": 6.623901403444711e-07,
+      "loss": 0.4668,
+      "step": 10417
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.819241628806296,
+      "learning_rate": 6.617413036489178e-07,
+      "loss": 0.49,
+      "step": 10418
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.877263096452448,
+      "learning_rate": 6.610927623695989e-07,
+      "loss": 0.485,
+      "step": 10419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.902262995217633,
+      "learning_rate": 6.604445165506795e-07,
+      "loss": 0.423,
+      "step": 10420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.46277417799673,
+      "learning_rate": 6.597965662363004e-07,
+      "loss": 0.5634,
+      "step": 10421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.0928206569027274,
+      "learning_rate": 6.591489114705879e-07,
+      "loss": 0.3962,
+      "step": 10422
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.644519373885853,
+      "learning_rate": 6.585015522976391e-07,
+      "loss": 0.4484,
+      "step": 10423
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.38200643162342,
+      "learning_rate": 6.578544887615395e-07,
+      "loss": 0.4331,
+      "step": 10424
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5790643445621144,
+      "learning_rate": 6.572077209063499e-07,
+      "loss": 0.6697,
+      "step": 10425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.20860314569165,
+      "learning_rate": 6.565612487761142e-07,
+      "loss": 0.4202,
+      "step": 10426
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.373508338834129,
+      "learning_rate": 6.559150724148528e-07,
+      "loss": 0.3884,
+      "step": 10427
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.127403023287075,
+      "learning_rate": 6.552691918665665e-07,
+      "loss": 0.4336,
+      "step": 10428
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.564735750919693,
+      "learning_rate": 6.546236071752399e-07,
+      "loss": 0.4196,
+      "step": 10429
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.363927999130109,
+      "learning_rate": 6.539783183848313e-07,
+      "loss": 0.4329,
+      "step": 10430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.793828327410748,
+      "learning_rate": 6.53333325539286e-07,
+      "loss": 0.4253,
+      "step": 10431
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 15.828882838160908,
+      "learning_rate": 6.526886286825201e-07,
+      "loss": 0.5414,
+      "step": 10432
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 49.610585017670225,
+      "learning_rate": 6.520442278584382e-07,
+      "loss": 0.4746,
+      "step": 10433
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7842474538733941,
+      "learning_rate": 6.514001231109196e-07,
+      "loss": 0.7361,
+      "step": 10434
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.481754580210738,
+      "learning_rate": 6.507563144838259e-07,
+      "loss": 0.4839,
+      "step": 10435
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.048944534120905,
+      "learning_rate": 6.501128020209974e-07,
+      "loss": 0.4873,
+      "step": 10436
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.18381810302493,
+      "learning_rate": 6.494695857662541e-07,
+      "loss": 0.4507,
+      "step": 10437
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 23.795221319176193,
+      "learning_rate": 6.488266657633957e-07,
+      "loss": 0.44,
+      "step": 10438
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.177167194125789,
+      "learning_rate": 6.481840420562025e-07,
+      "loss": 0.4658,
+      "step": 10439
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.085546552305263,
+      "learning_rate": 6.475417146884367e-07,
+      "loss": 0.3795,
+      "step": 10440
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.106987846158051,
+      "learning_rate": 6.468996837038338e-07,
+      "loss": 0.6599,
+      "step": 10441
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.11061882799167,
+      "learning_rate": 6.462579491461169e-07,
+      "loss": 0.4724,
+      "step": 10442
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.450664556198857,
+      "learning_rate": 6.456165110589818e-07,
+      "loss": 0.4458,
+      "step": 10443
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.036159816627623,
+      "learning_rate": 6.449753694861117e-07,
+      "loss": 0.4719,
+      "step": 10444
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.640198128217225,
+      "learning_rate": 6.443345244711629e-07,
+      "loss": 0.5371,
+      "step": 10445
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.9048793364893415,
+      "learning_rate": 6.436939760577743e-07,
+      "loss": 0.3962,
+      "step": 10446
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.134299122800509,
+      "learning_rate": 6.43053724289564e-07,
+      "loss": 0.558,
+      "step": 10447
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.878904770515035,
+      "learning_rate": 6.424137692101313e-07,
+      "loss": 0.406,
+      "step": 10448
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.45600773697066,
+      "learning_rate": 6.417741108630543e-07,
+      "loss": 0.3575,
+      "step": 10449
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.429540583375902,
+      "learning_rate": 6.411347492918912e-07,
+      "loss": 0.421,
+      "step": 10450
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.02694991510082,
+      "learning_rate": 6.404956845401794e-07,
+      "loss": 0.6605,
+      "step": 10451
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.72749692016468,
+      "learning_rate": 6.398569166514346e-07,
+      "loss": 0.4317,
+      "step": 10452
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.344754980063061,
+      "learning_rate": 6.392184456691569e-07,
+      "loss": 0.3977,
+      "step": 10453
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.070677224811537,
+      "learning_rate": 6.385802716368217e-07,
+      "loss": 0.425,
+      "step": 10454
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.216642006076766,
+      "learning_rate": 6.379423945978864e-07,
+      "loss": 0.4047,
+      "step": 10455
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.057270079530803,
+      "learning_rate": 6.373048145957867e-07,
+      "loss": 0.41,
+      "step": 10456
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.0764699348909135,
+      "learning_rate": 6.366675316739401e-07,
+      "loss": 0.4207,
+      "step": 10457
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.471625908486779,
+      "learning_rate": 6.360305458757415e-07,
+      "loss": 0.5192,
+      "step": 10458
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.65769514783094,
+      "learning_rate": 6.353938572445684e-07,
+      "loss": 0.4707,
+      "step": 10459
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.9680495247418,
+      "learning_rate": 6.34757465823776e-07,
+      "loss": 0.3516,
+      "step": 10460
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.709621256499326,
+      "learning_rate": 6.341213716566985e-07,
+      "loss": 0.4636,
+      "step": 10461
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.730077981754459,
+      "learning_rate": 6.334855747866531e-07,
+      "loss": 0.4342,
+      "step": 10462
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.393925330807606,
+      "learning_rate": 6.328500752569333e-07,
+      "loss": 0.4195,
+      "step": 10463
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.441658316889725,
+      "learning_rate": 6.322148731108147e-07,
+      "loss": 0.48,
+      "step": 10464
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.657717763852207,
+      "learning_rate": 6.315799683915496e-07,
+      "loss": 0.5384,
+      "step": 10465
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.018945649799478,
+      "learning_rate": 6.309453611423755e-07,
+      "loss": 0.3502,
+      "step": 10466
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.233374181385393,
+      "learning_rate": 6.303110514065036e-07,
+      "loss": 0.4367,
+      "step": 10467
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.576343370994917,
+      "learning_rate": 6.296770392271296e-07,
+      "loss": 0.408,
+      "step": 10468
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.50484053446716,
+      "learning_rate": 6.290433246474259e-07,
+      "loss": 0.5014,
+      "step": 10469
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.871930058105138,
+      "learning_rate": 6.284099077105448e-07,
+      "loss": 0.4058,
+      "step": 10470
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.332588504834748,
+      "learning_rate": 6.277767884596214e-07,
+      "loss": 0.4484,
+      "step": 10471
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.018533776908991,
+      "learning_rate": 6.271439669377671e-07,
+      "loss": 0.5253,
+      "step": 10472
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.7858378987512324,
+      "learning_rate": 6.265114431880742e-07,
+      "loss": 0.4146,
+      "step": 10473
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.059601042655727,
+      "learning_rate": 6.258792172536137e-07,
+      "loss": 0.407,
+      "step": 10474
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.745302213653109,
+      "learning_rate": 6.252472891774397e-07,
+      "loss": 0.406,
+      "step": 10475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.44450002631151,
+      "learning_rate": 6.246156590025815e-07,
+      "loss": 0.3869,
+      "step": 10476
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.275719569753547,
+      "learning_rate": 6.23984326772053e-07,
+      "loss": 0.5188,
+      "step": 10477
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4647143602298516,
+      "learning_rate": 6.233532925288422e-07,
+      "loss": 0.3025,
+      "step": 10478
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6333310009587114,
+      "learning_rate": 6.227225563159206e-07,
+      "loss": 0.7024,
+      "step": 10479
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.369714968415725,
+      "learning_rate": 6.220921181762396e-07,
+      "loss": 0.4772,
+      "step": 10480
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.472240697679355,
+      "learning_rate": 6.214619781527293e-07,
+      "loss": 0.3879,
+      "step": 10481
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.66462377309414,
+      "learning_rate": 6.208321362882985e-07,
+      "loss": 0.4666,
+      "step": 10482
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.329097465948058,
+      "learning_rate": 6.202025926258359e-07,
+      "loss": 0.4244,
+      "step": 10483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7902124542421685,
+      "learning_rate": 6.195733472082127e-07,
+      "loss": 0.741,
+      "step": 10484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.494432957112424,
+      "learning_rate": 6.189444000782752e-07,
+      "loss": 0.3316,
+      "step": 10485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.9638002063396565,
+      "learning_rate": 6.183157512788551e-07,
+      "loss": 0.3813,
+      "step": 10486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.230527884068608,
+      "learning_rate": 6.176874008527583e-07,
+      "loss": 0.4893,
+      "step": 10487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.477796505538793,
+      "learning_rate": 6.170593488427728e-07,
+      "loss": 0.3981,
+      "step": 10488
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.875107521617704,
+      "learning_rate": 6.164315952916677e-07,
+      "loss": 0.4985,
+      "step": 10489
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9753191717632488,
+      "learning_rate": 6.158041402421888e-07,
+      "loss": 0.7172,
+      "step": 10490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.515297873856066,
+      "learning_rate": 6.151769837370636e-07,
+      "loss": 0.4744,
+      "step": 10491
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.5947307441104455,
+      "learning_rate": 6.145501258189979e-07,
+      "loss": 0.4355,
+      "step": 10492
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.571888597590798,
+      "learning_rate": 6.139235665306786e-07,
+      "loss": 0.4754,
+      "step": 10493
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.699481595849329,
+      "learning_rate": 6.132973059147712e-07,
+      "loss": 0.4645,
+      "step": 10494
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.6090012196192,
+      "learning_rate": 6.126713440139226e-07,
+      "loss": 0.5443,
+      "step": 10495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.346696421477366,
+      "learning_rate": 6.120456808707575e-07,
+      "loss": 0.4894,
+      "step": 10496
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.273189595110079,
+      "learning_rate": 6.114203165278798e-07,
+      "loss": 0.4114,
+      "step": 10497
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.022577290644518,
+      "learning_rate": 6.107952510278736e-07,
+      "loss": 0.4698,
+      "step": 10498
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.773585516970096,
+      "learning_rate": 6.101704844133055e-07,
+      "loss": 0.4261,
+      "step": 10499
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.326206586332924,
+      "learning_rate": 6.095460167267175e-07,
+      "loss": 0.4971,
+      "step": 10500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.961090408207323,
+      "learning_rate": 6.08921848010633e-07,
+      "loss": 0.3934,
+      "step": 10501
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.975215670234234,
+      "learning_rate": 6.082979783075565e-07,
+      "loss": 0.3904,
+      "step": 10502
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.28384711478378,
+      "learning_rate": 6.076744076599688e-07,
+      "loss": 0.3587,
+      "step": 10503
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.411003772937734,
+      "learning_rate": 6.070511361103348e-07,
+      "loss": 0.324,
+      "step": 10504
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.681954790801855,
+      "learning_rate": 6.064281637010954e-07,
+      "loss": 0.4361,
+      "step": 10505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.45762604595088,
+      "learning_rate": 6.058054904746713e-07,
+      "loss": 0.4495,
+      "step": 10506
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.369677309095275,
+      "learning_rate": 6.05183116473464e-07,
+      "loss": 0.4039,
+      "step": 10507
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.024507325705816,
+      "learning_rate": 6.04561041739855e-07,
+      "loss": 0.4596,
+      "step": 10508
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.464954700552731,
+      "learning_rate": 6.039392663162069e-07,
+      "loss": 0.4589,
+      "step": 10509
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.439730177993359,
+      "learning_rate": 6.033177902448556e-07,
+      "loss": 0.4181,
+      "step": 10510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.86057477051127,
+      "learning_rate": 6.026966135681239e-07,
+      "loss": 0.3628,
+      "step": 10511
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.666350894404334,
+      "learning_rate": 6.020757363283098e-07,
+      "loss": 0.4504,
+      "step": 10512
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.315913861145313,
+      "learning_rate": 6.014551585676931e-07,
+      "loss": 0.428,
+      "step": 10513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 17.88735693660534,
+      "learning_rate": 6.00834880328533e-07,
+      "loss": 0.4327,
+      "step": 10514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 22.51094004791489,
+      "learning_rate": 6.002149016530662e-07,
+      "loss": 0.4953,
+      "step": 10515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.449801683543392,
+      "learning_rate": 5.9959522258351e-07,
+      "loss": 0.4884,
+      "step": 10516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.730961833493761,
+      "learning_rate": 5.989758431620646e-07,
+      "loss": 0.4281,
+      "step": 10517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.915384938437162,
+      "learning_rate": 5.983567634309051e-07,
+      "loss": 0.4116,
+      "step": 10518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.415743549982414,
+      "learning_rate": 5.977379834321867e-07,
+      "loss": 0.4741,
+      "step": 10519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.459178062831374,
+      "learning_rate": 5.97119503208049e-07,
+      "loss": 0.4778,
+      "step": 10520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.862146562694111,
+      "learning_rate": 5.965013228006045e-07,
+      "loss": 0.5099,
+      "step": 10521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.4952218393571615,
+      "learning_rate": 5.958834422519516e-07,
+      "loss": 0.4924,
+      "step": 10522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.446890860286992,
+      "learning_rate": 5.952658616041635e-07,
+      "loss": 0.4784,
+      "step": 10523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6503021061553684,
+      "learning_rate": 5.946485808992947e-07,
+      "loss": 0.6038,
+      "step": 10524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.28669604833972,
+      "learning_rate": 5.940316001793783e-07,
+      "loss": 0.4145,
+      "step": 10525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 34.24830137411749,
+      "learning_rate": 5.934149194864303e-07,
+      "loss": 0.4233,
+      "step": 10526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.968731316103216,
+      "learning_rate": 5.927985388624424e-07,
+      "loss": 0.4472,
+      "step": 10527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.791979088393134,
+      "learning_rate": 5.921824583493885e-07,
+      "loss": 0.3858,
+      "step": 10528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.177924549723611,
+      "learning_rate": 5.915666779892188e-07,
+      "loss": 0.523,
+      "step": 10529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.390426889101976,
+      "learning_rate": 5.909511978238663e-07,
+      "loss": 0.3863,
+      "step": 10530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.138529311604701,
+      "learning_rate": 5.903360178952445e-07,
+      "loss": 0.5451,
+      "step": 10531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.041794326917803,
+      "learning_rate": 5.897211382452423e-07,
+      "loss": 0.4309,
+      "step": 10532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.681816330078794,
+      "learning_rate": 5.891065589157308e-07,
+      "loss": 0.4927,
+      "step": 10533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.452042254101567,
+      "learning_rate": 5.884922799485593e-07,
+      "loss": 0.6295,
+      "step": 10534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.704497220511971,
+      "learning_rate": 5.878783013855588e-07,
+      "loss": 0.4086,
+      "step": 10535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.434915643559609,
+      "learning_rate": 5.872646232685381e-07,
+      "loss": 0.4644,
+      "step": 10536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.361015864223399,
+      "learning_rate": 5.866512456392865e-07,
+      "loss": 0.3908,
+      "step": 10537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.8143472091678445,
+      "learning_rate": 5.8603816853957e-07,
+      "loss": 0.5001,
+      "step": 10538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.232346943962196,
+      "learning_rate": 5.85425392011138e-07,
+      "loss": 0.3843,
+      "step": 10539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.362148915737216,
+      "learning_rate": 5.848129160957189e-07,
+      "loss": 0.4249,
+      "step": 10540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.454817369321631,
+      "learning_rate": 5.842007408350193e-07,
+      "loss": 0.4936,
+      "step": 10541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.090600084519228,
+      "learning_rate": 5.835888662707245e-07,
+      "loss": 0.4989,
+      "step": 10542
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.264424627550849,
+      "learning_rate": 5.829772924444999e-07,
+      "loss": 0.3662,
+      "step": 10543
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.2413343676081885,
+      "learning_rate": 5.823660193979929e-07,
+      "loss": 0.4524,
+      "step": 10544
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.649777965859831,
+      "learning_rate": 5.817550471728267e-07,
+      "loss": 0.4595,
+      "step": 10545
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.122972532895861,
+      "learning_rate": 5.811443758106089e-07,
+      "loss": 0.3721,
+      "step": 10546
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.242410386452244,
+      "learning_rate": 5.80534005352919e-07,
+      "loss": 0.4029,
+      "step": 10547
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.234234257234384,
+      "learning_rate": 5.799239358413244e-07,
+      "loss": 0.4208,
+      "step": 10548
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.509007501815793,
+      "learning_rate": 5.793141673173652e-07,
+      "loss": 0.5254,
+      "step": 10549
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.021088863728215,
+      "learning_rate": 5.787046998225665e-07,
+      "loss": 0.4278,
+      "step": 10550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.067028903803406,
+      "learning_rate": 5.78095533398429e-07,
+      "loss": 0.4277,
+      "step": 10551
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8735426017753505,
+      "learning_rate": 5.774866680864333e-07,
+      "loss": 0.3431,
+      "step": 10552
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.951126589284329,
+      "learning_rate": 5.76878103928043e-07,
+      "loss": 0.4452,
+      "step": 10553
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.023814174653067,
+      "learning_rate": 5.762698409646961e-07,
+      "loss": 0.422,
+      "step": 10554
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.21191483433985,
+      "learning_rate": 5.756618792378155e-07,
+      "loss": 0.4562,
+      "step": 10555
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.03206063609711,
+      "learning_rate": 5.75054218788797e-07,
+      "loss": 0.5017,
+      "step": 10556
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.326253766500847,
+      "learning_rate": 5.744468596590224e-07,
+      "loss": 0.4164,
+      "step": 10557
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.903919900398411,
+      "learning_rate": 5.738398018898483e-07,
+      "loss": 0.4991,
+      "step": 10558
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.862560254243162,
+      "learning_rate": 5.732330455226142e-07,
+      "loss": 0.3662,
+      "step": 10559
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.943875648800218,
+      "learning_rate": 5.726265905986378e-07,
+      "loss": 0.4646,
+      "step": 10560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.868800781021657,
+      "learning_rate": 5.720204371592136e-07,
+      "loss": 0.4242,
+      "step": 10561
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 27.910364393570422,
+      "learning_rate": 5.714145852456204e-07,
+      "loss": 0.4415,
+      "step": 10562
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.407812159097988,
+      "learning_rate": 5.708090348991124e-07,
+      "loss": 0.3241,
+      "step": 10563
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.119516461166051,
+      "learning_rate": 5.702037861609277e-07,
+      "loss": 0.4211,
+      "step": 10564
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.73527707831943,
+      "learning_rate": 5.695988390722767e-07,
+      "loss": 0.5371,
+      "step": 10565
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.965301486136835,
+      "learning_rate": 5.68994193674357e-07,
+      "loss": 0.5002,
+      "step": 10566
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.054314984773207,
+      "learning_rate": 5.683898500083407e-07,
+      "loss": 0.3702,
+      "step": 10567
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.320257426155811,
+      "learning_rate": 5.677858081153819e-07,
+      "loss": 0.5924,
+      "step": 10568
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.484642592162053,
+      "learning_rate": 5.67182068036613e-07,
+      "loss": 0.3776,
+      "step": 10569
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.050630478167138,
+      "learning_rate": 5.665786298131448e-07,
+      "loss": 0.3754,
+      "step": 10570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 18.115932645013082,
+      "learning_rate": 5.659754934860706e-07,
+      "loss": 0.488,
+      "step": 10571
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.431850160234852,
+      "learning_rate": 5.653726590964592e-07,
+      "loss": 0.409,
+      "step": 10572
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.174932440008337,
+      "learning_rate": 5.647701266853645e-07,
+      "loss": 0.5219,
+      "step": 10573
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.664813134167478,
+      "learning_rate": 5.641678962938118e-07,
+      "loss": 0.4709,
+      "step": 10574
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 32.44358881370683,
+      "learning_rate": 5.635659679628136e-07,
+      "loss": 0.418,
+      "step": 10575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.949338494733757,
+      "learning_rate": 5.629643417333563e-07,
+      "loss": 0.4303,
+      "step": 10576
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.901421154271128,
+      "learning_rate": 5.623630176464107e-07,
+      "loss": 0.4862,
+      "step": 10577
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.062829005795516,
+      "learning_rate": 5.617619957429227e-07,
+      "loss": 0.3794,
+      "step": 10578
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.168940648348184,
+      "learning_rate": 5.61161276063818e-07,
+      "loss": 0.4372,
+      "step": 10579
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.259166323053457,
+      "learning_rate": 5.60560858650005e-07,
+      "loss": 0.3954,
+      "step": 10580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.200120828307264,
+      "learning_rate": 5.599607435423682e-07,
+      "loss": 0.2692,
+      "step": 10581
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.524330030364089,
+      "learning_rate": 5.593609307817743e-07,
+      "loss": 0.5801,
+      "step": 10582
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.988294681276513,
+      "learning_rate": 5.587614204090669e-07,
+      "loss": 0.464,
+      "step": 10583
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.787987217801216,
+      "learning_rate": 5.581622124650698e-07,
+      "loss": 0.3898,
+      "step": 10584
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.780380849694485,
+      "learning_rate": 5.575633069905856e-07,
+      "loss": 0.4272,
+      "step": 10585
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.362730329552003,
+      "learning_rate": 5.569647040263993e-07,
+      "loss": 0.4845,
+      "step": 10586
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.113769865592512,
+      "learning_rate": 5.56366403613272e-07,
+      "loss": 0.4353,
+      "step": 10587
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.826833235031161,
+      "learning_rate": 5.557684057919455e-07,
+      "loss": 0.4248,
+      "step": 10588
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.362556089977003,
+      "learning_rate": 5.551707106031395e-07,
+      "loss": 0.3486,
+      "step": 10589
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.797618355869321,
+      "learning_rate": 5.545733180875551e-07,
+      "loss": 0.3773,
+      "step": 10590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.01107611142629,
+      "learning_rate": 5.53976228285874e-07,
+      "loss": 0.5514,
+      "step": 10591
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.162456583203287,
+      "learning_rate": 5.533794412387533e-07,
+      "loss": 0.4487,
+      "step": 10592
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.0933716019434,
+      "learning_rate": 5.527829569868326e-07,
+      "loss": 0.5143,
+      "step": 10593
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.231545213481572,
+      "learning_rate": 5.521867755707277e-07,
+      "loss": 0.5251,
+      "step": 10594
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5084506673489155,
+      "learning_rate": 5.51590897031039e-07,
+      "loss": 0.5521,
+      "step": 10595
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.490522266137951,
+      "learning_rate": 5.509953214083418e-07,
+      "loss": 0.5126,
+      "step": 10596
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.445877556573639,
+      "learning_rate": 5.504000487431927e-07,
+      "loss": 0.4749,
+      "step": 10597
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.378046362118583,
+      "learning_rate": 5.498050790761245e-07,
+      "loss": 0.387,
+      "step": 10598
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.8209488686443,
+      "learning_rate": 5.492104124476549e-07,
+      "loss": 0.4755,
+      "step": 10599
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.127392756351078,
+      "learning_rate": 5.486160488982778e-07,
+      "loss": 0.4397,
+      "step": 10600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 16.125959966733014,
+      "learning_rate": 5.480219884684662e-07,
+      "loss": 0.4068,
+      "step": 10601
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.448959535432984,
+      "learning_rate": 5.474282311986728e-07,
+      "loss": 0.5494,
+      "step": 10602
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.989058332181912,
+      "learning_rate": 5.468347771293292e-07,
+      "loss": 0.4001,
+      "step": 10603
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 41.88134908038953,
+      "learning_rate": 5.462416263008491e-07,
+      "loss": 0.4241,
+      "step": 10604
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.63895812637827,
+      "learning_rate": 5.456487787536219e-07,
+      "loss": 0.4269,
+      "step": 10605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.2719498357567645,
+      "learning_rate": 5.45056234528018e-07,
+      "loss": 0.4665,
+      "step": 10606
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.423057318124826,
+      "learning_rate": 5.444639936643858e-07,
+      "loss": 0.3205,
+      "step": 10607
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.163844914709499,
+      "learning_rate": 5.438720562030575e-07,
+      "loss": 0.4926,
+      "step": 10608
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.114288388871236,
+      "learning_rate": 5.43280422184338e-07,
+      "loss": 0.4957,
+      "step": 10609
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.5939921924652785,
+      "learning_rate": 5.426890916485178e-07,
+      "loss": 0.4516,
+      "step": 10610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.369970996118125,
+      "learning_rate": 5.420980646358626e-07,
+      "loss": 0.4497,
+      "step": 10611
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.240005504554558,
+      "learning_rate": 5.415073411866178e-07,
+      "loss": 0.4027,
+      "step": 10612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.064222455010462,
+      "learning_rate": 5.409169213410115e-07,
+      "loss": 0.4822,
+      "step": 10613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.9583258998634685,
+      "learning_rate": 5.403268051392463e-07,
+      "loss": 0.5435,
+      "step": 10614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5485836200931624,
+      "learning_rate": 5.397369926215079e-07,
+      "loss": 0.644,
+      "step": 10615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.746410383735796,
+      "learning_rate": 5.391474838279586e-07,
+      "loss": 0.4453,
+      "step": 10616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.9914578534819825,
+      "learning_rate": 5.385582787987431e-07,
+      "loss": 0.6038,
+      "step": 10617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 20.74751637206439,
+      "learning_rate": 5.379693775739814e-07,
+      "loss": 0.482,
+      "step": 10618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.80974097161534,
+      "learning_rate": 5.37380780193778e-07,
+      "loss": 0.2976,
+      "step": 10619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.389139012527421,
+      "learning_rate": 5.367924866982116e-07,
+      "loss": 0.3923,
+      "step": 10620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.492018880228164,
+      "learning_rate": 5.362044971273422e-07,
+      "loss": 0.4659,
+      "step": 10621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.093066921895723,
+      "learning_rate": 5.356168115212112e-07,
+      "loss": 0.5049,
+      "step": 10622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.622516039092203,
+      "learning_rate": 5.35029429919836e-07,
+      "loss": 0.4337,
+      "step": 10623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.179845443198539,
+      "learning_rate": 5.344423523632153e-07,
+      "loss": 0.5001,
+      "step": 10624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.215527590636626,
+      "learning_rate": 5.338555788913247e-07,
+      "loss": 0.5096,
+      "step": 10625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4089754290698835,
+      "learning_rate": 5.332691095441234e-07,
+      "loss": 0.6809,
+      "step": 10626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.055060217066554,
+      "learning_rate": 5.326829443615456e-07,
+      "loss": 0.4361,
+      "step": 10627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.968131468454415,
+      "learning_rate": 5.320970833835082e-07,
+      "loss": 0.459,
+      "step": 10628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.351250541333334,
+      "learning_rate": 5.315115266499043e-07,
+      "loss": 0.4991,
+      "step": 10629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.795212865861329,
+      "learning_rate": 5.309262742006077e-07,
+      "loss": 0.382,
+      "step": 10630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.504829094760015,
+      "learning_rate": 5.303413260754731e-07,
+      "loss": 0.4696,
+      "step": 10631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.053142007994412,
+      "learning_rate": 5.297566823143313e-07,
+      "loss": 0.5184,
+      "step": 10632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.923054437262971,
+      "learning_rate": 5.291723429569951e-07,
+      "loss": 0.441,
+      "step": 10633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.410372291221869,
+      "learning_rate": 5.28588308043253e-07,
+      "loss": 0.5562,
+      "step": 10634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7671284399429794,
+      "learning_rate": 5.280045776128784e-07,
+      "loss": 0.6798,
+      "step": 10635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.316191102145133,
+      "learning_rate": 5.274211517056182e-07,
+      "loss": 0.4393,
+      "step": 10636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.995550665562453,
+      "learning_rate": 5.268380303612031e-07,
+      "loss": 0.3847,
+      "step": 10637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.7045790612007705,
+      "learning_rate": 5.262552136193405e-07,
+      "loss": 0.4685,
+      "step": 10638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.376571351745271,
+      "learning_rate": 5.256727015197166e-07,
+      "loss": 0.4468,
+      "step": 10639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.300894802956708,
+      "learning_rate": 5.250904941019991e-07,
+      "loss": 0.3768,
+      "step": 10640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.515710395059223,
+      "learning_rate": 5.245085914058329e-07,
+      "loss": 0.5589,
+      "step": 10641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.882667397048252,
+      "learning_rate": 5.239269934708452e-07,
+      "loss": 0.5009,
+      "step": 10642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.161619338729131,
+      "learning_rate": 5.233457003366365e-07,
+      "loss": 0.4496,
+      "step": 10643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.1806494694770295,
+      "learning_rate": 5.227647120427932e-07,
+      "loss": 0.3351,
+      "step": 10644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.265138139181818,
+      "learning_rate": 5.221840286288759e-07,
+      "loss": 0.4447,
+      "step": 10645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.713694450230529,
+      "learning_rate": 5.21603650134429e-07,
+      "loss": 0.4679,
+      "step": 10646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.355049034828219,
+      "learning_rate": 5.210235765989729e-07,
+      "loss": 0.5468,
+      "step": 10647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.344954167671103,
+      "learning_rate": 5.204438080620073e-07,
+      "loss": 0.3945,
+      "step": 10648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.325433295276057,
+      "learning_rate": 5.198643445630109e-07,
+      "loss": 0.4377,
+      "step": 10649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.689651919976178,
+      "learning_rate": 5.192851861414444e-07,
+      "loss": 0.5468,
+      "step": 10650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.28568886575952,
+      "learning_rate": 5.18706332836747e-07,
+      "loss": 0.4286,
+      "step": 10651
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.467971309791951,
+      "learning_rate": 5.181277846883326e-07,
+      "loss": 0.5122,
+      "step": 10652
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.275414118062395,
+      "learning_rate": 5.175495417356008e-07,
+      "loss": 0.5242,
+      "step": 10653
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.274336329983204,
+      "learning_rate": 5.169716040179251e-07,
+      "loss": 0.3001,
+      "step": 10654
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.090062017225855,
+      "learning_rate": 5.16393971574663e-07,
+      "loss": 0.5008,
+      "step": 10655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.175238684782013,
+      "learning_rate": 5.158166444451468e-07,
+      "loss": 0.4415,
+      "step": 10656
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.181327465621119,
+      "learning_rate": 5.152396226686907e-07,
+      "loss": 0.4764,
+      "step": 10657
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.1079277213990535,
+      "learning_rate": 5.146629062845865e-07,
+      "loss": 0.4756,
+      "step": 10658
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 18.745021599053356,
+      "learning_rate": 5.140864953321073e-07,
+      "loss": 0.3506,
+      "step": 10659
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.661401346128813,
+      "learning_rate": 5.135103898505034e-07,
+      "loss": 0.4279,
+      "step": 10660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.572166725976806,
+      "learning_rate": 5.129345898790044e-07,
+      "loss": 0.5078,
+      "step": 10661
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.842061043437645,
+      "learning_rate": 5.123590954568214e-07,
+      "loss": 0.4451,
+      "step": 10662
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.679483436027319,
+      "learning_rate": 5.117839066231412e-07,
+      "loss": 0.4783,
+      "step": 10663
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 32.83290081896928,
+      "learning_rate": 5.11209023417133e-07,
+      "loss": 0.504,
+      "step": 10664
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.729548267299872,
+      "learning_rate": 5.106344458779439e-07,
+      "loss": 0.4881,
+      "step": 10665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.573321824176097,
+      "learning_rate": 5.100601740446993e-07,
+      "loss": 0.438,
+      "step": 10666
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.657829838516268,
+      "learning_rate": 5.094862079565033e-07,
+      "loss": 0.3788,
+      "step": 10667
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.488232711893151,
+      "learning_rate": 5.089125476524436e-07,
+      "loss": 0.5003,
+      "step": 10668
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.659660114973933,
+      "learning_rate": 5.08339193171582e-07,
+      "loss": 0.5548,
+      "step": 10669
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.630346281583237,
+      "learning_rate": 5.077661445529603e-07,
+      "loss": 0.5039,
+      "step": 10670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.648789216753675,
+      "learning_rate": 5.071934018356034e-07,
+      "loss": 0.3663,
+      "step": 10671
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 16.027148787306245,
+      "learning_rate": 5.066209650585103e-07,
+      "loss": 0.5279,
+      "step": 10672
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.524240898158136,
+      "learning_rate": 5.060488342606629e-07,
+      "loss": 0.544,
+      "step": 10673
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.570250040273248,
+      "learning_rate": 5.054770094810196e-07,
+      "loss": 0.4602,
+      "step": 10674
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.072099191468243,
+      "learning_rate": 5.049054907585205e-07,
+      "loss": 0.3093,
+      "step": 10675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.111526913674862,
+      "learning_rate": 5.043342781320809e-07,
+      "loss": 0.5527,
+      "step": 10676
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.658736000955046,
+      "learning_rate": 5.037633716406009e-07,
+      "loss": 0.432,
+      "step": 10677
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.954639630073798,
+      "learning_rate": 5.031927713229545e-07,
+      "loss": 0.3815,
+      "step": 10678
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.296036933356441,
+      "learning_rate": 5.026224772179994e-07,
+      "loss": 0.4166,
+      "step": 10679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.727021064319421,
+      "learning_rate": 5.020524893645673e-07,
+      "loss": 0.4582,
+      "step": 10680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.285757714122597,
+      "learning_rate": 5.014828078014727e-07,
+      "loss": 0.4688,
+      "step": 10681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.888523603928049,
+      "learning_rate": 5.009134325675102e-07,
+      "loss": 0.4613,
+      "step": 10682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.107953495078551,
+      "learning_rate": 5.003443637014504e-07,
+      "loss": 0.5634,
+      "step": 10683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8031692014475968,
+      "learning_rate": 4.997756012420441e-07,
+      "loss": 0.7214,
+      "step": 10684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.559543062405458,
+      "learning_rate": 4.992071452280212e-07,
+      "loss": 0.5146,
+      "step": 10685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.0804720365631875,
+      "learning_rate": 4.986389956980925e-07,
+      "loss": 0.4277,
+      "step": 10686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.303899048249388,
+      "learning_rate": 4.980711526909443e-07,
+      "loss": 0.3644,
+      "step": 10687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.875049694891006,
+      "learning_rate": 4.975036162452479e-07,
+      "loss": 0.3585,
+      "step": 10688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.265749131022462,
+      "learning_rate": 4.969363863996451e-07,
+      "loss": 0.4614,
+      "step": 10689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.884101190603494,
+      "learning_rate": 4.963694631927651e-07,
+      "loss": 0.689,
+      "step": 10690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.279741273410284,
+      "learning_rate": 4.958028466632125e-07,
+      "loss": 0.3187,
+      "step": 10691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.325175833143459,
+      "learning_rate": 4.95236536849571e-07,
+      "loss": 0.4493,
+      "step": 10692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.886664124464545,
+      "learning_rate": 4.94670533790404e-07,
+      "loss": 0.4605,
+      "step": 10693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9224905076888383,
+      "learning_rate": 4.941048375242524e-07,
+      "loss": 0.4302,
+      "step": 10694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.120107442662274,
+      "learning_rate": 4.935394480896393e-07,
+      "loss": 0.4648,
+      "step": 10695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.033695519145241,
+      "learning_rate": 4.929743655250647e-07,
+      "loss": 0.441,
+      "step": 10696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.285687813143347,
+      "learning_rate": 4.92409589869009e-07,
+      "loss": 0.4723,
+      "step": 10697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.156066952529154,
+      "learning_rate": 4.91845121159929e-07,
+      "loss": 0.4575,
+      "step": 10698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.6631030089663135,
+      "learning_rate": 4.912809594362638e-07,
+      "loss": 0.4428,
+      "step": 10699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.446068126141591,
+      "learning_rate": 4.9071710473643e-07,
+      "loss": 0.411,
+      "step": 10700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.0012982431387165,
+      "learning_rate": 4.901535570988241e-07,
+      "loss": 0.3954,
+      "step": 10701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.625226728871294,
+      "learning_rate": 4.895903165618216e-07,
+      "loss": 0.4637,
+      "step": 10702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.751116548225735,
+      "learning_rate": 4.890273831637743e-07,
+      "loss": 0.4228,
+      "step": 10703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.920166767893321,
+      "learning_rate": 4.884647569430184e-07,
+      "loss": 0.4458,
+      "step": 10704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.458071937137205,
+      "learning_rate": 4.879024379378639e-07,
+      "loss": 0.4307,
+      "step": 10705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.3924697717994405,
+      "learning_rate": 4.873404261866055e-07,
+      "loss": 0.4633,
+      "step": 10706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.017826938905311,
+      "learning_rate": 4.867787217275094e-07,
+      "loss": 0.3791,
+      "step": 10707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.883857839556061,
+      "learning_rate": 4.862173245988283e-07,
+      "loss": 0.3686,
+      "step": 10708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.949989334774397,
+      "learning_rate": 4.856562348387889e-07,
+      "loss": 0.525,
+      "step": 10709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.20834595807049,
+      "learning_rate": 4.850954524856011e-07,
+      "loss": 0.4874,
+      "step": 10710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.638089829530923,
+      "learning_rate": 4.845349775774505e-07,
+      "loss": 0.4174,
+      "step": 10711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 39.52556230040799,
+      "learning_rate": 4.839748101525015e-07,
+      "loss": 0.3701,
+      "step": 10712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.262199741995789,
+      "learning_rate": 4.834149502489021e-07,
+      "loss": 0.4062,
+      "step": 10713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.142621771621624,
+      "learning_rate": 4.82855397904774e-07,
+      "loss": 0.4167,
+      "step": 10714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.1886520019466635,
+      "learning_rate": 4.822961531582221e-07,
+      "loss": 0.4305,
+      "step": 10715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.701561242314317,
+      "learning_rate": 4.81737216047326e-07,
+      "loss": 0.4144,
+      "step": 10716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.985216855748179,
+      "learning_rate": 4.811785866101487e-07,
+      "loss": 0.5119,
+      "step": 10717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.765244127163362,
+      "learning_rate": 4.806202648847291e-07,
+      "loss": 0.496,
+      "step": 10718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.156761023786018,
+      "learning_rate": 4.80062250909088e-07,
+      "loss": 0.3692,
+      "step": 10719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.771073368330338,
+      "learning_rate": 4.795045447212232e-07,
+      "loss": 0.491,
+      "step": 10720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.97180979766636,
+      "learning_rate": 4.789471463591105e-07,
+      "loss": 0.6033,
+      "step": 10721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.60964459015753,
+      "learning_rate": 4.783900558607091e-07,
+      "loss": 0.4156,
+      "step": 10722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.519415500063144,
+      "learning_rate": 4.778332732639512e-07,
+      "loss": 0.4302,
+      "step": 10723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.913380753309882,
+      "learning_rate": 4.772767986067539e-07,
+      "loss": 0.4161,
+      "step": 10724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.44965113776685,
+      "learning_rate": 4.767206319270096e-07,
+      "loss": 0.3997,
+      "step": 10725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.441633663943001,
+      "learning_rate": 4.761647732625907e-07,
+      "loss": 0.488,
+      "step": 10726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8854975023944767,
+      "learning_rate": 4.756092226513481e-07,
+      "loss": 0.4566,
+      "step": 10727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.938744691676085,
+      "learning_rate": 4.750539801311138e-07,
+      "loss": 0.432,
+      "step": 10728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.768052710784065,
+      "learning_rate": 4.744990457396964e-07,
+      "loss": 0.7107,
+      "step": 10729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.529752085982504,
+      "learning_rate": 4.73944419514884e-07,
+      "loss": 0.4742,
+      "step": 10730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.085335563108044,
+      "learning_rate": 4.733901014944459e-07,
+      "loss": 0.4654,
+      "step": 10731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7757396093664743,
+      "learning_rate": 4.7283609171612676e-07,
+      "loss": 0.4276,
+      "step": 10732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.355860242179544,
+      "learning_rate": 4.722823902176538e-07,
+      "loss": 0.4122,
+      "step": 10733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.082079355356597,
+      "learning_rate": 4.717289970367311e-07,
+      "loss": 0.5939,
+      "step": 10734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.5312296859393575,
+      "learning_rate": 4.711759122110421e-07,
+      "loss": 0.4246,
+      "step": 10735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.374196497793377,
+      "learning_rate": 4.7062313577824816e-07,
+      "loss": 0.4431,
+      "step": 10736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.537644791498512,
+      "learning_rate": 4.7007066777599363e-07,
+      "loss": 0.4259,
+      "step": 10737
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.253013704185853,
+      "learning_rate": 4.695185082418979e-07,
+      "loss": 0.514,
+      "step": 10738
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.8118824651993215,
+      "learning_rate": 4.689666572135598e-07,
+      "loss": 0.5202,
+      "step": 10739
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.832311796506383,
+      "learning_rate": 4.6841511472855815e-07,
+      "loss": 0.4375,
+      "step": 10740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.778733577815581,
+      "learning_rate": 4.678638808244512e-07,
+      "loss": 0.4021,
+      "step": 10741
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.708013219727888,
+      "learning_rate": 4.6731295553877563e-07,
+      "loss": 0.453,
+      "step": 10742
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.7747802656818275,
+      "learning_rate": 4.6676233890904697e-07,
+      "loss": 0.4258,
+      "step": 10743
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 20.764102684544728,
+      "learning_rate": 4.662120309727597e-07,
+      "loss": 0.497,
+      "step": 10744
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.932100448321487,
+      "learning_rate": 4.6566203176738603e-07,
+      "loss": 0.4431,
+      "step": 10745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.586141942808512,
+      "learning_rate": 4.651123413303804e-07,
+      "loss": 0.4191,
+      "step": 10746
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.833916908335269,
+      "learning_rate": 4.645629596991741e-07,
+      "loss": 0.4485,
+      "step": 10747
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.631890845998997,
+      "learning_rate": 4.6401388691117653e-07,
+      "loss": 0.6791,
+      "step": 10748
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6176693614038833,
+      "learning_rate": 4.6346512300377676e-07,
+      "loss": 0.6735,
+      "step": 10749
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.474248144654915,
+      "learning_rate": 4.6291666801434376e-07,
+      "loss": 0.5265,
+      "step": 10750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.32619429546847,
+      "learning_rate": 4.623685219802265e-07,
+      "loss": 0.557,
+      "step": 10751
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.769202338944879,
+      "learning_rate": 4.6182068493874966e-07,
+      "loss": 0.4649,
+      "step": 10752
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.378793401773381,
+      "learning_rate": 4.6127315692721887e-07,
+      "loss": 0.451,
+      "step": 10753
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.458177991412856,
+      "learning_rate": 4.6072593798291763e-07,
+      "loss": 0.4736,
+      "step": 10754
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 39.45903614383265,
+      "learning_rate": 4.6017902814311067e-07,
+      "loss": 0.5802,
+      "step": 10755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.552022090100657,
+      "learning_rate": 4.596324274450387e-07,
+      "loss": 0.3853,
+      "step": 10756
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.556203368499456,
+      "learning_rate": 4.590861359259241e-07,
+      "loss": 0.5647,
+      "step": 10757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.563601173377471,
+      "learning_rate": 4.58540153622965e-07,
+      "loss": 0.3566,
+      "step": 10758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.413259594793036,
+      "learning_rate": 4.579944805733422e-07,
+      "loss": 0.359,
+      "step": 10759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.22948988318578,
+      "learning_rate": 4.5744911681421257e-07,
+      "loss": 0.3775,
+      "step": 10760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.776549870226388,
+      "learning_rate": 4.5690406238271425e-07,
+      "loss": 0.5381,
+      "step": 10761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.564490076414635,
+      "learning_rate": 4.563593173159619e-07,
+      "loss": 0.3839,
+      "step": 10762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.460030119684658,
+      "learning_rate": 4.5581488165104935e-07,
+      "loss": 0.5577,
+      "step": 10763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.796834101520815,
+      "learning_rate": 4.5527075542505295e-07,
+      "loss": 0.4775,
+      "step": 10764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.556445111917718,
+      "learning_rate": 4.547269386750236e-07,
+      "loss": 0.4617,
+      "step": 10765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.3340283718354,
+      "learning_rate": 4.541834314379928e-07,
+      "loss": 0.5029,
+      "step": 10766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.68465028559678,
+      "learning_rate": 4.5364023375097044e-07,
+      "loss": 0.412,
+      "step": 10767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.709799522484268,
+      "learning_rate": 4.530973456509474e-07,
+      "loss": 0.4464,
+      "step": 10768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.740945849909605,
+      "learning_rate": 4.525547671748909e-07,
+      "loss": 0.4426,
+      "step": 10769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.98978659228991,
+      "learning_rate": 4.520124983597485e-07,
+      "loss": 0.3873,
+      "step": 10770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.635940073371716,
+      "learning_rate": 4.5147053924244677e-07,
+      "loss": 0.4667,
+      "step": 10771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.45166408447843,
+      "learning_rate": 4.50928889859889e-07,
+      "loss": 0.4876,
+      "step": 10772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.652851594869193,
+      "learning_rate": 4.503875502489613e-07,
+      "loss": 0.4919,
+      "step": 10773
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.833775435670056,
+      "learning_rate": 4.4984652044652575e-07,
+      "loss": 0.3172,
+      "step": 10774
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.088264263943826,
+      "learning_rate": 4.4930580048942353e-07,
+      "loss": 0.4229,
+      "step": 10775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.795397732358448,
+      "learning_rate": 4.487653904144745e-07,
+      "loss": 0.5011,
+      "step": 10776
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.6856196289972205,
+      "learning_rate": 4.482252902584805e-07,
+      "loss": 0.4212,
+      "step": 10777
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.684523720754935,
+      "learning_rate": 4.476855000582181e-07,
+      "loss": 0.5298,
+      "step": 10778
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.942362864615242,
+      "learning_rate": 4.471460198504457e-07,
+      "loss": 0.458,
+      "step": 10779
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.35742827914608,
+      "learning_rate": 4.4660684967189947e-07,
+      "loss": 0.4924,
+      "step": 10780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.263442699404069,
+      "learning_rate": 4.4606798955929287e-07,
+      "loss": 0.4694,
+      "step": 10781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.080446964723087,
+      "learning_rate": 4.455294395493226e-07,
+      "loss": 0.4626,
+      "step": 10782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.94754888075403,
+      "learning_rate": 4.4499119967865876e-07,
+      "loss": 0.4083,
+      "step": 10783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.653518480977802,
+      "learning_rate": 4.4445326998395597e-07,
+      "loss": 0.3958,
+      "step": 10784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.451152677055248,
+      "learning_rate": 4.4391565050184216e-07,
+      "loss": 0.4809,
+      "step": 10785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.749676960371959,
+      "learning_rate": 4.433783412689285e-07,
+      "loss": 0.4491,
+      "step": 10786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.731646986703298,
+      "learning_rate": 4.4284134232180254e-07,
+      "loss": 0.4762,
+      "step": 10787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.419696062274603,
+      "learning_rate": 4.423046536970321e-07,
+      "loss": 0.4884,
+      "step": 10788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.847384440357955,
+      "learning_rate": 4.4176827543116365e-07,
+      "loss": 0.4131,
+      "step": 10789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.492008541316841,
+      "learning_rate": 4.4123220756072115e-07,
+      "loss": 0.3892,
+      "step": 10790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.760496627954598,
+      "learning_rate": 4.406964501222083e-07,
+      "loss": 0.7969,
+      "step": 10791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.430897236155411,
+      "learning_rate": 4.401610031521086e-07,
+      "loss": 0.4191,
+      "step": 10792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.802456996310606,
+      "learning_rate": 4.3962586668688523e-07,
+      "loss": 0.3864,
+      "step": 10793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.946638866629089,
+      "learning_rate": 4.3909104076297447e-07,
+      "loss": 0.3887,
+      "step": 10794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5620539776727196,
+      "learning_rate": 4.385565254167995e-07,
+      "loss": 0.6623,
+      "step": 10795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.505061851018018,
+      "learning_rate": 4.3802232068475557e-07,
+      "loss": 0.4238,
+      "step": 10796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 16.494531819229117,
+      "learning_rate": 4.3748842660322255e-07,
+      "loss": 0.4302,
+      "step": 10797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.603236034992963,
+      "learning_rate": 4.3695484320855407e-07,
+      "loss": 0.5413,
+      "step": 10798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.36769752159648,
+      "learning_rate": 4.3642157053708556e-07,
+      "loss": 0.3402,
+      "step": 10799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.168035052717034,
+      "learning_rate": 4.358886086251296e-07,
+      "loss": 0.4239,
+      "step": 10800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.650667234915651,
+      "learning_rate": 4.353559575089794e-07,
+      "loss": 0.4863,
+      "step": 10801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.4850402729663905,
+      "learning_rate": 4.348236172249082e-07,
+      "loss": 0.4279,
+      "step": 10802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.7209821008136865,
+      "learning_rate": 4.3429158780916194e-07,
+      "loss": 0.3929,
+      "step": 10803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.331975912324008,
+      "learning_rate": 4.337598692979722e-07,
+      "loss": 0.4427,
+      "step": 10804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.459983597695571,
+      "learning_rate": 4.332284617275451e-07,
+      "loss": 0.4429,
+      "step": 10805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.019164734363127,
+      "learning_rate": 4.3269736513406877e-07,
+      "loss": 0.475,
+      "step": 10806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.611326180133482,
+      "learning_rate": 4.321665795537083e-07,
+      "loss": 0.4129,
+      "step": 10807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.823623080802726,
+      "learning_rate": 4.31636105022607e-07,
+      "loss": 0.5339,
+      "step": 10808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 30.401134588456987,
+      "learning_rate": 4.3110594157688703e-07,
+      "loss": 0.5355,
+      "step": 10809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.928038329272696,
+      "learning_rate": 4.305760892526523e-07,
+      "loss": 0.395,
+      "step": 10810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.279509995272969,
+      "learning_rate": 4.3004654808598225e-07,
+      "loss": 0.4309,
+      "step": 10811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 15.66900582815444,
+      "learning_rate": 4.295173181129353e-07,
+      "loss": 0.4414,
+      "step": 10812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.624025605948546,
+      "learning_rate": 4.2898839936955205e-07,
+      "loss": 0.4733,
+      "step": 10813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.932831874562298,
+      "learning_rate": 4.284597918918471e-07,
+      "loss": 0.472,
+      "step": 10814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.702110052107703,
+      "learning_rate": 4.279314957158187e-07,
+      "loss": 0.4687,
+      "step": 10815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 42.81111569767203,
+      "learning_rate": 4.274035108774399e-07,
+      "loss": 0.337,
+      "step": 10816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.023273289622709,
+      "learning_rate": 4.268758374126647e-07,
+      "loss": 0.5128,
+      "step": 10817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.555675134896525,
+      "learning_rate": 4.263484753574243e-07,
+      "loss": 0.4285,
+      "step": 10818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.343234942867332,
+      "learning_rate": 4.258214247476317e-07,
+      "loss": 0.5156,
+      "step": 10819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.802834435342943,
+      "learning_rate": 4.252946856191742e-07,
+      "loss": 0.5031,
+      "step": 10820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.150415730107991,
+      "learning_rate": 4.2476825800792266e-07,
+      "loss": 0.571,
+      "step": 10821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.438872302581652,
+      "learning_rate": 4.2424214194972333e-07,
+      "loss": 0.4278,
+      "step": 10822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.139452462710724,
+      "learning_rate": 4.237163374804021e-07,
+      "loss": 0.4042,
+      "step": 10823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.237472400505637,
+      "learning_rate": 4.231908446357652e-07,
+      "loss": 0.3409,
+      "step": 10824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.807284209062312,
+      "learning_rate": 4.226656634515958e-07,
+      "loss": 0.5454,
+      "step": 10825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.378265857456565,
+      "learning_rate": 4.221407939636557e-07,
+      "loss": 0.4541,
+      "step": 10826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.7375043134151555,
+      "learning_rate": 4.2161623620768643e-07,
+      "loss": 0.4066,
+      "step": 10827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.256669578826651,
+      "learning_rate": 4.2109199021940884e-07,
+      "loss": 0.5171,
+      "step": 10828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.006917273332901,
+      "learning_rate": 4.2056805603452046e-07,
+      "loss": 0.4459,
+      "step": 10829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.931760376602904,
+      "learning_rate": 4.200444336887016e-07,
+      "loss": 0.347,
+      "step": 10830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.16546099796053,
+      "learning_rate": 4.1952112321760497e-07,
+      "loss": 0.3991,
+      "step": 10831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.825023710841377,
+      "learning_rate": 4.189981246568675e-07,
+      "loss": 0.4941,
+      "step": 10832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.361637497311877,
+      "learning_rate": 4.1847543804210357e-07,
+      "loss": 0.5067,
+      "step": 10833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.594046812247018,
+      "learning_rate": 4.1795306340890574e-07,
+      "loss": 0.5623,
+      "step": 10834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.163861980420671,
+      "learning_rate": 4.1743100079284447e-07,
+      "loss": 0.5376,
+      "step": 10835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.266415094291172,
+      "learning_rate": 4.1690925022947025e-07,
+      "loss": 0.4951,
+      "step": 10836
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.20364726916136,
+      "learning_rate": 4.163878117543124e-07,
+      "loss": 0.4789,
+      "step": 10837
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.746606637609538,
+      "learning_rate": 4.1586668540287753e-07,
+      "loss": 0.5348,
+      "step": 10838
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.562735173981608,
+      "learning_rate": 4.1534587121065506e-07,
+      "loss": 0.4898,
+      "step": 10839
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.915597681410611,
+      "learning_rate": 4.1482536921310544e-07,
+      "loss": 0.5077,
+      "step": 10840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.871566928337263,
+      "learning_rate": 4.143051794456754e-07,
+      "loss": 0.4586,
+      "step": 10841
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.270112272184984,
+      "learning_rate": 4.137853019437882e-07,
+      "loss": 0.4286,
+      "step": 10842
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.6962804650225465,
+      "learning_rate": 4.1326573674284386e-07,
+      "loss": 0.4649,
+      "step": 10843
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.297673164934822,
+      "learning_rate": 4.12746483878223e-07,
+      "loss": 0.4749,
+      "step": 10844
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 14.066619029975456,
+      "learning_rate": 4.122275433852835e-07,
+      "loss": 0.4579,
+      "step": 10845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.181680228825294,
+      "learning_rate": 4.1170891529936474e-07,
+      "loss": 0.3372,
+      "step": 10846
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.852806467236524,
+      "learning_rate": 4.1119059965578023e-07,
+      "loss": 0.4299,
+      "step": 10847
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.413369731478919,
+      "learning_rate": 4.1067259648982894e-07,
+      "loss": 0.491,
+      "step": 10848
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.302277921560424,
+      "learning_rate": 4.101549058367804e-07,
+      "loss": 0.5395,
+      "step": 10849
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.957941862627621,
+      "learning_rate": 4.0963752773189034e-07,
+      "loss": 0.3808,
+      "step": 10850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4359707526948706,
+      "learning_rate": 4.0912046221038724e-07,
+      "loss": 0.4806,
+      "step": 10851
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.067659550362333,
+      "learning_rate": 4.0860370930748297e-07,
+      "loss": 0.4368,
+      "step": 10852
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.2704983071496585,
+      "learning_rate": 4.080872690583659e-07,
+      "loss": 0.4683,
+      "step": 10853
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.0327653896874684,
+      "learning_rate": 4.0757114149820143e-07,
+      "loss": 0.3553,
+      "step": 10854
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.757806482111254,
+      "learning_rate": 4.0705532666213855e-07,
+      "loss": 0.4681,
+      "step": 10855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.217701488799861,
+      "learning_rate": 4.0653982458529917e-07,
+      "loss": 0.3338,
+      "step": 10856
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.108885735155422,
+      "learning_rate": 4.0602463530278856e-07,
+      "loss": 0.4349,
+      "step": 10857
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.375382429465272,
+      "learning_rate": 4.0550975884968867e-07,
+      "loss": 0.3522,
+      "step": 10858
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.100449845590948,
+      "learning_rate": 4.049951952610598e-07,
+      "loss": 0.5277,
+      "step": 10859
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.032686460557215,
+      "learning_rate": 4.0448094457194054e-07,
+      "loss": 0.4983,
+      "step": 10860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.21548138215765,
+      "learning_rate": 4.0396700681735066e-07,
+      "loss": 0.4324,
+      "step": 10861
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8372135124687943,
+      "learning_rate": 4.034533820322872e-07,
+      "loss": 0.3633,
+      "step": 10862
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.238164455859039,
+      "learning_rate": 4.0294007025172334e-07,
+      "loss": 0.4189,
+      "step": 10863
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.0994529517552944,
+      "learning_rate": 4.0242707151061597e-07,
+      "loss": 0.3981,
+      "step": 10864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.214600613218313,
+      "learning_rate": 4.0191438584389676e-07,
+      "loss": 0.4092,
+      "step": 10865
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.757577984404282,
+      "learning_rate": 4.014020132864782e-07,
+      "loss": 0.2753,
+      "step": 10866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.333774639345625,
+      "learning_rate": 4.008899538732497e-07,
+      "loss": 0.4565,
+      "step": 10867
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.434880220660935,
+      "learning_rate": 4.0037820763908106e-07,
+      "loss": 0.4685,
+      "step": 10868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 14.851034927657935,
+      "learning_rate": 3.998667746188184e-07,
+      "loss": 0.5054,
+      "step": 10869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.553400479676204,
+      "learning_rate": 3.9935565484728987e-07,
+      "loss": 0.4285,
+      "step": 10870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.313422831938844,
+      "learning_rate": 3.9884484835929994e-07,
+      "loss": 0.4633,
+      "step": 10871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.104261596747852,
+      "learning_rate": 3.983343551896307e-07,
+      "loss": 0.4432,
+      "step": 10872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.535794140842849,
+      "learning_rate": 3.978241753730472e-07,
+      "loss": 0.3831,
+      "step": 10873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.09734775202337,
+      "learning_rate": 3.973143089442877e-07,
+      "loss": 0.5293,
+      "step": 10874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.9612733037729795,
+      "learning_rate": 3.9680475593807443e-07,
+      "loss": 0.4796,
+      "step": 10875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.600239047406565,
+      "learning_rate": 3.9629551638910413e-07,
+      "loss": 0.4786,
+      "step": 10876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.991315318331405,
+      "learning_rate": 3.95786590332054e-07,
+      "loss": 0.5144,
+      "step": 10877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.472694629164636,
+      "learning_rate": 3.952779778015792e-07,
+      "loss": 0.5519,
+      "step": 10878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.610623397258935,
+      "learning_rate": 3.947696788323152e-07,
+      "loss": 0.4778,
+      "step": 10879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.763534904059215,
+      "learning_rate": 3.942616934588744e-07,
+      "loss": 0.4166,
+      "step": 10880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.928678040346305,
+      "learning_rate": 3.937540217158475e-07,
+      "loss": 0.528,
+      "step": 10881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.815353201920224,
+      "learning_rate": 3.932466636378063e-07,
+      "loss": 0.393,
+      "step": 10882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.05360176295947,
+      "learning_rate": 3.927396192592975e-07,
+      "loss": 0.566,
+      "step": 10883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.512812439589581,
+      "learning_rate": 3.9223288861485074e-07,
+      "loss": 0.3815,
+      "step": 10884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.40170557159473,
+      "learning_rate": 3.917264717389713e-07,
+      "loss": 0.4919,
+      "step": 10885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.005813480211444,
+      "learning_rate": 3.9122036866614365e-07,
+      "loss": 0.4362,
+      "step": 10886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.801566599852274,
+      "learning_rate": 3.907145794308309e-07,
+      "loss": 0.4393,
+      "step": 10887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.037808334072869,
+      "learning_rate": 3.9020910406747605e-07,
+      "loss": 0.5139,
+      "step": 10888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.832971411029142,
+      "learning_rate": 3.897039426104998e-07,
+      "loss": 0.454,
+      "step": 10889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.174264875237135,
+      "learning_rate": 3.891990950943003e-07,
+      "loss": 0.515,
+      "step": 10890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.6556550436110005,
+      "learning_rate": 3.88694561553255e-07,
+      "loss": 0.4943,
+      "step": 10891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.801165864619257,
+      "learning_rate": 3.881903420217215e-07,
+      "loss": 0.5469,
+      "step": 10892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6608541024192185,
+      "learning_rate": 3.8768643653403605e-07,
+      "loss": 0.7025,
+      "step": 10893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.474994567263621,
+      "learning_rate": 3.8718284512451077e-07,
+      "loss": 0.4605,
+      "step": 10894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.6120914353771,
+      "learning_rate": 3.866795678274382e-07,
+      "loss": 0.49,
+      "step": 10895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.749482532625963,
+      "learning_rate": 3.8617660467708875e-07,
+      "loss": 0.4669,
+      "step": 10896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.6707584367817745,
+      "learning_rate": 3.856739557077138e-07,
+      "loss": 0.4746,
+      "step": 10897
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.606825408362953,
+      "learning_rate": 3.8517162095354e-07,
+      "loss": 0.5119,
+      "step": 10898
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.446823961653113,
+      "learning_rate": 3.846696004487749e-07,
+      "loss": 0.4615,
+      "step": 10899
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.881081379997098,
+      "learning_rate": 3.841678942276028e-07,
+      "loss": 0.4917,
+      "step": 10900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.6294545457098035,
+      "learning_rate": 3.836665023241881e-07,
+      "loss": 0.4865,
+      "step": 10901
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.275425984974795,
+      "learning_rate": 3.831654247726746e-07,
+      "loss": 0.4672,
+      "step": 10902
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.210737465659836,
+      "learning_rate": 3.8266466160718274e-07,
+      "loss": 0.4094,
+      "step": 10903
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.935573930633282,
+      "learning_rate": 3.821642128618114e-07,
+      "loss": 0.4466,
+      "step": 10904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.820464965066469,
+      "learning_rate": 3.816640785706394e-07,
+      "loss": 0.3731,
+      "step": 10905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.87340542519831,
+      "learning_rate": 3.8116425876772445e-07,
+      "loss": 0.4599,
+      "step": 10906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.62391911643223,
+      "learning_rate": 3.8066475348710165e-07,
+      "loss": 0.4635,
+      "step": 10907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.148896358715954,
+      "learning_rate": 3.801655627627848e-07,
+      "loss": 0.4585,
+      "step": 10908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.639443193382423,
+      "learning_rate": 3.796666866287663e-07,
+      "loss": 0.5151,
+      "step": 10909
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8234179420210863,
+      "learning_rate": 3.7916812511901824e-07,
+      "loss": 0.7576,
+      "step": 10910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.534130783916726,
+      "learning_rate": 3.786698782674891e-07,
+      "loss": 0.329,
+      "step": 10911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.135765353150342,
+      "learning_rate": 3.781719461081096e-07,
+      "loss": 0.514,
+      "step": 10912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.525039443447166,
+      "learning_rate": 3.776743286747847e-07,
+      "loss": 0.5429,
+      "step": 10913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6635802899660728,
+      "learning_rate": 3.7717702600140015e-07,
+      "loss": 0.7478,
+      "step": 10914
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.292973381492155,
+      "learning_rate": 3.7668003812182106e-07,
+      "loss": 0.5072,
+      "step": 10915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.546382319492142,
+      "learning_rate": 3.7618336506988927e-07,
+      "loss": 0.4319,
+      "step": 10916
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.202226225591817,
+      "learning_rate": 3.756870068794277e-07,
+      "loss": 0.4492,
+      "step": 10917
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.740761192123163,
+      "learning_rate": 3.7519096358423326e-07,
+      "loss": 0.5628,
+      "step": 10918
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.878474398709826,
+      "learning_rate": 3.7469523521808606e-07,
+      "loss": 0.3613,
+      "step": 10919
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.596953353371123,
+      "learning_rate": 3.7419982181474245e-07,
+      "loss": 0.5527,
+      "step": 10920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.051441416502198,
+      "learning_rate": 3.7370472340793885e-07,
+      "loss": 0.445,
+      "step": 10921
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.49905179747874,
+      "learning_rate": 3.7320994003138875e-07,
+      "loss": 0.355,
+      "step": 10922
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.64959376800349,
+      "learning_rate": 3.7271547171878364e-07,
+      "loss": 0.4369,
+      "step": 10923
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.841744513516915,
+      "learning_rate": 3.722213185037965e-07,
+      "loss": 0.4696,
+      "step": 10924
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.336700657089786,
+      "learning_rate": 3.717274804200749e-07,
+      "loss": 0.4248,
+      "step": 10925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.638255536945171,
+      "learning_rate": 3.7123395750125025e-07,
+      "loss": 0.4928,
+      "step": 10926
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.267561653928529,
+      "learning_rate": 3.707407497809257e-07,
+      "loss": 0.4175,
+      "step": 10927
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.03584307345662,
+      "learning_rate": 3.702478572926882e-07,
+      "loss": 0.4197,
+      "step": 10928
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.093315359535303,
+      "learning_rate": 3.6975528007010095e-07,
+      "loss": 0.4628,
+      "step": 10929
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.190314954269939,
+      "learning_rate": 3.6926301814670717e-07,
+      "loss": 0.4385,
+      "step": 10930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.894658449858312,
+      "learning_rate": 3.6877107155602774e-07,
+      "loss": 0.586,
+      "step": 10931
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.215180096523787,
+      "learning_rate": 3.682794403315604e-07,
+      "loss": 0.5249,
+      "step": 10932
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.834800464673028,
+      "learning_rate": 3.677881245067855e-07,
+      "loss": 0.5293,
+      "step": 10933
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.831642666163464,
+      "learning_rate": 3.6729712411515684e-07,
+      "loss": 0.4718,
+      "step": 10934
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.91907331103056,
+      "learning_rate": 3.668064391901127e-07,
+      "loss": 0.3467,
+      "step": 10935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.293851499467353,
+      "learning_rate": 3.663160697650625e-07,
+      "loss": 0.4373,
+      "step": 10936
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.275605950452577,
+      "learning_rate": 3.658260158734017e-07,
+      "loss": 0.5214,
+      "step": 10937
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.240063206294181,
+      "learning_rate": 3.653362775484981e-07,
+      "loss": 0.4444,
+      "step": 10938
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.130281497178174,
+      "learning_rate": 3.648468548237033e-07,
+      "loss": 0.4737,
+      "step": 10939
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.410274262467666,
+      "learning_rate": 3.6435774773234346e-07,
+      "loss": 0.5071,
+      "step": 10940
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.191116284793196,
+      "learning_rate": 3.638689563077241e-07,
+      "loss": 0.3855,
+      "step": 10941
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.0848795043228,
+      "learning_rate": 3.6338048058313037e-07,
+      "loss": 0.455,
+      "step": 10942
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.902173791305412,
+      "learning_rate": 3.62892320591825e-07,
+      "loss": 0.4727,
+      "step": 10943
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.593995219600115,
+      "learning_rate": 3.6240447636705146e-07,
+      "loss": 0.4515,
+      "step": 10944
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.624663454936721,
+      "learning_rate": 3.6191694794202647e-07,
+      "loss": 0.4878,
+      "step": 10945
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.8922939673954,
+      "learning_rate": 3.6142973534995073e-07,
+      "loss": 0.495,
+      "step": 10946
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.176271922211677,
+      "learning_rate": 3.6094283862400046e-07,
+      "loss": 0.4436,
+      "step": 10947
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.493271074550496,
+      "learning_rate": 3.60456257797332e-07,
+      "loss": 0.5205,
+      "step": 10948
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.765613542339109,
+      "learning_rate": 3.599699929030792e-07,
+      "loss": 0.5378,
+      "step": 10949
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 19.314806882430457,
+      "learning_rate": 3.594840439743541e-07,
+      "loss": 0.4918,
+      "step": 10950
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.631703766707844,
+      "learning_rate": 3.589984110442468e-07,
+      "loss": 0.4816,
+      "step": 10951
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.454092894231007,
+      "learning_rate": 3.585130941458276e-07,
+      "loss": 0.5213,
+      "step": 10952
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.1944297495536915,
+      "learning_rate": 3.580280933121455e-07,
+      "loss": 0.4559,
+      "step": 10953
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 18.737961441689237,
+      "learning_rate": 3.575434085762258e-07,
+      "loss": 0.4253,
+      "step": 10954
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.542728405748668,
+      "learning_rate": 3.5705903997107383e-07,
+      "loss": 0.5342,
+      "step": 10955
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.718991129749867,
+      "learning_rate": 3.56574987529672e-07,
+      "loss": 0.5233,
+      "step": 10956
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.8768266722472475,
+      "learning_rate": 3.5609125128498343e-07,
+      "loss": 0.458,
+      "step": 10957
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.23155810547426,
+      "learning_rate": 3.5560783126994735e-07,
+      "loss": 0.4216,
+      "step": 10958
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.620746801174681,
+      "learning_rate": 3.551247275174835e-07,
+      "loss": 0.4582,
+      "step": 10959
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.471977737880629,
+      "learning_rate": 3.546419400604878e-07,
+      "loss": 0.397,
+      "step": 10960
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.242895661932133,
+      "learning_rate": 3.5415946893183727e-07,
+      "loss": 0.5254,
+      "step": 10961
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.723759508304411,
+      "learning_rate": 3.536773141643851e-07,
+      "loss": 0.4923,
+      "step": 10962
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.93541460696663,
+      "learning_rate": 3.5319547579096446e-07,
+      "loss": 0.4001,
+      "step": 10963
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.506788873418714,
+      "learning_rate": 3.527139538443869e-07,
+      "loss": 0.4679,
+      "step": 10964
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.254290432704881,
+      "learning_rate": 3.5223274835744006e-07,
+      "loss": 0.4043,
+      "step": 10965
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.071344777962342,
+      "learning_rate": 3.5175185936289436e-07,
+      "loss": 0.4458,
+      "step": 10966
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.659349394928798,
+      "learning_rate": 3.5127128689349477e-07,
+      "loss": 0.3822,
+      "step": 10967
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 15.568878601910507,
+      "learning_rate": 3.507910309819668e-07,
+      "loss": 0.4072,
+      "step": 10968
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.073034785844458,
+      "learning_rate": 3.5031109166101197e-07,
+      "loss": 0.5315,
+      "step": 10969
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.286609425948818,
+      "learning_rate": 3.498314689633142e-07,
+      "loss": 0.4388,
+      "step": 10970
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.082485361787202,
+      "learning_rate": 3.493521629215324e-07,
+      "loss": 0.5451,
+      "step": 10971
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.582246947768323,
+      "learning_rate": 3.4887317356830653e-07,
+      "loss": 0.3847,
+      "step": 10972
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.016025214344552,
+      "learning_rate": 3.483945009362527e-07,
+      "loss": 0.4842,
+      "step": 10973
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.802333249359825,
+      "learning_rate": 3.4791614505796543e-07,
+      "loss": 0.4828,
+      "step": 10974
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1715534697088663,
+      "learning_rate": 3.474381059660209e-07,
+      "loss": 0.6965,
+      "step": 10975
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.445051426547326,
+      "learning_rate": 3.469603836929697e-07,
+      "loss": 0.4406,
+      "step": 10976
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.516334502185812,
+      "learning_rate": 3.464829782713436e-07,
+      "loss": 0.5373,
+      "step": 10977
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.686922627727934,
+      "learning_rate": 3.4600588973365047e-07,
+      "loss": 0.5109,
+      "step": 10978
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.376521021032957,
+      "learning_rate": 3.4552911811237933e-07,
+      "loss": 0.3572,
+      "step": 10979
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.701057115554137,
+      "learning_rate": 3.4505266343999535e-07,
+      "loss": 0.5517,
+      "step": 10980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.707984219904,
+      "learning_rate": 3.4457652574894476e-07,
+      "loss": 0.3916,
+      "step": 10981
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.269509860050409,
+      "learning_rate": 3.4410070507164716e-07,
+      "loss": 0.5325,
+      "step": 10982
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.491881821360939,
+      "learning_rate": 3.436252014405062e-07,
+      "loss": 0.4056,
+      "step": 10983
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.201413535150215,
+      "learning_rate": 3.4315001488790133e-07,
+      "loss": 0.4618,
+      "step": 10984
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.822796617374138,
+      "learning_rate": 3.4267514544619076e-07,
+      "loss": 0.5132,
+      "step": 10985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.965600546906309,
+      "learning_rate": 3.422005931477107e-07,
+      "loss": 0.4218,
+      "step": 10986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.485137711700249,
+      "learning_rate": 3.417263580247748e-07,
+      "loss": 0.395,
+      "step": 10987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.716127882777165,
+      "learning_rate": 3.412524401096784e-07,
+      "loss": 0.4375,
+      "step": 10988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.782483742710626,
+      "learning_rate": 3.407788394346917e-07,
+      "loss": 0.4092,
+      "step": 10989
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.070039062863939,
+      "learning_rate": 3.4030555603206674e-07,
+      "loss": 0.4042,
+      "step": 10990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.448496125168358,
+      "learning_rate": 3.398325899340299e-07,
+      "loss": 0.493,
+      "step": 10991
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.108493297733262,
+      "learning_rate": 3.3935994117278883e-07,
+      "loss": 0.4019,
+      "step": 10992
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.34366433327203,
+      "learning_rate": 3.3888760978052937e-07,
+      "loss": 0.4623,
+      "step": 10993
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.938359063676468,
+      "learning_rate": 3.384155957894153e-07,
+      "loss": 0.5539,
+      "step": 10994
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.990224920523837,
+      "learning_rate": 3.3794389923158757e-07,
+      "loss": 0.4994,
+      "step": 10995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.952485158719312,
+      "learning_rate": 3.3747252013916656e-07,
+      "loss": 0.5145,
+      "step": 10996
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.623701122951101,
+      "learning_rate": 3.370014585442527e-07,
+      "loss": 0.4234,
+      "step": 10997
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.510683866352606,
+      "learning_rate": 3.365307144789215e-07,
+      "loss": 0.5204,
+      "step": 10998
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7502476540584715,
+      "learning_rate": 3.3606028797523005e-07,
+      "loss": 0.6592,
+      "step": 10999
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.259364191762195,
+      "learning_rate": 3.3559017906521107e-07,
+      "loss": 0.4139,
+      "step": 11000
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 15.141524259388397,
+      "learning_rate": 3.3512038778087843e-07,
+      "loss": 0.4645,
+      "step": 11001
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.298969120608211,
+      "learning_rate": 3.346509141542204e-07,
+      "loss": 0.407,
+      "step": 11002
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.922001513548508,
+      "learning_rate": 3.34181758217208e-07,
+      "loss": 0.4666,
+      "step": 11003
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.181528384048249,
+      "learning_rate": 3.3371292000178856e-07,
+      "loss": 0.4869,
+      "step": 11004
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7883511918442634,
+      "learning_rate": 3.332443995398865e-07,
+      "loss": 0.7634,
+      "step": 11005
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.261488162319514,
+      "learning_rate": 3.32776196863408e-07,
+      "loss": 0.4283,
+      "step": 11006
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.714799372827598,
+      "learning_rate": 3.3230831200423364e-07,
+      "loss": 0.4322,
+      "step": 11007
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6555138861373946,
+      "learning_rate": 3.3184074499422624e-07,
+      "loss": 0.4798,
+      "step": 11008
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.543895657988152,
+      "learning_rate": 3.3137349586522373e-07,
+      "loss": 0.4847,
+      "step": 11009
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.319199962625523,
+      "learning_rate": 3.3090656464904394e-07,
+      "loss": 0.4701,
+      "step": 11010
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.177440578796677,
+      "learning_rate": 3.3043995137748253e-07,
+      "loss": 0.4273,
+      "step": 11011
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.78015174490008,
+      "learning_rate": 3.2997365608231414e-07,
+      "loss": 0.4832,
+      "step": 11012
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 16.539549032660155,
+      "learning_rate": 3.2950767879529277e-07,
+      "loss": 0.4481,
+      "step": 11013
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.545500292913674,
+      "learning_rate": 3.290420195481464e-07,
+      "loss": 0.5143,
+      "step": 11014
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.4846424327466945,
+      "learning_rate": 3.285766783725869e-07,
+      "loss": 0.5614,
+      "step": 11015
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8300525318007896,
+      "learning_rate": 3.281116553003005e-07,
+      "loss": 0.7169,
+      "step": 11016
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.00993356120638,
+      "learning_rate": 3.276469503629548e-07,
+      "loss": 0.3867,
+      "step": 11017
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.282580663198048,
+      "learning_rate": 3.271825635921927e-07,
+      "loss": 0.4655,
+      "step": 11018
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.323523519628132,
+      "learning_rate": 3.267184950196378e-07,
+      "loss": 0.5421,
+      "step": 11019
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.4395230500807505,
+      "learning_rate": 3.2625474467688987e-07,
+      "loss": 0.4146,
+      "step": 11020
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.628784549081233,
+      "learning_rate": 3.2579131259552923e-07,
+      "loss": 0.4366,
+      "step": 11021
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.943884117033685,
+      "learning_rate": 3.2532819880711386e-07,
+      "loss": 0.4929,
+      "step": 11022
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.478916542349054,
+      "learning_rate": 3.248654033431786e-07,
+      "loss": 0.447,
+      "step": 11023
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.08262687689295,
+      "learning_rate": 3.2440292623523883e-07,
+      "loss": 0.4735,
+      "step": 11024
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.441685820791323,
+      "learning_rate": 3.2394076751478655e-07,
+      "loss": 0.4704,
+      "step": 11025
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.195361576381819,
+      "learning_rate": 3.2347892721329376e-07,
+      "loss": 0.4242,
+      "step": 11026
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3959204593764836,
+      "learning_rate": 3.230174053622087e-07,
+      "loss": 0.6512,
+      "step": 11027
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.5270022965112,
+      "learning_rate": 3.2255620199295955e-07,
+      "loss": 0.4864,
+      "step": 11028
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.386712873221448,
+      "learning_rate": 3.2209531713695063e-07,
+      "loss": 0.4481,
+      "step": 11029
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.898743290626272,
+      "learning_rate": 3.216347508255685e-07,
+      "loss": 0.4346,
+      "step": 11030
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.90082438730631,
+      "learning_rate": 3.2117450309017474e-07,
+      "loss": 0.5383,
+      "step": 11031
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 24.161244170351964,
+      "learning_rate": 3.2071457396210924e-07,
+      "loss": 0.4875,
+      "step": 11032
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.240866425787467,
+      "learning_rate": 3.202549634726931e-07,
+      "loss": 0.4273,
+      "step": 11033
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.352921517204521,
+      "learning_rate": 3.1979567165322134e-07,
+      "loss": 0.3834,
+      "step": 11034
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.019126641547583,
+      "learning_rate": 3.193366985349722e-07,
+      "loss": 0.4517,
+      "step": 11035
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.0967201986863735,
+      "learning_rate": 3.1887804414919845e-07,
+      "loss": 0.4776,
+      "step": 11036
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.176225601533948,
+      "learning_rate": 3.1841970852713233e-07,
+      "loss": 0.4623,
+      "step": 11037
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.105397954916534,
+      "learning_rate": 3.179616916999839e-07,
+      "loss": 0.506,
+      "step": 11038
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.734702959629869,
+      "learning_rate": 3.175039936989438e-07,
+      "loss": 0.4613,
+      "step": 11039
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9410117128848121,
+      "learning_rate": 3.1704661455517816e-07,
+      "loss": 0.7677,
+      "step": 11040
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.709083592629169,
+      "learning_rate": 3.1658955429983264e-07,
+      "loss": 0.477,
+      "step": 11041
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 12.729362724293344,
+      "learning_rate": 3.161328129640301e-07,
+      "loss": 0.3897,
+      "step": 11042
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.59820965293514,
+      "learning_rate": 3.1567639057887403e-07,
+      "loss": 0.4015,
+      "step": 11043
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.744327123138189,
+      "learning_rate": 3.1522028717544453e-07,
+      "loss": 0.5151,
+      "step": 11044
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.4011068039420636,
+      "learning_rate": 3.1476450278479955e-07,
+      "loss": 0.4823,
+      "step": 11045
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.356701095454373,
+      "learning_rate": 3.1430903743797703e-07,
+      "loss": 0.464,
+      "step": 11046
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.4801501690902565,
+      "learning_rate": 3.1385389116599e-07,
+      "loss": 0.457,
+      "step": 11047
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.4363141246220135,
+      "learning_rate": 3.1339906399983465e-07,
+      "loss": 0.6033,
+      "step": 11048
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.682026629900814,
+      "learning_rate": 3.1294455597048133e-07,
+      "loss": 0.4545,
+      "step": 11049
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.29575420819048,
+      "learning_rate": 3.1249036710888023e-07,
+      "loss": 0.3928,
+      "step": 11050
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.848593745285642,
+      "learning_rate": 3.1203649744595886e-07,
+      "loss": 0.5262,
+      "step": 11051
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.586981132577979,
+      "learning_rate": 3.1158294701262416e-07,
+      "loss": 0.4711,
+      "step": 11052
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.974787557699436,
+      "learning_rate": 3.1112971583976194e-07,
+      "loss": 0.4013,
+      "step": 11053
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.531867813771012,
+      "learning_rate": 3.106768039582342e-07,
+      "loss": 0.4264,
+      "step": 11054
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.658535851263383,
+      "learning_rate": 3.102242113988829e-07,
+      "loss": 0.391,
+      "step": 11055
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.45299817609381,
+      "learning_rate": 3.097719381925263e-07,
+      "loss": 0.4844,
+      "step": 11056
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.973386225110465,
+      "learning_rate": 3.0931998436996347e-07,
+      "loss": 0.5071,
+      "step": 11057
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.621889038569362,
+      "learning_rate": 3.088683499619699e-07,
+      "loss": 0.5334,
+      "step": 11058
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.461548127237281,
+      "learning_rate": 3.08417034999301e-07,
+      "loss": 0.5123,
+      "step": 11059
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.807709815160718,
+      "learning_rate": 3.0796603951268713e-07,
+      "loss": 0.3654,
+      "step": 11060
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.543052617080572,
+      "learning_rate": 3.0751536353284096e-07,
+      "loss": 0.5551,
+      "step": 11061
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.613493507852378,
+      "learning_rate": 3.0706500709045015e-07,
+      "loss": 0.3584,
+      "step": 11062
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.163840948212084,
+      "learning_rate": 3.066149702161836e-07,
+      "loss": 0.5504,
+      "step": 11063
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.744654434781156,
+      "learning_rate": 3.0616525294068613e-07,
+      "loss": 0.4319,
+      "step": 11064
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.224190766773867,
+      "learning_rate": 3.057158552945805e-07,
+      "loss": 0.6166,
+      "step": 11065
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.309606837091902,
+      "learning_rate": 3.0526677730847054e-07,
+      "loss": 0.4127,
+      "step": 11066
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.38715291169965,
+      "learning_rate": 3.0481801901293405e-07,
+      "loss": 0.4674,
+      "step": 11067
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.18377570891604,
+      "learning_rate": 3.043695804385333e-07,
+      "loss": 0.3557,
+      "step": 11068
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.7299690473858185,
+      "learning_rate": 3.039214616158004e-07,
+      "loss": 0.4493,
+      "step": 11069
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.991147734021401,
+      "learning_rate": 3.034736625752532e-07,
+      "loss": 0.5637,
+      "step": 11070
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.800923681648986,
+      "learning_rate": 3.030261833473835e-07,
+      "loss": 0.4472,
+      "step": 11071
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.567717681315841,
+      "learning_rate": 3.025790239626641e-07,
+      "loss": 0.3448,
+      "step": 11072
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.768713636112588,
+      "learning_rate": 3.02132184451544e-07,
+      "loss": 0.3719,
+      "step": 11073
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.144509007086821,
+      "learning_rate": 3.016856648444494e-07,
+      "loss": 0.4949,
+      "step": 11074
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.251633133340724,
+      "learning_rate": 3.012394651717887e-07,
+      "loss": 0.5476,
+      "step": 11075
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.82162792171854,
+      "learning_rate": 3.007935854639443e-07,
+      "loss": 0.5642,
+      "step": 11076
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.968777976353014,
+      "learning_rate": 3.0034802575128084e-07,
+      "loss": 0.5116,
+      "step": 11077
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.026989551865115,
+      "learning_rate": 2.9990278606413625e-07,
+      "loss": 0.4347,
+      "step": 11078
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.309514651898961,
+      "learning_rate": 2.9945786643283124e-07,
+      "loss": 0.3309,
+      "step": 11079
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.972949458178303,
+      "learning_rate": 2.9901326688766106e-07,
+      "loss": 0.4299,
+      "step": 11080
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.940520002315536,
+      "learning_rate": 2.9856898745890375e-07,
+      "loss": 0.5035,
+      "step": 11081
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.658009793051552,
+      "learning_rate": 2.9812502817681056e-07,
+      "loss": 0.4932,
+      "step": 11082
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.772578137799124,
+      "learning_rate": 2.9768138907161347e-07,
+      "loss": 0.5213,
+      "step": 11083
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.473548734558501,
+      "learning_rate": 2.9723807017352335e-07,
+      "loss": 0.4284,
+      "step": 11084
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 28.515728278896276,
+      "learning_rate": 2.967950715127266e-07,
+      "loss": 0.3868,
+      "step": 11085
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 11.219139402493171,
+      "learning_rate": 2.963523931193918e-07,
+      "loss": 0.5029,
+      "step": 11086
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.633766690523634,
+      "learning_rate": 2.95910035023661e-07,
+      "loss": 0.4518,
+      "step": 11087
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.593509834150169,
+      "learning_rate": 2.9546799725565844e-07,
+      "loss": 0.4152,
+      "step": 11088
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.731898931521043,
+      "learning_rate": 2.9502627984548395e-07,
+      "loss": 0.4668,
+      "step": 11089
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.56861963636143,
+      "learning_rate": 2.945848828232173e-07,
+      "loss": 0.4297,
+      "step": 11090
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.146050519847777,
+      "learning_rate": 2.9414380621891556e-07,
+      "loss": 0.4743,
+      "step": 11091
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.239327569431737,
+      "learning_rate": 2.9370305006261367e-07,
+      "loss": 0.5574,
+      "step": 11092
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.43506535424057,
+      "learning_rate": 2.932626143843248e-07,
+      "loss": 0.5285,
+      "step": 11093
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.88139302653387,
+      "learning_rate": 2.928224992140416e-07,
+      "loss": 0.4413,
+      "step": 11094
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.012126067970409,
+      "learning_rate": 2.92382704581734e-07,
+      "loss": 0.4035,
+      "step": 11095
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.154975881326154,
+      "learning_rate": 2.9194323051734975e-07,
+      "loss": 0.5268,
+      "step": 11096
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.042636360157579,
+      "learning_rate": 2.9150407705081483e-07,
+      "loss": 0.5312,
+      "step": 11097
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.6873575135514125,
+      "learning_rate": 2.910652442120332e-07,
+      "loss": 0.4531,
+      "step": 11098
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.743780840713411,
+      "learning_rate": 2.9062673203088865e-07,
+      "loss": 0.4653,
+      "step": 11099
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.758900077434269,
+      "learning_rate": 2.901885405372418e-07,
+      "loss": 0.4392,
+      "step": 11100
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.607872723079883,
+      "learning_rate": 2.8975066976093094e-07,
+      "loss": 0.4394,
+      "step": 11101
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.193069031207871,
+      "learning_rate": 2.893131197317722e-07,
+      "loss": 0.4851,
+      "step": 11102
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.511058439390941,
+      "learning_rate": 2.888758904795624e-07,
+      "loss": 0.4638,
+      "step": 11103
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.50303363610532,
+      "learning_rate": 2.8843898203407494e-07,
+      "loss": 0.3786,
+      "step": 11104
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.938282455397818,
+      "learning_rate": 2.8800239442506097e-07,
+      "loss": 0.4259,
+      "step": 11105
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.83008404925349,
+      "learning_rate": 2.8756612768225e-07,
+      "loss": 0.458,
+      "step": 11106
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.068763473817459,
+      "learning_rate": 2.871301818353495e-07,
+      "loss": 0.4246,
+      "step": 11107
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.642642731426982,
+      "learning_rate": 2.866945569140461e-07,
+      "loss": 0.4265,
+      "step": 11108
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4664288439194868,
+      "learning_rate": 2.862592529480035e-07,
+      "loss": 0.695,
+      "step": 11109
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.174032832850336,
+      "learning_rate": 2.8582426996686453e-07,
+      "loss": 0.5047,
+      "step": 11110
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.626272151024884,
+      "learning_rate": 2.8538960800024887e-07,
+      "loss": 0.4002,
+      "step": 11111
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.290315452728261,
+      "learning_rate": 2.849552670777556e-07,
+      "loss": 0.4369,
+      "step": 11112
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.822767214752867,
+      "learning_rate": 2.8452124722896103e-07,
+      "loss": 0.3807,
+      "step": 11113
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.862250424111432,
+      "learning_rate": 2.8408754848342044e-07,
+      "loss": 0.4165,
+      "step": 11114
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.166435094794241,
+      "learning_rate": 2.8365417087066693e-07,
+      "loss": 0.5018,
+      "step": 11115
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 31.31521338112398,
+      "learning_rate": 2.8322111442021017e-07,
+      "loss": 0.434,
+      "step": 11116
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.134876895192364,
+      "learning_rate": 2.8278837916154157e-07,
+      "loss": 0.4616,
+      "step": 11117
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.723480902570563,
+      "learning_rate": 2.8235596512412757e-07,
+      "loss": 0.4988,
+      "step": 11118
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.244784872881556,
+      "learning_rate": 2.819238723374135e-07,
+      "loss": 0.4682,
+      "step": 11119
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.765299092152621,
+      "learning_rate": 2.814921008308219e-07,
+      "loss": 0.7125,
+      "step": 11120
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.75163811659303,
+      "learning_rate": 2.8106065063375596e-07,
+      "loss": 0.4659,
+      "step": 11121
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.296974995331538,
+      "learning_rate": 2.80629521775595e-07,
+      "loss": 0.5055,
+      "step": 11122
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.449873204025407,
+      "learning_rate": 2.801987142856977e-07,
+      "loss": 0.4451,
+      "step": 11123
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.380979060075322,
+      "learning_rate": 2.7976822819339943e-07,
+      "loss": 0.3657,
+      "step": 11124
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.604518732223517,
+      "learning_rate": 2.7933806352801405e-07,
+      "loss": 0.4925,
+      "step": 11125
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.52995966924385,
+      "learning_rate": 2.789082203188348e-07,
+      "loss": 0.4655,
+      "step": 11126
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.968400151844385,
+      "learning_rate": 2.784786985951315e-07,
+      "loss": 0.4081,
+      "step": 11127
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.504937179172753,
+      "learning_rate": 2.780494983861531e-07,
+      "loss": 0.4941,
+      "step": 11128
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.611881159284488,
+      "learning_rate": 2.776206197211251e-07,
+      "loss": 0.4745,
+      "step": 11129
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.790712014471988,
+      "learning_rate": 2.77192062629254e-07,
+      "loss": 0.5265,
+      "step": 11130
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.608549567507776,
+      "learning_rate": 2.7676382713972106e-07,
+      "loss": 0.4428,
+      "step": 11131
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.080628735955828,
+      "learning_rate": 2.763359132816884e-07,
+      "loss": 0.5175,
+      "step": 11132
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.68460510502026,
+      "learning_rate": 2.7590832108429443e-07,
+      "loss": 0.5048,
+      "step": 11133
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5105254420442997,
+      "learning_rate": 2.754810505766559e-07,
+      "loss": 0.6628,
+      "step": 11134
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.7606670563080895,
+      "learning_rate": 2.750541017878694e-07,
+      "loss": 0.4256,
+      "step": 11135
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.865027807556842,
+      "learning_rate": 2.7462747474700737e-07,
+      "loss": 0.4376,
+      "step": 11136
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.969776577078873,
+      "learning_rate": 2.7420116948312145e-07,
+      "loss": 0.5146,
+      "step": 11137
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.751277445982658,
+      "learning_rate": 2.737751860252397e-07,
+      "loss": 0.4701,
+      "step": 11138
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.8081411908067055,
+      "learning_rate": 2.7334952440237206e-07,
+      "loss": 0.4223,
+      "step": 11139
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.44729652863358,
+      "learning_rate": 2.7292418464350265e-07,
+      "loss": 0.4539,
+      "step": 11140
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.398124542535656,
+      "learning_rate": 2.7249916677759613e-07,
+      "loss": 0.4665,
+      "step": 11141
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.435862457782296,
+      "learning_rate": 2.7207447083359375e-07,
+      "loss": 0.5381,
+      "step": 11142
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.54314186111477,
+      "learning_rate": 2.716500968404151e-07,
+      "loss": 0.3931,
+      "step": 11143
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.110041120736164,
+      "learning_rate": 2.7122604482695935e-07,
+      "loss": 0.5653,
+      "step": 11144
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.880950060882165,
+      "learning_rate": 2.7080231482210174e-07,
+      "loss": 0.4293,
+      "step": 11145
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.16247870144748,
+      "learning_rate": 2.703789068546969e-07,
+      "loss": 0.4803,
+      "step": 11146
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.439920477518887,
+      "learning_rate": 2.6995582095357576e-07,
+      "loss": 0.4758,
+      "step": 11147
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.345436945896317,
+      "learning_rate": 2.695330571475502e-07,
+      "loss": 0.5204,
+      "step": 11148
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.386709529681008,
+      "learning_rate": 2.691106154654072e-07,
+      "loss": 0.3803,
+      "step": 11149
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.545983846984111,
+      "learning_rate": 2.686884959359148e-07,
+      "loss": 0.428,
+      "step": 11150
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 38.94694319640496,
+      "learning_rate": 2.682666985878163e-07,
+      "loss": 0.5143,
+      "step": 11151
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.691770021051591,
+      "learning_rate": 2.678452234498341e-07,
+      "loss": 0.4355,
+      "step": 11152
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.237560660727679,
+      "learning_rate": 2.6742407055066924e-07,
+      "loss": 0.5205,
+      "step": 11153
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.2791959677721305,
+      "learning_rate": 2.6700323991899977e-07,
+      "loss": 0.3827,
+      "step": 11154
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.875929266072708,
+      "learning_rate": 2.665827315834846e-07,
+      "loss": 0.5127,
+      "step": 11155
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.118011127402325,
+      "learning_rate": 2.661625455727551e-07,
+      "loss": 0.472,
+      "step": 11156
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.215090104654741,
+      "learning_rate": 2.6574268191542687e-07,
+      "loss": 0.4196,
+      "step": 11157
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.672380040963694,
+      "learning_rate": 2.6532314064008856e-07,
+      "loss": 0.5163,
+      "step": 11158
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.037624020216674,
+      "learning_rate": 2.649039217753108e-07,
+      "loss": 0.475,
+      "step": 11159
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 19.013726339796705,
+      "learning_rate": 2.644850253496406e-07,
+      "loss": 0.3906,
+      "step": 11160
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.916864092991442,
+      "learning_rate": 2.640664513916019e-07,
+      "loss": 0.4775,
+      "step": 11161
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.089930221180044,
+      "learning_rate": 2.6364819992969747e-07,
+      "loss": 0.4142,
+      "step": 11162
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.126234311333005,
+      "learning_rate": 2.6323027099240893e-07,
+      "loss": 0.4263,
+      "step": 11163
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5818079181249538,
+      "learning_rate": 2.6281266460819685e-07,
+      "loss": 0.6792,
+      "step": 11164
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.618604435993939,
+      "learning_rate": 2.623953808054963e-07,
+      "loss": 0.4118,
+      "step": 11165
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.615265021729274,
+      "learning_rate": 2.619784196127234e-07,
+      "loss": 0.4176,
+      "step": 11166
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.922820552560597,
+      "learning_rate": 2.6156178105827044e-07,
+      "loss": 0.5061,
+      "step": 11167
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.217919085936318,
+      "learning_rate": 2.611454651705103e-07,
+      "loss": 0.4428,
+      "step": 11168
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.462207676085954,
+      "learning_rate": 2.6072947197779087e-07,
+      "loss": 0.4335,
+      "step": 11169
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.744462159699722,
+      "learning_rate": 2.6031380150844053e-07,
+      "loss": 0.5056,
+      "step": 11170
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.280272988450974,
+      "learning_rate": 2.5989845379076285e-07,
+      "loss": 0.4755,
+      "step": 11171
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 31.806973614040835,
+      "learning_rate": 2.594834288530429e-07,
+      "loss": 0.494,
+      "step": 11172
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.1910664634334776,
+      "learning_rate": 2.59068726723542e-07,
+      "loss": 0.4131,
+      "step": 11173
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.586091411073373,
+      "learning_rate": 2.586543474304981e-07,
+      "loss": 0.4648,
+      "step": 11174
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.929748341710359,
+      "learning_rate": 2.5824029100212977e-07,
+      "loss": 0.3856,
+      "step": 11175
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 34.17851887047427,
+      "learning_rate": 2.5782655746663165e-07,
+      "loss": 0.4421,
+      "step": 11176
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.578998621982564,
+      "learning_rate": 2.5741314685217846e-07,
+      "loss": 0.5433,
+      "step": 11177
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.309939623130838,
+      "learning_rate": 2.5700005918692094e-07,
+      "loss": 0.4487,
+      "step": 11178
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.967103580918945,
+      "learning_rate": 2.5658729449898835e-07,
+      "loss": 0.4622,
+      "step": 11179
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8132866322343668,
+      "learning_rate": 2.561748528164876e-07,
+      "loss": 0.6217,
+      "step": 11180
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.293871997799272,
+      "learning_rate": 2.557627341675051e-07,
+      "loss": 0.3595,
+      "step": 11181
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.364949803319018,
+      "learning_rate": 2.553509385801045e-07,
+      "loss": 0.4543,
+      "step": 11182
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.591562640000382,
+      "learning_rate": 2.549394660823257e-07,
+      "loss": 0.4783,
+      "step": 11183
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 15.250181840646642,
+      "learning_rate": 2.545283167021895e-07,
+      "loss": 0.5081,
+      "step": 11184
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.254293990340845,
+      "learning_rate": 2.5411749046769295e-07,
+      "loss": 0.5174,
+      "step": 11185
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.810213900311329,
+      "learning_rate": 2.537069874068121e-07,
+      "loss": 0.4889,
+      "step": 11186
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.252098917294296,
+      "learning_rate": 2.5329680754749953e-07,
+      "loss": 0.43,
+      "step": 11187
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.187291029251357,
+      "learning_rate": 2.5288695091768734e-07,
+      "loss": 0.3397,
+      "step": 11188
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.038189339465995,
+      "learning_rate": 2.5247741754528376e-07,
+      "loss": 0.4056,
+      "step": 11189
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.3325300324872345,
+      "learning_rate": 2.520682074581782e-07,
+      "loss": 0.4146,
+      "step": 11190
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6735502050911335,
+      "learning_rate": 2.5165932068423384e-07,
+      "loss": 0.4017,
+      "step": 11191
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.893914687511943,
+      "learning_rate": 2.5125075725129675e-07,
+      "loss": 0.4126,
+      "step": 11192
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.952471338630371,
+      "learning_rate": 2.5084251718718535e-07,
+      "loss": 0.6034,
+      "step": 11193
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.899699102725438,
+      "learning_rate": 2.5043460051970004e-07,
+      "loss": 0.4003,
+      "step": 11194
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.8636681824313985,
+      "learning_rate": 2.500270072766192e-07,
+      "loss": 0.5363,
+      "step": 11195
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.70453661083232,
+      "learning_rate": 2.4961973748569735e-07,
+      "loss": 0.4786,
+      "step": 11196
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.829287461104516,
+      "learning_rate": 2.4921279117466836e-07,
+      "loss": 0.4137,
+      "step": 11197
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.193949744525236,
+      "learning_rate": 2.488061683712417e-07,
+      "loss": 0.4736,
+      "step": 11198
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 22.851400529827398,
+      "learning_rate": 2.483998691031081e-07,
+      "loss": 0.4031,
+      "step": 11199
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.073551141977198,
+      "learning_rate": 2.4799389339793425e-07,
+      "loss": 0.4367,
+      "step": 11200
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.897425798476869,
+      "learning_rate": 2.475882412833669e-07,
+      "loss": 0.5001,
+      "step": 11201
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.99579517258518,
+      "learning_rate": 2.471829127870262e-07,
+      "loss": 0.4974,
+      "step": 11202
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.835972391119814,
+      "learning_rate": 2.467779079365151e-07,
+      "loss": 0.5658,
+      "step": 11203
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.615061854617615,
+      "learning_rate": 2.463732267594121e-07,
+      "loss": 0.4447,
+      "step": 11204
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.261233622534124,
+      "learning_rate": 2.4596886928327513e-07,
+      "loss": 0.364,
+      "step": 11205
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.868919145039918,
+      "learning_rate": 2.455648355356383e-07,
+      "loss": 0.5422,
+      "step": 11206
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8663499398344046,
+      "learning_rate": 2.451611255440134e-07,
+      "loss": 0.4107,
+      "step": 11207
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.526647685201222,
+      "learning_rate": 2.4475773933589407e-07,
+      "loss": 0.365,
+      "step": 11208
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.299748637285886,
+      "learning_rate": 2.4435467693874614e-07,
+      "loss": 0.4538,
+      "step": 11209
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8859953881853337,
+      "learning_rate": 2.4395193838001985e-07,
+      "loss": 0.6806,
+      "step": 11210
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.646639736611304,
+      "learning_rate": 2.4354952368713605e-07,
+      "loss": 0.4476,
+      "step": 11211
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8766550875061423,
+      "learning_rate": 2.4314743288750053e-07,
+      "loss": 0.4461,
+      "step": 11212
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.614364183337272,
+      "learning_rate": 2.427456660084915e-07,
+      "loss": 0.5154,
+      "step": 11213
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.446288308949917,
+      "learning_rate": 2.4234422307746916e-07,
+      "loss": 0.5512,
+      "step": 11214
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.251106179989053,
+      "learning_rate": 2.4194310412176955e-07,
+      "loss": 0.4706,
+      "step": 11215
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 46.94818969398241,
+      "learning_rate": 2.4154230916870635e-07,
+      "loss": 0.5166,
+      "step": 11216
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.446622209338635,
+      "learning_rate": 2.411418382455732e-07,
+      "loss": 0.437,
+      "step": 11217
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.53184104018517,
+      "learning_rate": 2.4074169137963955e-07,
+      "loss": 0.4586,
+      "step": 11218
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.969227729244851,
+      "learning_rate": 2.403418685981551e-07,
+      "loss": 0.4246,
+      "step": 11219
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.404884486967389,
+      "learning_rate": 2.3994236992834317e-07,
+      "loss": 0.5492,
+      "step": 11220
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.810614053354511,
+      "learning_rate": 2.395431953974098e-07,
+      "loss": 0.5379,
+      "step": 11221
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.248221297327244,
+      "learning_rate": 2.391443450325365e-07,
+      "loss": 0.3634,
+      "step": 11222
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.257606088443147,
+      "learning_rate": 2.3874581886088375e-07,
+      "loss": 0.477,
+      "step": 11223
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.25165231092015,
+      "learning_rate": 2.383476169095894e-07,
+      "loss": 0.4107,
+      "step": 11224
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 19.952399117689943,
+      "learning_rate": 2.3794973920576835e-07,
+      "loss": 0.4839,
+      "step": 11225
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.238833899091113,
+      "learning_rate": 2.3755218577651506e-07,
+      "loss": 0.4728,
+      "step": 11226
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.580534490130866,
+      "learning_rate": 2.3715495664890065e-07,
+      "loss": 0.3506,
+      "step": 11227
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.801100219013474,
+      "learning_rate": 2.3675805184997512e-07,
+      "loss": 0.4384,
+      "step": 11228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.912488381203095,
+      "learning_rate": 2.3636147140676634e-07,
+      "loss": 0.4763,
+      "step": 11229
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7047180900778383,
+      "learning_rate": 2.3596521534627882e-07,
+      "loss": 0.677,
+      "step": 11230
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.990159408357375,
+      "learning_rate": 2.3556928369549537e-07,
+      "loss": 0.4796,
+      "step": 11231
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.919713846011853,
+      "learning_rate": 2.351736764813789e-07,
+      "loss": 0.4561,
+      "step": 11232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.065161746680585,
+      "learning_rate": 2.3477839373086787e-07,
+      "loss": 0.5326,
+      "step": 11233
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.455600312803747,
+      "learning_rate": 2.3438343547087793e-07,
+      "loss": 0.3976,
+      "step": 11234
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.269296518362269,
+      "learning_rate": 2.3398880172830596e-07,
+      "loss": 0.4236,
+      "step": 11235
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.891150145328712,
+      "learning_rate": 2.335944925300232e-07,
+      "loss": 0.5039,
+      "step": 11236
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.836085832682652,
+      "learning_rate": 2.3320050790288152e-07,
+      "loss": 0.4185,
+      "step": 11237
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.8314628530025985,
+      "learning_rate": 2.3280684787370888e-07,
+      "loss": 0.5294,
+      "step": 11238
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.113212155904026,
+      "learning_rate": 2.3241351246931222e-07,
+      "loss": 0.4114,
+      "step": 11239
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.497031473260663,
+      "learning_rate": 2.3202050171647505e-07,
+      "loss": 0.651,
+      "step": 11240
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.411573693303693,
+      "learning_rate": 2.31627815641961e-07,
+      "loss": 0.488,
+      "step": 11241
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6069597620197664,
+      "learning_rate": 2.3123545427250926e-07,
+      "loss": 0.6759,
+      "step": 11242
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.356064200488767,
+      "learning_rate": 2.308434176348384e-07,
+      "loss": 0.5243,
+      "step": 11243
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.207465003880337,
+      "learning_rate": 2.304517057556438e-07,
+      "loss": 0.3411,
+      "step": 11244
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 14.766777509901718,
+      "learning_rate": 2.3006031866159906e-07,
+      "loss": 0.555,
+      "step": 11245
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.337948337253251,
+      "learning_rate": 2.2966925637935789e-07,
+      "loss": 0.4764,
+      "step": 11246
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.726714497905624,
+      "learning_rate": 2.2927851893554786e-07,
+      "loss": 0.4263,
+      "step": 11247
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.00995656227278,
+      "learning_rate": 2.288881063567777e-07,
+      "loss": 0.4764,
+      "step": 11248
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.469355051131018,
+      "learning_rate": 2.2849801866963116e-07,
+      "loss": 0.397,
+      "step": 11249
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 15.280282958025603,
+      "learning_rate": 2.2810825590067366e-07,
+      "loss": 0.4394,
+      "step": 11250
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.012699887556675,
+      "learning_rate": 2.2771881807644502e-07,
+      "loss": 0.4169,
+      "step": 11251
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.234006300597439,
+      "learning_rate": 2.273297052234641e-07,
+      "loss": 0.593,
+      "step": 11252
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.499220699596461,
+      "learning_rate": 2.26940917368228e-07,
+      "loss": 0.4443,
+      "step": 11253
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.62713750110424,
+      "learning_rate": 2.265524545372111e-07,
+      "loss": 0.438,
+      "step": 11254
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.206362060277845,
+      "learning_rate": 2.2616431675686724e-07,
+      "loss": 0.4874,
+      "step": 11255
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.670340713316513,
+      "learning_rate": 2.257765040536264e-07,
+      "loss": 0.3688,
+      "step": 11256
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.691344285164969,
+      "learning_rate": 2.2538901645389633e-07,
+      "loss": 0.5253,
+      "step": 11257
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.81907749393082,
+      "learning_rate": 2.250018539840626e-07,
+      "loss": 0.3687,
+      "step": 11258
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.499218357296148,
+      "learning_rate": 2.2461501667049136e-07,
+      "loss": 0.3958,
+      "step": 11259
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.896942835597738,
+      "learning_rate": 2.2422850453952316e-07,
+      "loss": 0.4742,
+      "step": 11260
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.139217545171032,
+      "learning_rate": 2.2384231761747755e-07,
+      "loss": 0.4725,
+      "step": 11261
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.699861892585537,
+      "learning_rate": 2.2345645593065235e-07,
+      "loss": 0.3869,
+      "step": 11262
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.708588386041593,
+      "learning_rate": 2.2307091950532379e-07,
+      "loss": 0.5299,
+      "step": 11263
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.173071292844721,
+      "learning_rate": 2.226857083677436e-07,
+      "loss": 0.4817,
+      "step": 11264
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.540237569351094,
+      "learning_rate": 2.2230082254414475e-07,
+      "loss": 0.4684,
+      "step": 11265
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.97991953635067,
+      "learning_rate": 2.219162620607357e-07,
+      "loss": 0.5033,
+      "step": 11266
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.625539542474722,
+      "learning_rate": 2.215320269437027e-07,
+      "loss": 0.4684,
+      "step": 11267
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.455956563853172,
+      "learning_rate": 2.2114811721921093e-07,
+      "loss": 0.4339,
+      "step": 11268
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.708622196568424,
+      "learning_rate": 2.2076453291340338e-07,
+      "loss": 0.4986,
+      "step": 11269
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.049210722732215,
+      "learning_rate": 2.2038127405239974e-07,
+      "loss": 0.5392,
+      "step": 11270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 20.61905293133141,
+      "learning_rate": 2.1999834066229797e-07,
+      "loss": 0.4322,
+      "step": 11271
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.586074764924544,
+      "learning_rate": 2.1961573276917504e-07,
+      "loss": 0.5207,
+      "step": 11272
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.976746841163992,
+      "learning_rate": 2.1923345039908395e-07,
+      "loss": 0.4548,
+      "step": 11273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.061184898463614,
+      "learning_rate": 2.1885149357805723e-07,
+      "loss": 0.4576,
+      "step": 11274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.912103724882778,
+      "learning_rate": 2.1846986233210465e-07,
+      "loss": 0.38,
+      "step": 11275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.893752756670171,
+      "learning_rate": 2.1808855668721262e-07,
+      "loss": 0.5045,
+      "step": 11276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.874894163249511,
+      "learning_rate": 2.1770757666934706e-07,
+      "loss": 0.4992,
+      "step": 11277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.576326103145101,
+      "learning_rate": 2.1732692230445051e-07,
+      "loss": 0.3682,
+      "step": 11278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9494565162067028,
+      "learning_rate": 2.1694659361844506e-07,
+      "loss": 0.6687,
+      "step": 11279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.901527269907655,
+      "learning_rate": 2.1656659063722718e-07,
+      "loss": 0.4844,
+      "step": 11280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.061704974299484,
+      "learning_rate": 2.1618691338667563e-07,
+      "loss": 0.5517,
+      "step": 11281
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.375017545593619,
+      "learning_rate": 2.1580756189264307e-07,
+      "loss": 0.4882,
+      "step": 11282
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.20849862042661,
+      "learning_rate": 2.1542853618096327e-07,
+      "loss": 0.6062,
+      "step": 11283
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.426508576794566,
+      "learning_rate": 2.1504983627744559e-07,
+      "loss": 0.3518,
+      "step": 11284
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.594955782841906,
+      "learning_rate": 2.1467146220787661e-07,
+      "loss": 0.4218,
+      "step": 11285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.206534250328508,
+      "learning_rate": 2.142934139980235e-07,
+      "loss": 0.3566,
+      "step": 11286
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.824904162919672,
+      "learning_rate": 2.1391569167362848e-07,
+      "loss": 0.3717,
+      "step": 11287
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5397035289302379,
+      "learning_rate": 2.135382952604148e-07,
+      "loss": 0.6585,
+      "step": 11288
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.78025479174655,
+      "learning_rate": 2.1316122478407863e-07,
+      "loss": 0.478,
+      "step": 11289
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.6817339083716325,
+      "learning_rate": 2.127844802702994e-07,
+      "loss": 0.4159,
+      "step": 11290
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.800514896352382,
+      "learning_rate": 2.1240806174472938e-07,
+      "loss": 0.7626,
+      "step": 11291
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.416353514744097,
+      "learning_rate": 2.1203196923300306e-07,
+      "loss": 0.4596,
+      "step": 11292
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.708240609444541,
+      "learning_rate": 2.1165620276072997e-07,
+      "loss": 0.4603,
+      "step": 11293
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.928825129363538,
+      "learning_rate": 2.1128076235349738e-07,
+      "loss": 0.4895,
+      "step": 11294
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.3634458304368975,
+      "learning_rate": 2.1090564803687208e-07,
+      "loss": 0.4708,
+      "step": 11295
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.193536914605929,
+      "learning_rate": 2.1053085983639698e-07,
+      "loss": 0.383,
+      "step": 11296
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.5230970984997265,
+      "learning_rate": 2.1015639777759555e-07,
+      "loss": 0.4749,
+      "step": 11297
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.575948235592673,
+      "learning_rate": 2.0978226188596352e-07,
+      "loss": 0.4484,
+      "step": 11298
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.503330809969674,
+      "learning_rate": 2.0940845218698047e-07,
+      "loss": 0.481,
+      "step": 11299
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6130425979904857,
+      "learning_rate": 2.0903496870609996e-07,
+      "loss": 0.6814,
+      "step": 11300
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.079761234784539,
+      "learning_rate": 2.0866181146875552e-07,
+      "loss": 0.5186,
+      "step": 11301
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.445964050960354,
+      "learning_rate": 2.0828898050035684e-07,
+      "loss": 0.377,
+      "step": 11302
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.689495444033852,
+      "learning_rate": 2.079164758262925e-07,
+      "loss": 0.3381,
+      "step": 11303
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.865095670406068,
+      "learning_rate": 2.0754429747192717e-07,
+      "loss": 0.4356,
+      "step": 11304
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.300021298659329,
+      "learning_rate": 2.0717244546260617e-07,
+      "loss": 0.4741,
+      "step": 11305
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.23807557519792,
+      "learning_rate": 2.0680091982365147e-07,
+      "loss": 0.4745,
+      "step": 11306
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.699197517607991,
+      "learning_rate": 2.064297205803595e-07,
+      "loss": 0.4761,
+      "step": 11307
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.355384947974976,
+      "learning_rate": 2.0605884775801e-07,
+      "loss": 0.3795,
+      "step": 11308
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7825249208023366,
+      "learning_rate": 2.0568830138185613e-07,
+      "loss": 0.7145,
+      "step": 11309
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.848936487272054,
+      "learning_rate": 2.053180814771316e-07,
+      "loss": 0.4993,
+      "step": 11310
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.492420864640309,
+      "learning_rate": 2.0494818806904627e-07,
+      "loss": 0.5001,
+      "step": 11311
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.334026424067172,
+      "learning_rate": 2.0457862118278825e-07,
+      "loss": 0.5073,
+      "step": 11312
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.648079248575381,
+      "learning_rate": 2.042093808435225e-07,
+      "loss": 0.5123,
+      "step": 11313
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.827262044006549,
+      "learning_rate": 2.0384046707639383e-07,
+      "loss": 0.443,
+      "step": 11314
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.555259476882311,
+      "learning_rate": 2.034718799065244e-07,
+      "loss": 0.4654,
+      "step": 11315
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.929739902169473,
+      "learning_rate": 2.0310361935901135e-07,
+      "loss": 0.5575,
+      "step": 11316
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.505719408192287,
+      "learning_rate": 2.0273568545893295e-07,
+      "loss": 0.5617,
+      "step": 11317
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.83843710776294,
+      "learning_rate": 2.0236807823134253e-07,
+      "loss": 0.4706,
+      "step": 11318
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.259264638654124,
+      "learning_rate": 2.0200079770127445e-07,
+      "loss": 0.5488,
+      "step": 11319
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.709585760151824,
+      "learning_rate": 2.016338438937382e-07,
+      "loss": 0.4249,
+      "step": 11320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.25039737976286,
+      "learning_rate": 2.0126721683372097e-07,
+      "loss": 0.5595,
+      "step": 11321
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.30170935960185,
+      "learning_rate": 2.0090091654618838e-07,
+      "loss": 0.4966,
+      "step": 11322
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.663335935458789,
+      "learning_rate": 2.005349430560849e-07,
+      "loss": 0.3741,
+      "step": 11323
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.717981835166699,
+      "learning_rate": 2.001692963883306e-07,
+      "loss": 0.3711,
+      "step": 11324
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.1073823140489605,
+      "learning_rate": 1.9980397656782558e-07,
+      "loss": 0.3786,
+      "step": 11325
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.741141320138104,
+      "learning_rate": 1.9943898361944603e-07,
+      "loss": 0.4464,
+      "step": 11326
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.359542648243201,
+      "learning_rate": 1.990743175680454e-07,
+      "loss": 0.4578,
+      "step": 11327
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.358462849829406,
+      "learning_rate": 1.987099784384572e-07,
+      "loss": 0.5799,
+      "step": 11328
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.936658098730845,
+      "learning_rate": 1.9834596625549098e-07,
+      "loss": 0.5325,
+      "step": 11329
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.993337537254432,
+      "learning_rate": 1.9798228104393414e-07,
+      "loss": 0.3619,
+      "step": 11330
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.786973775257433,
+      "learning_rate": 1.9761892282855188e-07,
+      "loss": 0.3953,
+      "step": 11331
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.187426684647622,
+      "learning_rate": 1.972558916340872e-07,
+      "loss": 0.4751,
+      "step": 11332
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.310726307113148,
+      "learning_rate": 1.9689318748526087e-07,
+      "loss": 0.3688,
+      "step": 11333
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.666932846014268,
+      "learning_rate": 1.9653081040677312e-07,
+      "loss": 0.4514,
+      "step": 11334
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.51441784290712,
+      "learning_rate": 1.9616876042329759e-07,
+      "loss": 0.4348,
+      "step": 11335
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.325757009137726,
+      "learning_rate": 1.9580703755948948e-07,
+      "loss": 0.3934,
+      "step": 11336
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.567496231514079,
+      "learning_rate": 1.9544564183998138e-07,
+      "loss": 0.4996,
+      "step": 11337
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.050856603581868,
+      "learning_rate": 1.9508457328938191e-07,
+      "loss": 0.4231,
+      "step": 11338
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 23.91179959599125,
+      "learning_rate": 1.9472383193227806e-07,
+      "loss": 0.3579,
+      "step": 11339
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.436841909170104,
+      "learning_rate": 1.9436341779323464e-07,
+      "loss": 0.4581,
+      "step": 11340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.221567782407266,
+      "learning_rate": 1.9400333089679478e-07,
+      "loss": 0.5167,
+      "step": 11341
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1717113011290876,
+      "learning_rate": 1.9364357126747836e-07,
+      "loss": 0.759,
+      "step": 11342
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.147737308305491,
+      "learning_rate": 1.932841389297846e-07,
+      "loss": 0.39,
+      "step": 11343
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 73.20161924550452,
+      "learning_rate": 1.9292503390818674e-07,
+      "loss": 0.5189,
+      "step": 11344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.283626779073568,
+      "learning_rate": 1.925662562271402e-07,
+      "loss": 0.4956,
+      "step": 11345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.225258335048545,
+      "learning_rate": 1.9220780591107658e-07,
+      "loss": 0.3468,
+      "step": 11346
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 41.51828280832801,
+      "learning_rate": 1.9184968298440353e-07,
+      "loss": 0.4889,
+      "step": 11347
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.510703170064492,
+      "learning_rate": 1.9149188747150827e-07,
+      "loss": 0.521,
+      "step": 11348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.281908238045267,
+      "learning_rate": 1.9113441939675402e-07,
+      "loss": 0.5035,
+      "step": 11349
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.986218384676774,
+      "learning_rate": 1.9077727878448416e-07,
+      "loss": 0.4607,
+      "step": 11350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 19.207473048266714,
+      "learning_rate": 1.904204656590175e-07,
+      "loss": 0.3774,
+      "step": 11351
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.365746827320601,
+      "learning_rate": 1.90063980044653e-07,
+      "loss": 0.4562,
+      "step": 11352
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.8728957642378505,
+      "learning_rate": 1.8970782196566339e-07,
+      "loss": 0.4408,
+      "step": 11353
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.283156344290727,
+      "learning_rate": 1.8935199144630323e-07,
+      "loss": 0.4255,
+      "step": 11354
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 16.973255681053647,
+      "learning_rate": 1.8899648851080142e-07,
+      "loss": 0.4826,
+      "step": 11355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.447424600072097,
+      "learning_rate": 1.886413131833681e-07,
+      "loss": 0.4414,
+      "step": 11356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.111684063430292,
+      "learning_rate": 1.8828646548818775e-07,
+      "loss": 0.5205,
+      "step": 11357
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.3411713815612565,
+      "learning_rate": 1.8793194544942383e-07,
+      "loss": 0.5152,
+      "step": 11358
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.13441033982416,
+      "learning_rate": 1.8757775309121872e-07,
+      "loss": 0.3318,
+      "step": 11359
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.091814435894198,
+      "learning_rate": 1.8722388843769035e-07,
+      "loss": 0.4347,
+      "step": 11360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.038900310873267,
+      "learning_rate": 1.8687035151293664e-07,
+      "loss": 0.583,
+      "step": 11361
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.8757672663438125,
+      "learning_rate": 1.8651714234102945e-07,
+      "loss": 0.3898,
+      "step": 11362
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.814204389560874,
+      "learning_rate": 1.8616426094602348e-07,
+      "loss": 0.3695,
+      "step": 11363
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.970931917629398,
+      "learning_rate": 1.858117073519461e-07,
+      "loss": 0.4018,
+      "step": 11364
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.603393541408445,
+      "learning_rate": 1.8545948158280646e-07,
+      "loss": 0.532,
+      "step": 11365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.627491471613471,
+      "learning_rate": 1.8510758366258873e-07,
+      "loss": 0.5244,
+      "step": 11366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.637492744986652,
+      "learning_rate": 1.8475601361525542e-07,
+      "loss": 0.3894,
+      "step": 11367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.027780668304597,
+      "learning_rate": 1.844047714647479e-07,
+      "loss": 0.4819,
+      "step": 11368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.463014711791086,
+      "learning_rate": 1.8405385723498258e-07,
+      "loss": 0.4244,
+      "step": 11369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.538591784701949,
+      "learning_rate": 1.8370327094985651e-07,
+      "loss": 0.4802,
+      "step": 11370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.032101473513803,
+      "learning_rate": 1.833530126332428e-07,
+      "loss": 0.4153,
+      "step": 11371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.832687642706837,
+      "learning_rate": 1.8300308230899232e-07,
+      "loss": 0.5169,
+      "step": 11372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.598500028619876,
+      "learning_rate": 1.826534800009333e-07,
+      "loss": 0.5046,
+      "step": 11373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.797695531397839,
+      "learning_rate": 1.8230420573287333e-07,
+      "loss": 0.5511,
+      "step": 11374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.623853247616913,
+      "learning_rate": 1.819552595285956e-07,
+      "loss": 0.486,
+      "step": 11375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0748940159315903,
+      "learning_rate": 1.8160664141186113e-07,
+      "loss": 0.7392,
+      "step": 11376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.130624405649426,
+      "learning_rate": 1.812583514064109e-07,
+      "loss": 0.3439,
+      "step": 11377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.49232961991837,
+      "learning_rate": 1.8091038953596042e-07,
+      "loss": 0.5012,
+      "step": 11378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.42230304039193,
+      "learning_rate": 1.8056275582420569e-07,
+      "loss": 0.4966,
+      "step": 11379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.58698811117311,
+      "learning_rate": 1.8021545029481836e-07,
+      "loss": 0.502,
+      "step": 11380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.177103392261301,
+      "learning_rate": 1.7986847297144782e-07,
+      "loss": 0.5003,
+      "step": 11381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.461069516418512,
+      "learning_rate": 1.7952182387772244e-07,
+      "loss": 0.462,
+      "step": 11382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8711685125769684,
+      "learning_rate": 1.7917550303724773e-07,
+      "loss": 0.4742,
+      "step": 11383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.74878585466319,
+      "learning_rate": 1.7882951047360597e-07,
+      "loss": 0.409,
+      "step": 11384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.741859830566998,
+      "learning_rate": 1.784838462103572e-07,
+      "loss": 0.4156,
+      "step": 11385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.660350761791588,
+      "learning_rate": 1.7813851027104146e-07,
+      "loss": 0.3935,
+      "step": 11386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.469460155273441,
+      "learning_rate": 1.7779350267917273e-07,
+      "loss": 0.4566,
+      "step": 11387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.093922906342798,
+      "learning_rate": 1.7744882345824555e-07,
+      "loss": 0.4027,
+      "step": 11388
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.482348665677883,
+      "learning_rate": 1.7710447263173114e-07,
+      "loss": 0.4849,
+      "step": 11389
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.404027226922579,
+      "learning_rate": 1.7676045022307798e-07,
+      "loss": 0.4446,
+      "step": 11390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.456626323254357,
+      "learning_rate": 1.7641675625571175e-07,
+      "loss": 0.4402,
+      "step": 11391
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.218424671105299,
+      "learning_rate": 1.7607339075303765e-07,
+      "loss": 0.3994,
+      "step": 11392
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.701407215039443,
+      "learning_rate": 1.7573035373843695e-07,
+      "loss": 0.4627,
+      "step": 11393
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.627471748979173,
+      "learning_rate": 1.7538764523526874e-07,
+      "loss": 0.4713,
+      "step": 11394
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.173425281018418,
+      "learning_rate": 1.7504526526686938e-07,
+      "loss": 0.4142,
+      "step": 11395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 11.02585520152784,
+      "learning_rate": 1.7470321385655408e-07,
+      "loss": 0.5513,
+      "step": 11396
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.509627176758444,
+      "learning_rate": 1.743614910276159e-07,
+      "loss": 0.4945,
+      "step": 11397
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.138290579919941,
+      "learning_rate": 1.7402009680332398e-07,
+      "loss": 0.3919,
+      "step": 11398
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.127385607585135,
+      "learning_rate": 1.7367903120692531e-07,
+      "loss": 0.364,
+      "step": 11399
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.954504335057699,
+      "learning_rate": 1.7333829426164516e-07,
+      "loss": 0.4219,
+      "step": 11400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.347898013106509,
+      "learning_rate": 1.7299788599068611e-07,
+      "loss": 0.4602,
+      "step": 11401
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6985404939853521,
+      "learning_rate": 1.7265780641722907e-07,
+      "loss": 0.6854,
+      "step": 11402
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.974314958640094,
+      "learning_rate": 1.7231805556443158e-07,
+      "loss": 0.4393,
+      "step": 11403
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.028333301531822,
+      "learning_rate": 1.7197863345542855e-07,
+      "loss": 0.4184,
+      "step": 11404
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.3234522389154035,
+      "learning_rate": 1.716395401133336e-07,
+      "loss": 0.4432,
+      "step": 11405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.416894831072497,
+      "learning_rate": 1.7130077556123782e-07,
+      "loss": 0.4591,
+      "step": 11406
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.126297506831954,
+      "learning_rate": 1.7096233982220988e-07,
+      "loss": 0.582,
+      "step": 11407
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.634289113228043,
+      "learning_rate": 1.7062423291929476e-07,
+      "loss": 0.4808,
+      "step": 11408
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.130489965739695,
+      "learning_rate": 1.702864548755162e-07,
+      "loss": 0.4686,
+      "step": 11409
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.641693837913127,
+      "learning_rate": 1.699490057138764e-07,
+      "loss": 0.4663,
+      "step": 11410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.5002470627020505,
+      "learning_rate": 1.696118854573531e-07,
+      "loss": 0.4609,
+      "step": 11411
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.740564760338744,
+      "learning_rate": 1.6927509412890342e-07,
+      "loss": 0.4251,
+      "step": 11412
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 21.434808986652794,
+      "learning_rate": 1.689386317514602e-07,
+      "loss": 0.524,
+      "step": 11413
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.155170264481313,
+      "learning_rate": 1.6860249834793617e-07,
+      "loss": 0.5573,
+      "step": 11414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.256884385825666,
+      "learning_rate": 1.6826669394122032e-07,
+      "loss": 0.6095,
+      "step": 11415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.59942196950006,
+      "learning_rate": 1.6793121855417927e-07,
+      "loss": 0.5243,
+      "step": 11416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.081261965412626,
+      "learning_rate": 1.6759607220965758e-07,
+      "loss": 0.5343,
+      "step": 11417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.013944454395471,
+      "learning_rate": 1.672612549304764e-07,
+      "loss": 0.5782,
+      "step": 11418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.239372296817901,
+      "learning_rate": 1.6692676673943697e-07,
+      "loss": 0.4184,
+      "step": 11419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.300791089249775,
+      "learning_rate": 1.6659260765931495e-07,
+      "loss": 0.4378,
+      "step": 11420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7065178515862014,
+      "learning_rate": 1.6625877771286602e-07,
+      "loss": 0.682,
+      "step": 11421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8784912781257763,
+      "learning_rate": 1.6592527692282145e-07,
+      "loss": 0.7269,
+      "step": 11422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.6719539737959686,
+      "learning_rate": 1.6559210531189251e-07,
+      "loss": 0.542,
+      "step": 11423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.299158838765616,
+      "learning_rate": 1.652592629027655e-07,
+      "loss": 0.3391,
+      "step": 11424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.939094261749924,
+      "learning_rate": 1.6492674971810675e-07,
+      "loss": 0.4423,
+      "step": 11425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.905281335184566,
+      "learning_rate": 1.645945657805581e-07,
+      "loss": 0.4136,
+      "step": 11426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.456775071589688,
+      "learning_rate": 1.6426271111273982e-07,
+      "loss": 0.3773,
+      "step": 11427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.930818899919905,
+      "learning_rate": 1.6393118573724998e-07,
+      "loss": 0.5065,
+      "step": 11428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.550805641956851,
+      "learning_rate": 1.635999896766638e-07,
+      "loss": 0.6215,
+      "step": 11429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.481832663210772,
+      "learning_rate": 1.6326912295353492e-07,
+      "loss": 0.4315,
+      "step": 11430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.55919266531699,
+      "learning_rate": 1.6293858559039256e-07,
+      "loss": 0.4512,
+      "step": 11431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.418835121687177,
+      "learning_rate": 1.6260837760974645e-07,
+      "loss": 0.4854,
+      "step": 11432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.132333646813521,
+      "learning_rate": 1.6227849903408088e-07,
+      "loss": 0.486,
+      "step": 11433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.934057697428099,
+      "learning_rate": 1.6194894988586008e-07,
+      "loss": 0.5179,
+      "step": 11434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.314329509573874,
+      "learning_rate": 1.6161973018752443e-07,
+      "loss": 0.3701,
+      "step": 11435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.6295103854596835,
+      "learning_rate": 1.612908399614921e-07,
+      "loss": 0.4433,
+      "step": 11436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6928666517087585,
+      "learning_rate": 1.6096227923016018e-07,
+      "loss": 0.6611,
+      "step": 11437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.592686816929904,
+      "learning_rate": 1.6063404801590076e-07,
+      "loss": 0.4754,
+      "step": 11438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.094046902486669,
+      "learning_rate": 1.6030614634106656e-07,
+      "loss": 0.3692,
+      "step": 11439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.052906016445789,
+      "learning_rate": 1.5997857422798413e-07,
+      "loss": 0.4245,
+      "step": 11440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.196399924116581,
+      "learning_rate": 1.5965133169896176e-07,
+      "loss": 0.3972,
+      "step": 11441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.8914956341525455,
+      "learning_rate": 1.5932441877628157e-07,
+      "loss": 0.557,
+      "step": 11442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.687496373603961,
+      "learning_rate": 1.5899783548220637e-07,
+      "loss": 0.456,
+      "step": 11443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.657293288708691,
+      "learning_rate": 1.586715818389739e-07,
+      "loss": 0.3892,
+      "step": 11444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.226217310782827,
+      "learning_rate": 1.5834565786880085e-07,
+      "loss": 0.4912,
+      "step": 11445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.232358882909963,
+      "learning_rate": 1.5802006359388166e-07,
+      "loss": 0.4893,
+      "step": 11446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.477832535108857,
+      "learning_rate": 1.5769479903638697e-07,
+      "loss": 0.3476,
+      "step": 11447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.074120640624336,
+      "learning_rate": 1.5736986421846734e-07,
+      "loss": 0.3887,
+      "step": 11448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.0440362298410015,
+      "learning_rate": 1.5704525916224734e-07,
+      "loss": 0.3736,
+      "step": 11449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.6865610498237285,
+      "learning_rate": 1.5672098388983313e-07,
+      "loss": 0.5009,
+      "step": 11450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.824429270798614,
+      "learning_rate": 1.563970384233049e-07,
+      "loss": 0.3909,
+      "step": 11451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.8859387164569945,
+      "learning_rate": 1.5607342278472326e-07,
+      "loss": 0.4065,
+      "step": 11452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.710821911160771,
+      "learning_rate": 1.557501369961245e-07,
+      "loss": 0.4643,
+      "step": 11453
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.38734099718092,
+      "learning_rate": 1.5542718107952215e-07,
+      "loss": 0.4115,
+      "step": 11454
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.017298452591955,
+      "learning_rate": 1.5510455505690858e-07,
+      "loss": 0.4055,
+      "step": 11455
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.735156642917309,
+      "learning_rate": 1.5478225895025344e-07,
+      "loss": 0.3683,
+      "step": 11456
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3293653262423266,
+      "learning_rate": 1.544602927815042e-07,
+      "loss": 0.4616,
+      "step": 11457
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.160752525358305,
+      "learning_rate": 1.5413865657258387e-07,
+      "loss": 0.3697,
+      "step": 11458
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.617650854767634,
+      "learning_rate": 1.5381735034539603e-07,
+      "loss": 0.5269,
+      "step": 11459
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.693895873010239,
+      "learning_rate": 1.5349637412181818e-07,
+      "loss": 0.3721,
+      "step": 11460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 74.74321229781519,
+      "learning_rate": 1.5317572792371006e-07,
+      "loss": 0.5935,
+      "step": 11461
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.52281796406001,
+      "learning_rate": 1.528554117729042e-07,
+      "loss": 0.4193,
+      "step": 11462
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 14.051160659530696,
+      "learning_rate": 1.525354256912137e-07,
+      "loss": 0.4573,
+      "step": 11463
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.794900318127523,
+      "learning_rate": 1.5221576970042674e-07,
+      "loss": 0.4873,
+      "step": 11464
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.3477073108092155,
+      "learning_rate": 1.518964438223125e-07,
+      "loss": 0.5226,
+      "step": 11465
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.10301947182925,
+      "learning_rate": 1.5157744807861473e-07,
+      "loss": 0.4535,
+      "step": 11466
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.382574739412098,
+      "learning_rate": 1.5125878249105551e-07,
+      "loss": 0.3088,
+      "step": 11467
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7889422610167662,
+      "learning_rate": 1.5094044708133526e-07,
+      "loss": 0.6589,
+      "step": 11468
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.044256067109898,
+      "learning_rate": 1.506224418711294e-07,
+      "loss": 0.4759,
+      "step": 11469
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.364292841325155,
+      "learning_rate": 1.5030476688209505e-07,
+      "loss": 0.3583,
+      "step": 11470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.256530579184556,
+      "learning_rate": 1.4998742213586326e-07,
+      "loss": 0.4623,
+      "step": 11471
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.861577848773808,
+      "learning_rate": 1.4967040765404394e-07,
+      "loss": 0.52,
+      "step": 11472
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.581231505999119,
+      "learning_rate": 1.4935372345822318e-07,
+      "loss": 0.3893,
+      "step": 11473
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.8793556793495245,
+      "learning_rate": 1.4903736956996818e-07,
+      "loss": 0.4527,
+      "step": 11474
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.103262024816199,
+      "learning_rate": 1.4872134601081945e-07,
+      "loss": 0.5737,
+      "step": 11475
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.076326756394448,
+      "learning_rate": 1.4840565280229758e-07,
+      "loss": 0.3749,
+      "step": 11476
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1060517146822573,
+      "learning_rate": 1.480902899658998e-07,
+      "loss": 0.4059,
+      "step": 11477
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.678496479656054,
+      "learning_rate": 1.4777525752310006e-07,
+      "loss": 0.6845,
+      "step": 11478
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.627685042101258,
+      "learning_rate": 1.4746055549535178e-07,
+      "loss": 0.4232,
+      "step": 11479
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.539091388187131,
+      "learning_rate": 1.4714618390408442e-07,
+      "loss": 0.5205,
+      "step": 11480
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.820774014716035,
+      "learning_rate": 1.468321427707059e-07,
+      "loss": 0.5032,
+      "step": 11481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.629263160079217,
+      "learning_rate": 1.465184321165991e-07,
+      "loss": 0.4702,
+      "step": 11482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.670743884771571,
+      "learning_rate": 1.462050519631286e-07,
+      "loss": 0.41,
+      "step": 11483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.218015007941308,
+      "learning_rate": 1.4589200233163235e-07,
+      "loss": 0.4454,
+      "step": 11484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.468163111775843,
+      "learning_rate": 1.4557928324342997e-07,
+      "loss": 0.5428,
+      "step": 11485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.174723766057577,
+      "learning_rate": 1.4526689471981324e-07,
+      "loss": 0.481,
+      "step": 11486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.508124693412551,
+      "learning_rate": 1.4495483678205634e-07,
+      "loss": 0.5067,
+      "step": 11487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 22.097816992977098,
+      "learning_rate": 1.446431094514089e-07,
+      "loss": 0.4421,
+      "step": 11488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.57854671819371,
+      "learning_rate": 1.443317127490984e-07,
+      "loss": 0.4172,
+      "step": 11489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.688015211841718,
+      "learning_rate": 1.4402064669632842e-07,
+      "loss": 0.4461,
+      "step": 11490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.854151113976407,
+      "learning_rate": 1.4370991131428202e-07,
+      "loss": 0.4578,
+      "step": 11491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.179892912096825,
+      "learning_rate": 1.4339950662411895e-07,
+      "loss": 0.3831,
+      "step": 11492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.135317502199106,
+      "learning_rate": 1.4308943264697618e-07,
+      "loss": 0.5101,
+      "step": 11493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.889462224458908,
+      "learning_rate": 1.4277968940396901e-07,
+      "loss": 0.4809,
+      "step": 11494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.830262469502505,
+      "learning_rate": 1.4247027691618842e-07,
+      "loss": 0.508,
+      "step": 11495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.0150600386095,
+      "learning_rate": 1.4216119520470416e-07,
+      "loss": 0.3814,
+      "step": 11496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.7303713601981885,
+      "learning_rate": 1.4185244429056445e-07,
+      "loss": 0.448,
+      "step": 11497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.4467122368176595,
+      "learning_rate": 1.4154402419479296e-07,
+      "loss": 0.4572,
+      "step": 11498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.543374432208434,
+      "learning_rate": 1.412359349383924e-07,
+      "loss": 0.4953,
+      "step": 11499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 20.987994851535728,
+      "learning_rate": 1.4092817654234094e-07,
+      "loss": 0.4561,
+      "step": 11500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.537917603030027,
+      "learning_rate": 1.4062074902759738e-07,
+      "loss": 0.4636,
+      "step": 11501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.237051861201826,
+      "learning_rate": 1.4031365241509443e-07,
+      "loss": 0.5215,
+      "step": 11502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.530051878587837,
+      "learning_rate": 1.400068867257459e-07,
+      "loss": 0.5097,
+      "step": 11503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.063082294735661,
+      "learning_rate": 1.3970045198043959e-07,
+      "loss": 0.4532,
+      "step": 11504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.2062191334217784,
+      "learning_rate": 1.3939434820004316e-07,
+      "loss": 0.4279,
+      "step": 11505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.940366592631442,
+      "learning_rate": 1.390885754054e-07,
+      "loss": 0.331,
+      "step": 11506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.66706666991078,
+      "learning_rate": 1.3878313361733287e-07,
+      "loss": 0.4966,
+      "step": 11507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.31417157704154,
+      "learning_rate": 1.3847802285664124e-07,
+      "loss": 0.4676,
+      "step": 11508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.897416388104626,
+      "learning_rate": 1.381732431441002e-07,
+      "loss": 0.4568,
+      "step": 11509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.104192142570299,
+      "learning_rate": 1.378687945004653e-07,
+      "loss": 0.4496,
+      "step": 11510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.235856703505518,
+      "learning_rate": 1.375646769464678e-07,
+      "loss": 0.4228,
+      "step": 11511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.943573015385022,
+      "learning_rate": 1.372608905028172e-07,
+      "loss": 0.5341,
+      "step": 11512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.380271820142145,
+      "learning_rate": 1.3695743519019923e-07,
+      "loss": 0.4739,
+      "step": 11513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7676416862495046,
+      "learning_rate": 1.366543110292784e-07,
+      "loss": 0.7479,
+      "step": 11514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.3841688250043935,
+      "learning_rate": 1.3635151804069546e-07,
+      "loss": 0.4375,
+      "step": 11515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.840960144297037,
+      "learning_rate": 1.3604905624507058e-07,
+      "loss": 0.5918,
+      "step": 11516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 18.25998734955402,
+      "learning_rate": 1.3574692566299841e-07,
+      "loss": 0.2993,
+      "step": 11517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.714792553373541,
+      "learning_rate": 1.3544512631505357e-07,
+      "loss": 0.3864,
+      "step": 11518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.749513805878944,
+      "learning_rate": 1.35143658221788e-07,
+      "loss": 0.4706,
+      "step": 11519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.748405067221822,
+      "learning_rate": 1.3484252140372856e-07,
+      "loss": 0.7587,
+      "step": 11520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.140184136238671,
+      "learning_rate": 1.3454171588138277e-07,
+      "loss": 0.5179,
+      "step": 11521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.077648271388964,
+      "learning_rate": 1.3424124167523367e-07,
+      "loss": 0.475,
+      "step": 11522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.897996807994439,
+      "learning_rate": 1.339410988057427e-07,
+      "loss": 0.389,
+      "step": 11523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.71601384670913,
+      "learning_rate": 1.336412872933468e-07,
+      "loss": 0.4747,
+      "step": 11524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.614718554277049,
+      "learning_rate": 1.3334180715846358e-07,
+      "loss": 0.468,
+      "step": 11525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.957672455729346,
+      "learning_rate": 1.3304265842148557e-07,
+      "loss": 0.4933,
+      "step": 11526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2287648011707293,
+      "learning_rate": 1.3274384110278315e-07,
+      "loss": 0.4114,
+      "step": 11527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 14.585411188749157,
+      "learning_rate": 1.3244535522270506e-07,
+      "loss": 0.4604,
+      "step": 11528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.42444763789108,
+      "learning_rate": 1.3214720080157617e-07,
+      "loss": 0.3728,
+      "step": 11529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.293872542784051,
+      "learning_rate": 1.3184937785970074e-07,
+      "loss": 0.4386,
+      "step": 11530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.732729652779555,
+      "learning_rate": 1.3155188641735817e-07,
+      "loss": 0.4683,
+      "step": 11531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.863819890790948,
+      "learning_rate": 1.3125472649480608e-07,
+      "loss": 0.4451,
+      "step": 11532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.015054341289198,
+      "learning_rate": 1.3095789811228e-07,
+      "loss": 0.5063,
+      "step": 11533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.110789702672132,
+      "learning_rate": 1.3066140128999317e-07,
+      "loss": 0.4468,
+      "step": 11534
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.593702640889502,
+      "learning_rate": 1.3036523604813556e-07,
+      "loss": 0.5151,
+      "step": 11535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.052646207964693,
+      "learning_rate": 1.300694024068738e-07,
+      "loss": 0.4766,
+      "step": 11536
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.479433519550708,
+      "learning_rate": 1.2977390038635406e-07,
+      "loss": 0.5402,
+      "step": 11537
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.395452984448614,
+      "learning_rate": 1.2947873000669798e-07,
+      "loss": 0.3366,
+      "step": 11538
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.603838289541558,
+      "learning_rate": 1.2918389128800558e-07,
+      "loss": 0.3516,
+      "step": 11539
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.467609671545336,
+      "learning_rate": 1.288893842503547e-07,
+      "loss": 0.5718,
+      "step": 11540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.144040196068328,
+      "learning_rate": 1.2859520891379874e-07,
+      "loss": 0.5072,
+      "step": 11541
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.273814494286781,
+      "learning_rate": 1.2830136529837e-07,
+      "loss": 0.4266,
+      "step": 11542
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.87184675276764,
+      "learning_rate": 1.2800785342407918e-07,
+      "loss": 0.4655,
+      "step": 11543
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 17.04459316026223,
+      "learning_rate": 1.2771467331091192e-07,
+      "loss": 0.5645,
+      "step": 11544
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.130831559224933,
+      "learning_rate": 1.2742182497883282e-07,
+      "loss": 0.4177,
+      "step": 11545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.0997495760788,
+      "learning_rate": 1.2712930844778315e-07,
+      "loss": 0.3719,
+      "step": 11546
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 11.450200159785012,
+      "learning_rate": 1.2683712373768198e-07,
+      "loss": 0.446,
+      "step": 11547
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.509943036240171,
+      "learning_rate": 1.2654527086842728e-07,
+      "loss": 0.5559,
+      "step": 11548
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.320524088963724,
+      "learning_rate": 1.2625374985989147e-07,
+      "loss": 0.526,
+      "step": 11549
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.285996625998535,
+      "learning_rate": 1.2596256073192648e-07,
+      "loss": 0.4424,
+      "step": 11550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.067923359357742,
+      "learning_rate": 1.256717035043603e-07,
+      "loss": 0.3724,
+      "step": 11551
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.717384373267102,
+      "learning_rate": 1.2538117819699992e-07,
+      "loss": 0.7521,
+      "step": 11552
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.615511182473085,
+      "learning_rate": 1.2509098482962778e-07,
+      "loss": 0.4346,
+      "step": 11553
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.88654541241245,
+      "learning_rate": 1.248011234220059e-07,
+      "loss": 0.4105,
+      "step": 11554
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.467884898231196,
+      "learning_rate": 1.245115939938718e-07,
+      "loss": 0.4503,
+      "step": 11555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.034443170005988,
+      "learning_rate": 1.2422239656494085e-07,
+      "loss": 0.3671,
+      "step": 11556
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.853555836025838,
+      "learning_rate": 1.2393353115490782e-07,
+      "loss": 0.4089,
+      "step": 11557
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.740410946006138,
+      "learning_rate": 1.236449977834414e-07,
+      "loss": 0.4869,
+      "step": 11558
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.173426389842368,
+      "learning_rate": 1.2335679647019038e-07,
+      "loss": 0.4879,
+      "step": 11559
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.1794008852242,
+      "learning_rate": 1.2306892723477903e-07,
+      "loss": 0.4173,
+      "step": 11560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.472772247364588,
+      "learning_rate": 1.227813900968111e-07,
+      "loss": 0.4332,
+      "step": 11561
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 9.64415098814743,
+      "learning_rate": 1.224941850758654e-07,
+      "loss": 0.5438,
+      "step": 11562
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.839596532075706,
+      "learning_rate": 1.2220731219150128e-07,
+      "loss": 0.4563,
+      "step": 11563
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.491903144323414,
+      "learning_rate": 1.2192077146325088e-07,
+      "loss": 0.4402,
+      "step": 11564
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.512482033941751,
+      "learning_rate": 1.216345629106286e-07,
+      "loss": 0.4406,
+      "step": 11565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.7880116193403195,
+      "learning_rate": 1.2134868655312271e-07,
+      "loss": 0.4012,
+      "step": 11566
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.520452408549216,
+      "learning_rate": 1.2106314241020044e-07,
+      "loss": 0.4645,
+      "step": 11567
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.57660545526302,
+      "learning_rate": 1.2077793050130682e-07,
+      "loss": 0.3948,
+      "step": 11568
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.479558062815338,
+      "learning_rate": 1.2049305084586183e-07,
+      "loss": 0.4309,
+      "step": 11569
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.268191368035996,
+      "learning_rate": 1.2020850346326607e-07,
+      "loss": 0.4449,
+      "step": 11570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.069469384421107,
+      "learning_rate": 1.1992428837289516e-07,
+      "loss": 0.5782,
+      "step": 11571
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 12.836519483913808,
+      "learning_rate": 1.1964040559410416e-07,
+      "loss": 0.3187,
+      "step": 11572
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.786284508469446,
+      "learning_rate": 1.1935685514622264e-07,
+      "loss": 0.4632,
+      "step": 11573
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.361485912948419,
+      "learning_rate": 1.1907363704855957e-07,
+      "loss": 0.3831,
+      "step": 11574
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.945630770189836,
+      "learning_rate": 1.1879075132040063e-07,
+      "loss": 0.5053,
+      "step": 11575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.682974647887133,
+      "learning_rate": 1.1850819798101043e-07,
+      "loss": 0.6894,
+      "step": 11576
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6886679136584326,
+      "learning_rate": 1.1822597704962802e-07,
+      "loss": 0.4227,
+      "step": 11577
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.495829886071331,
+      "learning_rate": 1.179440885454719e-07,
+      "loss": 0.4835,
+      "step": 11578
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.4422172003981,
+      "learning_rate": 1.1766253248773785e-07,
+      "loss": 0.5204,
+      "step": 11579
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.106123042094938,
+      "learning_rate": 1.1738130889559829e-07,
+      "loss": 0.4131,
+      "step": 11580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.122820332981326,
+      "learning_rate": 1.1710041778820347e-07,
+      "loss": 0.48,
+      "step": 11581
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.670800492033629,
+      "learning_rate": 1.168198591846803e-07,
+      "loss": 0.6123,
+      "step": 11582
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.6505676422816835,
+      "learning_rate": 1.1653963310413407e-07,
+      "loss": 0.4285,
+      "step": 11583
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.779418661107704,
+      "learning_rate": 1.1625973956564674e-07,
+      "loss": 0.6103,
+      "step": 11584
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.914073565823525,
+      "learning_rate": 1.1598017858827803e-07,
+      "loss": 0.5109,
+      "step": 11585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.8501319916968795,
+      "learning_rate": 1.1570095019106497e-07,
+      "loss": 0.4084,
+      "step": 11586
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.874954839201474,
+      "learning_rate": 1.1542205439302068e-07,
+      "loss": 0.4472,
+      "step": 11587
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.881326098825039,
+      "learning_rate": 1.1514349121313829e-07,
+      "loss": 0.5632,
+      "step": 11588
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.271106721593955,
+      "learning_rate": 1.148652606703854e-07,
+      "loss": 0.6421,
+      "step": 11589
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.220234432395323,
+      "learning_rate": 1.1458736278370963e-07,
+      "loss": 0.4395,
+      "step": 11590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.864056325063293,
+      "learning_rate": 1.1430979757203309e-07,
+      "loss": 0.4982,
+      "step": 11591
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 19.80117494117158,
+      "learning_rate": 1.1403256505425786e-07,
+      "loss": 0.4696,
+      "step": 11592
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.575486774841123,
+      "learning_rate": 1.1375566524926163e-07,
+      "loss": 0.5106,
+      "step": 11593
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.856467332465616,
+      "learning_rate": 1.1347909817590041e-07,
+      "loss": 0.6064,
+      "step": 11594
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.3261040195575395,
+      "learning_rate": 1.1320286385300694e-07,
+      "loss": 0.4642,
+      "step": 11595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.306774299655076,
+      "learning_rate": 1.1292696229939171e-07,
+      "loss": 0.4019,
+      "step": 11596
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 21.21110289128163,
+      "learning_rate": 1.1265139353384247e-07,
+      "loss": 0.4075,
+      "step": 11597
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.3736195901239,
+      "learning_rate": 1.1237615757512366e-07,
+      "loss": 0.4065,
+      "step": 11598
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.006070145718774,
+      "learning_rate": 1.1210125444197862e-07,
+      "loss": 0.4231,
+      "step": 11599
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.849113188858378,
+      "learning_rate": 1.1182668415312625e-07,
+      "loss": 0.428,
+      "step": 11600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9796239168120482,
+      "learning_rate": 1.115524467272644e-07,
+      "loss": 0.7386,
+      "step": 11601
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.836209854166928,
+      "learning_rate": 1.1127854218306589e-07,
+      "loss": 0.4229,
+      "step": 11602
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.879354686154559,
+      "learning_rate": 1.1100497053918413e-07,
+      "loss": 0.4026,
+      "step": 11603
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.245587424273831,
+      "learning_rate": 1.1073173181424756e-07,
+      "loss": 0.4123,
+      "step": 11604
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.602866600599965,
+      "learning_rate": 1.1045882602686186e-07,
+      "loss": 0.4156,
+      "step": 11605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.353981803884436,
+      "learning_rate": 1.1018625319561105e-07,
+      "loss": 0.3922,
+      "step": 11606
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.49324590457674,
+      "learning_rate": 1.0991401333905638e-07,
+      "loss": 0.5071,
+      "step": 11607
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.155664623773674,
+      "learning_rate": 1.0964210647573637e-07,
+      "loss": 0.3746,
+      "step": 11608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.406806273885073,
+      "learning_rate": 1.0937053262416675e-07,
+      "loss": 0.3908,
+      "step": 11609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.305815054389938,
+      "learning_rate": 1.0909929180283996e-07,
+      "loss": 0.508,
+      "step": 11610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9190119327814643,
+      "learning_rate": 1.0882838403022566e-07,
+      "loss": 0.7784,
+      "step": 11611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.0047919824014455,
+      "learning_rate": 1.0855780932477356e-07,
+      "loss": 0.4723,
+      "step": 11612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.416649290837379,
+      "learning_rate": 1.0828756770490667e-07,
+      "loss": 0.4773,
+      "step": 11613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.085741213403297,
+      "learning_rate": 1.0801765918902807e-07,
+      "loss": 0.406,
+      "step": 11614
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.604354392143282,
+      "learning_rate": 1.0774808379551694e-07,
+      "loss": 0.4485,
+      "step": 11615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.05342128408013,
+      "learning_rate": 1.0747884154273024e-07,
+      "loss": 0.5495,
+      "step": 11616
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.181194402757416,
+      "learning_rate": 1.0720993244900279e-07,
+      "loss": 0.3867,
+      "step": 11617
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.92514507797389,
+      "learning_rate": 1.0694135653264548e-07,
+      "loss": 0.3675,
+      "step": 11618
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.858891131858335,
+      "learning_rate": 1.0667311381194756e-07,
+      "loss": 0.3594,
+      "step": 11619
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.511763218246652,
+      "learning_rate": 1.0640520430517442e-07,
+      "loss": 0.4653,
+      "step": 11620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.113984813074643,
+      "learning_rate": 1.0613762803057092e-07,
+      "loss": 0.5366,
+      "step": 11621
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.943441205490572,
+      "learning_rate": 1.0587038500635637e-07,
+      "loss": 0.4251,
+      "step": 11622
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.07780165102275,
+      "learning_rate": 1.0560347525073012e-07,
+      "loss": 0.3694,
+      "step": 11623
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.0686547923286565,
+      "learning_rate": 1.0533689878186592e-07,
+      "loss": 0.3882,
+      "step": 11624
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.438810579511,
+      "learning_rate": 1.0507065561791762e-07,
+      "loss": 0.4764,
+      "step": 11625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.968860155117332,
+      "learning_rate": 1.048047457770146e-07,
+      "loss": 0.4784,
+      "step": 11626
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.296580694936535,
+      "learning_rate": 1.0453916927726516e-07,
+      "loss": 0.4707,
+      "step": 11627
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.484454545304505,
+      "learning_rate": 1.0427392613675313e-07,
+      "loss": 0.436,
+      "step": 11628
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.333680365772887,
+      "learning_rate": 1.0400901637354021e-07,
+      "loss": 0.434,
+      "step": 11629
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.861437873228187,
+      "learning_rate": 1.0374444000566586e-07,
+      "loss": 0.4218,
+      "step": 11630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.431182097359568,
+      "learning_rate": 1.0348019705114731e-07,
+      "loss": 0.5332,
+      "step": 11631
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.501466480615937,
+      "learning_rate": 1.0321628752797685e-07,
+      "loss": 0.4254,
+      "step": 11632
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.63024211916897,
+      "learning_rate": 1.0295271145412623e-07,
+      "loss": 0.4743,
+      "step": 11633
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.090853758282561,
+      "learning_rate": 1.0268946884754439e-07,
+      "loss": 0.538,
+      "step": 11634
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.689375880122282,
+      "learning_rate": 1.0242655972615533e-07,
+      "loss": 0.4765,
+      "step": 11635
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.005508787300998,
+      "learning_rate": 1.0216398410786477e-07,
+      "loss": 0.4368,
+      "step": 11636
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.496004376657689,
+      "learning_rate": 1.0190174201055003e-07,
+      "loss": 0.4669,
+      "step": 11637
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.655318258176362,
+      "learning_rate": 1.0163983345207018e-07,
+      "loss": 0.514,
+      "step": 11638
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.298075628620874,
+      "learning_rate": 1.0137825845026039e-07,
+      "loss": 0.415,
+      "step": 11639
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.461028802064989,
+      "learning_rate": 1.0111701702293197e-07,
+      "loss": 0.3666,
+      "step": 11640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.992741531181353,
+      "learning_rate": 1.0085610918787458e-07,
+      "loss": 0.4808,
+      "step": 11641
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.412063607225072,
+      "learning_rate": 1.0059553496285401e-07,
+      "loss": 0.4982,
+      "step": 11642
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.333196080538354,
+      "learning_rate": 1.0033529436561551e-07,
+      "loss": 0.4811,
+      "step": 11643
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.853011693872266,
+      "learning_rate": 1.000753874138799e-07,
+      "loss": 0.4277,
+      "step": 11644
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.314920151232669,
+      "learning_rate": 9.981581412534524e-08,
+      "loss": 0.5022,
+      "step": 11645
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.318320938917518,
+      "learning_rate": 9.955657451768852e-08,
+      "loss": 0.5461,
+      "step": 11646
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.400941430794082,
+      "learning_rate": 9.929766860856116e-08,
+      "loss": 0.3013,
+      "step": 11647
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.133436066310033,
+      "learning_rate": 9.903909641559462e-08,
+      "loss": 0.3487,
+      "step": 11648
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7376321093810712,
+      "learning_rate": 9.878085795639591e-08,
+      "loss": 0.7532,
+      "step": 11649
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.3718645813393,
+      "learning_rate": 9.8522953248551e-08,
+      "loss": 0.391,
+      "step": 11650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.8944088525779135,
+      "learning_rate": 9.826538230962024e-08,
+      "loss": 0.5513,
+      "step": 11651
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.273017427744852,
+      "learning_rate": 9.800814515714519e-08,
+      "loss": 0.449,
+      "step": 11652
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.722001814672729,
+      "learning_rate": 9.77512418086407e-08,
+      "loss": 0.7375,
+      "step": 11653
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.409704207907621,
+      "learning_rate": 9.749467228160226e-08,
+      "loss": 0.4192,
+      "step": 11654
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.089046689877757,
+      "learning_rate": 9.723843659349974e-08,
+      "loss": 0.5338,
+      "step": 11655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.882174539680255,
+      "learning_rate": 9.69825347617831e-08,
+      "loss": 0.605,
+      "step": 11656
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.469988896363211,
+      "learning_rate": 9.672696680387672e-08,
+      "loss": 0.4026,
+      "step": 11657
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.441577554493786,
+      "learning_rate": 9.64717327371839e-08,
+      "loss": 0.4505,
+      "step": 11658
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.664829303316385,
+      "learning_rate": 9.621683257908687e-08,
+      "loss": 0.5479,
+      "step": 11659
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.611181483042992,
+      "learning_rate": 9.596226634694006e-08,
+      "loss": 0.4741,
+      "step": 11660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 44.532584038659856,
+      "learning_rate": 9.570803405808071e-08,
+      "loss": 0.4757,
+      "step": 11661
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.362644115557529,
+      "learning_rate": 9.545413572981999e-08,
+      "loss": 0.4633,
+      "step": 11662
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.949152733869511,
+      "learning_rate": 9.520057137944739e-08,
+      "loss": 0.3502,
+      "step": 11663
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.177749581002315,
+      "learning_rate": 9.494734102422964e-08,
+      "loss": 0.4426,
+      "step": 11664
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.786991667878426,
+      "learning_rate": 9.469444468141075e-08,
+      "loss": 0.4987,
+      "step": 11665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.411780234482778,
+      "learning_rate": 9.444188236821084e-08,
+      "loss": 0.4363,
+      "step": 11666
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.714526051635842,
+      "learning_rate": 9.41896541018289e-08,
+      "loss": 0.4828,
+      "step": 11667
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7920656649605124,
+      "learning_rate": 9.393775989944176e-08,
+      "loss": 0.3577,
+      "step": 11668
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.291424844273159,
+      "learning_rate": 9.36861997782007e-08,
+      "loss": 0.5555,
+      "step": 11669
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.133224689595,
+      "learning_rate": 9.343497375523646e-08,
+      "loss": 0.4023,
+      "step": 11670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.71056597417416,
+      "learning_rate": 9.318408184765593e-08,
+      "loss": 0.3828,
+      "step": 11671
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.1213068777394986,
+      "learning_rate": 9.29335240725443e-08,
+      "loss": 0.5062,
+      "step": 11672
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.832893151922404,
+      "learning_rate": 9.26833004469635e-08,
+      "loss": 0.4888,
+      "step": 11673
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.723037983166167,
+      "learning_rate": 9.243341098795266e-08,
+      "loss": 0.5443,
+      "step": 11674
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.479361127028856,
+      "learning_rate": 9.218385571252764e-08,
+      "loss": 0.4287,
+      "step": 11675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.088008504626857,
+      "learning_rate": 9.193463463768204e-08,
+      "loss": 0.3953,
+      "step": 11676
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.164123108524629,
+      "learning_rate": 9.168574778038785e-08,
+      "loss": 0.4956,
+      "step": 11677
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.961801516882067,
+      "learning_rate": 9.143719515759153e-08,
+      "loss": 0.4136,
+      "step": 11678
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.313147818370453,
+      "learning_rate": 9.118897678621952e-08,
+      "loss": 0.4735,
+      "step": 11679
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.986636203576449,
+      "learning_rate": 9.094109268317386e-08,
+      "loss": 0.4982,
+      "step": 11680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.688288492261191,
+      "learning_rate": 9.069354286533494e-08,
+      "loss": 0.4367,
+      "step": 11681
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.725445920164778,
+      "learning_rate": 9.044632734955927e-08,
+      "loss": 0.5764,
+      "step": 11682
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.963422895500949,
+      "learning_rate": 9.019944615268173e-08,
+      "loss": 0.37,
+      "step": 11683
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.320088815410833,
+      "learning_rate": 8.995289929151274e-08,
+      "loss": 0.4259,
+      "step": 11684
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.935899788652921,
+      "learning_rate": 8.970668678284222e-08,
+      "loss": 0.3846,
+      "step": 11685
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.25408964263981,
+      "learning_rate": 8.94608086434362e-08,
+      "loss": 0.4205,
+      "step": 11686
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.137379223325498,
+      "learning_rate": 8.92152648900363e-08,
+      "loss": 0.4438,
+      "step": 11687
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.097457597373513,
+      "learning_rate": 8.897005553936466e-08,
+      "loss": 0.4078,
+      "step": 11688
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.492216743626361,
+      "learning_rate": 8.872518060811796e-08,
+      "loss": 0.5304,
+      "step": 11689
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.561108305665975,
+      "learning_rate": 8.848064011297175e-08,
+      "loss": 0.4685,
+      "step": 11690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.16386846436003,
+      "learning_rate": 8.823643407057769e-08,
+      "loss": 0.4663,
+      "step": 11691
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.093625597698394,
+      "learning_rate": 8.799256249756583e-08,
+      "loss": 0.511,
+      "step": 11692
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.864724087713076,
+      "learning_rate": 8.77490254105412e-08,
+      "loss": 0.4525,
+      "step": 11693
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.471812375052428,
+      "learning_rate": 8.750582282608944e-08,
+      "loss": 0.3821,
+      "step": 11694
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.865060094376531,
+      "learning_rate": 8.726295476077118e-08,
+      "loss": 0.4524,
+      "step": 11695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6886217982713139,
+      "learning_rate": 8.702042123112375e-08,
+      "loss": 0.6559,
+      "step": 11696
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.597312759739353,
+      "learning_rate": 8.677822225366228e-08,
+      "loss": 0.4678,
+      "step": 11697
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.232163093566827,
+      "learning_rate": 8.65363578448808e-08,
+      "loss": 0.2839,
+      "step": 11698
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.658303602776978,
+      "learning_rate": 8.629482802124944e-08,
+      "loss": 0.5048,
+      "step": 11699
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.426208121791598,
+      "learning_rate": 8.605363279921397e-08,
+      "loss": 0.4635,
+      "step": 11700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.114214121648248,
+      "learning_rate": 8.581277219519957e-08,
+      "loss": 0.362,
+      "step": 11701
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.650848020869388,
+      "learning_rate": 8.5572246225607e-08,
+      "loss": 0.5708,
+      "step": 11702
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.551791190667435,
+      "learning_rate": 8.533205490681595e-08,
+      "loss": 0.4955,
+      "step": 11703
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.439729776374647,
+      "learning_rate": 8.509219825518167e-08,
+      "loss": 0.3586,
+      "step": 11704
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.588066984166215,
+      "learning_rate": 8.48526762870383e-08,
+      "loss": 0.5677,
+      "step": 11705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.062911020715514,
+      "learning_rate": 8.461348901869449e-08,
+      "loss": 0.4146,
+      "step": 11706
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8679808878168855,
+      "learning_rate": 8.43746364664394e-08,
+      "loss": 0.6742,
+      "step": 11707
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.158427394563628,
+      "learning_rate": 8.413611864653726e-08,
+      "loss": 0.4848,
+      "step": 11708
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.83713285292366,
+      "learning_rate": 8.389793557523063e-08,
+      "loss": 0.4568,
+      "step": 11709
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.037240700252177,
+      "learning_rate": 8.36600872687382e-08,
+      "loss": 0.3741,
+      "step": 11710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 16.520449595801026,
+      "learning_rate": 8.34225737432559e-08,
+      "loss": 0.527,
+      "step": 11711
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.683358563091811,
+      "learning_rate": 8.318539501495804e-08,
+      "loss": 0.367,
+      "step": 11712
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.334278675880398,
+      "learning_rate": 8.294855109999556e-08,
+      "loss": 0.3948,
+      "step": 11713
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.369228180456841,
+      "learning_rate": 8.271204201449668e-08,
+      "loss": 0.5554,
+      "step": 11714
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.724346847546992,
+      "learning_rate": 8.247586777456517e-08,
+      "loss": 0.4009,
+      "step": 11715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.465290663147614,
+      "learning_rate": 8.224002839628486e-08,
+      "loss": 0.4804,
+      "step": 11716
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.425135878253242,
+      "learning_rate": 8.200452389571457e-08,
+      "loss": 0.4491,
+      "step": 11717
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 18.670736670983242,
+      "learning_rate": 8.176935428889199e-08,
+      "loss": 0.4798,
+      "step": 11718
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.723219558952866,
+      "learning_rate": 8.153451959183045e-08,
+      "loss": 0.4968,
+      "step": 11719
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.620130793777157,
+      "learning_rate": 8.13000198205216e-08,
+      "loss": 0.4304,
+      "step": 11720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.027562251952621,
+      "learning_rate": 8.106585499093322e-08,
+      "loss": 0.4928,
+      "step": 11721
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.51689608765787,
+      "learning_rate": 8.083202511901145e-08,
+      "loss": 0.4523,
+      "step": 11722
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.2755944032143915,
+      "learning_rate": 8.059853022067909e-08,
+      "loss": 0.4213,
+      "step": 11723
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.5833285735294735,
+      "learning_rate": 8.036537031183567e-08,
+      "loss": 0.5127,
+      "step": 11724
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.46778015424209,
+      "learning_rate": 8.013254540835846e-08,
+      "loss": 0.5562,
+      "step": 11725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.259231271229889,
+      "learning_rate": 7.990005552610203e-08,
+      "loss": 0.4563,
+      "step": 11726
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.95932020786489,
+      "learning_rate": 7.966790068089758e-08,
+      "loss": 0.4554,
+      "step": 11727
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.992069414181936,
+      "learning_rate": 7.94360808885547e-08,
+      "loss": 0.4544,
+      "step": 11728
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.953119251910372,
+      "learning_rate": 7.920459616485799e-08,
+      "loss": 0.4377,
+      "step": 11729
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.167684751140357,
+      "learning_rate": 7.897344652557148e-08,
+      "loss": 0.5344,
+      "step": 11730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.2195353551522405,
+      "learning_rate": 7.874263198643484e-08,
+      "loss": 0.5451,
+      "step": 11731
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.719443841337295,
+      "learning_rate": 7.851215256316657e-08,
+      "loss": 0.4016,
+      "step": 11732
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.072635834158064,
+      "learning_rate": 7.828200827145971e-08,
+      "loss": 0.4508,
+      "step": 11733
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.038924103800183,
+      "learning_rate": 7.80521991269878e-08,
+      "loss": 0.5478,
+      "step": 11734
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.959800633893229,
+      "learning_rate": 7.782272514539834e-08,
+      "loss": 0.5224,
+      "step": 11735
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.217889825353766,
+      "learning_rate": 7.759358634231829e-08,
+      "loss": 0.4992,
+      "step": 11736
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.86336530588187,
+      "learning_rate": 7.736478273335069e-08,
+      "loss": 0.4174,
+      "step": 11737
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.185793398123485,
+      "learning_rate": 7.713631433407587e-08,
+      "loss": 0.4695,
+      "step": 11738
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 53.96950139933418,
+      "learning_rate": 7.69081811600525e-08,
+      "loss": 0.5031,
+      "step": 11739
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.286541091644835,
+      "learning_rate": 7.668038322681426e-08,
+      "loss": 0.4283,
+      "step": 11740
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.947859413541335,
+      "learning_rate": 7.645292054987374e-08,
+      "loss": 0.421,
+      "step": 11741
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.107827510968,
+      "learning_rate": 7.622579314472023e-08,
+      "loss": 0.4461,
+      "step": 11742
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.443453030149002,
+      "learning_rate": 7.599900102682023e-08,
+      "loss": 0.526,
+      "step": 11743
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.496877265399613,
+      "learning_rate": 7.57725442116164e-08,
+      "loss": 0.4142,
+      "step": 11744
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.41919403746649,
+      "learning_rate": 7.554642271453028e-08,
+      "loss": 0.4274,
+      "step": 11745
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.329408131854739,
+      "learning_rate": 7.532063655095956e-08,
+      "loss": 0.482,
+      "step": 11746
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.627528435098732,
+      "learning_rate": 7.509518573627972e-08,
+      "loss": 0.5141,
+      "step": 11747
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.149797854912007,
+      "learning_rate": 7.487007028584126e-08,
+      "loss": 0.5134,
+      "step": 11748
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.119288549601406,
+      "learning_rate": 7.464529021497524e-08,
+      "loss": 0.4164,
+      "step": 11749
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 14.141081383934424,
+      "learning_rate": 7.442084553898831e-08,
+      "loss": 0.4165,
+      "step": 11750
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.8438880570212,
+      "learning_rate": 7.419673627316326e-08,
+      "loss": 0.5176,
+      "step": 11751
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.253843154200025,
+      "learning_rate": 7.39729624327612e-08,
+      "loss": 0.5629,
+      "step": 11752
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.847712236673394,
+      "learning_rate": 7.374952403301994e-08,
+      "loss": 0.431,
+      "step": 11753
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.636764116097556,
+      "learning_rate": 7.352642108915564e-08,
+      "loss": 0.4777,
+      "step": 11754
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.275400487237704,
+      "learning_rate": 7.330365361635949e-08,
+      "loss": 0.5649,
+      "step": 11755
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.677949484020699,
+      "learning_rate": 7.308122162980102e-08,
+      "loss": 0.4999,
+      "step": 11756
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.376891064176104,
+      "learning_rate": 7.2859125144627e-08,
+      "loss": 0.5388,
+      "step": 11757
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.993489953845217,
+      "learning_rate": 7.263736417596202e-08,
+      "loss": 0.3739,
+      "step": 11758
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.246930166350284,
+      "learning_rate": 7.241593873890618e-08,
+      "loss": 0.5339,
+      "step": 11759
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.550124337840629,
+      "learning_rate": 7.2194848848538e-08,
+      "loss": 0.426,
+      "step": 11760
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.451840007772068,
+      "learning_rate": 7.197409451991211e-08,
+      "loss": 0.4591,
+      "step": 11761
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.235168016294588,
+      "learning_rate": 7.175367576806147e-08,
+      "loss": 0.4508,
+      "step": 11762
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.929621167701433,
+      "learning_rate": 7.153359260799575e-08,
+      "loss": 0.4805,
+      "step": 11763
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.426311836845426,
+      "learning_rate": 7.131384505470129e-08,
+      "loss": 0.4591,
+      "step": 11764
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.686532963418099,
+      "learning_rate": 7.109443312314224e-08,
+      "loss": 0.3822,
+      "step": 11765
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.48341910773895,
+      "learning_rate": 7.087535682825886e-08,
+      "loss": 0.4643,
+      "step": 11766
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.055185860444162,
+      "learning_rate": 7.065661618496977e-08,
+      "loss": 0.4648,
+      "step": 11767
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.133156907273454,
+      "learning_rate": 7.04382112081703e-08,
+      "loss": 0.4485,
+      "step": 11768
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0539016061012703,
+      "learning_rate": 7.022014191273296e-08,
+      "loss": 0.7569,
+      "step": 11769
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.080333379180113,
+      "learning_rate": 7.000240831350757e-08,
+      "loss": 0.5143,
+      "step": 11770
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.263270626218241,
+      "learning_rate": 6.978501042531948e-08,
+      "loss": 0.3655,
+      "step": 11771
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9376365155508872,
+      "learning_rate": 6.956794826297409e-08,
+      "loss": 0.473,
+      "step": 11772
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.693810137454668,
+      "learning_rate": 6.935122184125175e-08,
+      "loss": 0.7032,
+      "step": 11773
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.727422475759975,
+      "learning_rate": 6.913483117491016e-08,
+      "loss": 0.47,
+      "step": 11774
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.401556533200457,
+      "learning_rate": 6.891877627868527e-08,
+      "loss": 0.4774,
+      "step": 11775
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.7379606532631104,
+      "learning_rate": 6.87030571672892e-08,
+      "loss": 0.4742,
+      "step": 11776
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 13.20319975264263,
+      "learning_rate": 6.848767385541078e-08,
+      "loss": 0.4825,
+      "step": 11777
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.167096496537795,
+      "learning_rate": 6.82726263577177e-08,
+      "loss": 0.459,
+      "step": 11778
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.624255136207435,
+      "learning_rate": 6.805791468885381e-08,
+      "loss": 0.4968,
+      "step": 11779
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 34.005411824130505,
+      "learning_rate": 6.784353886343909e-08,
+      "loss": 0.5446,
+      "step": 11780
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.897474028813699,
+      "learning_rate": 6.762949889607185e-08,
+      "loss": 0.458,
+      "step": 11781
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.35868865305301,
+      "learning_rate": 6.741579480132821e-08,
+      "loss": 0.4316,
+      "step": 11782
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.535807692287301,
+      "learning_rate": 6.72024265937593e-08,
+      "loss": 0.467,
+      "step": 11783
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.308048556475582,
+      "learning_rate": 6.698939428789463e-08,
+      "loss": 0.501,
+      "step": 11784
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.771792227307047,
+      "learning_rate": 6.677669789824148e-08,
+      "loss": 0.5168,
+      "step": 11785
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.035804998794564,
+      "learning_rate": 6.656433743928326e-08,
+      "loss": 0.5016,
+      "step": 11786
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.563648787906457,
+      "learning_rate": 6.635231292548062e-08,
+      "loss": 0.4115,
+      "step": 11787
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.493891718302013,
+      "learning_rate": 6.614062437127145e-08,
+      "loss": 0.4564,
+      "step": 11788
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 16.406312448764464,
+      "learning_rate": 6.59292717910709e-08,
+      "loss": 0.399,
+      "step": 11789
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.106929499222087,
+      "learning_rate": 6.571825519927134e-08,
+      "loss": 0.3611,
+      "step": 11790
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.082576294164208,
+      "learning_rate": 6.550757461024183e-08,
+      "loss": 0.4562,
+      "step": 11791
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.444249330713934,
+      "learning_rate": 6.529723003832866e-08,
+      "loss": 0.4985,
+      "step": 11792
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.089595417752494,
+      "learning_rate": 6.508722149785541e-08,
+      "loss": 0.4947,
+      "step": 11793
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.161797815248942,
+      "learning_rate": 6.487754900312338e-08,
+      "loss": 0.4517,
+      "step": 11794
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5703754224519004,
+      "learning_rate": 6.466821256840949e-08,
+      "loss": 0.6449,
+      "step": 11795
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.828970393044963,
+      "learning_rate": 6.445921220796902e-08,
+      "loss": 0.5033,
+      "step": 11796
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.509171361654681,
+      "learning_rate": 6.425054793603391e-08,
+      "loss": 0.3878,
+      "step": 11797
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.421939649864106,
+      "learning_rate": 6.404221976681335e-08,
+      "loss": 0.5539,
+      "step": 11798
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 11.927585207210262,
+      "learning_rate": 6.383422771449376e-08,
+      "loss": 0.4042,
+      "step": 11799
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.116346666729761,
+      "learning_rate": 6.362657179323772e-08,
+      "loss": 0.4522,
+      "step": 11800
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.838705418902189,
+      "learning_rate": 6.341925201718668e-08,
+      "loss": 0.4111,
+      "step": 11801
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.991068766670715,
+      "learning_rate": 6.321226840045713e-08,
+      "loss": 0.4848,
+      "step": 11802
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.011055466136114,
+      "learning_rate": 6.300562095714502e-08,
+      "loss": 0.4464,
+      "step": 11803
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.455693885718079,
+      "learning_rate": 6.279930970132076e-08,
+      "loss": 0.4273,
+      "step": 11804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.482739730916527,
+      "learning_rate": 6.25933346470342e-08,
+      "loss": 0.3912,
+      "step": 11805
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6000907838681024,
+      "learning_rate": 6.238769580831083e-08,
+      "loss": 0.6938,
+      "step": 11806
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.78244904594326,
+      "learning_rate": 6.218239319915442e-08,
+      "loss": 0.548,
+      "step": 11807
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.762522515024522,
+      "learning_rate": 6.19774268335438e-08,
+      "loss": 0.4623,
+      "step": 11808
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.158184993823155,
+      "learning_rate": 6.177279672543723e-08,
+      "loss": 0.3876,
+      "step": 11809
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.650562801376682,
+      "learning_rate": 6.156850288877025e-08,
+      "loss": 0.3511,
+      "step": 11810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.744442334570657,
+      "learning_rate": 6.136454533745229e-08,
+      "loss": 0.3606,
+      "step": 11811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.288292509408494,
+      "learning_rate": 6.116092408537277e-08,
+      "loss": 0.4322,
+      "step": 11812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.731658393832755,
+      "learning_rate": 6.095763914639729e-08,
+      "loss": 0.5083,
+      "step": 11813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.411274420994035,
+      "learning_rate": 6.075469053436921e-08,
+      "loss": 0.3959,
+      "step": 11814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.654261272153555,
+      "learning_rate": 6.055207826310805e-08,
+      "loss": 0.404,
+      "step": 11815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.5482934188056605,
+      "learning_rate": 6.034980234641108e-08,
+      "loss": 0.4309,
+      "step": 11816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.808176228786444,
+      "learning_rate": 6.014786279805174e-08,
+      "loss": 0.4399,
+      "step": 11817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.241561444048052,
+      "learning_rate": 5.994625963178124e-08,
+      "loss": 0.6024,
+      "step": 11818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.704754139862108,
+      "learning_rate": 5.974499286132918e-08,
+      "loss": 0.506,
+      "step": 11819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8809091639224274,
+      "learning_rate": 5.9544062500399014e-08,
+      "loss": 0.6975,
+      "step": 11820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.4233210497543265,
+      "learning_rate": 5.934346856267481e-08,
+      "loss": 0.4309,
+      "step": 11821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.438788096181495,
+      "learning_rate": 5.914321106181509e-08,
+      "loss": 0.4026,
+      "step": 11822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.819033869468508,
+      "learning_rate": 5.8943290011457266e-08,
+      "loss": 0.6668,
+      "step": 11823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.192248293390169,
+      "learning_rate": 5.8743705425214346e-08,
+      "loss": 0.4639,
+      "step": 11824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.919429157242018,
+      "learning_rate": 5.854445731667768e-08,
+      "loss": 0.4173,
+      "step": 11825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.597663679219671,
+      "learning_rate": 5.834554569941531e-08,
+      "loss": 0.4408,
+      "step": 11826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.437674549260568,
+      "learning_rate": 5.814697058697139e-08,
+      "loss": 0.4895,
+      "step": 11827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.007447593158226,
+      "learning_rate": 5.794873199286899e-08,
+      "loss": 0.4727,
+      "step": 11828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.302618442181678,
+      "learning_rate": 5.7750829930606764e-08,
+      "loss": 0.3762,
+      "step": 11829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.10987232250198,
+      "learning_rate": 5.7553264413661157e-08,
+      "loss": 0.422,
+      "step": 11830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.388598766140799,
+      "learning_rate": 5.735603545548529e-08,
+      "loss": 0.535,
+      "step": 11831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.319442799596357,
+      "learning_rate": 5.715914306951009e-08,
+      "loss": 0.4025,
+      "step": 11832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.774003057651901,
+      "learning_rate": 5.696258726914261e-08,
+      "loss": 0.4509,
+      "step": 11833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.255880113579401,
+      "learning_rate": 5.676636806776714e-08,
+      "loss": 0.4155,
+      "step": 11834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.811120655318735,
+      "learning_rate": 5.657048547874578e-08,
+      "loss": 0.398,
+      "step": 11835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.370512215526295,
+      "learning_rate": 5.637493951541728e-08,
+      "loss": 0.3738,
+      "step": 11836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.302709029774789,
+      "learning_rate": 5.617973019109768e-08,
+      "loss": 0.396,
+      "step": 11837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.758067672402363,
+      "learning_rate": 5.5984857519079113e-08,
+      "loss": 0.5213,
+      "step": 11838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.387393254091603,
+      "learning_rate": 5.579032151263209e-08,
+      "loss": 0.4043,
+      "step": 11839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.414739632236423,
+      "learning_rate": 5.5596122185003764e-08,
+      "loss": 0.4471,
+      "step": 11840
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.346888552712554,
+      "learning_rate": 5.5402259549418023e-08,
+      "loss": 0.4377,
+      "step": 11841
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.5539252585517795,
+      "learning_rate": 5.520873361907597e-08,
+      "loss": 0.4641,
+      "step": 11842
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.38224470851627,
+      "learning_rate": 5.5015544407155954e-08,
+      "loss": 0.4138,
+      "step": 11843
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.909856906693069,
+      "learning_rate": 5.4822691926813556e-08,
+      "loss": 0.4305,
+      "step": 11844
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.6784862277814,
+      "learning_rate": 5.4630176191180495e-08,
+      "loss": 0.5041,
+      "step": 11845
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.590514351112655,
+      "learning_rate": 5.443799721336684e-08,
+      "loss": 0.4734,
+      "step": 11846
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.155607025963272,
+      "learning_rate": 5.424615500645991e-08,
+      "loss": 0.5169,
+      "step": 11847
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.768094162304882,
+      "learning_rate": 5.405464958352091e-08,
+      "loss": 0.4695,
+      "step": 11848
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.614275032606107,
+      "learning_rate": 5.386348095759275e-08,
+      "loss": 0.4257,
+      "step": 11849
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.537328439997272,
+      "learning_rate": 5.367264914169223e-08,
+      "loss": 0.5228,
+      "step": 11850
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.249073849957462,
+      "learning_rate": 5.348215414881452e-08,
+      "loss": 0.4387,
+      "step": 11851
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.868767734574256,
+      "learning_rate": 5.3291995991931466e-08,
+      "loss": 0.4152,
+      "step": 11852
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.124852217162465,
+      "learning_rate": 5.310217468399104e-08,
+      "loss": 0.4359,
+      "step": 11853
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.114389887886277,
+      "learning_rate": 5.2912690237920675e-08,
+      "loss": 0.4132,
+      "step": 11854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.1228327941792795,
+      "learning_rate": 5.2723542666622826e-08,
+      "loss": 0.4889,
+      "step": 11855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.680433586525278,
+      "learning_rate": 5.253473198297776e-08,
+      "loss": 0.4271,
+      "step": 11856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5306368463354252,
+      "learning_rate": 5.234625819984185e-08,
+      "loss": 0.7247,
+      "step": 11857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.271338386034314,
+      "learning_rate": 5.215812133004983e-08,
+      "loss": 0.5779,
+      "step": 11858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.97589604868857,
+      "learning_rate": 5.197032138641367e-08,
+      "loss": 0.4781,
+      "step": 11859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.202443063401861,
+      "learning_rate": 5.1782858381720925e-08,
+      "loss": 0.4231,
+      "step": 11860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.192308384219843,
+      "learning_rate": 5.1595732328736936e-08,
+      "loss": 0.4566,
+      "step": 11861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.293261612465952,
+      "learning_rate": 5.140894324020429e-08,
+      "loss": 0.4459,
+      "step": 11862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.141597982885352,
+      "learning_rate": 5.122249112884281e-08,
+      "loss": 0.4955,
+      "step": 11863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 144.48035169013636,
+      "learning_rate": 5.103637600734901e-08,
+      "loss": 0.4677,
+      "step": 11864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.690089322354425,
+      "learning_rate": 5.0850597888396634e-08,
+      "loss": 0.5953,
+      "step": 11865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.012969376328263,
+      "learning_rate": 5.066515678463557e-08,
+      "loss": 0.4956,
+      "step": 11866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5095717513347453,
+      "learning_rate": 5.04800527086946e-08,
+      "loss": 0.692,
+      "step": 11867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.65001058475019,
+      "learning_rate": 5.029528567317754e-08,
+      "loss": 0.6212,
+      "step": 11868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.2684750543959815,
+      "learning_rate": 5.011085569066654e-08,
+      "loss": 0.4564,
+      "step": 11869
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.417465341931219,
+      "learning_rate": 4.992676277372099e-08,
+      "loss": 0.5445,
+      "step": 11870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5960063885726243,
+      "learning_rate": 4.974300693487588e-08,
+      "loss": 0.4098,
+      "step": 11871
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.67366271821907,
+      "learning_rate": 4.955958818664508e-08,
+      "loss": 0.5636,
+      "step": 11872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4767857794672292,
+      "learning_rate": 4.937650654151749e-08,
+      "loss": 0.691,
+      "step": 11873
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.787731154621455,
+      "learning_rate": 4.919376201196146e-08,
+      "loss": 0.4327,
+      "step": 11874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.469960647382713,
+      "learning_rate": 4.901135461042039e-08,
+      "loss": 0.4459,
+      "step": 11875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.99179617125659,
+      "learning_rate": 4.882928434931544e-08,
+      "loss": 0.4494,
+      "step": 11876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.698020616419825,
+      "learning_rate": 4.864755124104448e-08,
+      "loss": 0.5299,
+      "step": 11877
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.451007736883973,
+      "learning_rate": 4.846615529798371e-08,
+      "loss": 0.4048,
+      "step": 11878
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.193135235121703,
+      "learning_rate": 4.8285096532484367e-08,
+      "loss": 0.3986,
+      "step": 11879
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.524402061781414,
+      "learning_rate": 4.810437495687603e-08,
+      "loss": 0.4213,
+      "step": 11880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.214454951405807,
+      "learning_rate": 4.7923990583465526e-08,
+      "loss": 0.4268,
+      "step": 11881
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.915079560177084,
+      "learning_rate": 4.7743943424535255e-08,
+      "loss": 0.4122,
+      "step": 11882
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.035567107840432,
+      "learning_rate": 4.7564233492347066e-08,
+      "loss": 0.4592,
+      "step": 11883
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 15.569597964963183,
+      "learning_rate": 4.738486079913729e-08,
+      "loss": 0.3705,
+      "step": 11884
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.934754835692306,
+      "learning_rate": 4.720582535712059e-08,
+      "loss": 0.4256,
+      "step": 11885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.017158319143487,
+      "learning_rate": 4.70271271784889e-08,
+      "loss": 0.4436,
+      "step": 11886
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.441256741908014,
+      "learning_rate": 4.684876627541024e-08,
+      "loss": 0.5793,
+      "step": 11887
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 32.32261310575091,
+      "learning_rate": 4.6670742660030465e-08,
+      "loss": 0.4622,
+      "step": 11888
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.681146519987356,
+      "learning_rate": 4.6493056344472096e-08,
+      "loss": 0.5294,
+      "step": 11889
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.090377503852615,
+      "learning_rate": 4.6315707340834906e-08,
+      "loss": 0.4296,
+      "step": 11890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.77511638808844,
+      "learning_rate": 4.613869566119589e-08,
+      "loss": 0.4046,
+      "step": 11891
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.715788724742003,
+      "learning_rate": 4.596202131760819e-08,
+      "loss": 0.4761,
+      "step": 11892
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.000260682642713,
+      "learning_rate": 4.578568432210273e-08,
+      "loss": 0.4471,
+      "step": 11893
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.606069498086299,
+      "learning_rate": 4.560968468668714e-08,
+      "loss": 0.4072,
+      "step": 11894
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.310735677543716,
+      "learning_rate": 4.5434022423346824e-08,
+      "loss": 0.5483,
+      "step": 11895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.3645047453442105,
+      "learning_rate": 4.525869754404277e-08,
+      "loss": 0.457,
+      "step": 11896
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.727158323455887,
+      "learning_rate": 4.508371006071488e-08,
+      "loss": 0.4956,
+      "step": 11897
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.04658367261442,
+      "learning_rate": 4.49090599852775e-08,
+      "loss": 0.4595,
+      "step": 11898
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.8474294985139315,
+      "learning_rate": 4.4734747329625014e-08,
+      "loss": 0.5718,
+      "step": 11899
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.78308289120337,
+      "learning_rate": 4.456077210562626e-08,
+      "loss": 0.5178,
+      "step": 11900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.510819066710182,
+      "learning_rate": 4.438713432512898e-08,
+      "loss": 0.4347,
+      "step": 11901
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.529011208120253,
+      "learning_rate": 4.421383399995705e-08,
+      "loss": 0.4912,
+      "step": 11902
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.242665413775984,
+      "learning_rate": 4.4040871141911025e-08,
+      "loss": 0.4787,
+      "step": 11903
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 21.2263793043964,
+      "learning_rate": 4.3868245762768715e-08,
+      "loss": 0.4631,
+      "step": 11904
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.05682315188813,
+      "learning_rate": 4.3695957874286265e-08,
+      "loss": 0.4771,
+      "step": 11905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.997794595837972,
+      "learning_rate": 4.352400748819485e-08,
+      "loss": 0.4573,
+      "step": 11906
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.194299931430457,
+      "learning_rate": 4.335239461620344e-08,
+      "loss": 0.3904,
+      "step": 11907
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.664574041282564,
+      "learning_rate": 4.318111926999824e-08,
+      "loss": 0.5234,
+      "step": 11908
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.300606497679658,
+      "learning_rate": 4.301018146124214e-08,
+      "loss": 0.5784,
+      "step": 11909
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.788144200783936,
+      "learning_rate": 4.2839581201576385e-08,
+      "loss": 0.5816,
+      "step": 11910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.595317733373496,
+      "learning_rate": 4.2669318502616684e-08,
+      "loss": 0.4829,
+      "step": 11911
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.75793381151776,
+      "learning_rate": 4.2499393375958206e-08,
+      "loss": 0.5451,
+      "step": 11912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.118124739407931,
+      "learning_rate": 4.2329805833171144e-08,
+      "loss": 0.5371,
+      "step": 11913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.388805636803316,
+      "learning_rate": 4.2160555885804587e-08,
+      "loss": 0.4638,
+      "step": 11914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.8074294733993295,
+      "learning_rate": 4.199164354538377e-08,
+      "loss": 0.4631,
+      "step": 11915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.011996182192922,
+      "learning_rate": 4.1823068823409475e-08,
+      "loss": 0.5677,
+      "step": 11916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.239638627227559,
+      "learning_rate": 4.1654831731361976e-08,
+      "loss": 0.4214,
+      "step": 11917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.870114131830911,
+      "learning_rate": 4.148693228069767e-08,
+      "loss": 0.491,
+      "step": 11918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 14.49322268529905,
+      "learning_rate": 4.1319370482849616e-08,
+      "loss": 0.4414,
+      "step": 11919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.749730436217057,
+      "learning_rate": 4.11521463492276e-08,
+      "loss": 0.5481,
+      "step": 11920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 42.08650211288601,
+      "learning_rate": 4.09852598912186e-08,
+      "loss": 0.4623,
+      "step": 11921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.237303269116277,
+      "learning_rate": 4.081871112018798e-08,
+      "loss": 0.4729,
+      "step": 11922
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.962487478943116,
+      "learning_rate": 4.065250004747612e-08,
+      "loss": 0.4199,
+      "step": 11923
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.285569983949233,
+      "learning_rate": 4.0486626684401176e-08,
+      "loss": 0.3272,
+      "step": 11924
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.648338255420256,
+      "learning_rate": 4.032109104225912e-08,
+      "loss": 0.3792,
+      "step": 11925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.080919608142058,
+      "learning_rate": 4.015589313232149e-08,
+      "loss": 0.4729,
+      "step": 11926
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.357305983403695,
+      "learning_rate": 3.999103296583817e-08,
+      "loss": 0.436,
+      "step": 11927
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.4277961975093385,
+      "learning_rate": 3.9826510554034635e-08,
+      "loss": 0.5521,
+      "step": 11928
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.670656095355808,
+      "learning_rate": 3.96623259081147e-08,
+      "loss": 0.4239,
+      "step": 11929
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.696580713687895,
+      "learning_rate": 3.9498479039258294e-08,
+      "loss": 0.4709,
+      "step": 11930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.616446269178685,
+      "learning_rate": 3.9334969958622624e-08,
+      "loss": 0.5802,
+      "step": 11931
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.652956892654295,
+      "learning_rate": 3.9171798677342665e-08,
+      "loss": 0.5098,
+      "step": 11932
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.506685588481018,
+      "learning_rate": 3.900896520652897e-08,
+      "loss": 0.4607,
+      "step": 11933
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.894175349658938,
+      "learning_rate": 3.8846469557269894e-08,
+      "loss": 0.5111,
+      "step": 11934
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.9826785985924955,
+      "learning_rate": 3.8684311740630476e-08,
+      "loss": 0.4591,
+      "step": 11935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7272098056964231,
+      "learning_rate": 3.8522491767653546e-08,
+      "loss": 0.7605,
+      "step": 11936
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.92435172408022,
+      "learning_rate": 3.8361009649357515e-08,
+      "loss": 0.4432,
+      "step": 11937
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.726392838068984,
+      "learning_rate": 3.81998653967397e-08,
+      "loss": 0.4134,
+      "step": 11938
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8638540603653717,
+      "learning_rate": 3.803905902077243e-08,
+      "loss": 0.7214,
+      "step": 11939
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.122951561094371,
+      "learning_rate": 3.7878590532405834e-08,
+      "loss": 0.4168,
+      "step": 11940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.145211239651552,
+      "learning_rate": 3.771845994256784e-08,
+      "loss": 0.5603,
+      "step": 11941
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 19.99069636246576,
+      "learning_rate": 3.75586672621614e-08,
+      "loss": 0.4796,
+      "step": 11942
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 13.393281867934649,
+      "learning_rate": 3.739921250206946e-08,
+      "loss": 0.462,
+      "step": 11943
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.612807979796039,
+      "learning_rate": 3.7240095673148904e-08,
+      "loss": 0.4834,
+      "step": 11944
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.752789106674938,
+      "learning_rate": 3.70813167862355e-08,
+      "loss": 0.3823,
+      "step": 11945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.667236580780397,
+      "learning_rate": 3.692287585214005e-08,
+      "loss": 0.4485,
+      "step": 11946
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.74698681840303,
+      "learning_rate": 3.676477288165392e-08,
+      "loss": 0.4864,
+      "step": 11947
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8491209110849995,
+      "learning_rate": 3.660700788554128e-08,
+      "loss": 0.3681,
+      "step": 11948
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.16700517962048,
+      "learning_rate": 3.644958087454631e-08,
+      "loss": 0.3622,
+      "step": 11949
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.014217838159361,
+      "learning_rate": 3.629249185938877e-08,
+      "loss": 0.4667,
+      "step": 11950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.6523016351744815,
+      "learning_rate": 3.61357408507651e-08,
+      "loss": 0.4125,
+      "step": 11951
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.668634099775164,
+      "learning_rate": 3.597932785935121e-08,
+      "loss": 0.4783,
+      "step": 11952
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.2635577538016065,
+      "learning_rate": 3.5823252895795804e-08,
+      "loss": 0.4491,
+      "step": 11953
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.90798488170239,
+      "learning_rate": 3.566751597072815e-08,
+      "loss": 0.5171,
+      "step": 11954
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.946847280823654,
+      "learning_rate": 3.5512117094753104e-08,
+      "loss": 0.5157,
+      "step": 11955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.22534958128641,
+      "learning_rate": 3.5357056278453315e-08,
+      "loss": 0.3538,
+      "step": 11956
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.0552010408236505,
+      "learning_rate": 3.520233353238645e-08,
+      "loss": 0.3562,
+      "step": 11957
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.929464673417368,
+      "learning_rate": 3.504794886708907e-08,
+      "loss": 0.4153,
+      "step": 11958
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.787052087953398,
+      "learning_rate": 3.489390229307388e-08,
+      "loss": 0.4673,
+      "step": 11959
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8848615587779038,
+      "learning_rate": 3.474019382083083e-08,
+      "loss": 0.6077,
+      "step": 11960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.975136999779235,
+      "learning_rate": 3.4586823460827643e-08,
+      "loss": 0.4481,
+      "step": 11961
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.115352289162367,
+      "learning_rate": 3.443379122350654e-08,
+      "loss": 0.4948,
+      "step": 11962
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.867056954768568,
+      "learning_rate": 3.428109711928973e-08,
+      "loss": 0.3959,
+      "step": 11963
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.918346881362464,
+      "learning_rate": 3.4128741158574455e-08,
+      "loss": 0.4542,
+      "step": 11964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.019702347970657,
+      "learning_rate": 3.397672335173574e-08,
+      "loss": 0.4683,
+      "step": 11965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.4767519866026335,
+      "learning_rate": 3.382504370912476e-08,
+      "loss": 0.4839,
+      "step": 11966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.09306301324595,
+      "learning_rate": 3.367370224107047e-08,
+      "loss": 0.5115,
+      "step": 11967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.937866296692283,
+      "learning_rate": 3.352269895787907e-08,
+      "loss": 0.3629,
+      "step": 11968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.05993199520233,
+      "learning_rate": 3.3372033869832344e-08,
+      "loss": 0.4517,
+      "step": 11969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.95747630948206,
+      "learning_rate": 3.322170698719096e-08,
+      "loss": 0.4037,
+      "step": 11970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.685295854325606,
+      "learning_rate": 3.307171832019007e-08,
+      "loss": 0.5286,
+      "step": 11971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.272336986369703,
+      "learning_rate": 3.292206787904484e-08,
+      "loss": 0.5358,
+      "step": 11972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.848719092253018,
+      "learning_rate": 3.277275567394489e-08,
+      "loss": 0.4065,
+      "step": 11973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.733356853257411,
+      "learning_rate": 3.2623781715057645e-08,
+      "loss": 0.5011,
+      "step": 11974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 20.281553126134778,
+      "learning_rate": 3.2475146012528324e-08,
+      "loss": 0.4563,
+      "step": 11975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9363413185055895,
+      "learning_rate": 3.2326848576477164e-08,
+      "loss": 0.4826,
+      "step": 11976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.396246156080043,
+      "learning_rate": 3.217888941700387e-08,
+      "loss": 0.4002,
+      "step": 11977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.794852979151059,
+      "learning_rate": 3.203126854418315e-08,
+      "loss": 0.4444,
+      "step": 11978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.934705243511784,
+      "learning_rate": 3.188398596806697e-08,
+      "loss": 0.452,
+      "step": 11979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.7964369267574245,
+      "learning_rate": 3.173704169868508e-08,
+      "loss": 0.5069,
+      "step": 11980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.492824449822018,
+      "learning_rate": 3.1590435746043925e-08,
+      "loss": 0.5038,
+      "step": 11981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.324295437704676,
+      "learning_rate": 3.144416812012663e-08,
+      "loss": 0.4344,
+      "step": 11982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.243113009374217,
+      "learning_rate": 3.1298238830893e-08,
+      "loss": 0.5389,
+      "step": 11983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.641573643839173,
+      "learning_rate": 3.115264788828065e-08,
+      "loss": 0.358,
+      "step": 11984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.039497712383215,
+      "learning_rate": 3.1007395302203314e-08,
+      "loss": 0.4537,
+      "step": 11985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.490531669770009,
+      "learning_rate": 3.086248108255196e-08,
+      "loss": 0.3059,
+      "step": 11986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.718750080787586,
+      "learning_rate": 3.071790523919482e-08,
+      "loss": 0.4369,
+      "step": 11987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.678259120328175,
+      "learning_rate": 3.057366778197679e-08,
+      "loss": 0.478,
+      "step": 11988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.622016968387827,
+      "learning_rate": 3.042976872072056e-08,
+      "loss": 0.3934,
+      "step": 11989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.524226340143262,
+      "learning_rate": 3.0286208065223864e-08,
+      "loss": 0.4259,
+      "step": 11990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.612865238945881,
+      "learning_rate": 3.0142985825262764e-08,
+      "loss": 0.3672,
+      "step": 11991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.567793054098077,
+      "learning_rate": 3.000010201059056e-08,
+      "loss": 0.3867,
+      "step": 11992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6168149436673835,
+      "learning_rate": 2.98575566309367e-08,
+      "loss": 0.7652,
+      "step": 11993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.692753822353628,
+      "learning_rate": 2.9715349696008422e-08,
+      "loss": 0.4499,
+      "step": 11994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.987282803339749,
+      "learning_rate": 2.957348121548853e-08,
+      "loss": 0.5257,
+      "step": 11995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.539815013457304,
+      "learning_rate": 2.943195119903819e-08,
+      "loss": 0.397,
+      "step": 11996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.7556123311347305,
+      "learning_rate": 2.929075965629413e-08,
+      "loss": 0.4729,
+      "step": 11997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.801317179721986,
+      "learning_rate": 2.9149906596872555e-08,
+      "loss": 0.5329,
+      "step": 11998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.846749956669222,
+      "learning_rate": 2.900939203036357e-08,
+      "loss": 0.4226,
+      "step": 11999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.476759307380837,
+      "learning_rate": 2.886921596633563e-08,
+      "loss": 0.4524,
+      "step": 12000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.260411724521363,
+      "learning_rate": 2.8729378414334984e-08,
+      "loss": 0.4129,
+      "step": 12001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.072322569919052,
+      "learning_rate": 2.8589879383882913e-08,
+      "loss": 0.4575,
+      "step": 12002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.202620251541477,
+      "learning_rate": 2.845071888447959e-08,
+      "loss": 0.4753,
+      "step": 12003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.889891382332123,
+      "learning_rate": 2.8311896925600212e-08,
+      "loss": 0.6914,
+      "step": 12004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.999789541862202,
+      "learning_rate": 2.8173413516699444e-08,
+      "loss": 0.4472,
+      "step": 12005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.171670609755644,
+      "learning_rate": 2.803526866720585e-08,
+      "loss": 0.4038,
+      "step": 12006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.471574945885305,
+      "learning_rate": 2.7897462386527465e-08,
+      "loss": 0.4152,
+      "step": 12007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.615544032847585,
+      "learning_rate": 2.7759994684047886e-08,
+      "loss": 0.5017,
+      "step": 12008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.6315881576899445,
+      "learning_rate": 2.7622865569127967e-08,
+      "loss": 0.4225,
+      "step": 12009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.206475975159607,
+      "learning_rate": 2.748607505110579e-08,
+      "loss": 0.3904,
+      "step": 12010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.701011732676849,
+      "learning_rate": 2.7349623139296122e-08,
+      "loss": 0.5431,
+      "step": 12011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.366948469063783,
+      "learning_rate": 2.7213509842990982e-08,
+      "loss": 0.4035,
+      "step": 12012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 39.15451061183386,
+      "learning_rate": 2.7077735171458508e-08,
+      "loss": 0.5422,
+      "step": 12013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.614355235307364,
+      "learning_rate": 2.6942299133945194e-08,
+      "loss": 0.7532,
+      "step": 12014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.385203898655167,
+      "learning_rate": 2.680720173967255e-08,
+      "loss": 0.4242,
+      "step": 12015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.953928226940668,
+      "learning_rate": 2.6672442997840997e-08,
+      "loss": 0.3591,
+      "step": 12016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9566925621619053,
+      "learning_rate": 2.653802291762708e-08,
+      "loss": 0.4374,
+      "step": 12017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.750675316346013,
+      "learning_rate": 2.6403941508183485e-08,
+      "loss": 0.3768,
+      "step": 12018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.788829834120566,
+      "learning_rate": 2.6270198778640678e-08,
+      "loss": 0.379,
+      "step": 12019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4288801747184947,
+      "learning_rate": 2.6136794738106375e-08,
+      "loss": 0.6742,
+      "step": 12020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.883695111532398,
+      "learning_rate": 2.6003729395664977e-08,
+      "loss": 0.4154,
+      "step": 12021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.653281489748066,
+      "learning_rate": 2.5871002760377018e-08,
+      "loss": 0.3866,
+      "step": 12022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.547543094765025,
+      "learning_rate": 2.573861484128082e-08,
+      "loss": 0.5084,
+      "step": 12023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.047811808062089,
+      "learning_rate": 2.560656564739139e-08,
+      "loss": 0.4314,
+      "step": 12024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.95402899463519,
+      "learning_rate": 2.5474855187700987e-08,
+      "loss": 0.4961,
+      "step": 12025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.748334665224216,
+      "learning_rate": 2.5343483471178544e-08,
+      "loss": 0.4507,
+      "step": 12026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.3896329874555615,
+      "learning_rate": 2.5212450506769127e-08,
+      "loss": 0.548,
+      "step": 12027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.035135106756535,
+      "learning_rate": 2.5081756303396155e-08,
+      "loss": 0.4697,
+      "step": 12028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.518464496762412,
+      "learning_rate": 2.4951400869959176e-08,
+      "loss": 0.3842,
+      "step": 12029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.479710230754511,
+      "learning_rate": 2.4821384215335533e-08,
+      "loss": 0.4651,
+      "step": 12030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8124503388048985,
+      "learning_rate": 2.469170634837703e-08,
+      "loss": 0.4296,
+      "step": 12031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.555292705552382,
+      "learning_rate": 2.4562367277916056e-08,
+      "loss": 0.3634,
+      "step": 12032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.406557689195445,
+      "learning_rate": 2.4433367012758337e-08,
+      "loss": 0.3642,
+      "step": 12033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.492778569905039,
+      "learning_rate": 2.4304705561689623e-08,
+      "loss": 0.4307,
+      "step": 12034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.0880550762317505,
+      "learning_rate": 2.4176382933470688e-08,
+      "loss": 0.5341,
+      "step": 12035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.739273261505951,
+      "learning_rate": 2.404839913683954e-08,
+      "loss": 0.4346,
+      "step": 12036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.126361959530628,
+      "learning_rate": 2.392075418051143e-08,
+      "loss": 0.4554,
+      "step": 12037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 15.386754697501376,
+      "learning_rate": 2.3793448073178294e-08,
+      "loss": 0.4968,
+      "step": 12038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.689847281887338,
+      "learning_rate": 2.3666480823509862e-08,
+      "loss": 0.4101,
+      "step": 12039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.249344851574446,
+      "learning_rate": 2.353985244015089e-08,
+      "loss": 0.492,
+      "step": 12040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.67048619435994,
+      "learning_rate": 2.3413562931725032e-08,
+      "loss": 0.468,
+      "step": 12041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.792513856428386,
+      "learning_rate": 2.328761230683152e-08,
+      "loss": 0.5294,
+      "step": 12042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.494555914855442,
+      "learning_rate": 2.3162000574047383e-08,
+      "loss": 0.3983,
+      "step": 12043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.320413359587946,
+      "learning_rate": 2.3036727741926335e-08,
+      "loss": 0.4292,
+      "step": 12044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.6410299100261945,
+      "learning_rate": 2.291179381899822e-08,
+      "loss": 0.4338,
+      "step": 12045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8554520462756527,
+      "learning_rate": 2.278719881377123e-08,
+      "loss": 0.7483,
+      "step": 12046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.917776879373452,
+      "learning_rate": 2.2662942734729687e-08,
+      "loss": 0.5017,
+      "step": 12047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.669134260505604,
+      "learning_rate": 2.2539025590334605e-08,
+      "loss": 0.501,
+      "step": 12048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.947459249820652,
+      "learning_rate": 2.2415447389024236e-08,
+      "loss": 0.4397,
+      "step": 12049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.0479310132995,
+      "learning_rate": 2.229220813921351e-08,
+      "loss": 0.4598,
+      "step": 12050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.648468942478255,
+      "learning_rate": 2.2169307849294608e-08,
+      "loss": 0.4381,
+      "step": 12051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.355367970359024,
+      "learning_rate": 2.204674652763694e-08,
+      "loss": 0.4422,
+      "step": 12052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.29311714802318,
+      "learning_rate": 2.1924524182585506e-08,
+      "loss": 0.4724,
+      "step": 12053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.920211298224478,
+      "learning_rate": 2.1802640822464195e-08,
+      "loss": 0.4488,
+      "step": 12054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.460687113279303,
+      "learning_rate": 2.1681096455571928e-08,
+      "loss": 0.5618,
+      "step": 12055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.037606586186053,
+      "learning_rate": 2.1559891090185415e-08,
+      "loss": 0.7349,
+      "step": 12056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.610699731389902,
+      "learning_rate": 2.143902473455861e-08,
+      "loss": 0.4934,
+      "step": 12057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.424155519593487,
+      "learning_rate": 2.13184973969216e-08,
+      "loss": 0.582,
+      "step": 12058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.3785931487727305,
+      "learning_rate": 2.119830908548226e-08,
+      "loss": 0.4805,
+      "step": 12059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.232448194564889,
+      "learning_rate": 2.1078459808424047e-08,
+      "loss": 0.4921,
+      "step": 12060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.62038877430078,
+      "learning_rate": 2.095894957390876e-08,
+      "loss": 0.4007,
+      "step": 12061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.405348287393177,
+      "learning_rate": 2.0839778390074338e-08,
+      "loss": 0.4406,
+      "step": 12062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.463062463093248,
+      "learning_rate": 2.0720946265035957e-08,
+      "loss": 0.4579,
+      "step": 12063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 32.97406011533392,
+      "learning_rate": 2.0602453206885475e-08,
+      "loss": 0.4117,
+      "step": 12064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.5050855843612645,
+      "learning_rate": 2.0484299223691438e-08,
+      "loss": 0.4494,
+      "step": 12065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.0953003516669515,
+      "learning_rate": 2.036648432350019e-08,
+      "loss": 0.4354,
+      "step": 12066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.695213547670246,
+      "learning_rate": 2.0249008514334202e-08,
+      "loss": 0.3921,
+      "step": 12067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.965065297656398,
+      "learning_rate": 2.013187180419318e-08,
+      "loss": 0.4224,
+      "step": 12068
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.533899892061783,
+      "learning_rate": 2.0015074201052976e-08,
+      "loss": 0.4793,
+      "step": 12069
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.7777710169605845,
+      "learning_rate": 1.989861571286722e-08,
+      "loss": 0.418,
+      "step": 12070
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.58161282104225,
+      "learning_rate": 1.9782496347566792e-08,
+      "loss": 0.478,
+      "step": 12071
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.646311502664988,
+      "learning_rate": 1.9666716113058704e-08,
+      "loss": 0.7292,
+      "step": 12072
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.393536945982943,
+      "learning_rate": 1.9551275017226644e-08,
+      "loss": 0.497,
+      "step": 12073
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.39757692823315,
+      "learning_rate": 1.9436173067932107e-08,
+      "loss": 0.3905,
+      "step": 12074
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 10.953113834462844,
+      "learning_rate": 1.9321410273012708e-08,
+      "loss": 0.4461,
+      "step": 12075
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.315459000882422,
+      "learning_rate": 1.9206986640283865e-08,
+      "loss": 0.4467,
+      "step": 12076
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.649786159999267,
+      "learning_rate": 1.909290217753601e-08,
+      "loss": 0.4244,
+      "step": 12077
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.278335095232435,
+      "learning_rate": 1.8979156892539598e-08,
+      "loss": 0.3889,
+      "step": 12078
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.887375269156932,
+      "learning_rate": 1.886575079303843e-08,
+      "loss": 0.4504,
+      "step": 12079
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.485076187213427,
+      "learning_rate": 1.8752683886756328e-08,
+      "loss": 0.4736,
+      "step": 12080
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.214718165144905,
+      "learning_rate": 1.8639956181391582e-08,
+      "loss": 0.4511,
+      "step": 12081
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.918617026046256,
+      "learning_rate": 1.8527567684620828e-08,
+      "loss": 0.3826,
+      "step": 12082
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.372756369572135,
+      "learning_rate": 1.8415518404097942e-08,
+      "loss": 0.4665,
+      "step": 12083
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.182527830007951,
+      "learning_rate": 1.830380834745238e-08,
+      "loss": 0.4249,
+      "step": 12084
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.153376087383311,
+      "learning_rate": 1.8192437522290827e-08,
+      "loss": 0.3691,
+      "step": 12085
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.786001036287271,
+      "learning_rate": 1.808140593619723e-08,
+      "loss": 0.482,
+      "step": 12086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.697915496944337,
+      "learning_rate": 1.7970713596733302e-08,
+      "loss": 0.5227,
+      "step": 12087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.159855446210151,
+      "learning_rate": 1.786036051143525e-08,
+      "loss": 0.4147,
+      "step": 12088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.00105719037766,
+      "learning_rate": 1.775034668781872e-08,
+      "loss": 0.395,
+      "step": 12089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.336320503163185,
+      "learning_rate": 1.7640672133374394e-08,
+      "loss": 0.4145,
+      "step": 12090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.116468780237897,
+      "learning_rate": 1.7531336855571292e-08,
+      "loss": 0.4217,
+      "step": 12091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.28763383224543,
+      "learning_rate": 1.7422340861854013e-08,
+      "loss": 0.4487,
+      "step": 12092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.90335397320926,
+      "learning_rate": 1.7313684159645515e-08,
+      "loss": 0.4543,
+      "step": 12093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.997216891156546,
+      "learning_rate": 1.720536675634432e-08,
+      "loss": 0.3941,
+      "step": 12094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.42125656026848,
+      "learning_rate": 1.7097388659326197e-08,
+      "loss": 0.3994,
+      "step": 12095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.825173906910768,
+      "learning_rate": 1.6989749875944704e-08,
+      "loss": 0.4481,
+      "step": 12096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9920233203527518,
+      "learning_rate": 1.688245041352843e-08,
+      "loss": 0.7071,
+      "step": 12097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.98574242339956,
+      "learning_rate": 1.6775490279384855e-08,
+      "loss": 0.5102,
+      "step": 12098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.229470266404867,
+      "learning_rate": 1.66688694807976e-08,
+      "loss": 0.3883,
+      "step": 12099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.2160060711070075,
+      "learning_rate": 1.6562588025026417e-08,
+      "loss": 0.4602,
+      "step": 12100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.587305850774353,
+      "learning_rate": 1.64566459193094e-08,
+      "loss": 0.4635,
+      "step": 12101
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.884292030611133,
+      "learning_rate": 1.635104317085967e-08,
+      "loss": 0.4765,
+      "step": 12102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.848807338249659,
+      "learning_rate": 1.6245779786869254e-08,
+      "loss": 0.4659,
+      "step": 12103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.207352467462423,
+      "learning_rate": 1.6140855774505747e-08,
+      "loss": 0.4228,
+      "step": 12104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.603705532499202,
+      "learning_rate": 1.6036271140913995e-08,
+      "loss": 0.4085,
+      "step": 12105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.747934421920564,
+      "learning_rate": 1.5932025893215518e-08,
+      "loss": 0.5425,
+      "step": 12106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.47517798896968,
+      "learning_rate": 1.582812003850964e-08,
+      "loss": 0.4874,
+      "step": 12107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.605512706557024,
+      "learning_rate": 1.5724553583871262e-08,
+      "loss": 0.4763,
+      "step": 12108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.201188619592232,
+      "learning_rate": 1.562132653635362e-08,
+      "loss": 0.3539,
+      "step": 12109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.165639699260832,
+      "learning_rate": 1.5518438902984434e-08,
+      "loss": 0.546,
+      "step": 12110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.870622839749599,
+      "learning_rate": 1.5415890690771428e-08,
+      "loss": 0.4892,
+      "step": 12111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.285489042449012,
+      "learning_rate": 1.5313681906697354e-08,
+      "loss": 0.4969,
+      "step": 12112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5544714669734572,
+      "learning_rate": 1.5211812557721638e-08,
+      "loss": 0.6685,
+      "step": 12113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.9334193832444155,
+      "learning_rate": 1.5110282650781516e-08,
+      "loss": 0.4199,
+      "step": 12114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.696732556059748,
+      "learning_rate": 1.5009092192790342e-08,
+      "loss": 0.4592,
+      "step": 12115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.790812900509665,
+      "learning_rate": 1.4908241190639827e-08,
+      "loss": 0.4306,
+      "step": 12116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.554206552275598,
+      "learning_rate": 1.4807729651196146e-08,
+      "loss": 0.4642,
+      "step": 12117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.073967345719604,
+      "learning_rate": 1.4707557581304376e-08,
+      "loss": 0.5976,
+      "step": 12118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.514741827187157,
+      "learning_rate": 1.4607724987785731e-08,
+      "loss": 0.4511,
+      "step": 12119
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.885203138814121,
+      "learning_rate": 1.4508231877438106e-08,
+      "loss": 0.5573,
+      "step": 12120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 30.47152368656504,
+      "learning_rate": 1.440907825703719e-08,
+      "loss": 0.5334,
+      "step": 12121
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4796542796979595,
+      "learning_rate": 1.431026413333425e-08,
+      "loss": 0.7535,
+      "step": 12122
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.446026352156819,
+      "learning_rate": 1.4211789513058349e-08,
+      "loss": 0.5315,
+      "step": 12123
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.533269296978145,
+      "learning_rate": 1.4113654402914679e-08,
+      "loss": 0.4641,
+      "step": 12124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.05563435256815,
+      "learning_rate": 1.401585880958678e-08,
+      "loss": 0.5215,
+      "step": 12125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.5148517020514545,
+      "learning_rate": 1.3918402739733772e-08,
+      "loss": 0.4774,
+      "step": 12126
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.051055072721827,
+      "learning_rate": 1.3821286199991457e-08,
+      "loss": 0.5013,
+      "step": 12127
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.959106054154223,
+      "learning_rate": 1.3724509196972879e-08,
+      "loss": 0.501,
+      "step": 12128
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.598248227175035,
+      "learning_rate": 1.3628071737269433e-08,
+      "loss": 0.382,
+      "step": 12129
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.5126801857517895,
+      "learning_rate": 1.3531973827446976e-08,
+      "loss": 0.5682,
+      "step": 12130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5318592241941944,
+      "learning_rate": 1.3436215474049719e-08,
+      "loss": 0.6534,
+      "step": 12131
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.602213422773175,
+      "learning_rate": 1.3340796683598001e-08,
+      "loss": 0.3774,
+      "step": 12132
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.569565964754153,
+      "learning_rate": 1.3245717462589957e-08,
+      "loss": 0.4196,
+      "step": 12133
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.947769553019833,
+      "learning_rate": 1.3150977817499856e-08,
+      "loss": 0.4715,
+      "step": 12134
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.235300553771234,
+      "learning_rate": 1.3056577754779199e-08,
+      "loss": 0.4453,
+      "step": 12135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.844147598912116,
+      "learning_rate": 1.296251728085618e-08,
+      "loss": 0.5103,
+      "step": 12136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.977680793821477,
+      "learning_rate": 1.2868796402135119e-08,
+      "loss": 0.4682,
+      "step": 12137
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.685811811004337,
+      "learning_rate": 1.2775415124998692e-08,
+      "loss": 0.494,
+      "step": 12138
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.0539464145939705,
+      "learning_rate": 1.2682373455806252e-08,
+      "loss": 0.5025,
+      "step": 12139
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.132757197757231,
+      "learning_rate": 1.2589671400892733e-08,
+      "loss": 0.4685,
+      "step": 12140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.874192209370747,
+      "learning_rate": 1.2497308966570864e-08,
+      "loss": 0.5057,
+      "step": 12141
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.676899092207474,
+      "learning_rate": 1.240528615913006e-08,
+      "loss": 0.3756,
+      "step": 12142
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.082877696844944,
+      "learning_rate": 1.231360298483697e-08,
+      "loss": 0.4407,
+      "step": 12143
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.308775629288685,
+      "learning_rate": 1.2222259449934937e-08,
+      "loss": 0.4593,
+      "step": 12144
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.176385229371494,
+      "learning_rate": 1.2131255560643984e-08,
+      "loss": 0.4304,
+      "step": 12145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.262898930619765,
+      "learning_rate": 1.2040591323160267e-08,
+      "loss": 0.4969,
+      "step": 12146
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.755737664537999,
+      "learning_rate": 1.195026674365829e-08,
+      "loss": 0.4554,
+      "step": 12147
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.216821505370321,
+      "learning_rate": 1.1860281828288689e-08,
+      "loss": 0.3692,
+      "step": 12148
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.376970658924414,
+      "learning_rate": 1.177063658317934e-08,
+      "loss": 0.4649,
+      "step": 12149
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.700790021964995,
+      "learning_rate": 1.1681331014434805e-08,
+      "loss": 0.3189,
+      "step": 12150
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6512747288555474,
+      "learning_rate": 1.159236512813522e-08,
+      "loss": 0.6818,
+      "step": 12151
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.908183652622873,
+      "learning_rate": 1.1503738930339624e-08,
+      "loss": 0.4238,
+      "step": 12152
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.171125951031674,
+      "learning_rate": 1.1415452427083195e-08,
+      "loss": 0.4494,
+      "step": 12153
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.181543814501135,
+      "learning_rate": 1.1327505624377789e-08,
+      "loss": 0.4465,
+      "step": 12154
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.44715744072198,
+      "learning_rate": 1.1239898528211945e-08,
+      "loss": 0.5579,
+      "step": 12155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.991884221966761,
+      "learning_rate": 1.1152631144551452e-08,
+      "loss": 0.4631,
+      "step": 12156
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.244009051311519,
+      "learning_rate": 1.1065703479338219e-08,
+      "loss": 0.4505,
+      "step": 12157
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.229001567499674,
+      "learning_rate": 1.0979115538493069e-08,
+      "loss": 0.3959,
+      "step": 12158
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.949371436456738,
+      "learning_rate": 1.0892867327911283e-08,
+      "loss": 0.4898,
+      "step": 12159
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.143098340915358,
+      "learning_rate": 1.0806958853465943e-08,
+      "loss": 0.4692,
+      "step": 12160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6101120700036846,
+      "learning_rate": 1.0721390121006813e-08,
+      "loss": 0.7098,
+      "step": 12161
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.83418156020922,
+      "learning_rate": 1.0636161136361455e-08,
+      "loss": 0.4681,
+      "step": 12162
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8585533583145482,
+      "learning_rate": 1.055127190533356e-08,
+      "loss": 0.7372,
+      "step": 12163
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.855058277164124,
+      "learning_rate": 1.0466722433702948e-08,
+      "loss": 0.4134,
+      "step": 12164
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.590275176410862,
+      "learning_rate": 1.038251272722779e-08,
+      "loss": 0.4567,
+      "step": 12165
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.33417857065082,
+      "learning_rate": 1.0298642791641833e-08,
+      "loss": 0.4526,
+      "step": 12166
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.754594761400509,
+      "learning_rate": 1.0215112632656621e-08,
+      "loss": 0.4707,
+      "step": 12167
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.487238021619286,
+      "learning_rate": 1.013192225596038e-08,
+      "loss": 0.448,
+      "step": 12168
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.962659604533155,
+      "learning_rate": 1.0049071667218024e-08,
+      "loss": 0.5019,
+      "step": 12169
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.307679722431318,
+      "learning_rate": 9.966560872070041e-09,
+      "loss": 0.4825,
+      "step": 12170
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.612264460990371,
+      "learning_rate": 9.884389876136936e-09,
+      "loss": 0.4731,
+      "step": 12171
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7016220058365215,
+      "learning_rate": 9.802558685013119e-09,
+      "loss": 0.6881,
+      "step": 12172
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.518163824533031,
+      "learning_rate": 9.721067304270804e-09,
+      "loss": 0.5277,
+      "step": 12173
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.385881009840993,
+      "learning_rate": 9.639915739459993e-09,
+      "loss": 0.5202,
+      "step": 12174
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.872339816989136,
+      "learning_rate": 9.559103996105711e-09,
+      "loss": 0.3984,
+      "step": 12175
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.86672164734473,
+      "learning_rate": 9.478632079711892e-09,
+      "loss": 0.4286,
+      "step": 12176
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.825661141698118,
+      "learning_rate": 9.398499995757481e-09,
+      "loss": 0.4216,
+      "step": 12177
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.227678520067817,
+      "learning_rate": 9.318707749699784e-09,
+      "loss": 0.4855,
+      "step": 12178
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.410934169921568,
+      "learning_rate": 9.239255346971676e-09,
+      "loss": 0.4672,
+      "step": 12179
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.21607556866497,
+      "learning_rate": 9.160142792984383e-09,
+      "loss": 0.4613,
+      "step": 12180
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.151957009367798,
+      "learning_rate": 9.081370093124153e-09,
+      "loss": 0.4771,
+      "step": 12181
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.804774195913412,
+      "learning_rate": 9.002937252755584e-09,
+      "loss": 0.4354,
+      "step": 12182
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.954506181798749,
+      "learning_rate": 8.924844277219402e-09,
+      "loss": 0.3247,
+      "step": 12183
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.829938117426024,
+      "learning_rate": 8.847091171833578e-09,
+      "loss": 0.3374,
+      "step": 12184
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.905872067702361,
+      "learning_rate": 8.769677941892762e-09,
+      "loss": 0.4676,
+      "step": 12185
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.960512431113013,
+      "learning_rate": 8.692604592668296e-09,
+      "loss": 0.4704,
+      "step": 12186
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.656561382222711,
+      "learning_rate": 8.61587112940876e-09,
+      "loss": 0.43,
+      "step": 12187
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.248537039470732,
+      "learning_rate": 8.53947755733886e-09,
+      "loss": 0.4483,
+      "step": 12188
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.079960436555146,
+      "learning_rate": 8.463423881661104e-09,
+      "loss": 0.5275,
+      "step": 12189
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.566265092310262,
+      "learning_rate": 8.387710107554681e-09,
+      "loss": 0.5193,
+      "step": 12190
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.122266821233486,
+      "learning_rate": 8.312336240174357e-09,
+      "loss": 0.5133,
+      "step": 12191
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.67446583839315,
+      "learning_rate": 8.237302284654359e-09,
+      "loss": 0.4438,
+      "step": 12192
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.059280309400231,
+      "learning_rate": 8.162608246102265e-09,
+      "loss": 0.4726,
+      "step": 12193
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.8527282170341195,
+      "learning_rate": 8.088254129605677e-09,
+      "loss": 0.4903,
+      "step": 12194
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.785454981548458,
+      "learning_rate": 8.014239940227764e-09,
+      "loss": 0.4276,
+      "step": 12195
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.31588705254961,
+      "learning_rate": 7.940565683007828e-09,
+      "loss": 0.4455,
+      "step": 12196
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.040124038892478,
+      "learning_rate": 7.867231362963524e-09,
+      "loss": 0.7046,
+      "step": 12197
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9712716679760063,
+      "learning_rate": 7.794236985088077e-09,
+      "loss": 0.4078,
+      "step": 12198
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.9115116293666246,
+      "learning_rate": 7.721582554352513e-09,
+      "loss": 0.3809,
+      "step": 12199
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9268582113514423,
+      "learning_rate": 7.649268075703986e-09,
+      "loss": 0.7015,
+      "step": 12200
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.400924523202882,
+      "learning_rate": 7.577293554066335e-09,
+      "loss": 0.4281,
+      "step": 12201
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.240541515639728,
+      "learning_rate": 7.505658994341747e-09,
+      "loss": 0.48,
+      "step": 12202
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.758407609702367,
+      "learning_rate": 7.434364401407435e-09,
+      "loss": 0.3892,
+      "step": 12203
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.740061762871127,
+      "learning_rate": 7.363409780118402e-09,
+      "loss": 0.3389,
+      "step": 12204
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.853647816955954,
+      "learning_rate": 7.292795135306896e-09,
+      "loss": 0.3372,
+      "step": 12205
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6494041688485654,
+      "learning_rate": 7.222520471780181e-09,
+      "loss": 0.7246,
+      "step": 12206
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5944751104945598,
+      "learning_rate": 7.152585794325539e-09,
+      "loss": 0.676,
+      "step": 12207
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.717075302123017,
+      "learning_rate": 7.0829911077030526e-09,
+      "loss": 0.4776,
+      "step": 12208
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.34044148882599,
+      "learning_rate": 7.013736416653927e-09,
+      "loss": 0.4363,
+      "step": 12209
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.34921045860124,
+      "learning_rate": 6.94482172589217e-09,
+      "loss": 0.3638,
+      "step": 12210
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.2008702483904905,
+      "learning_rate": 6.876247040111805e-09,
+      "loss": 0.3424,
+      "step": 12211
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.287054738234311,
+      "learning_rate": 6.8080123639824306e-09,
+      "loss": 0.3811,
+      "step": 12212
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.50411758835782,
+      "learning_rate": 6.7401177021503286e-09,
+      "loss": 0.5115,
+      "step": 12213
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.074254682523527,
+      "learning_rate": 6.6725630592384685e-09,
+      "loss": 0.479,
+      "step": 12214
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.405239265554464,
+      "learning_rate": 6.605348439847059e-09,
+      "loss": 0.4316,
+      "step": 12215
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.111201081824599,
+      "learning_rate": 6.538473848553551e-09,
+      "loss": 0.5461,
+      "step": 12216
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.90304689316303,
+      "learning_rate": 6.471939289911522e-09,
+      "loss": 0.4506,
+      "step": 12217
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.171742954143355,
+      "learning_rate": 6.405744768451794e-09,
+      "loss": 0.5405,
+      "step": 12218
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.666945330105323,
+      "learning_rate": 6.339890288681871e-09,
+      "loss": 0.5597,
+      "step": 12219
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.5894589245268715,
+      "learning_rate": 6.274375855085946e-09,
+      "loss": 0.4841,
+      "step": 12220
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.93855570033258,
+      "learning_rate": 6.209201472126003e-09,
+      "loss": 0.4164,
+      "step": 12221
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.925559855455822,
+      "learning_rate": 6.144367144239049e-09,
+      "loss": 0.5381,
+      "step": 12222
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.459957344595565,
+      "learning_rate": 6.079872875840997e-09,
+      "loss": 0.4407,
+      "step": 12223
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.0726345368386045,
+      "learning_rate": 6.015718671323334e-09,
+      "loss": 0.425,
+      "step": 12224
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.467544764163453,
+      "learning_rate": 5.951904535054231e-09,
+      "loss": 0.3633,
+      "step": 12225
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 11.593211448884324,
+      "learning_rate": 5.888430471379103e-09,
+      "loss": 0.4331,
+      "step": 12226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.814237846027247,
+      "learning_rate": 5.825296484621157e-09,
+      "loss": 0.5173,
+      "step": 12227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.854472025999147,
+      "learning_rate": 5.7625025790791766e-09,
+      "loss": 0.4736,
+      "step": 12228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.597613711084895,
+      "learning_rate": 5.700048759028631e-09,
+      "loss": 0.4251,
+      "step": 12229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.214730849407977,
+      "learning_rate": 5.63793502872223e-09,
+      "loss": 0.379,
+      "step": 12230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.803274124424318,
+      "learning_rate": 5.576161392390478e-09,
+      "loss": 0.3998,
+      "step": 12231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.494409240377395,
+      "learning_rate": 5.51472785424001e-09,
+      "loss": 0.4856,
+      "step": 12232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.294933688118913,
+      "learning_rate": 5.453634418453035e-09,
+      "loss": 0.4664,
+      "step": 12233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.494414697097111,
+      "learning_rate": 5.392881089190671e-09,
+      "loss": 0.3684,
+      "step": 12234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.198795371677408,
+      "learning_rate": 5.332467870589608e-09,
+      "loss": 0.5338,
+      "step": 12235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.317305866045283,
+      "learning_rate": 5.272394766763777e-09,
+      "loss": 0.4223,
+      "step": 12236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.1988304285028155,
+      "learning_rate": 5.2126617818037965e-09,
+      "loss": 0.4433,
+      "step": 12237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.287551702489482,
+      "learning_rate": 5.153268919777521e-09,
+      "loss": 0.5084,
+      "step": 12238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.984992947011204,
+      "learning_rate": 5.094216184728939e-09,
+      "loss": 0.4238,
+      "step": 12239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.433556922590417,
+      "learning_rate": 5.0355035806792795e-09,
+      "loss": 0.4641,
+      "step": 12240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.930945566159448,
+      "learning_rate": 4.97713111162701e-09,
+      "loss": 0.3116,
+      "step": 12241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.752677661646599,
+      "learning_rate": 4.919098781546727e-09,
+      "loss": 0.3732,
+      "step": 12242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.28504934951518,
+      "learning_rate": 4.861406594390827e-09,
+      "loss": 0.4408,
+      "step": 12243
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.848660351753154,
+      "learning_rate": 4.8040545540867234e-09,
+      "loss": 0.4477,
+      "step": 12244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.811198086960836,
+      "learning_rate": 4.747042664540736e-09,
+      "loss": 0.372,
+      "step": 12245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.3049064971901965,
+      "learning_rate": 4.690370929634758e-09,
+      "loss": 0.4259,
+      "step": 12246
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.045773347424767,
+      "learning_rate": 4.6340393532279256e-09,
+      "loss": 0.349,
+      "step": 12247
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.1359873730894945,
+      "learning_rate": 4.578047939156616e-09,
+      "loss": 0.4907,
+      "step": 12248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.162559477737076,
+      "learning_rate": 4.522396691232778e-09,
+      "loss": 0.3987,
+      "step": 12249
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.4808020561769215,
+      "learning_rate": 4.467085613246713e-09,
+      "loss": 0.4411,
+      "step": 12250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.685432327067899,
+      "learning_rate": 4.412114708964299e-09,
+      "loss": 0.3827,
+      "step": 12251
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.882858382333568,
+      "learning_rate": 4.357483982129207e-09,
+      "loss": 0.5442,
+      "step": 12252
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.460206689676475,
+      "learning_rate": 4.303193436461239e-09,
+      "loss": 0.5823,
+      "step": 12253
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.127927931170022,
+      "learning_rate": 4.249243075657439e-09,
+      "loss": 0.4317,
+      "step": 12254
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.797515907813668,
+      "learning_rate": 4.195632903392089e-09,
+      "loss": 0.5107,
+      "step": 12255
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.2203825657106,
+      "learning_rate": 4.142362923315602e-09,
+      "loss": 0.4931,
+      "step": 12256
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.482162339391323,
+      "learning_rate": 4.089433139054521e-09,
+      "loss": 0.4633,
+      "step": 12257
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.8839693760960285,
+      "learning_rate": 4.036843554214298e-09,
+      "loss": 0.457,
+      "step": 12258
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.0930648432316294,
+      "learning_rate": 3.9845941723759555e-09,
+      "loss": 0.5259,
+      "step": 12259
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.023038710426348,
+      "learning_rate": 3.932684997096647e-09,
+      "loss": 0.3615,
+      "step": 12260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.097279229148997,
+      "learning_rate": 3.881116031911325e-09,
+      "loss": 0.4972,
+      "step": 12261
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.9282455471510636,
+      "learning_rate": 3.8298872803321785e-09,
+      "loss": 0.5006,
+      "step": 12262
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.54324973186495,
+      "learning_rate": 3.778998745847528e-09,
+      "loss": 0.4883,
+      "step": 12263
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.57176205311789,
+      "learning_rate": 3.728450431922381e-09,
+      "loss": 0.4867,
+      "step": 12264
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.527273534057132,
+      "learning_rate": 3.6782423419989833e-09,
+      "loss": 0.551,
+      "step": 12265
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.434788931838995,
+      "learning_rate": 3.628374479495711e-09,
+      "loss": 0.3936,
+      "step": 12266
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.145965304595382,
+      "learning_rate": 3.578846847809292e-09,
+      "loss": 0.3902,
+      "step": 12267
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.17482534739161,
+      "learning_rate": 3.5296594503120285e-09,
+      "loss": 0.4579,
+      "step": 12268
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.117923033092591,
+      "learning_rate": 3.4808122903534635e-09,
+      "loss": 0.5387,
+      "step": 12269
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.092798106724094,
+      "learning_rate": 3.4323053712592704e-09,
+      "loss": 0.4168,
+      "step": 12270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.982675901930011,
+      "learning_rate": 3.384138696333472e-09,
+      "loss": 0.4213,
+      "step": 12271
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.886294098132506,
+      "learning_rate": 3.336312268855113e-09,
+      "loss": 0.4665,
+      "step": 12272
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.461605776308228,
+      "learning_rate": 3.2888260920815875e-09,
+      "loss": 0.5274,
+      "step": 12273
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.13896232784939,
+      "learning_rate": 3.2416801692458645e-09,
+      "loss": 0.4866,
+      "step": 12274
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.947571612060741,
+      "learning_rate": 3.1948745035592642e-09,
+      "loss": 0.4649,
+      "step": 12275
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.240342809449452,
+      "learning_rate": 3.148409098208127e-09,
+      "loss": 0.4131,
+      "step": 12276
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.886619480749809,
+      "learning_rate": 3.1022839563571438e-09,
+      "loss": 0.5102,
+      "step": 12277
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.225794893829645,
+      "learning_rate": 3.0564990811471353e-09,
+      "loss": 0.5353,
+      "step": 12278
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.2886694392670615,
+      "learning_rate": 3.0110544756956072e-09,
+      "loss": 0.4646,
+      "step": 12279
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.441763768730558,
+      "learning_rate": 2.965950143097307e-09,
+      "loss": 0.3963,
+      "step": 12280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.859922522170238,
+      "learning_rate": 2.9211860864236665e-09,
+      "loss": 0.5379,
+      "step": 12281
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.842588542895735,
+      "learning_rate": 2.8767623087228026e-09,
+      "loss": 0.497,
+      "step": 12282
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.337032680123096,
+      "learning_rate": 2.8326788130200733e-09,
+      "loss": 0.4227,
+      "step": 12283
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.822565924833635,
+      "learning_rate": 2.7889356023169667e-09,
+      "loss": 0.5308,
+      "step": 12284
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.039067071889835,
+      "learning_rate": 2.7455326795927662e-09,
+      "loss": 0.4261,
+      "step": 12285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.499668684749894,
+      "learning_rate": 2.70247004780233e-09,
+      "loss": 0.5775,
+      "step": 12286
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.947896139310315,
+      "learning_rate": 2.659747709878313e-09,
+      "loss": 0.4531,
+      "step": 12287
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.031917662806817,
+      "learning_rate": 2.617365668730054e-09,
+      "loss": 0.3903,
+      "step": 12288
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.8249845145650765,
+      "learning_rate": 2.5753239272435783e-09,
+      "loss": 0.3945,
+      "step": 12289
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.334496172094954,
+      "learning_rate": 2.533622488281595e-09,
+      "loss": 0.4577,
+      "step": 12290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9115285498681274,
+      "learning_rate": 2.4922613546835006e-09,
+      "loss": 0.7891,
+      "step": 12291
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.793963484732989,
+      "learning_rate": 2.451240529266485e-09,
+      "loss": 0.4428,
+      "step": 12292
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.743203575208449,
+      "learning_rate": 2.4105600148233156e-09,
+      "loss": 0.6066,
+      "step": 12293
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.627486093573117,
+      "learning_rate": 2.3702198141245524e-09,
+      "loss": 0.5144,
+      "step": 12294
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.934564968592557,
+      "learning_rate": 2.3302199299168883e-09,
+      "loss": 0.4736,
+      "step": 12295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.881913182300282,
+      "learning_rate": 2.29056036492481e-09,
+      "loss": 0.5465,
+      "step": 12296
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.746945243633245,
+      "learning_rate": 2.251241121847825e-09,
+      "loss": 0.525,
+      "step": 12297
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.518501469600612,
+      "learning_rate": 2.212262203363791e-09,
+      "loss": 0.3953,
+      "step": 12298
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.483806376673023,
+      "learning_rate": 2.1736236121278067e-09,
+      "loss": 0.4335,
+      "step": 12299
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.006042066257485,
+      "learning_rate": 2.13532535076999e-09,
+      "loss": 0.5649,
+      "step": 12300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.49642577777653,
+      "learning_rate": 2.0973674218988103e-09,
+      "loss": 0.3224,
+      "step": 12301
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.165521041985531,
+      "learning_rate": 2.059749828098867e-09,
+      "loss": 0.5051,
+      "step": 12302
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.051886705351754,
+      "learning_rate": 2.0224725719319995e-09,
+      "loss": 0.4117,
+      "step": 12303
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.84493230609287,
+      "learning_rate": 1.9855356559361773e-09,
+      "loss": 0.4882,
+      "step": 12304
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.595186054251168,
+      "learning_rate": 1.948939082626611e-09,
+      "loss": 0.405,
+      "step": 12305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.200728877622803,
+      "learning_rate": 1.9126828544963062e-09,
+      "loss": 0.4283,
+      "step": 12306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.029166947025014,
+      "learning_rate": 1.8767669740127337e-09,
+      "loss": 0.4187,
+      "step": 12307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7487520359611703,
+      "learning_rate": 1.8411914436228252e-09,
+      "loss": 0.656,
+      "step": 12308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.9722681072880075,
+      "learning_rate": 1.805956265748532e-09,
+      "loss": 0.453,
+      "step": 12309
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.757432029849586,
+      "learning_rate": 1.7710614427890461e-09,
+      "loss": 0.4119,
+      "step": 12310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.48764497970712,
+      "learning_rate": 1.736506977121355e-09,
+      "loss": 0.5534,
+      "step": 12311
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8394747359644712,
+      "learning_rate": 1.7022928710974663e-09,
+      "loss": 0.6834,
+      "step": 12312
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.141206763024278,
+      "learning_rate": 1.668419127047738e-09,
+      "loss": 0.3661,
+      "step": 12313
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.971716325014953,
+      "learning_rate": 1.6348857472792134e-09,
+      "loss": 0.4339,
+      "step": 12314
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.356697073428689,
+      "learning_rate": 1.6016927340745114e-09,
+      "loss": 0.3791,
+      "step": 12315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.786190702323146,
+      "learning_rate": 1.5688400896940459e-09,
+      "loss": 0.4281,
+      "step": 12316
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.388340499683361,
+      "learning_rate": 1.5363278163754714e-09,
+      "loss": 0.3862,
+      "step": 12317
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.299726057160669,
+      "learning_rate": 1.5041559163331277e-09,
+      "loss": 0.3684,
+      "step": 12318
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.391417466264691,
+      "learning_rate": 1.4723243917563746e-09,
+      "loss": 0.3973,
+      "step": 12319
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.968998835027378,
+      "learning_rate": 1.4408332448134776e-09,
+      "loss": 0.4672,
+      "step": 12320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.59396877558612,
+      "learning_rate": 1.4096824776493877e-09,
+      "loss": 0.4897,
+      "step": 12321
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.131365350024623,
+      "learning_rate": 1.378872092384631e-09,
+      "loss": 0.4579,
+      "step": 12322
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.9808187289774235,
+      "learning_rate": 1.3484020911175289e-09,
+      "loss": 0.4392,
+      "step": 12323
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.081660770813718,
+      "learning_rate": 1.318272475923088e-09,
+      "loss": 0.5392,
+      "step": 12324
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.657027165415673,
+      "learning_rate": 1.2884832488524456e-09,
+      "loss": 0.4498,
+      "step": 12325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.49336807394651,
+      "learning_rate": 1.2590344119350895e-09,
+      "loss": 0.5406,
+      "step": 12326
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.759749190030008,
+      "learning_rate": 1.229925967174972e-09,
+      "loss": 0.5469,
+      "step": 12327
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.014157706437048,
+      "learning_rate": 1.2011579165555065e-09,
+      "loss": 0.4574,
+      "step": 12328
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9120254926692795,
+      "learning_rate": 1.1727302620356818e-09,
+      "loss": 0.3917,
+      "step": 12329
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.4398585112747,
+      "learning_rate": 1.1446430055500612e-09,
+      "loss": 0.4346,
+      "step": 12330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.100516743513857,
+      "learning_rate": 1.1168961490121145e-09,
+      "loss": 0.4109,
+      "step": 12331
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.333058839336728,
+      "learning_rate": 1.0894896943114407e-09,
+      "loss": 0.4285,
+      "step": 12332
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5657820259573567,
+      "learning_rate": 1.0624236433143254e-09,
+      "loss": 0.6991,
+      "step": 12333
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.622281104928992,
+      "learning_rate": 1.0356979978631832e-09,
+      "loss": 0.491,
+      "step": 12334
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.874232743685954,
+      "learning_rate": 1.0093127597782248e-09,
+      "loss": 0.4195,
+      "step": 12335
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.0880233308739085,
+      "learning_rate": 9.832679308563463e-10,
+      "loss": 0.3731,
+      "step": 12336
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.0350086703700665,
+      "learning_rate": 9.57563512871129e-10,
+      "loss": 0.4866,
+      "step": 12337
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.499963968640344,
+      "learning_rate": 9.321995075728395e-10,
+      "loss": 0.5812,
+      "step": 12338
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.788885173418302,
+      "learning_rate": 9.071759166884298e-10,
+      "loss": 0.4631,
+      "step": 12339
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.318155459471333,
+      "learning_rate": 8.824927419220919e-10,
+      "loss": 0.5079,
+      "step": 12340
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.471391704598835,
+      "learning_rate": 8.581499849547037e-10,
+      "loss": 0.4406,
+      "step": 12341
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.7475772819405675,
+      "learning_rate": 8.341476474432731e-10,
+      "loss": 0.4882,
+      "step": 12342
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.0124839553561475,
+      "learning_rate": 8.104857310237135e-10,
+      "loss": 0.3861,
+      "step": 12343
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.972921771562594,
+      "learning_rate": 7.871642373058485e-10,
+      "loss": 0.4335,
+      "step": 12344
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.90403364920596,
+      "learning_rate": 7.641831678784073e-10,
+      "loss": 0.4644,
+      "step": 12345
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.608122185646707,
+      "learning_rate": 7.415425243056939e-10,
+      "loss": 0.4665,
+      "step": 12346
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.340127626879913,
+      "learning_rate": 7.192423081303634e-10,
+      "loss": 0.3871,
+      "step": 12347
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.676554347777364,
+      "learning_rate": 6.972825208706457e-10,
+      "loss": 0.4373,
+      "step": 12348
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.843444284526281,
+      "learning_rate": 6.756631640214561e-10,
+      "loss": 0.3519,
+      "step": 12349
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.414388695770437,
+      "learning_rate": 6.543842390555056e-10,
+      "loss": 0.4983,
+      "step": 12350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.167791792891203,
+      "learning_rate": 6.334457474216349e-10,
+      "loss": 0.4892,
+      "step": 12351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.47005270795603,
+      "learning_rate": 6.128476905459257e-10,
+      "loss": 0.4552,
+      "step": 12352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.205803681804681,
+      "learning_rate": 5.925900698300347e-10,
+      "loss": 0.4035,
+      "step": 12353
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.296808004755775,
+      "learning_rate": 5.72672886654524e-10,
+      "loss": 0.2997,
+      "step": 12354
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.075252901435729,
+      "learning_rate": 5.53096142375531e-10,
+      "loss": 0.4931,
+      "step": 12355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.14516768059169,
+      "learning_rate": 5.338598383253235e-10,
+      "loss": 0.3261,
+      "step": 12356
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 18.253184614549028,
+      "learning_rate": 5.149639758145198e-10,
+      "loss": 0.4664,
+      "step": 12357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.203298166587041,
+      "learning_rate": 4.96408556129313e-10,
+      "loss": 0.5269,
+      "step": 12358
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 25.12724600191014,
+      "learning_rate": 4.781935805342474e-10,
+      "loss": 0.5631,
+      "step": 12359
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 14.05092039188108,
+      "learning_rate": 4.603190502683319e-10,
+      "loss": 0.5639,
+      "step": 12360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.660507348551832,
+      "learning_rate": 4.427849665500361e-10,
+      "loss": 0.5075,
+      "step": 12361
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 16.709982227658116,
+      "learning_rate": 4.2559133057229474e-10,
+      "loss": 0.4075,
+      "step": 12362
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8720465076322488,
+      "learning_rate": 4.087381435069482e-10,
+      "loss": 0.7225,
+      "step": 12363
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.966644779921784,
+      "learning_rate": 3.9222540650030173e-10,
+      "loss": 0.4653,
+      "step": 12364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6154198090060472,
+      "learning_rate": 3.760531206781215e-10,
+      "loss": 0.7492,
+      "step": 12365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.759541871820733,
+      "learning_rate": 3.6022128714119363e-10,
+      "loss": 0.4919,
+      "step": 12366
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.4693793976798695,
+      "learning_rate": 3.447299069669896e-10,
+      "loss": 0.4528,
+      "step": 12367
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.3703917343865175,
+      "learning_rate": 3.295789812113315e-10,
+      "loss": 0.5828,
+      "step": 12368
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.699994494042367,
+      "learning_rate": 3.1476851090506134e-10,
+      "loss": 0.4092,
+      "step": 12369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.74674183806656,
+      "learning_rate": 3.0029849705792704e-10,
+      "loss": 0.5297,
+      "step": 12370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.089689107461484,
+      "learning_rate": 2.8616894065358615e-10,
+      "loss": 0.4738,
+      "step": 12371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.856554706101609,
+      "learning_rate": 2.7237984265571227e-10,
+      "loss": 0.7561,
+      "step": 12372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 23.81961901065677,
+      "learning_rate": 2.5893120400244385e-10,
+      "loss": 0.4011,
+      "step": 12373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.865166424208086,
+      "learning_rate": 2.4582302561027004e-10,
+      "loss": 0.444,
+      "step": 12374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9236683400685333,
+      "learning_rate": 2.330553083706999e-10,
+      "loss": 0.7075,
+      "step": 12375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.786657651958321,
+      "learning_rate": 2.2062805315414826e-10,
+      "loss": 0.5838,
+      "step": 12376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.243731787392494,
+      "learning_rate": 2.0854126080660508e-10,
+      "loss": 0.4098,
+      "step": 12377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.9338341026982295,
+      "learning_rate": 1.9679493215074563e-10,
+      "loss": 0.5228,
+      "step": 12378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.950445255468274,
+      "learning_rate": 1.8538906798704071e-10,
+      "loss": 0.5075,
+      "step": 12379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.237364645548257,
+      "learning_rate": 1.743236690920913e-10,
+      "loss": 0.4039,
+      "step": 12380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.072695592912051,
+      "learning_rate": 1.6359873621862865e-10,
+      "loss": 0.5244,
+      "step": 12381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.179674157826277,
+      "learning_rate": 1.532142700977346e-10,
+      "loss": 0.5309,
+      "step": 12382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.19834446952693,
+      "learning_rate": 1.431702714360661e-10,
+      "loss": 0.415,
+      "step": 12383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.907986494887362,
+      "learning_rate": 1.3346674091863076e-10,
+      "loss": 0.4985,
+      "step": 12384
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.63832979306293,
+      "learning_rate": 1.2410367920490107e-10,
+      "loss": 0.4207,
+      "step": 12385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9875917234176934,
+      "learning_rate": 1.1508108693270015e-10,
+      "loss": 0.5165,
+      "step": 12386
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.43690725386349,
+      "learning_rate": 1.0639896471709154e-10,
+      "loss": 0.5201,
+      "step": 12387
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.692586057739486,
+      "learning_rate": 9.805731314871392e-11,
+      "loss": 0.4849,
+      "step": 12388
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.123771362679009,
+      "learning_rate": 9.005613279600145e-11,
+      "loss": 0.4406,
+      "step": 12389
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.7261874573314895,
+      "learning_rate": 8.239542420351854e-11,
+      "loss": 0.4873,
+      "step": 12390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.354978695859381,
+      "learning_rate": 7.507518789307e-11,
+      "loss": 0.4405,
+      "step": 12391
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.020770047788513,
+      "learning_rate": 6.809542436259087e-11,
+      "loss": 0.4168,
+      "step": 12392
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.696415014218201,
+      "learning_rate": 6.145613408836682e-11,
+      "loss": 0.4714,
+      "step": 12393
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.2274374231956555,
+      "learning_rate": 5.51573175217035e-11,
+      "loss": 0.4267,
+      "step": 12394
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.046595541681103,
+      "learning_rate": 4.9198975091702126e-11,
+      "loss": 0.4596,
+      "step": 12395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.970154818912543,
+      "learning_rate": 4.358110720414921e-11,
+      "loss": 0.4091,
+      "step": 12396
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.4056769323205165,
+      "learning_rate": 3.830371424207169e-11,
+      "loss": 0.5022,
+      "step": 12397
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.7233332385401035,
+      "learning_rate": 3.33667965640716e-11,
+      "loss": 0.4959,
+      "step": 12398
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6779419133167717,
+      "learning_rate": 2.8770354506546526e-11,
+      "loss": 0.7547,
+      "step": 12399
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 12.624429850560057,
+      "learning_rate": 2.4514388382579358e-11,
+      "loss": 0.4525,
+      "step": 12400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.632699475444129,
+      "learning_rate": 2.0598898482493413e-11,
+      "loss": 0.4062,
+      "step": 12401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1917548113968803,
+      "learning_rate": 1.702388507218711e-11,
+      "loss": 0.3561,
+      "step": 12402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.764426881773953,
+      "learning_rate": 1.3789348394799285e-11,
+      "loss": 0.4728,
+      "step": 12403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.706806377711919,
+      "learning_rate": 1.0895288671819437e-11,
+      "loss": 0.4535,
+      "step": 12404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.6896722432023425,
+      "learning_rate": 8.341706099201929e-12,
+      "loss": 0.4449,
+      "step": 12405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 15.64899679006073,
+      "learning_rate": 6.128600851806887e-12,
+      "loss": 0.3812,
+      "step": 12406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.685677141325396,
+      "learning_rate": 4.255973079514419e-12,
+      "loss": 0.5073,
+      "step": 12407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.942863939641343,
+      "learning_rate": 2.7238229100001734e-12,
+      "loss": 0.541,
+      "step": 12408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.697896757010862,
+      "learning_rate": 1.5321504476251137e-12,
+      "loss": 0.4455,
+      "step": 12409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.000634344501936,
+      "learning_rate": 6.809557739906325e-13,
+      "loss": 0.4571,
+      "step": 12410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.1632245360669256,
+      "learning_rate": 1.702389462732157e-13,
+      "loss": 0.4696,
+      "step": 12411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 13.299971383956429,
+      "learning_rate": 0.0,
+      "loss": 0.4474,
+      "step": 12412
+    },
+    {
+      "epoch": 1.0,
+      "step": 12412,
+      "total_flos": 1698080928153600.0,
+      "train_loss": 0.5148781449865649,
+      "train_runtime": 128086.7117,
+      "train_samples_per_second": 6.202,
+      "train_steps_per_second": 0.097
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12412,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3000,
+  "total_flos": 1698080928153600.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}