diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,56021 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9928021841648051,
+  "eval_steps": 500,
+  "global_step": 8000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.415610232722863,
+      "learning_rate": 4.132231404958678e-08,
+      "loss": 0.7865,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.7507651462908065,
+      "learning_rate": 8.264462809917357e-08,
+      "loss": 0.8337,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.697050383152881,
+      "learning_rate": 1.2396694214876034e-07,
+      "loss": 0.7606,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.833288022383022,
+      "learning_rate": 1.6528925619834713e-07,
+      "loss": 0.7577,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.6610300424792745,
+      "learning_rate": 2.066115702479339e-07,
+      "loss": 0.8072,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.361957249992969,
+      "learning_rate": 2.4793388429752067e-07,
+      "loss": 0.787,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.178361424391466,
+      "learning_rate": 2.892561983471075e-07,
+      "loss": 0.8219,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7176108393289844,
+      "learning_rate": 3.3057851239669426e-07,
+      "loss": 0.4465,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6557419046963692,
+      "learning_rate": 3.7190082644628103e-07,
+      "loss": 0.4455,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.388562834977926,
+      "learning_rate": 4.132231404958678e-07,
+      "loss": 0.7729,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.305265496796701,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 0.8012,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.2907850336775555,
+      "learning_rate": 4.958677685950413e-07,
+      "loss": 0.8273,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.601527102678475,
+      "learning_rate": 5.371900826446281e-07,
+      "loss": 0.793,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.968476594018755,
+      "learning_rate": 5.78512396694215e-07,
+      "loss": 0.788,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.524793672412605,
+      "learning_rate": 6.198347107438018e-07,
+      "loss": 0.7818,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.443521791810294,
+      "learning_rate": 6.611570247933885e-07,
+      "loss": 0.7434,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.773132453512853,
+      "learning_rate": 7.024793388429753e-07,
+      "loss": 0.7548,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7482263437220364,
+      "learning_rate": 7.438016528925621e-07,
+      "loss": 0.7902,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.836441534543556,
+      "learning_rate": 7.851239669421488e-07,
+      "loss": 0.7889,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9739204476729437,
+      "learning_rate": 8.264462809917356e-07,
+      "loss": 0.7762,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.4474365552708,
+      "learning_rate": 8.677685950413224e-07,
+      "loss": 0.7679,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4580858839397326,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.8085,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.413169939123294,
+      "learning_rate": 9.50413223140496e-07,
+      "loss": 0.6965,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8731977460065075,
+      "learning_rate": 9.917355371900827e-07,
+      "loss": 0.7589,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7087501271234804,
+      "learning_rate": 1.0330578512396695e-06,
+      "loss": 0.6862,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0679975335981493,
+      "learning_rate": 1.0743801652892562e-06,
+      "loss": 0.6426,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.616920787492988,
+      "learning_rate": 1.115702479338843e-06,
+      "loss": 0.7291,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.431583405470523,
+      "learning_rate": 1.15702479338843e-06,
+      "loss": 0.6693,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.759569588989731,
+      "learning_rate": 1.1983471074380167e-06,
+      "loss": 0.6582,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4037335332718492,
+      "learning_rate": 1.2396694214876035e-06,
+      "loss": 0.6811,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4900842129537524,
+      "learning_rate": 1.28099173553719e-06,
+      "loss": 0.7467,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.028361911401686,
+      "learning_rate": 1.322314049586777e-06,
+      "loss": 0.6178,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.571997398821945,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 0.6036,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.276076194538056,
+      "learning_rate": 1.4049586776859506e-06,
+      "loss": 0.6242,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2059372728177826,
+      "learning_rate": 1.4462809917355372e-06,
+      "loss": 0.6995,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5259678899939773,
+      "learning_rate": 1.4876033057851241e-06,
+      "loss": 0.641,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.099201309749444,
+      "learning_rate": 1.5289256198347107e-06,
+      "loss": 0.6355,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0199781281044813,
+      "learning_rate": 1.5702479338842977e-06,
+      "loss": 0.6039,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.172474850706783,
+      "learning_rate": 1.6115702479338842e-06,
+      "loss": 0.6257,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.051256221405437,
+      "learning_rate": 1.6528925619834712e-06,
+      "loss": 0.6485,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8784132131735476,
+      "learning_rate": 1.694214876033058e-06,
+      "loss": 0.6131,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3302814905160405,
+      "learning_rate": 1.7355371900826448e-06,
+      "loss": 0.6148,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.675021092939445,
+      "learning_rate": 1.7768595041322315e-06,
+      "loss": 0.618,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9396764176262677,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.607,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.162903769359593,
+      "learning_rate": 1.859504132231405e-06,
+      "loss": 0.6073,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8725730728879908,
+      "learning_rate": 1.900826446280992e-06,
+      "loss": 0.6039,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1190143535626333,
+      "learning_rate": 1.9421487603305786e-06,
+      "loss": 0.5924,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.831058773283454,
+      "learning_rate": 1.9834710743801654e-06,
+      "loss": 0.6027,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7497304742654913,
+      "learning_rate": 2.024793388429752e-06,
+      "loss": 0.5308,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.031101871069113,
+      "learning_rate": 2.066115702479339e-06,
+      "loss": 0.5773,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0924155453957884,
+      "learning_rate": 2.1074380165289257e-06,
+      "loss": 0.5808,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1410226694005132,
+      "learning_rate": 2.1487603305785124e-06,
+      "loss": 0.6618,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9601583251234524,
+      "learning_rate": 2.1900826446280992e-06,
+      "loss": 0.5147,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.217739964833312,
+      "learning_rate": 2.231404958677686e-06,
+      "loss": 0.6055,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4871845999980895,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 0.5695,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.728315202799985,
+      "learning_rate": 2.31404958677686e-06,
+      "loss": 0.5878,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1838581200045892,
+      "learning_rate": 2.3553719008264463e-06,
+      "loss": 0.5239,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2711658012889107,
+      "learning_rate": 2.3966942148760335e-06,
+      "loss": 0.595,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9191083300460372,
+      "learning_rate": 2.43801652892562e-06,
+      "loss": 0.597,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.232113453296245,
+      "learning_rate": 2.479338842975207e-06,
+      "loss": 0.5439,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.247069820214734,
+      "learning_rate": 2.5206611570247934e-06,
+      "loss": 0.6307,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1081482509346166,
+      "learning_rate": 2.56198347107438e-06,
+      "loss": 0.6352,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.038799318139236,
+      "learning_rate": 2.6033057851239673e-06,
+      "loss": 0.6116,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2767344300217505,
+      "learning_rate": 2.644628099173554e-06,
+      "loss": 0.5667,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1918624104732034,
+      "learning_rate": 2.6859504132231405e-06,
+      "loss": 0.6243,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9263213735038611,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.5455,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.079115837709929,
+      "learning_rate": 2.7685950413223144e-06,
+      "loss": 0.5963,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.422611576288611,
+      "learning_rate": 2.809917355371901e-06,
+      "loss": 0.5641,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6977361968463944,
+      "learning_rate": 2.851239669421488e-06,
+      "loss": 0.5915,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6310705010720266,
+      "learning_rate": 2.8925619834710743e-06,
+      "loss": 0.5116,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9813718105958915,
+      "learning_rate": 2.9338842975206615e-06,
+      "loss": 0.5831,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.533914016058211,
+      "learning_rate": 2.9752066115702483e-06,
+      "loss": 0.5544,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.749797864900883,
+      "learning_rate": 3.016528925619835e-06,
+      "loss": 0.6231,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2739710963750737,
+      "learning_rate": 3.0578512396694214e-06,
+      "loss": 0.5658,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6293489910001722,
+      "learning_rate": 3.0991735537190086e-06,
+      "loss": 0.5888,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.27649736387961,
+      "learning_rate": 3.1404958677685953e-06,
+      "loss": 0.566,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.971009882984369,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 0.5731,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7269504201622956,
+      "learning_rate": 3.2231404958677685e-06,
+      "loss": 0.5792,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.398280082914536,
+      "learning_rate": 3.264462809917356e-06,
+      "loss": 0.5617,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.239264649971885,
+      "learning_rate": 3.3057851239669424e-06,
+      "loss": 0.6041,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1614348411341076,
+      "learning_rate": 3.347107438016529e-06,
+      "loss": 0.6224,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5115529692047422,
+      "learning_rate": 3.388429752066116e-06,
+      "loss": 0.5348,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.669544917763194,
+      "learning_rate": 3.429752066115703e-06,
+      "loss": 0.5508,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9449410268347895,
+      "learning_rate": 3.4710743801652895e-06,
+      "loss": 0.6149,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2476415793258164,
+      "learning_rate": 3.5123966942148763e-06,
+      "loss": 0.566,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.284326985907656,
+      "learning_rate": 3.553719008264463e-06,
+      "loss": 0.5919,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.329487557124964,
+      "learning_rate": 3.5950413223140502e-06,
+      "loss": 0.5753,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6915372745929467,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.5402,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.840436747417264,
+      "learning_rate": 3.6776859504132234e-06,
+      "loss": 0.5935,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0556986278371054,
+      "learning_rate": 3.71900826446281e-06,
+      "loss": 0.5469,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.216467905440008,
+      "learning_rate": 3.7603305785123973e-06,
+      "loss": 0.62,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7956144664342064,
+      "learning_rate": 3.801652892561984e-06,
+      "loss": 0.5953,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.052474595357274,
+      "learning_rate": 3.842975206611571e-06,
+      "loss": 0.5889,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6229806435376872,
+      "learning_rate": 3.884297520661157e-06,
+      "loss": 0.5417,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.529529144898349,
+      "learning_rate": 3.925619834710744e-06,
+      "loss": 0.462,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.104085886441274,
+      "learning_rate": 3.966942148760331e-06,
+      "loss": 0.5942,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7522377211377986,
+      "learning_rate": 4.008264462809918e-06,
+      "loss": 0.5685,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.433678798953759,
+      "learning_rate": 4.049586776859504e-06,
+      "loss": 0.5783,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2162277574010516,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 0.5774,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.1328075855667015,
+      "learning_rate": 4.132231404958678e-06,
+      "loss": 0.4919,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1548005092718774,
+      "learning_rate": 4.173553719008265e-06,
+      "loss": 0.5601,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.275791310124109,
+      "learning_rate": 4.214876033057851e-06,
+      "loss": 0.5364,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.729405960454034,
+      "learning_rate": 4.2561983471074386e-06,
+      "loss": 0.5553,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8846153766860623,
+      "learning_rate": 4.297520661157025e-06,
+      "loss": 0.5103,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0246132215342603,
+      "learning_rate": 4.338842975206612e-06,
+      "loss": 0.5289,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7353792905572192,
+      "learning_rate": 4.3801652892561984e-06,
+      "loss": 0.5145,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.167202823008346,
+      "learning_rate": 4.421487603305786e-06,
+      "loss": 0.5981,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2486808689824627,
+      "learning_rate": 4.462809917355372e-06,
+      "loss": 0.584,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7923020421622617,
+      "learning_rate": 4.504132231404959e-06,
+      "loss": 0.5168,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.998942559674158,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.5632,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1861973024624484,
+      "learning_rate": 4.586776859504133e-06,
+      "loss": 0.5556,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.902236458891177,
+      "learning_rate": 4.62809917355372e-06,
+      "loss": 0.5957,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3472123997091527,
+      "learning_rate": 4.669421487603306e-06,
+      "loss": 0.5574,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.826111048367796,
+      "learning_rate": 4.710743801652893e-06,
+      "loss": 0.5195,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8812962425133646,
+      "learning_rate": 4.75206611570248e-06,
+      "loss": 0.5703,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8624556818394709,
+      "learning_rate": 4.793388429752067e-06,
+      "loss": 0.5656,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.766670998630178,
+      "learning_rate": 4.834710743801653e-06,
+      "loss": 0.5282,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.039610136010939,
+      "learning_rate": 4.87603305785124e-06,
+      "loss": 0.5132,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5079262458353866,
+      "learning_rate": 4.917355371900827e-06,
+      "loss": 0.5289,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4038638893489335,
+      "learning_rate": 4.958677685950414e-06,
+      "loss": 0.5659,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6960722697241173,
+      "learning_rate": 5e-06,
+      "loss": 0.5051,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2161861436132058,
+      "learning_rate": 5.041322314049587e-06,
+      "loss": 0.5622,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0070358384779747,
+      "learning_rate": 5.082644628099174e-06,
+      "loss": 0.569,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.072153444096047,
+      "learning_rate": 5.12396694214876e-06,
+      "loss": 0.5898,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.961455431362149,
+      "learning_rate": 5.165289256198347e-06,
+      "loss": 0.5278,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7520151908460355,
+      "learning_rate": 5.206611570247935e-06,
+      "loss": 0.5762,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6835662297556955,
+      "learning_rate": 5.247933884297521e-06,
+      "loss": 0.5071,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.702178064947002,
+      "learning_rate": 5.289256198347108e-06,
+      "loss": 0.5928,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2345051265285494,
+      "learning_rate": 5.3305785123966946e-06,
+      "loss": 0.5699,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0642219152001933,
+      "learning_rate": 5.371900826446281e-06,
+      "loss": 0.5713,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0108826296235645,
+      "learning_rate": 5.413223140495868e-06,
+      "loss": 0.5459,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1627936472454423,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.516,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6647761146847948,
+      "learning_rate": 5.495867768595042e-06,
+      "loss": 0.4862,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7119580663489593,
+      "learning_rate": 5.537190082644629e-06,
+      "loss": 0.5691,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5852970473911814,
+      "learning_rate": 5.578512396694216e-06,
+      "loss": 0.5285,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4735294501089675,
+      "learning_rate": 5.619834710743802e-06,
+      "loss": 0.5869,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.04141538102587,
+      "learning_rate": 5.661157024793389e-06,
+      "loss": 0.5159,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.724866138626163,
+      "learning_rate": 5.702479338842976e-06,
+      "loss": 0.5488,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7139656285038183,
+      "learning_rate": 5.743801652892562e-06,
+      "loss": 0.5026,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.847574100805026,
+      "learning_rate": 5.785123966942149e-06,
+      "loss": 0.5027,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.230329154971066,
+      "learning_rate": 5.826446280991736e-06,
+      "loss": 0.5595,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.795342282430791,
+      "learning_rate": 5.867768595041323e-06,
+      "loss": 0.5263,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8716269720617247,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 0.5301,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9299922283974924,
+      "learning_rate": 5.9504132231404965e-06,
+      "loss": 0.5687,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7136333319419209,
+      "learning_rate": 5.991735537190083e-06,
+      "loss": 0.5037,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7337829913060052,
+      "learning_rate": 6.03305785123967e-06,
+      "loss": 0.4986,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6919406868411878,
+      "learning_rate": 6.074380165289256e-06,
+      "loss": 0.4855,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.328246418923426,
+      "learning_rate": 6.115702479338843e-06,
+      "loss": 0.5565,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.441722885984954,
+      "learning_rate": 6.15702479338843e-06,
+      "loss": 0.5145,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1257138893693006,
+      "learning_rate": 6.198347107438017e-06,
+      "loss": 0.5299,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1770823656371974,
+      "learning_rate": 6.239669421487604e-06,
+      "loss": 0.5518,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1037313109160927,
+      "learning_rate": 6.280991735537191e-06,
+      "loss": 0.5186,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.095149553016966,
+      "learning_rate": 6.322314049586777e-06,
+      "loss": 0.5334,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6095163348330389,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.4232,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.260362966455831,
+      "learning_rate": 6.404958677685951e-06,
+      "loss": 0.5246,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.18109831500874,
+      "learning_rate": 6.446280991735537e-06,
+      "loss": 0.5689,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6992471311150057,
+      "learning_rate": 6.487603305785124e-06,
+      "loss": 0.5432,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1996411301077297,
+      "learning_rate": 6.528925619834712e-06,
+      "loss": 0.5471,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.683198647229002,
+      "learning_rate": 6.5702479338842985e-06,
+      "loss": 0.4844,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9315335319997355,
+      "learning_rate": 6.611570247933885e-06,
+      "loss": 0.5114,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.034007691224993,
+      "learning_rate": 6.652892561983472e-06,
+      "loss": 0.5095,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.933206220453076,
+      "learning_rate": 6.694214876033058e-06,
+      "loss": 0.5584,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7594153622700035,
+      "learning_rate": 6.735537190082645e-06,
+      "loss": 0.5267,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.128978895295904,
+      "learning_rate": 6.776859504132232e-06,
+      "loss": 0.4973,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8752257215193229,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 0.5569,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7552075283979744,
+      "learning_rate": 6.859504132231406e-06,
+      "loss": 0.5137,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7116607536756927,
+      "learning_rate": 6.900826446280993e-06,
+      "loss": 0.4852,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8310325583717713,
+      "learning_rate": 6.942148760330579e-06,
+      "loss": 0.512,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.051351087670528,
+      "learning_rate": 6.983471074380166e-06,
+      "loss": 0.5072,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6691184220184736,
+      "learning_rate": 7.0247933884297525e-06,
+      "loss": 0.553,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.472591362280194,
+      "learning_rate": 7.066115702479339e-06,
+      "loss": 0.5742,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.97874128709382,
+      "learning_rate": 7.107438016528926e-06,
+      "loss": 0.5504,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.968673756150883,
+      "learning_rate": 7.1487603305785124e-06,
+      "loss": 0.5672,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.58437763050975,
+      "learning_rate": 7.1900826446281005e-06,
+      "loss": 0.5689,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1221625942102254,
+      "learning_rate": 7.231404958677687e-06,
+      "loss": 0.551,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7421455025898023,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.5774,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8920356174568798,
+      "learning_rate": 7.31404958677686e-06,
+      "loss": 0.5269,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9144789767737893,
+      "learning_rate": 7.355371900826447e-06,
+      "loss": 0.5161,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.34326962708711,
+      "learning_rate": 7.396694214876033e-06,
+      "loss": 0.5565,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.702683434570155,
+      "learning_rate": 7.43801652892562e-06,
+      "loss": 0.5342,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4440904350216632,
+      "learning_rate": 7.479338842975207e-06,
+      "loss": 0.4631,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2025744403847805,
+      "learning_rate": 7.520661157024795e-06,
+      "loss": 0.5517,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3152746565304083,
+      "learning_rate": 7.561983471074381e-06,
+      "loss": 0.5385,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8138121988036877,
+      "learning_rate": 7.603305785123968e-06,
+      "loss": 0.5312,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9696695332798093,
+      "learning_rate": 7.644628099173555e-06,
+      "loss": 0.4975,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1638549313150577,
+      "learning_rate": 7.685950413223142e-06,
+      "loss": 0.4598,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.679763180841708,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 0.5828,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.755753338347303,
+      "learning_rate": 7.768595041322314e-06,
+      "loss": 0.6424,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0607035699738367,
+      "learning_rate": 7.809917355371902e-06,
+      "loss": 0.6084,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.563843230786939,
+      "learning_rate": 7.851239669421489e-06,
+      "loss": 0.5493,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7007150791601604,
+      "learning_rate": 7.892561983471076e-06,
+      "loss": 0.5373,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6296232588728798,
+      "learning_rate": 7.933884297520661e-06,
+      "loss": 0.5412,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.642761568125003,
+      "learning_rate": 7.975206611570249e-06,
+      "loss": 0.5568,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.062493720572721,
+      "learning_rate": 8.016528925619836e-06,
+      "loss": 0.4958,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.840113636335198,
+      "learning_rate": 8.057851239669421e-06,
+      "loss": 0.5443,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2357413647918993,
+      "learning_rate": 8.099173553719009e-06,
+      "loss": 0.5351,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.220715219963027,
+      "learning_rate": 8.140495867768596e-06,
+      "loss": 0.5326,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9436370507926042,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.4906,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.233288939881527,
+      "learning_rate": 8.22314049586777e-06,
+      "loss": 0.4961,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.064536487822645,
+      "learning_rate": 8.264462809917356e-06,
+      "loss": 0.5768,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7690467508481755,
+      "learning_rate": 8.305785123966943e-06,
+      "loss": 0.55,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.137594915236406,
+      "learning_rate": 8.34710743801653e-06,
+      "loss": 0.4988,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9534259558411042,
+      "learning_rate": 8.388429752066116e-06,
+      "loss": 0.5566,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5020717347328643,
+      "learning_rate": 8.429752066115703e-06,
+      "loss": 0.5513,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.660014453532568,
+      "learning_rate": 8.47107438016529e-06,
+      "loss": 0.4613,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6983360001377713,
+      "learning_rate": 8.512396694214877e-06,
+      "loss": 0.4163,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.38401204497113,
+      "learning_rate": 8.553719008264464e-06,
+      "loss": 0.5223,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.657695910352119,
+      "learning_rate": 8.59504132231405e-06,
+      "loss": 0.6005,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1393074553357745,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 0.5133,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.524402023924675,
+      "learning_rate": 8.677685950413224e-06,
+      "loss": 0.5466,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.064634815950332,
+      "learning_rate": 8.71900826446281e-06,
+      "loss": 0.4988,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.96453185900645,
+      "learning_rate": 8.760330578512397e-06,
+      "loss": 0.5043,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3011311824481058,
+      "learning_rate": 8.801652892561984e-06,
+      "loss": 0.5443,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9545398206419267,
+      "learning_rate": 8.842975206611571e-06,
+      "loss": 0.5573,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2270890144831594,
+      "learning_rate": 8.884297520661158e-06,
+      "loss": 0.5475,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.367048313677594,
+      "learning_rate": 8.925619834710744e-06,
+      "loss": 0.5026,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3731067718776466,
+      "learning_rate": 8.966942148760331e-06,
+      "loss": 0.5645,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7771313439516687,
+      "learning_rate": 9.008264462809918e-06,
+      "loss": 0.5453,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8002080571572299,
+      "learning_rate": 9.049586776859506e-06,
+      "loss": 0.495,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.302513893194134,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.5282,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.695747565735532,
+      "learning_rate": 9.132231404958678e-06,
+      "loss": 0.5554,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.083286425298984,
+      "learning_rate": 9.173553719008265e-06,
+      "loss": 0.5131,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.330380712840821,
+      "learning_rate": 9.214876033057853e-06,
+      "loss": 0.5469,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8673689557327797,
+      "learning_rate": 9.25619834710744e-06,
+      "loss": 0.5133,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7630221758042226,
+      "learning_rate": 9.297520661157025e-06,
+      "loss": 0.5031,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.421818000908611,
+      "learning_rate": 9.338842975206613e-06,
+      "loss": 0.4737,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1159305242439594,
+      "learning_rate": 9.3801652892562e-06,
+      "loss": 0.5586,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.542872602570653,
+      "learning_rate": 9.421487603305785e-06,
+      "loss": 0.5837,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.8866498424994997,
+      "learning_rate": 9.462809917355372e-06,
+      "loss": 0.5472,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8632223477667107,
+      "learning_rate": 9.50413223140496e-06,
+      "loss": 0.4915,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.769130478817958,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 0.5311,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5927783003609426,
+      "learning_rate": 9.586776859504134e-06,
+      "loss": 0.4975,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.346701530645649,
+      "learning_rate": 9.62809917355372e-06,
+      "loss": 0.5696,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8552511574832422,
+      "learning_rate": 9.669421487603307e-06,
+      "loss": 0.5047,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8612115784851542,
+      "learning_rate": 9.710743801652894e-06,
+      "loss": 0.5683,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7957162557860462,
+      "learning_rate": 9.75206611570248e-06,
+      "loss": 0.5231,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6128091928743173,
+      "learning_rate": 9.793388429752067e-06,
+      "loss": 0.5223,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7441349517290226,
+      "learning_rate": 9.834710743801654e-06,
+      "loss": 0.5417,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7518601110115872,
+      "learning_rate": 9.876033057851241e-06,
+      "loss": 0.5866,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1033851691625745,
+      "learning_rate": 9.917355371900828e-06,
+      "loss": 0.5297,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9910098722645073,
+      "learning_rate": 9.958677685950414e-06,
+      "loss": 0.4461,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8174366976581458,
+      "learning_rate": 1e-05,
+      "loss": 0.5063,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.674372365245035,
+      "learning_rate": 9.999999596102982e-06,
+      "loss": 0.5256,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7487857604376191,
+      "learning_rate": 9.999998384411993e-06,
+      "loss": 0.5951,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6303611245059892,
+      "learning_rate": 9.99999636492723e-06,
+      "loss": 0.6027,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0701765426010597,
+      "learning_rate": 9.999993537649018e-06,
+      "loss": 0.5304,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1090228708257226,
+      "learning_rate": 9.999989902577813e-06,
+      "loss": 0.5487,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2097391973231413,
+      "learning_rate": 9.999985459714203e-06,
+      "loss": 0.5548,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6819193332040362,
+      "learning_rate": 9.999980209058907e-06,
+      "loss": 0.4655,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7160171625750975,
+      "learning_rate": 9.999974150612773e-06,
+      "loss": 0.5536,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.044131904537645,
+      "learning_rate": 9.999967284376777e-06,
+      "loss": 0.5704,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8803170341600512,
+      "learning_rate": 9.999959610352032e-06,
+      "loss": 0.5053,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1350599019041843,
+      "learning_rate": 9.999951128539776e-06,
+      "loss": 0.5145,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.959337484545383,
+      "learning_rate": 9.999941838941378e-06,
+      "loss": 0.5465,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.263615164770816,
+      "learning_rate": 9.999931741558342e-06,
+      "loss": 0.5284,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9392426887684562,
+      "learning_rate": 9.999920836392298e-06,
+      "loss": 0.5087,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0601038827538445,
+      "learning_rate": 9.999909123445006e-06,
+      "loss": 0.5072,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9443319471352343,
+      "learning_rate": 9.99989660271836e-06,
+      "loss": 0.5114,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9823465986216011,
+      "learning_rate": 9.999883274214383e-06,
+      "loss": 0.5244,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5894097797953115,
+      "learning_rate": 9.99986913793523e-06,
+      "loss": 0.5607,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1034456786503797,
+      "learning_rate": 9.99985419388318e-06,
+      "loss": 0.5246,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.925708553269999,
+      "learning_rate": 9.999838442060652e-06,
+      "loss": 0.5578,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.031790416570794,
+      "learning_rate": 9.999821882470188e-06,
+      "loss": 0.5533,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.299536220707557,
+      "learning_rate": 9.999804515114465e-06,
+      "loss": 0.4861,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.339327687472147,
+      "learning_rate": 9.999786339996288e-06,
+      "loss": 0.5479,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7242515927078733,
+      "learning_rate": 9.999767357118594e-06,
+      "loss": 0.4831,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.910605648741558,
+      "learning_rate": 9.99974756648445e-06,
+      "loss": 0.5574,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0505930996455395,
+      "learning_rate": 9.99972696809705e-06,
+      "loss": 0.5484,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3422061192597923,
+      "learning_rate": 9.999705561959727e-06,
+      "loss": 0.5685,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.685857192457201,
+      "learning_rate": 9.999683348075937e-06,
+      "loss": 0.4722,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9628329354004277,
+      "learning_rate": 9.999660326449267e-06,
+      "loss": 0.5987,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2789243043528704,
+      "learning_rate": 9.99963649708344e-06,
+      "loss": 0.5422,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.421312306756197,
+      "learning_rate": 9.999611859982304e-06,
+      "loss": 0.5501,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0296521584425005,
+      "learning_rate": 9.99958641514984e-06,
+      "loss": 0.5765,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5470558706281141,
+      "learning_rate": 9.999560162590157e-06,
+      "loss": 0.5157,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4700139418770868,
+      "learning_rate": 9.999533102307497e-06,
+      "loss": 0.506,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6813756337253796,
+      "learning_rate": 9.999505234306232e-06,
+      "loss": 0.513,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6349444581628045,
+      "learning_rate": 9.999476558590865e-06,
+      "loss": 0.5735,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9746160718526162,
+      "learning_rate": 9.99944707516603e-06,
+      "loss": 0.5774,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.4811608488552475,
+      "learning_rate": 9.999416784036488e-06,
+      "loss": 0.4944,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5957764732299413,
+      "learning_rate": 9.999385685207135e-06,
+      "loss": 0.5006,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9472233687602192,
+      "learning_rate": 9.999353778682992e-06,
+      "loss": 0.5112,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9484798031195745,
+      "learning_rate": 9.999321064469216e-06,
+      "loss": 0.4938,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.920263709004281,
+      "learning_rate": 9.999287542571092e-06,
+      "loss": 0.4956,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6468350417563766,
+      "learning_rate": 9.999253212994035e-06,
+      "loss": 0.4424,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.844929706278709,
+      "learning_rate": 9.999218075743594e-06,
+      "loss": 0.5706,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.482273640778281,
+      "learning_rate": 9.999182130825443e-06,
+      "loss": 0.5308,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9083396990379513,
+      "learning_rate": 9.99914537824539e-06,
+      "loss": 0.5654,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7555024837914865,
+      "learning_rate": 9.99910781800937e-06,
+      "loss": 0.5698,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.563462775984113,
+      "learning_rate": 9.999069450123458e-06,
+      "loss": 0.5425,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.537480565574436,
+      "learning_rate": 9.999030274593845e-06,
+      "loss": 0.5425,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.201318330504299,
+      "learning_rate": 9.998990291426864e-06,
+      "loss": 0.5718,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0509238907273457,
+      "learning_rate": 9.998949500628977e-06,
+      "loss": 0.4952,
+      "step": 293
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6658988414947071,
+      "learning_rate": 9.998907902206769e-06,
+      "loss": 0.5382,
+      "step": 294
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.076364793574101,
+      "learning_rate": 9.998865496166963e-06,
+      "loss": 0.5792,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6504825769137341,
+      "learning_rate": 9.99882228251641e-06,
+      "loss": 0.5384,
+      "step": 296
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.195943188299471,
+      "learning_rate": 9.998778261262093e-06,
+      "loss": 0.4845,
+      "step": 297
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1865330279591957,
+      "learning_rate": 9.99873343241112e-06,
+      "loss": 0.5822,
+      "step": 298
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.583521959332929,
+      "learning_rate": 9.998687795970739e-06,
+      "loss": 0.5497,
+      "step": 299
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.867600307176823,
+      "learning_rate": 9.998641351948319e-06,
+      "loss": 0.5321,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9599824750180688,
+      "learning_rate": 9.998594100351362e-06,
+      "loss": 0.5741,
+      "step": 301
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6829182091305455,
+      "learning_rate": 9.998546041187507e-06,
+      "loss": 0.5211,
+      "step": 302
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0686970869720076,
+      "learning_rate": 9.998497174464517e-06,
+      "loss": 0.525,
+      "step": 303
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7658227546806848,
+      "learning_rate": 9.998447500190283e-06,
+      "loss": 0.5234,
+      "step": 304
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1642748578874906,
+      "learning_rate": 9.998397018372833e-06,
+      "loss": 0.5805,
+      "step": 305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8098658826161074,
+      "learning_rate": 9.998345729020324e-06,
+      "loss": 0.5115,
+      "step": 306
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0808928845396357,
+      "learning_rate": 9.998293632141042e-06,
+      "loss": 0.5643,
+      "step": 307
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.062037007760145,
+      "learning_rate": 9.998240727743401e-06,
+      "loss": 0.5168,
+      "step": 308
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1539325790326656,
+      "learning_rate": 9.99818701583595e-06,
+      "loss": 0.4943,
+      "step": 309
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.820375964588249,
+      "learning_rate": 9.998132496427366e-06,
+      "loss": 0.5088,
+      "step": 310
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.954096590969464,
+      "learning_rate": 9.998077169526459e-06,
+      "loss": 0.5659,
+      "step": 311
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.285658810614788,
+      "learning_rate": 9.998021035142166e-06,
+      "loss": 0.5142,
+      "step": 312
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9304285977081468,
+      "learning_rate": 9.997964093283555e-06,
+      "loss": 0.5098,
+      "step": 313
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4164241091216825,
+      "learning_rate": 9.997906343959827e-06,
+      "loss": 0.5105,
+      "step": 314
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6374326074383707,
+      "learning_rate": 9.997847787180313e-06,
+      "loss": 0.5503,
+      "step": 315
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9372827013095961,
+      "learning_rate": 9.99778842295447e-06,
+      "loss": 0.5462,
+      "step": 316
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5169237820059158,
+      "learning_rate": 9.997728251291891e-06,
+      "loss": 0.5151,
+      "step": 317
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.56778689015433,
+      "learning_rate": 9.997667272202297e-06,
+      "loss": 0.4994,
+      "step": 318
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7363415337809248,
+      "learning_rate": 9.99760548569554e-06,
+      "loss": 0.4856,
+      "step": 319
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6500642022053966,
+      "learning_rate": 9.997542891781602e-06,
+      "loss": 0.541,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.646446903693105,
+      "learning_rate": 9.997479490470594e-06,
+      "loss": 0.5681,
+      "step": 321
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6120661264680305,
+      "learning_rate": 9.997415281772762e-06,
+      "loss": 0.5224,
+      "step": 322
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6969882678190478,
+      "learning_rate": 9.997350265698477e-06,
+      "loss": 0.563,
+      "step": 323
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.702569911701666,
+      "learning_rate": 9.997284442258244e-06,
+      "loss": 0.5087,
+      "step": 324
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5080891548286657,
+      "learning_rate": 9.997217811462698e-06,
+      "loss": 0.5377,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.292261492859581,
+      "learning_rate": 9.997150373322602e-06,
+      "loss": 0.5686,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.150320693707667,
+      "learning_rate": 9.997082127848853e-06,
+      "loss": 0.5537,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8924546749988123,
+      "learning_rate": 9.997013075052476e-06,
+      "loss": 0.5463,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.654314133243649,
+      "learning_rate": 9.996943214944626e-06,
+      "loss": 0.5565,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.923442603059371,
+      "learning_rate": 9.99687254753659e-06,
+      "loss": 0.5881,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8293811891706615,
+      "learning_rate": 9.996801072839789e-06,
+      "loss": 0.5718,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.480332829339136,
+      "learning_rate": 9.996728790865763e-06,
+      "loss": 0.503,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6201449901428477,
+      "learning_rate": 9.996655701626195e-06,
+      "loss": 0.4973,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2742138495119044,
+      "learning_rate": 9.99658180513289e-06,
+      "loss": 0.4997,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.120557910370504,
+      "learning_rate": 9.99650710139779e-06,
+      "loss": 0.5397,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5861768219312984,
+      "learning_rate": 9.996431590432962e-06,
+      "loss": 0.543,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.614805525586774,
+      "learning_rate": 9.996355272250607e-06,
+      "loss": 0.55,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6053554102899765,
+      "learning_rate": 9.996278146863054e-06,
+      "loss": 0.5516,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.796024780854161,
+      "learning_rate": 9.996200214282762e-06,
+      "loss": 0.5283,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6948255023691408,
+      "learning_rate": 9.996121474522323e-06,
+      "loss": 0.5753,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6740506396070718,
+      "learning_rate": 9.996041927594457e-06,
+      "loss": 0.5164,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6315151778674875,
+      "learning_rate": 9.995961573512018e-06,
+      "loss": 0.5759,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.701863204355339,
+      "learning_rate": 9.995880412287985e-06,
+      "loss": 0.5348,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.103113802438708,
+      "learning_rate": 9.995798443935473e-06,
+      "loss": 0.54,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8491565354564938,
+      "learning_rate": 9.995715668467722e-06,
+      "loss": 0.5071,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0318284949244294,
+      "learning_rate": 9.995632085898106e-06,
+      "loss": 0.5683,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1338428368044196,
+      "learning_rate": 9.995547696240132e-06,
+      "loss": 0.531,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9904163800757648,
+      "learning_rate": 9.995462499507427e-06,
+      "loss": 0.5595,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5269130775268465,
+      "learning_rate": 9.99537649571376e-06,
+      "loss": 0.517,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6240532368533307,
+      "learning_rate": 9.995289684873027e-06,
+      "loss": 0.5192,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3651319032052907,
+      "learning_rate": 9.99520206699925e-06,
+      "loss": 0.4983,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.342350856871222,
+      "learning_rate": 9.995113642106584e-06,
+      "loss": 0.5554,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6168109185854025,
+      "learning_rate": 9.995024410209316e-06,
+      "loss": 0.5243,
+      "step": 353
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6988720397199879,
+      "learning_rate": 9.994934371321862e-06,
+      "loss": 0.5005,
+      "step": 354
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.159118467801112,
+      "learning_rate": 9.994843525458771e-06,
+      "loss": 0.5256,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7501348381594886,
+      "learning_rate": 9.994751872634717e-06,
+      "loss": 0.5444,
+      "step": 356
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.558673168866606,
+      "learning_rate": 9.994659412864508e-06,
+      "loss": 0.5795,
+      "step": 357
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3712882222338771,
+      "learning_rate": 9.994566146163082e-06,
+      "loss": 0.4981,
+      "step": 358
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.917448635976479,
+      "learning_rate": 9.994472072545506e-06,
+      "loss": 0.5797,
+      "step": 359
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5699481449625465,
+      "learning_rate": 9.994377192026981e-06,
+      "loss": 0.5329,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.349005454005592,
+      "learning_rate": 9.994281504622831e-06,
+      "loss": 0.5051,
+      "step": 361
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.434754572817806,
+      "learning_rate": 9.99418501034852e-06,
+      "loss": 0.4765,
+      "step": 362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7721504794383782,
+      "learning_rate": 9.994087709219637e-06,
+      "loss": 0.603,
+      "step": 363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2223558544616,
+      "learning_rate": 9.993989601251899e-06,
+      "loss": 0.5589,
+      "step": 364
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1127193527497323,
+      "learning_rate": 9.993890686461159e-06,
+      "loss": 0.5474,
+      "step": 365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1347486367959934,
+      "learning_rate": 9.993790964863394e-06,
+      "loss": 0.5455,
+      "step": 366
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7495898246789374,
+      "learning_rate": 9.993690436474719e-06,
+      "loss": 0.4887,
+      "step": 367
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.783308092584701,
+      "learning_rate": 9.993589101311373e-06,
+      "loss": 0.5348,
+      "step": 368
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5856712387472398,
+      "learning_rate": 9.993486959389728e-06,
+      "loss": 0.5714,
+      "step": 369
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6477109895625353,
+      "learning_rate": 9.993384010726288e-06,
+      "loss": 0.5417,
+      "step": 370
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.309065931402148,
+      "learning_rate": 9.99328025533768e-06,
+      "loss": 0.5701,
+      "step": 371
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6912380894441247,
+      "learning_rate": 9.993175693240673e-06,
+      "loss": 0.5116,
+      "step": 372
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1409674989933656,
+      "learning_rate": 9.993070324452154e-06,
+      "loss": 0.5192,
+      "step": 373
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.778627892390217,
+      "learning_rate": 9.99296414898915e-06,
+      "loss": 0.5501,
+      "step": 374
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.570838939085222,
+      "learning_rate": 9.992857166868814e-06,
+      "loss": 0.4959,
+      "step": 375
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0505514157390987,
+      "learning_rate": 9.99274937810843e-06,
+      "loss": 0.5285,
+      "step": 376
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6209806009721568,
+      "learning_rate": 9.99264078272541e-06,
+      "loss": 0.5205,
+      "step": 377
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.561951117823356,
+      "learning_rate": 9.992531380737303e-06,
+      "loss": 0.5162,
+      "step": 378
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6842346047481311,
+      "learning_rate": 9.99242117216178e-06,
+      "loss": 0.5459,
+      "step": 379
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5206441075953183,
+      "learning_rate": 9.992310157016647e-06,
+      "loss": 0.5212,
+      "step": 380
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.613170143739462,
+      "learning_rate": 9.992198335319839e-06,
+      "loss": 0.5296,
+      "step": 381
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7097419358184256,
+      "learning_rate": 9.992085707089424e-06,
+      "loss": 0.5358,
+      "step": 382
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8302659825726482,
+      "learning_rate": 9.991972272343595e-06,
+      "loss": 0.4922,
+      "step": 383
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.3884596685854786,
+      "learning_rate": 9.991858031100682e-06,
+      "loss": 0.5421,
+      "step": 384
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8222840733689352,
+      "learning_rate": 9.991742983379137e-06,
+      "loss": 0.5477,
+      "step": 385
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8910852787106245,
+      "learning_rate": 9.991627129197552e-06,
+      "loss": 0.5638,
+      "step": 386
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5426420623594577,
+      "learning_rate": 9.991510468574642e-06,
+      "loss": 0.5331,
+      "step": 387
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6218897462067299,
+      "learning_rate": 9.991393001529255e-06,
+      "loss": 0.4998,
+      "step": 388
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4803448874576803,
+      "learning_rate": 9.991274728080368e-06,
+      "loss": 0.5687,
+      "step": 389
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.817078050217724,
+      "learning_rate": 9.991155648247086e-06,
+      "loss": 0.5589,
+      "step": 390
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.56274673389566,
+      "learning_rate": 9.991035762048655e-06,
+      "loss": 0.5402,
+      "step": 391
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6700379174795714,
+      "learning_rate": 9.990915069504438e-06,
+      "loss": 0.4653,
+      "step": 392
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6793263712894418,
+      "learning_rate": 9.990793570633935e-06,
+      "loss": 0.4721,
+      "step": 393
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6887319698367116,
+      "learning_rate": 9.990671265456778e-06,
+      "loss": 0.5377,
+      "step": 394
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4642447699000192,
+      "learning_rate": 9.990548153992721e-06,
+      "loss": 0.552,
+      "step": 395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8643006750012217,
+      "learning_rate": 9.990424236261657e-06,
+      "loss": 0.5422,
+      "step": 396
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7130101747622972,
+      "learning_rate": 9.990299512283608e-06,
+      "loss": 0.4625,
+      "step": 397
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.509124542655514,
+      "learning_rate": 9.990173982078721e-06,
+      "loss": 0.5449,
+      "step": 398
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7287882922623734,
+      "learning_rate": 9.990047645667277e-06,
+      "loss": 0.525,
+      "step": 399
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.597759493266317,
+      "learning_rate": 9.98992050306969e-06,
+      "loss": 0.5109,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7164045972462163,
+      "learning_rate": 9.989792554306496e-06,
+      "loss": 0.4961,
+      "step": 401
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7717464858573455,
+      "learning_rate": 9.98966379939837e-06,
+      "loss": 0.4837,
+      "step": 402
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9481512152040525,
+      "learning_rate": 9.989534238366113e-06,
+      "loss": 0.5405,
+      "step": 403
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6987543219058931,
+      "learning_rate": 9.989403871230654e-06,
+      "loss": 0.5504,
+      "step": 404
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4308349414603916,
+      "learning_rate": 9.989272698013058e-06,
+      "loss": 0.5327,
+      "step": 405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6876272721147005,
+      "learning_rate": 9.989140718734515e-06,
+      "loss": 0.5607,
+      "step": 406
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.729149900507196,
+      "learning_rate": 9.989007933416348e-06,
+      "loss": 0.4808,
+      "step": 407
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7700639883038296,
+      "learning_rate": 9.988874342080011e-06,
+      "loss": 0.5669,
+      "step": 408
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5781092289283492,
+      "learning_rate": 9.988739944747086e-06,
+      "loss": 0.505,
+      "step": 409
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.557526972913807,
+      "learning_rate": 9.988604741439288e-06,
+      "loss": 0.5129,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8408561874260345,
+      "learning_rate": 9.988468732178456e-06,
+      "loss": 0.5595,
+      "step": 411
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0413512443099053,
+      "learning_rate": 9.988331916986565e-06,
+      "loss": 0.5793,
+      "step": 412
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6768995403140314,
+      "learning_rate": 9.988194295885721e-06,
+      "loss": 0.5092,
+      "step": 413
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4892709861156581,
+      "learning_rate": 9.988055868898156e-06,
+      "loss": 0.5342,
+      "step": 414
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6539830627005252,
+      "learning_rate": 9.987916636046234e-06,
+      "loss": 0.5679,
+      "step": 415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6412699630465084,
+      "learning_rate": 9.987776597352451e-06,
+      "loss": 0.531,
+      "step": 416
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5955019956242151,
+      "learning_rate": 9.987635752839429e-06,
+      "loss": 0.4732,
+      "step": 417
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4359477688178175,
+      "learning_rate": 9.987494102529924e-06,
+      "loss": 0.4767,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6153040652848387,
+      "learning_rate": 9.987351646446824e-06,
+      "loss": 0.4845,
+      "step": 419
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.592003559782074,
+      "learning_rate": 9.987208384613136e-06,
+      "loss": 0.5956,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7946167267388238,
+      "learning_rate": 9.987064317052013e-06,
+      "loss": 0.6168,
+      "step": 421
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7008916654016195,
+      "learning_rate": 9.986919443786727e-06,
+      "loss": 0.5003,
+      "step": 422
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.2870924806038975,
+      "learning_rate": 9.986773764840684e-06,
+      "loss": 0.5361,
+      "step": 423
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.484401416940637,
+      "learning_rate": 9.98662728023742e-06,
+      "loss": 0.5721,
+      "step": 424
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.547533213628365,
+      "learning_rate": 9.986479990000598e-06,
+      "loss": 0.5358,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.673741921515085,
+      "learning_rate": 9.98633189415402e-06,
+      "loss": 0.504,
+      "step": 426
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4089500017727525,
+      "learning_rate": 9.986182992721606e-06,
+      "loss": 0.5495,
+      "step": 427
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2003603241293996,
+      "learning_rate": 9.986033285727418e-06,
+      "loss": 0.5278,
+      "step": 428
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.784641006031379,
+      "learning_rate": 9.985882773195638e-06,
+      "loss": 0.5357,
+      "step": 429
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7317785476187197,
+      "learning_rate": 9.985731455150584e-06,
+      "loss": 0.5567,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.648485799879626,
+      "learning_rate": 9.985579331616705e-06,
+      "loss": 0.4999,
+      "step": 431
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9685222126772437,
+      "learning_rate": 9.985426402618574e-06,
+      "loss": 0.5034,
+      "step": 432
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4689771404993912,
+      "learning_rate": 9.985272668180901e-06,
+      "loss": 0.537,
+      "step": 433
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.688785913533373,
+      "learning_rate": 9.985118128328522e-06,
+      "loss": 0.5454,
+      "step": 434
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.48562468774087,
+      "learning_rate": 9.984962783086403e-06,
+      "loss": 0.4896,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.459814886971303,
+      "learning_rate": 9.984806632479643e-06,
+      "loss": 0.4912,
+      "step": 436
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5872647895748622,
+      "learning_rate": 9.98464967653347e-06,
+      "loss": 0.5714,
+      "step": 437
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9004557876468335,
+      "learning_rate": 9.98449191527324e-06,
+      "loss": 0.494,
+      "step": 438
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.985375325252717,
+      "learning_rate": 9.984333348724442e-06,
+      "loss": 0.5637,
+      "step": 439
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8027928544018976,
+      "learning_rate": 9.984173976912695e-06,
+      "loss": 0.5001,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5841099007415989,
+      "learning_rate": 9.984013799863744e-06,
+      "loss": 0.5864,
+      "step": 441
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5646937500034126,
+      "learning_rate": 9.983852817603468e-06,
+      "loss": 0.574,
+      "step": 442
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.97296027361365,
+      "learning_rate": 9.983691030157876e-06,
+      "loss": 0.4592,
+      "step": 443
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7128940493354311,
+      "learning_rate": 9.983528437553106e-06,
+      "loss": 0.5014,
+      "step": 444
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3553160387069387,
+      "learning_rate": 9.983365039815425e-06,
+      "loss": 0.5138,
+      "step": 445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7551202165825732,
+      "learning_rate": 9.983200836971234e-06,
+      "loss": 0.4924,
+      "step": 446
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.783065671146444,
+      "learning_rate": 9.983035829047058e-06,
+      "loss": 0.4881,
+      "step": 447
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6346481923047431,
+      "learning_rate": 9.982870016069557e-06,
+      "loss": 0.5086,
+      "step": 448
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.269760431343647,
+      "learning_rate": 9.98270339806552e-06,
+      "loss": 0.4659,
+      "step": 449
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9208644340466066,
+      "learning_rate": 9.982535975061866e-06,
+      "loss": 0.4691,
+      "step": 450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.700303977995299,
+      "learning_rate": 9.982367747085642e-06,
+      "loss": 0.4813,
+      "step": 451
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4580808287839546,
+      "learning_rate": 9.982198714164029e-06,
+      "loss": 0.5482,
+      "step": 452
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6395785682570598,
+      "learning_rate": 9.982028876324334e-06,
+      "loss": 0.5674,
+      "step": 453
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.089402865370069,
+      "learning_rate": 9.981858233593996e-06,
+      "loss": 0.5959,
+      "step": 454
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7647424329197148,
+      "learning_rate": 9.981686786000584e-06,
+      "loss": 0.5567,
+      "step": 455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6426304996228964,
+      "learning_rate": 9.981514533571797e-06,
+      "loss": 0.5418,
+      "step": 456
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.124650738441062,
+      "learning_rate": 9.981341476335464e-06,
+      "loss": 0.602,
+      "step": 457
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.379438226743473,
+      "learning_rate": 9.981167614319542e-06,
+      "loss": 0.4779,
+      "step": 458
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0204003882969204,
+      "learning_rate": 9.980992947552124e-06,
+      "loss": 0.5301,
+      "step": 459
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6188537242497516,
+      "learning_rate": 9.980817476061426e-06,
+      "loss": 0.4723,
+      "step": 460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.636311850641546,
+      "learning_rate": 9.980641199875797e-06,
+      "loss": 0.5494,
+      "step": 461
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5196244609813645,
+      "learning_rate": 9.980464119023718e-06,
+      "loss": 0.5092,
+      "step": 462
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.59037993007294,
+      "learning_rate": 9.980286233533795e-06,
+      "loss": 0.5209,
+      "step": 463
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4369151490083365,
+      "learning_rate": 9.980107543434769e-06,
+      "loss": 0.5809,
+      "step": 464
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0160569481227903,
+      "learning_rate": 9.97992804875551e-06,
+      "loss": 0.5392,
+      "step": 465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7213586451768685,
+      "learning_rate": 9.979747749525014e-06,
+      "loss": 0.5294,
+      "step": 466
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5676191987425458,
+      "learning_rate": 9.979566645772412e-06,
+      "loss": 0.4995,
+      "step": 467
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2941976901857233,
+      "learning_rate": 9.97938473752696e-06,
+      "loss": 0.5625,
+      "step": 468
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3231610487610865,
+      "learning_rate": 9.979202024818052e-06,
+      "loss": 0.561,
+      "step": 469
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0242057322321005,
+      "learning_rate": 9.979018507675202e-06,
+      "loss": 0.5524,
+      "step": 470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8226219382548836,
+      "learning_rate": 9.978834186128063e-06,
+      "loss": 0.512,
+      "step": 471
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4955935257507191,
+      "learning_rate": 9.97864906020641e-06,
+      "loss": 0.5431,
+      "step": 472
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.680011950409537,
+      "learning_rate": 9.978463129940153e-06,
+      "loss": 0.5267,
+      "step": 473
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.565469276971012,
+      "learning_rate": 9.978276395359332e-06,
+      "loss": 0.5174,
+      "step": 474
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6714958132903396,
+      "learning_rate": 9.978088856494115e-06,
+      "loss": 0.5051,
+      "step": 475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8389831857433339,
+      "learning_rate": 9.977900513374799e-06,
+      "loss": 0.5401,
+      "step": 476
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5499532118888697,
+      "learning_rate": 9.977711366031813e-06,
+      "loss": 0.544,
+      "step": 477
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.727431677360426,
+      "learning_rate": 9.977521414495716e-06,
+      "loss": 0.518,
+      "step": 478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.157775699970845,
+      "learning_rate": 9.977330658797199e-06,
+      "loss": 0.5556,
+      "step": 479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5612974363773897,
+      "learning_rate": 9.977139098967075e-06,
+      "loss": 0.5161,
+      "step": 480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7175998885815322,
+      "learning_rate": 9.976946735036296e-06,
+      "loss": 0.5357,
+      "step": 481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.406772227901931,
+      "learning_rate": 9.97675356703594e-06,
+      "loss": 0.5333,
+      "step": 482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1451551555311283,
+      "learning_rate": 9.976559594997211e-06,
+      "loss": 0.5144,
+      "step": 483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8879155418167197,
+      "learning_rate": 9.976364818951452e-06,
+      "loss": 0.5342,
+      "step": 484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4034381097267934,
+      "learning_rate": 9.976169238930128e-06,
+      "loss": 0.5143,
+      "step": 485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6253361682506393,
+      "learning_rate": 9.975972854964837e-06,
+      "loss": 0.5532,
+      "step": 486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0486162168827273,
+      "learning_rate": 9.975775667087308e-06,
+      "loss": 0.5694,
+      "step": 487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5977042020706673,
+      "learning_rate": 9.975577675329396e-06,
+      "loss": 0.6012,
+      "step": 488
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6581212962368017,
+      "learning_rate": 9.97537887972309e-06,
+      "loss": 0.5014,
+      "step": 489
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.450762817262192,
+      "learning_rate": 9.975179280300507e-06,
+      "loss": 0.5355,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.496453897715201,
+      "learning_rate": 9.974978877093892e-06,
+      "loss": 0.5749,
+      "step": 491
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5955663116958974,
+      "learning_rate": 9.974777670135627e-06,
+      "loss": 0.5485,
+      "step": 492
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6511586849526274,
+      "learning_rate": 9.974575659458214e-06,
+      "loss": 0.528,
+      "step": 493
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8173706390670754,
+      "learning_rate": 9.97437284509429e-06,
+      "loss": 0.5043,
+      "step": 494
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.450248038097281,
+      "learning_rate": 9.974169227076623e-06,
+      "loss": 0.56,
+      "step": 495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3759328275504474,
+      "learning_rate": 9.97396480543811e-06,
+      "loss": 0.519,
+      "step": 496
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6118263272522013,
+      "learning_rate": 9.973759580211776e-06,
+      "loss": 0.4886,
+      "step": 497
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4398888457780274,
+      "learning_rate": 9.973553551430775e-06,
+      "loss": 0.5334,
+      "step": 498
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.469914373917854,
+      "learning_rate": 9.973346719128395e-06,
+      "loss": 0.4947,
+      "step": 499
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.570025029324263,
+      "learning_rate": 9.973139083338052e-06,
+      "loss": 0.5887,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7122866183202354,
+      "learning_rate": 9.97293064409329e-06,
+      "loss": 0.5287,
+      "step": 501
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4599374691955285,
+      "learning_rate": 9.972721401427785e-06,
+      "loss": 0.5127,
+      "step": 502
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4856231479658881,
+      "learning_rate": 9.972511355375341e-06,
+      "loss": 0.5805,
+      "step": 503
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0164015385758502,
+      "learning_rate": 9.972300505969896e-06,
+      "loss": 0.5619,
+      "step": 504
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.2997701887534436,
+      "learning_rate": 9.97208885324551e-06,
+      "loss": 0.4729,
+      "step": 505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.756468060556157,
+      "learning_rate": 9.97187639723638e-06,
+      "loss": 0.5284,
+      "step": 506
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7198355693077105,
+      "learning_rate": 9.971663137976829e-06,
+      "loss": 0.5327,
+      "step": 507
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7937945953770489,
+      "learning_rate": 9.971449075501313e-06,
+      "loss": 0.5882,
+      "step": 508
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4428157798894228,
+      "learning_rate": 9.971234209844413e-06,
+      "loss": 0.516,
+      "step": 509
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6343343346489136,
+      "learning_rate": 9.971018541040844e-06,
+      "loss": 0.4583,
+      "step": 510
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6346268222138216,
+      "learning_rate": 9.970802069125449e-06,
+      "loss": 0.5788,
+      "step": 511
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.236311818205272,
+      "learning_rate": 9.9705847941332e-06,
+      "loss": 0.5473,
+      "step": 512
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.186404431476861,
+      "learning_rate": 9.970366716099203e-06,
+      "loss": 0.5021,
+      "step": 513
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.448273391917764,
+      "learning_rate": 9.970147835058686e-06,
+      "loss": 0.4891,
+      "step": 514
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6707135102073185,
+      "learning_rate": 9.969928151047012e-06,
+      "loss": 0.5128,
+      "step": 515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9157604861759812,
+      "learning_rate": 9.969707664099677e-06,
+      "loss": 0.5314,
+      "step": 516
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4641731434697254,
+      "learning_rate": 9.969486374252297e-06,
+      "loss": 0.5648,
+      "step": 517
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6951811673512422,
+      "learning_rate": 9.969264281540627e-06,
+      "loss": 0.5429,
+      "step": 518
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5351361206003573,
+      "learning_rate": 9.969041386000547e-06,
+      "loss": 0.5203,
+      "step": 519
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8922924188464132,
+      "learning_rate": 9.968817687668067e-06,
+      "loss": 0.5381,
+      "step": 520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5799451731489043,
+      "learning_rate": 9.96859318657933e-06,
+      "loss": 0.5668,
+      "step": 521
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0926399677961864,
+      "learning_rate": 9.968367882770601e-06,
+      "loss": 0.5635,
+      "step": 522
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9474775210488486,
+      "learning_rate": 9.968141776278287e-06,
+      "loss": 0.571,
+      "step": 523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7431408889119921,
+      "learning_rate": 9.96791486713891e-06,
+      "loss": 0.5826,
+      "step": 524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8072112193176124,
+      "learning_rate": 9.967687155389135e-06,
+      "loss": 0.6058,
+      "step": 525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5487646831257245,
+      "learning_rate": 9.967458641065748e-06,
+      "loss": 0.5161,
+      "step": 526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9345132321574223,
+      "learning_rate": 9.967229324205666e-06,
+      "loss": 0.5713,
+      "step": 527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9435970449322106,
+      "learning_rate": 9.966999204845942e-06,
+      "loss": 0.5563,
+      "step": 528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.43796673259639,
+      "learning_rate": 9.966768283023752e-06,
+      "loss": 0.5661,
+      "step": 529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6557336189733225,
+      "learning_rate": 9.966536558776399e-06,
+      "loss": 0.5597,
+      "step": 530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9099136381706796,
+      "learning_rate": 9.966304032141325e-06,
+      "loss": 0.4683,
+      "step": 531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5887496350127177,
+      "learning_rate": 9.966070703156096e-06,
+      "loss": 0.476,
+      "step": 532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2832288568597026,
+      "learning_rate": 9.965836571858408e-06,
+      "loss": 0.4918,
+      "step": 533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8915667838943808,
+      "learning_rate": 9.965601638286085e-06,
+      "loss": 0.4534,
+      "step": 534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5838608224606279,
+      "learning_rate": 9.965365902477085e-06,
+      "loss": 0.5287,
+      "step": 535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5069296219030366,
+      "learning_rate": 9.965129364469492e-06,
+      "loss": 0.522,
+      "step": 536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.575781289038582,
+      "learning_rate": 9.96489202430152e-06,
+      "loss": 0.5527,
+      "step": 537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.495415835304909,
+      "learning_rate": 9.964653882011516e-06,
+      "loss": 0.5092,
+      "step": 538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6262876945226945,
+      "learning_rate": 9.964414937637953e-06,
+      "loss": 0.4981,
+      "step": 539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6387936310785975,
+      "learning_rate": 9.964175191219436e-06,
+      "loss": 0.4891,
+      "step": 540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6217650667702996,
+      "learning_rate": 9.963934642794694e-06,
+      "loss": 0.5808,
+      "step": 541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4841632250821106,
+      "learning_rate": 9.963693292402592e-06,
+      "loss": 0.552,
+      "step": 542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2404882976865395,
+      "learning_rate": 9.963451140082124e-06,
+      "loss": 0.5235,
+      "step": 543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8579527183857778,
+      "learning_rate": 9.963208185872407e-06,
+      "loss": 0.5543,
+      "step": 544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9117719830923814,
+      "learning_rate": 9.962964429812697e-06,
+      "loss": 0.5189,
+      "step": 545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5305379419100715,
+      "learning_rate": 9.962719871942374e-06,
+      "loss": 0.5631,
+      "step": 546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.882149654586994,
+      "learning_rate": 9.962474512300948e-06,
+      "loss": 0.5624,
+      "step": 547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6646759237474957,
+      "learning_rate": 9.962228350928058e-06,
+      "loss": 0.5551,
+      "step": 548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5649208125553558,
+      "learning_rate": 9.961981387863476e-06,
+      "loss": 0.5388,
+      "step": 549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4445542695684628,
+      "learning_rate": 9.961733623147098e-06,
+      "loss": 0.5359,
+      "step": 550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.621089252780402,
+      "learning_rate": 9.961485056818957e-06,
+      "loss": 0.5501,
+      "step": 551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3600616481489842,
+      "learning_rate": 9.961235688919204e-06,
+      "loss": 0.5164,
+      "step": 552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5291096429900197,
+      "learning_rate": 9.960985519488133e-06,
+      "loss": 0.5464,
+      "step": 553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8444597766221988,
+      "learning_rate": 9.96073454856616e-06,
+      "loss": 0.5621,
+      "step": 554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6732011691950741,
+      "learning_rate": 9.96048277619383e-06,
+      "loss": 0.5049,
+      "step": 555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.438419778694226,
+      "learning_rate": 9.960230202411818e-06,
+      "loss": 0.5538,
+      "step": 556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6603929165553797,
+      "learning_rate": 9.95997682726093e-06,
+      "loss": 0.5469,
+      "step": 557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7668487791259488,
+      "learning_rate": 9.959722650782104e-06,
+      "loss": 0.555,
+      "step": 558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6218233851859314,
+      "learning_rate": 9.959467673016403e-06,
+      "loss": 0.5416,
+      "step": 559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0098143485676965,
+      "learning_rate": 9.959211894005018e-06,
+      "loss": 0.588,
+      "step": 560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5346689636597521,
+      "learning_rate": 9.958955313789277e-06,
+      "loss": 0.6114,
+      "step": 561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3691963998562475,
+      "learning_rate": 9.958697932410631e-06,
+      "loss": 0.5647,
+      "step": 562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5902891681020572,
+      "learning_rate": 9.958439749910658e-06,
+      "loss": 0.5241,
+      "step": 563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4317090557464656,
+      "learning_rate": 9.958180766331076e-06,
+      "loss": 0.5344,
+      "step": 564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6223451298186553,
+      "learning_rate": 9.957920981713723e-06,
+      "loss": 0.5354,
+      "step": 565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4911420362299912,
+      "learning_rate": 9.95766039610057e-06,
+      "loss": 0.5261,
+      "step": 566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.459595289704798,
+      "learning_rate": 9.957399009533716e-06,
+      "loss": 0.5441,
+      "step": 567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.50536016248566,
+      "learning_rate": 9.957136822055392e-06,
+      "loss": 0.5742,
+      "step": 568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.099600133776981,
+      "learning_rate": 9.956873833707958e-06,
+      "loss": 0.5347,
+      "step": 569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.823400575954244,
+      "learning_rate": 9.956610044533897e-06,
+      "loss": 0.5679,
+      "step": 570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.211244433828808,
+      "learning_rate": 9.956345454575831e-06,
+      "loss": 0.5144,
+      "step": 571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3813580483388566,
+      "learning_rate": 9.956080063876503e-06,
+      "loss": 0.5213,
+      "step": 572
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.547609620866616,
+      "learning_rate": 9.955813872478794e-06,
+      "loss": 0.5107,
+      "step": 573
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9698736531779897,
+      "learning_rate": 9.955546880425706e-06,
+      "loss": 0.5011,
+      "step": 574
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7132662201669262,
+      "learning_rate": 9.955279087760374e-06,
+      "loss": 0.4852,
+      "step": 575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1289820789250253,
+      "learning_rate": 9.955010494526062e-06,
+      "loss": 0.6084,
+      "step": 576
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6894981066518925,
+      "learning_rate": 9.954741100766167e-06,
+      "loss": 0.5068,
+      "step": 577
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.993185449507924,
+      "learning_rate": 9.95447090652421e-06,
+      "loss": 0.5744,
+      "step": 578
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9175405662571439,
+      "learning_rate": 9.954199911843842e-06,
+      "loss": 0.5143,
+      "step": 579
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.707912930910467,
+      "learning_rate": 9.953928116768848e-06,
+      "loss": 0.5874,
+      "step": 580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1196759791921496,
+      "learning_rate": 9.953655521343132e-06,
+      "loss": 0.5544,
+      "step": 581
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6578798781305177,
+      "learning_rate": 9.953382125610742e-06,
+      "loss": 0.5006,
+      "step": 582
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7117964965327253,
+      "learning_rate": 9.953107929615843e-06,
+      "loss": 0.5286,
+      "step": 583
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.718532476961154,
+      "learning_rate": 9.952832933402733e-06,
+      "loss": 0.5907,
+      "step": 584
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.874584366704575,
+      "learning_rate": 9.952557137015846e-06,
+      "loss": 0.5456,
+      "step": 585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.091952706607619,
+      "learning_rate": 9.952280540499733e-06,
+      "loss": 0.5807,
+      "step": 586
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8340961057465826,
+      "learning_rate": 9.952003143899082e-06,
+      "loss": 0.5044,
+      "step": 587
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4142321636334534,
+      "learning_rate": 9.951724947258713e-06,
+      "loss": 0.4952,
+      "step": 588
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5498823391862544,
+      "learning_rate": 9.951445950623565e-06,
+      "loss": 0.5229,
+      "step": 589
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.976631116773797,
+      "learning_rate": 9.951166154038716e-06,
+      "loss": 0.5212,
+      "step": 590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4043970725348118,
+      "learning_rate": 9.950885557549369e-06,
+      "loss": 0.5213,
+      "step": 591
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6411492750911545,
+      "learning_rate": 9.950604161200855e-06,
+      "loss": 0.508,
+      "step": 592
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6155280198944444,
+      "learning_rate": 9.950321965038638e-06,
+      "loss": 0.5176,
+      "step": 593
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6329573178476742,
+      "learning_rate": 9.950038969108312e-06,
+      "loss": 0.5381,
+      "step": 594
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5322368189387603,
+      "learning_rate": 9.949755173455591e-06,
+      "loss": 0.4737,
+      "step": 595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8966104227734712,
+      "learning_rate": 9.94947057812633e-06,
+      "loss": 0.5447,
+      "step": 596
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3255920186297554,
+      "learning_rate": 9.949185183166506e-06,
+      "loss": 0.4983,
+      "step": 597
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.514982661457202,
+      "learning_rate": 9.948898988622226e-06,
+      "loss": 0.4703,
+      "step": 598
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5578472451579808,
+      "learning_rate": 9.94861199453973e-06,
+      "loss": 0.5355,
+      "step": 599
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8462377204745257,
+      "learning_rate": 9.948324200965382e-06,
+      "loss": 0.5217,
+      "step": 600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0501987500355052,
+      "learning_rate": 9.948035607945678e-06,
+      "loss": 0.5216,
+      "step": 601
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7196396011246718,
+      "learning_rate": 9.947746215527246e-06,
+      "loss": 0.4568,
+      "step": 602
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2029295205716912,
+      "learning_rate": 9.947456023756834e-06,
+      "loss": 0.5113,
+      "step": 603
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4106620683922255,
+      "learning_rate": 9.94716503268133e-06,
+      "loss": 0.5382,
+      "step": 604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4704451672909378,
+      "learning_rate": 9.946873242347741e-06,
+      "loss": 0.552,
+      "step": 605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5073803415658742,
+      "learning_rate": 9.946580652803215e-06,
+      "loss": 0.4871,
+      "step": 606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9073111646355374,
+      "learning_rate": 9.946287264095016e-06,
+      "loss": 0.5451,
+      "step": 607
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4481931142536788,
+      "learning_rate": 9.945993076270548e-06,
+      "loss": 0.4984,
+      "step": 608
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5698280416999764,
+      "learning_rate": 9.945698089377338e-06,
+      "loss": 0.5163,
+      "step": 609
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5623181354974682,
+      "learning_rate": 9.945402303463045e-06,
+      "loss": 0.4943,
+      "step": 610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.545427331377954,
+      "learning_rate": 9.945105718575452e-06,
+      "loss": 0.5753,
+      "step": 611
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.519017117437064,
+      "learning_rate": 9.94480833476248e-06,
+      "loss": 0.4687,
+      "step": 612
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.320764969254525,
+      "learning_rate": 9.94451015207217e-06,
+      "loss": 0.5192,
+      "step": 613
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.585198566620101,
+      "learning_rate": 9.944211170552698e-06,
+      "loss": 0.5299,
+      "step": 614
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.065016309964868,
+      "learning_rate": 9.943911390252368e-06,
+      "loss": 0.5043,
+      "step": 615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.717156629974031,
+      "learning_rate": 9.943610811219608e-06,
+      "loss": 0.5314,
+      "step": 616
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3992808758110693,
+      "learning_rate": 9.943309433502985e-06,
+      "loss": 0.4816,
+      "step": 617
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4741157871118,
+      "learning_rate": 9.943007257151185e-06,
+      "loss": 0.5328,
+      "step": 618
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6625878117192396,
+      "learning_rate": 9.942704282213028e-06,
+      "loss": 0.5106,
+      "step": 619
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5814144236989829,
+      "learning_rate": 9.942400508737465e-06,
+      "loss": 0.5308,
+      "step": 620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5003023210361535,
+      "learning_rate": 9.942095936773568e-06,
+      "loss": 0.5584,
+      "step": 621
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5907340486009833,
+      "learning_rate": 9.941790566370547e-06,
+      "loss": 0.5498,
+      "step": 622
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2882804193460367,
+      "learning_rate": 9.941484397577737e-06,
+      "loss": 0.5108,
+      "step": 623
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.679829152208987,
+      "learning_rate": 9.941177430444601e-06,
+      "loss": 0.5687,
+      "step": 624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4950337923655184,
+      "learning_rate": 9.940869665020736e-06,
+      "loss": 0.583,
+      "step": 625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.692963812544074,
+      "learning_rate": 9.940561101355858e-06,
+      "loss": 0.507,
+      "step": 626
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5440986461401733,
+      "learning_rate": 9.940251739499824e-06,
+      "loss": 0.5577,
+      "step": 627
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6306303651228953,
+      "learning_rate": 9.93994157950261e-06,
+      "loss": 0.4947,
+      "step": 628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6663936494364988,
+      "learning_rate": 9.939630621414328e-06,
+      "loss": 0.467,
+      "step": 629
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8972238621140487,
+      "learning_rate": 9.939318865285214e-06,
+      "loss": 0.5485,
+      "step": 630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4247705518456582,
+      "learning_rate": 9.939006311165636e-06,
+      "loss": 0.5473,
+      "step": 631
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.718273329018105,
+      "learning_rate": 9.938692959106089e-06,
+      "loss": 0.5431,
+      "step": 632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6943576952650891,
+      "learning_rate": 9.9383788091572e-06,
+      "loss": 0.5074,
+      "step": 633
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6876591007539876,
+      "learning_rate": 9.938063861369718e-06,
+      "loss": 0.5159,
+      "step": 634
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7548691737757782,
+      "learning_rate": 9.93774811579453e-06,
+      "loss": 0.5381,
+      "step": 635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5154806990192187,
+      "learning_rate": 9.937431572482646e-06,
+      "loss": 0.5053,
+      "step": 636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5611688276113909,
+      "learning_rate": 9.937114231485207e-06,
+      "loss": 0.506,
+      "step": 637
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9093304638448643,
+      "learning_rate": 9.936796092853482e-06,
+      "loss": 0.482,
+      "step": 638
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.097493189282801,
+      "learning_rate": 9.936477156638868e-06,
+      "loss": 0.5737,
+      "step": 639
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.430398311443158,
+      "learning_rate": 9.936157422892892e-06,
+      "loss": 0.5013,
+      "step": 640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8149506986026083,
+      "learning_rate": 9.935836891667212e-06,
+      "loss": 0.5898,
+      "step": 641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6180185699234901,
+      "learning_rate": 9.93551556301361e-06,
+      "loss": 0.5698,
+      "step": 642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7222498111719983,
+      "learning_rate": 9.935193436984001e-06,
+      "loss": 0.5386,
+      "step": 643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7724001895597116,
+      "learning_rate": 9.934870513630428e-06,
+      "loss": 0.5139,
+      "step": 644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5072040003212988,
+      "learning_rate": 9.93454679300506e-06,
+      "loss": 0.5524,
+      "step": 645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5356579540974762,
+      "learning_rate": 9.934222275160199e-06,
+      "loss": 0.5048,
+      "step": 646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.108613917254826,
+      "learning_rate": 9.933896960148272e-06,
+      "loss": 0.5056,
+      "step": 647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6624497135490821,
+      "learning_rate": 9.933570848021837e-06,
+      "loss": 0.4759,
+      "step": 648
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8727451645721778,
+      "learning_rate": 9.933243938833581e-06,
+      "loss": 0.5638,
+      "step": 649
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4408274780775827,
+      "learning_rate": 9.932916232636318e-06,
+      "loss": 0.5385,
+      "step": 650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.488341298958083,
+      "learning_rate": 9.932587729482992e-06,
+      "loss": 0.5178,
+      "step": 651
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5293715713962739,
+      "learning_rate": 9.932258429426678e-06,
+      "loss": 0.4827,
+      "step": 652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.739811601628948,
+      "learning_rate": 9.931928332520573e-06,
+      "loss": 0.5144,
+      "step": 653
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7864609906352191,
+      "learning_rate": 9.93159743881801e-06,
+      "loss": 0.5343,
+      "step": 654
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5659000901067446,
+      "learning_rate": 9.931265748372447e-06,
+      "loss": 0.4829,
+      "step": 655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3856751502290265,
+      "learning_rate": 9.930933261237472e-06,
+      "loss": 0.5254,
+      "step": 656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6639494622011914,
+      "learning_rate": 9.930599977466802e-06,
+      "loss": 0.5489,
+      "step": 657
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5503942485192046,
+      "learning_rate": 9.93026589711428e-06,
+      "loss": 0.4999,
+      "step": 658
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6374283584517775,
+      "learning_rate": 9.929931020233878e-06,
+      "loss": 0.5678,
+      "step": 659
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4451768173598474,
+      "learning_rate": 9.929595346879704e-06,
+      "loss": 0.5326,
+      "step": 660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.771283449714104,
+      "learning_rate": 9.929258877105985e-06,
+      "loss": 0.5726,
+      "step": 661
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5995943876705156,
+      "learning_rate": 9.928921610967079e-06,
+      "loss": 0.5317,
+      "step": 662
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0357535708739407,
+      "learning_rate": 9.92858354851748e-06,
+      "loss": 0.5355,
+      "step": 663
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6543400868553626,
+      "learning_rate": 9.928244689811799e-06,
+      "loss": 0.5311,
+      "step": 664
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5531763061855233,
+      "learning_rate": 9.927905034904785e-06,
+      "loss": 0.4885,
+      "step": 665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6727414973141812,
+      "learning_rate": 9.927564583851313e-06,
+      "loss": 0.5728,
+      "step": 666
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5057542539112305,
+      "learning_rate": 9.927223336706384e-06,
+      "loss": 0.5613,
+      "step": 667
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3621577823437236,
+      "learning_rate": 9.926881293525129e-06,
+      "loss": 0.5133,
+      "step": 668
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3917135906447404,
+      "learning_rate": 9.92653845436281e-06,
+      "loss": 0.582,
+      "step": 669
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1837662859106874,
+      "learning_rate": 9.926194819274812e-06,
+      "loss": 0.5057,
+      "step": 670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5927160077016809,
+      "learning_rate": 9.925850388316657e-06,
+      "loss": 0.44,
+      "step": 671
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3077815833136432,
+      "learning_rate": 9.925505161543988e-06,
+      "loss": 0.5325,
+      "step": 672
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3547793282803282,
+      "learning_rate": 9.925159139012581e-06,
+      "loss": 0.4944,
+      "step": 673
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8707164450695393,
+      "learning_rate": 9.924812320778338e-06,
+      "loss": 0.497,
+      "step": 674
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5811520204557328,
+      "learning_rate": 9.92446470689729e-06,
+      "loss": 0.5824,
+      "step": 675
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6237920019107122,
+      "learning_rate": 9.924116297425598e-06,
+      "loss": 0.546,
+      "step": 676
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.300885812955191,
+      "learning_rate": 9.923767092419551e-06,
+      "loss": 0.543,
+      "step": 677
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4815799886567405,
+      "learning_rate": 9.923417091935564e-06,
+      "loss": 0.5462,
+      "step": 678
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7256114801909598,
+      "learning_rate": 9.923066296030185e-06,
+      "loss": 0.5211,
+      "step": 679
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5039129063851944,
+      "learning_rate": 9.922714704760088e-06,
+      "loss": 0.5257,
+      "step": 680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4942655542398324,
+      "learning_rate": 9.922362318182075e-06,
+      "loss": 0.534,
+      "step": 681
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6854065304986234,
+      "learning_rate": 9.922009136353077e-06,
+      "loss": 0.5436,
+      "step": 682
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6207158877626267,
+      "learning_rate": 9.921655159330154e-06,
+      "loss": 0.5232,
+      "step": 683
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9176833918479952,
+      "learning_rate": 9.921300387170494e-06,
+      "loss": 0.5185,
+      "step": 684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.222262958976972,
+      "learning_rate": 9.920944819931412e-06,
+      "loss": 0.5406,
+      "step": 685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5443955907847353,
+      "learning_rate": 9.920588457670357e-06,
+      "loss": 0.5517,
+      "step": 686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4342050400083974,
+      "learning_rate": 9.920231300444897e-06,
+      "loss": 0.5497,
+      "step": 687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.311067650265911,
+      "learning_rate": 9.91987334831274e-06,
+      "loss": 0.5226,
+      "step": 688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.471479637322531,
+      "learning_rate": 9.919514601331712e-06,
+      "loss": 0.5122,
+      "step": 689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7926523501870986,
+      "learning_rate": 9.919155059559772e-06,
+      "loss": 0.55,
+      "step": 690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5675007462680313,
+      "learning_rate": 9.918794723055009e-06,
+      "loss": 0.5821,
+      "step": 691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4604240106900952,
+      "learning_rate": 9.918433591875637e-06,
+      "loss": 0.564,
+      "step": 692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4861143775734027,
+      "learning_rate": 9.918071666080003e-06,
+      "loss": 0.5813,
+      "step": 693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.143330932505597,
+      "learning_rate": 9.917708945726574e-06,
+      "loss": 0.5196,
+      "step": 694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.44234958841204,
+      "learning_rate": 9.917345430873955e-06,
+      "loss": 0.5551,
+      "step": 695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.48817999565854,
+      "learning_rate": 9.916981121580873e-06,
+      "loss": 0.521,
+      "step": 696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.774381530446406,
+      "learning_rate": 9.916616017906185e-06,
+      "loss": 0.5086,
+      "step": 697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6229915734516575,
+      "learning_rate": 9.91625011990888e-06,
+      "loss": 0.4771,
+      "step": 698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.448594905519429,
+      "learning_rate": 9.915883427648066e-06,
+      "loss": 0.5662,
+      "step": 699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9591531503663824,
+      "learning_rate": 9.915515941182992e-06,
+      "loss": 0.4842,
+      "step": 700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.36950835978183,
+      "learning_rate": 9.915147660573024e-06,
+      "loss": 0.4944,
+      "step": 701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8648299110331485,
+      "learning_rate": 9.914778585877665e-06,
+      "loss": 0.5539,
+      "step": 702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7320586877113233,
+      "learning_rate": 9.914408717156538e-06,
+      "loss": 0.5153,
+      "step": 703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4658542485684234,
+      "learning_rate": 9.9140380544694e-06,
+      "loss": 0.572,
+      "step": 704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4362278910057131,
+      "learning_rate": 9.913666597876137e-06,
+      "loss": 0.5114,
+      "step": 705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0480313933332335,
+      "learning_rate": 9.913294347436758e-06,
+      "loss": 0.4573,
+      "step": 706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3863115172008518,
+      "learning_rate": 9.912921303211405e-06,
+      "loss": 0.5416,
+      "step": 707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.756096947460397,
+      "learning_rate": 9.912547465260347e-06,
+      "loss": 0.4997,
+      "step": 708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.681996815479112,
+      "learning_rate": 9.91217283364398e-06,
+      "loss": 0.5189,
+      "step": 709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8491451039304299,
+      "learning_rate": 9.91179740842283e-06,
+      "loss": 0.5712,
+      "step": 710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6846675664657432,
+      "learning_rate": 9.911421189657548e-06,
+      "loss": 0.52,
+      "step": 711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6256288323393315,
+      "learning_rate": 9.911044177408917e-06,
+      "loss": 0.5461,
+      "step": 712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4504352891549575,
+      "learning_rate": 9.910666371737848e-06,
+      "loss": 0.5386,
+      "step": 713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6943756408365623,
+      "learning_rate": 9.910287772705374e-06,
+      "loss": 0.51,
+      "step": 714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6456825439428284,
+      "learning_rate": 9.909908380372667e-06,
+      "loss": 0.5576,
+      "step": 715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3685240161884205,
+      "learning_rate": 9.909528194801019e-06,
+      "loss": 0.5561,
+      "step": 716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6513433815250442,
+      "learning_rate": 9.90914721605185e-06,
+      "loss": 0.4893,
+      "step": 717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2340929468815107,
+      "learning_rate": 9.908765444186713e-06,
+      "loss": 0.5007,
+      "step": 718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3786299592454312,
+      "learning_rate": 9.908382879267286e-06,
+      "loss": 0.5752,
+      "step": 719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4800630744877326,
+      "learning_rate": 9.907999521355377e-06,
+      "loss": 0.5015,
+      "step": 720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2232871580378,
+      "learning_rate": 9.907615370512919e-06,
+      "loss": 0.497,
+      "step": 721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2928878365401588,
+      "learning_rate": 9.907230426801975e-06,
+      "loss": 0.5447,
+      "step": 722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4967207824812627,
+      "learning_rate": 9.906844690284737e-06,
+      "loss": 0.5291,
+      "step": 723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.816746657284892,
+      "learning_rate": 9.906458161023522e-06,
+      "loss": 0.5342,
+      "step": 724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0861253042637484,
+      "learning_rate": 9.90607083908078e-06,
+      "loss": 0.5903,
+      "step": 725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5926144364257493,
+      "learning_rate": 9.905682724519086e-06,
+      "loss": 0.5445,
+      "step": 726
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3870599530402037,
+      "learning_rate": 9.90529381740114e-06,
+      "loss": 0.5318,
+      "step": 727
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6319587413630097,
+      "learning_rate": 9.904904117789777e-06,
+      "loss": 0.5497,
+      "step": 728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9513204246105857,
+      "learning_rate": 9.904513625747957e-06,
+      "loss": 0.5404,
+      "step": 729
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6492222551928712,
+      "learning_rate": 9.904122341338764e-06,
+      "loss": 0.5099,
+      "step": 730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7330336994895605,
+      "learning_rate": 9.903730264625415e-06,
+      "loss": 0.5237,
+      "step": 731
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.335780098202449,
+      "learning_rate": 9.903337395671255e-06,
+      "loss": 0.5068,
+      "step": 732
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6746985228306138,
+      "learning_rate": 9.902943734539752e-06,
+      "loss": 0.479,
+      "step": 733
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5679095009354174,
+      "learning_rate": 9.902549281294508e-06,
+      "loss": 0.5457,
+      "step": 734
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5978083443392859,
+      "learning_rate": 9.90215403599925e-06,
+      "loss": 0.5094,
+      "step": 735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3502793900824692,
+      "learning_rate": 9.901757998717833e-06,
+      "loss": 0.5287,
+      "step": 736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3767717373956716,
+      "learning_rate": 9.90136116951424e-06,
+      "loss": 0.5638,
+      "step": 737
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4100309801230537,
+      "learning_rate": 9.900963548452584e-06,
+      "loss": 0.5088,
+      "step": 738
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6723873034933143,
+      "learning_rate": 9.900565135597104e-06,
+      "loss": 0.5254,
+      "step": 739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4968579496803647,
+      "learning_rate": 9.900165931012164e-06,
+      "loss": 0.5401,
+      "step": 740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.512318360157006,
+      "learning_rate": 9.899765934762263e-06,
+      "loss": 0.4433,
+      "step": 741
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5025123815418007,
+      "learning_rate": 9.899365146912018e-06,
+      "loss": 0.511,
+      "step": 742
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6683442789870047,
+      "learning_rate": 9.898963567526188e-06,
+      "loss": 0.5479,
+      "step": 743
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6151733284076952,
+      "learning_rate": 9.898561196669645e-06,
+      "loss": 0.5266,
+      "step": 744
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4827781091080732,
+      "learning_rate": 9.8981580344074e-06,
+      "loss": 0.5435,
+      "step": 745
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3361012140606874,
+      "learning_rate": 9.897754080804583e-06,
+      "loss": 0.5011,
+      "step": 746
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.570149978353058,
+      "learning_rate": 9.89734933592646e-06,
+      "loss": 0.5391,
+      "step": 747
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4440488628055852,
+      "learning_rate": 9.896943799838419e-06,
+      "loss": 0.563,
+      "step": 748
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4722485313647773,
+      "learning_rate": 9.896537472605979e-06,
+      "loss": 0.5872,
+      "step": 749
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6458885186334251,
+      "learning_rate": 9.896130354294787e-06,
+      "loss": 0.4825,
+      "step": 750
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0870975798160076,
+      "learning_rate": 9.895722444970613e-06,
+      "loss": 0.5378,
+      "step": 751
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6713600482428013,
+      "learning_rate": 9.89531374469936e-06,
+      "loss": 0.5243,
+      "step": 752
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3666798769307291,
+      "learning_rate": 9.89490425354706e-06,
+      "loss": 0.537,
+      "step": 753
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6965904249779475,
+      "learning_rate": 9.894493971579864e-06,
+      "loss": 0.5512,
+      "step": 754
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3900037721001617,
+      "learning_rate": 9.894082898864062e-06,
+      "loss": 0.5467,
+      "step": 755
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.524197186398035,
+      "learning_rate": 9.893671035466065e-06,
+      "loss": 0.4735,
+      "step": 756
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7339854586343567,
+      "learning_rate": 9.893258381452412e-06,
+      "loss": 0.5089,
+      "step": 757
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3923524219523842,
+      "learning_rate": 9.892844936889772e-06,
+      "loss": 0.4909,
+      "step": 758
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5124594114302838,
+      "learning_rate": 9.89243070184494e-06,
+      "loss": 0.5394,
+      "step": 759
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8034130536722808,
+      "learning_rate": 9.892015676384838e-06,
+      "loss": 0.512,
+      "step": 760
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5762898788709665,
+      "learning_rate": 9.891599860576519e-06,
+      "loss": 0.5248,
+      "step": 761
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4714297837827943,
+      "learning_rate": 9.891183254487162e-06,
+      "loss": 0.5465,
+      "step": 762
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3214866314049687,
+      "learning_rate": 9.890765858184073e-06,
+      "loss": 0.516,
+      "step": 763
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9633591420432888,
+      "learning_rate": 9.890347671734684e-06,
+      "loss": 0.5366,
+      "step": 764
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3580228817655544,
+      "learning_rate": 9.889928695206559e-06,
+      "loss": 0.5315,
+      "step": 765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5198904764841936,
+      "learning_rate": 9.889508928667387e-06,
+      "loss": 0.5439,
+      "step": 766
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.239507833632809,
+      "learning_rate": 9.889088372184983e-06,
+      "loss": 0.549,
+      "step": 767
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0090601913369452,
+      "learning_rate": 9.888667025827295e-06,
+      "loss": 0.5388,
+      "step": 768
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7600992496202628,
+      "learning_rate": 9.888244889662394e-06,
+      "loss": 0.5286,
+      "step": 769
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.363528192366846,
+      "learning_rate": 9.887821963758478e-06,
+      "loss": 0.5206,
+      "step": 770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3125021258357175,
+      "learning_rate": 9.887398248183876e-06,
+      "loss": 0.5342,
+      "step": 771
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1185739451091274,
+      "learning_rate": 9.886973743007042e-06,
+      "loss": 0.559,
+      "step": 772
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3878073704129925,
+      "learning_rate": 9.88654844829656e-06,
+      "loss": 0.5069,
+      "step": 773
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.603486854404389,
+      "learning_rate": 9.88612236412114e-06,
+      "loss": 0.5484,
+      "step": 774
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5023961328735826,
+      "learning_rate": 9.885695490549616e-06,
+      "loss": 0.5537,
+      "step": 775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.345851732601472,
+      "learning_rate": 9.88526782765096e-06,
+      "loss": 0.5059,
+      "step": 776
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3519885963470128,
+      "learning_rate": 9.884839375494258e-06,
+      "loss": 0.4886,
+      "step": 777
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.48289100247411,
+      "learning_rate": 9.884410134148733e-06,
+      "loss": 0.5144,
+      "step": 778
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4008779870643109,
+      "learning_rate": 9.883980103683735e-06,
+      "loss": 0.5803,
+      "step": 779
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4562131509620257,
+      "learning_rate": 9.883549284168736e-06,
+      "loss": 0.5574,
+      "step": 780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4444072614610364,
+      "learning_rate": 9.883117675673339e-06,
+      "loss": 0.4858,
+      "step": 781
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5080506086390568,
+      "learning_rate": 9.882685278267277e-06,
+      "loss": 0.5444,
+      "step": 782
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7045840262615868,
+      "learning_rate": 9.882252092020403e-06,
+      "loss": 0.5505,
+      "step": 783
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7370813343274576,
+      "learning_rate": 9.881818117002706e-06,
+      "loss": 0.5415,
+      "step": 784
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.018153134156803,
+      "learning_rate": 9.881383353284295e-06,
+      "loss": 0.4996,
+      "step": 785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3253195899746142,
+      "learning_rate": 9.880947800935414e-06,
+      "loss": 0.5106,
+      "step": 786
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7024567402620956,
+      "learning_rate": 9.880511460026428e-06,
+      "loss": 0.5396,
+      "step": 787
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4091988414010796,
+      "learning_rate": 9.88007433062783e-06,
+      "loss": 0.555,
+      "step": 788
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7087771391854086,
+      "learning_rate": 9.879636412810244e-06,
+      "loss": 0.4847,
+      "step": 789
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7231897900894366,
+      "learning_rate": 9.879197706644423e-06,
+      "loss": 0.5219,
+      "step": 790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7027320981466387,
+      "learning_rate": 9.878758212201238e-06,
+      "loss": 0.5014,
+      "step": 791
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3482912883508638,
+      "learning_rate": 9.878317929551695e-06,
+      "loss": 0.4895,
+      "step": 792
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3152332995824816,
+      "learning_rate": 9.877876858766927e-06,
+      "loss": 0.5465,
+      "step": 793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4997609418658844,
+      "learning_rate": 9.877434999918192e-06,
+      "loss": 0.5262,
+      "step": 794
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6539265342268819,
+      "learning_rate": 9.876992353076875e-06,
+      "loss": 0.4821,
+      "step": 795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6249056812162253,
+      "learning_rate": 9.876548918314492e-06,
+      "loss": 0.534,
+      "step": 796
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5004396199830836,
+      "learning_rate": 9.876104695702682e-06,
+      "loss": 0.5699,
+      "step": 797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2826945141311428,
+      "learning_rate": 9.875659685313214e-06,
+      "loss": 0.4789,
+      "step": 798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5469724214523863,
+      "learning_rate": 9.875213887217983e-06,
+      "loss": 0.5576,
+      "step": 799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5415041213137937,
+      "learning_rate": 9.87476730148901e-06,
+      "loss": 0.5988,
+      "step": 800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7021497983380178,
+      "learning_rate": 9.874319928198448e-06,
+      "loss": 0.5403,
+      "step": 801
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3549275899238842,
+      "learning_rate": 9.873871767418572e-06,
+      "loss": 0.455,
+      "step": 802
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2532327534233543,
+      "learning_rate": 9.873422819221785e-06,
+      "loss": 0.5064,
+      "step": 803
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7360535633059955,
+      "learning_rate": 9.872973083680622e-06,
+      "loss": 0.5882,
+      "step": 804
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4059071965483354,
+      "learning_rate": 9.87252256086774e-06,
+      "loss": 0.5658,
+      "step": 805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.530355778895807,
+      "learning_rate": 9.872071250855925e-06,
+      "loss": 0.5147,
+      "step": 806
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3847353474560062,
+      "learning_rate": 9.87161915371809e-06,
+      "loss": 0.5407,
+      "step": 807
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2893581665896208,
+      "learning_rate": 9.871166269527275e-06,
+      "loss": 0.5172,
+      "step": 808
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4749327657449687,
+      "learning_rate": 9.870712598356647e-06,
+      "loss": 0.5339,
+      "step": 809
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4358631901359513,
+      "learning_rate": 9.870258140279503e-06,
+      "loss": 0.5119,
+      "step": 810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1722531501187503,
+      "learning_rate": 9.869802895369262e-06,
+      "loss": 0.5623,
+      "step": 811
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3154410118141262,
+      "learning_rate": 9.869346863699474e-06,
+      "loss": 0.5222,
+      "step": 812
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5619568170354916,
+      "learning_rate": 9.868890045343814e-06,
+      "loss": 0.5189,
+      "step": 813
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5104678506036293,
+      "learning_rate": 9.868432440376086e-06,
+      "loss": 0.5747,
+      "step": 814
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3470148837454545,
+      "learning_rate": 9.867974048870221e-06,
+      "loss": 0.5123,
+      "step": 815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3544857302522781,
+      "learning_rate": 9.867514870900274e-06,
+      "loss": 0.5256,
+      "step": 816
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1417064240265287,
+      "learning_rate": 9.867054906540432e-06,
+      "loss": 0.5228,
+      "step": 817
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.069195455278062,
+      "learning_rate": 9.866594155865004e-06,
+      "loss": 0.5605,
+      "step": 818
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6880607864614212,
+      "learning_rate": 9.866132618948428e-06,
+      "loss": 0.49,
+      "step": 819
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9067634496940928,
+      "learning_rate": 9.86567029586527e-06,
+      "loss": 0.5758,
+      "step": 820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3809878109560794,
+      "learning_rate": 9.865207186690225e-06,
+      "loss": 0.4819,
+      "step": 821
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4377570369203871,
+      "learning_rate": 9.86474329149811e-06,
+      "loss": 0.5409,
+      "step": 822
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2837566075142783,
+      "learning_rate": 9.864278610363869e-06,
+      "loss": 0.5082,
+      "step": 823
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.376197397673897,
+      "learning_rate": 9.863813143362579e-06,
+      "loss": 0.468,
+      "step": 824
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5091655795057206,
+      "learning_rate": 9.863346890569438e-06,
+      "loss": 0.5511,
+      "step": 825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2654935539583412,
+      "learning_rate": 9.862879852059775e-06,
+      "loss": 0.5588,
+      "step": 826
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.480429796301672,
+      "learning_rate": 9.862412027909045e-06,
+      "loss": 0.5077,
+      "step": 827
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4231689337630336,
+      "learning_rate": 9.861943418192825e-06,
+      "loss": 0.5572,
+      "step": 828
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6537058377816523,
+      "learning_rate": 9.861474022986828e-06,
+      "loss": 0.5611,
+      "step": 829
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4257922752948444,
+      "learning_rate": 9.861003842366886e-06,
+      "loss": 0.5383,
+      "step": 830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5238556011282056,
+      "learning_rate": 9.86053287640896e-06,
+      "loss": 0.5396,
+      "step": 831
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5593477658026424,
+      "learning_rate": 9.860061125189143e-06,
+      "loss": 0.5556,
+      "step": 832
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6877101213463102,
+      "learning_rate": 9.859588588783644e-06,
+      "loss": 0.5304,
+      "step": 833
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.52836473336756,
+      "learning_rate": 9.85911526726881e-06,
+      "loss": 0.5365,
+      "step": 834
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4292053696202944,
+      "learning_rate": 9.85864116072111e-06,
+      "loss": 0.524,
+      "step": 835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3567394691546826,
+      "learning_rate": 9.85816626921714e-06,
+      "loss": 0.5385,
+      "step": 836
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.813389252993085,
+      "learning_rate": 9.857690592833621e-06,
+      "loss": 0.5389,
+      "step": 837
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6373671449371334,
+      "learning_rate": 9.857214131647405e-06,
+      "loss": 0.517,
+      "step": 838
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.550185315855915,
+      "learning_rate": 9.856736885735467e-06,
+      "loss": 0.5129,
+      "step": 839
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2396005024409704,
+      "learning_rate": 9.856258855174912e-06,
+      "loss": 0.5535,
+      "step": 840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4116583979770747,
+      "learning_rate": 9.855780040042966e-06,
+      "loss": 0.4832,
+      "step": 841
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6187709239417065,
+      "learning_rate": 9.85530044041699e-06,
+      "loss": 0.5649,
+      "step": 842
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3743636030552622,
+      "learning_rate": 9.854820056374468e-06,
+      "loss": 0.548,
+      "step": 843
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4003641447914825,
+      "learning_rate": 9.854338887993006e-06,
+      "loss": 0.5033,
+      "step": 844
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6185345234958257,
+      "learning_rate": 9.853856935350345e-06,
+      "loss": 0.593,
+      "step": 845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8383993635732563,
+      "learning_rate": 9.853374198524347e-06,
+      "loss": 0.5019,
+      "step": 846
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4429440730187004,
+      "learning_rate": 9.852890677593003e-06,
+      "loss": 0.553,
+      "step": 847
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.502777317388333,
+      "learning_rate": 9.852406372634427e-06,
+      "loss": 0.5163,
+      "step": 848
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7042353791138922,
+      "learning_rate": 9.85192128372687e-06,
+      "loss": 0.5671,
+      "step": 849
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5670146983134932,
+      "learning_rate": 9.851435410948693e-06,
+      "loss": 0.5254,
+      "step": 850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6943773775700972,
+      "learning_rate": 9.8509487543784e-06,
+      "loss": 0.5858,
+      "step": 851
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7259489878374763,
+      "learning_rate": 9.850461314094611e-06,
+      "loss": 0.5266,
+      "step": 852
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4839090775806405,
+      "learning_rate": 9.84997309017608e-06,
+      "loss": 0.4945,
+      "step": 853
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6430132038958682,
+      "learning_rate": 9.849484082701679e-06,
+      "loss": 0.5514,
+      "step": 854
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.330689307629748,
+      "learning_rate": 9.848994291750416e-06,
+      "loss": 0.5068,
+      "step": 855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4367448150793294,
+      "learning_rate": 9.848503717401416e-06,
+      "loss": 0.4849,
+      "step": 856
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5435415775983266,
+      "learning_rate": 9.848012359733941e-06,
+      "loss": 0.5299,
+      "step": 857
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0779784325734822,
+      "learning_rate": 9.847520218827372e-06,
+      "loss": 0.5144,
+      "step": 858
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7886067252096047,
+      "learning_rate": 9.847027294761218e-06,
+      "loss": 0.5121,
+      "step": 859
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6874506009278324,
+      "learning_rate": 9.846533587615117e-06,
+      "loss": 0.5396,
+      "step": 860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6245535898960135,
+      "learning_rate": 9.846039097468829e-06,
+      "loss": 0.5138,
+      "step": 861
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5610689448816104,
+      "learning_rate": 9.845543824402245e-06,
+      "loss": 0.5456,
+      "step": 862
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.867756358837439,
+      "learning_rate": 9.84504776849538e-06,
+      "loss": 0.5296,
+      "step": 863
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6572342482634592,
+      "learning_rate": 9.844550929828375e-06,
+      "loss": 0.4763,
+      "step": 864
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5801874916299146,
+      "learning_rate": 9.844053308481504e-06,
+      "loss": 0.5699,
+      "step": 865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4583504122348614,
+      "learning_rate": 9.843554904535157e-06,
+      "loss": 0.5041,
+      "step": 866
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.611938008913365,
+      "learning_rate": 9.843055718069856e-06,
+      "loss": 0.5419,
+      "step": 867
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.43075395951784,
+      "learning_rate": 9.842555749166252e-06,
+      "loss": 0.5617,
+      "step": 868
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.380501740350102,
+      "learning_rate": 9.842054997905115e-06,
+      "loss": 0.4923,
+      "step": 869
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4650493645359763,
+      "learning_rate": 9.841553464367349e-06,
+      "loss": 0.5058,
+      "step": 870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.314374886920701,
+      "learning_rate": 9.84105114863398e-06,
+      "loss": 0.5574,
+      "step": 871
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.335405641847203,
+      "learning_rate": 9.840548050786162e-06,
+      "loss": 0.5267,
+      "step": 872
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.882923799877262,
+      "learning_rate": 9.840044170905175e-06,
+      "loss": 0.5018,
+      "step": 873
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.606183086211936,
+      "learning_rate": 9.839539509072425e-06,
+      "loss": 0.5532,
+      "step": 874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4106290303040485,
+      "learning_rate": 9.839034065369444e-06,
+      "loss": 0.4926,
+      "step": 875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8398370549040346,
+      "learning_rate": 9.83852783987789e-06,
+      "loss": 0.5319,
+      "step": 876
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8145118768900863,
+      "learning_rate": 9.838020832679552e-06,
+      "loss": 0.5236,
+      "step": 877
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.1562995225815056,
+      "learning_rate": 9.837513043856338e-06,
+      "loss": 0.513,
+      "step": 878
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5446941041707583,
+      "learning_rate": 9.837004473490286e-06,
+      "loss": 0.5432,
+      "step": 879
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7330545831926203,
+      "learning_rate": 9.83649512166356e-06,
+      "loss": 0.5111,
+      "step": 880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.822826636212995,
+      "learning_rate": 9.835984988458454e-06,
+      "loss": 0.5,
+      "step": 881
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.483699967216602,
+      "learning_rate": 9.835474073957379e-06,
+      "loss": 0.5181,
+      "step": 882
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8228914204587476,
+      "learning_rate": 9.834962378242882e-06,
+      "loss": 0.5745,
+      "step": 883
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7397105597518532,
+      "learning_rate": 9.834449901397628e-06,
+      "loss": 0.5306,
+      "step": 884
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3840024922502834,
+      "learning_rate": 9.833936643504417e-06,
+      "loss": 0.4915,
+      "step": 885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6649130295487007,
+      "learning_rate": 9.833422604646165e-06,
+      "loss": 0.5362,
+      "step": 886
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7078837515289695,
+      "learning_rate": 9.832907784905924e-06,
+      "loss": 0.4886,
+      "step": 887
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7511691661944424,
+      "learning_rate": 9.832392184366867e-06,
+      "loss": 0.4758,
+      "step": 888
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3751504432892512,
+      "learning_rate": 9.831875803112291e-06,
+      "loss": 0.5,
+      "step": 889
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4513862307154417,
+      "learning_rate": 9.831358641225624e-06,
+      "loss": 0.5377,
+      "step": 890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5931908846008063,
+      "learning_rate": 9.830840698790418e-06,
+      "loss": 0.5329,
+      "step": 891
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3425686954116223,
+      "learning_rate": 9.83032197589035e-06,
+      "loss": 0.4917,
+      "step": 892
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2865043128692821,
+      "learning_rate": 9.829802472609227e-06,
+      "loss": 0.4823,
+      "step": 893
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4653845961751846,
+      "learning_rate": 9.829282189030976e-06,
+      "loss": 0.5754,
+      "step": 894
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5380876117092384,
+      "learning_rate": 9.828761125239655e-06,
+      "loss": 0.5642,
+      "step": 895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6425053205832336,
+      "learning_rate": 9.828239281319447e-06,
+      "loss": 0.5546,
+      "step": 896
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.256609403421453,
+      "learning_rate": 9.827716657354658e-06,
+      "loss": 0.4601,
+      "step": 897
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.390734805023612,
+      "learning_rate": 9.827193253429726e-06,
+      "loss": 0.5261,
+      "step": 898
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7799958313656696,
+      "learning_rate": 9.826669069629209e-06,
+      "loss": 0.5457,
+      "step": 899
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4733773601073992,
+      "learning_rate": 9.826144106037794e-06,
+      "loss": 0.5611,
+      "step": 900
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9129884749417165,
+      "learning_rate": 9.825618362740295e-06,
+      "loss": 0.5339,
+      "step": 901
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.243230069846606,
+      "learning_rate": 9.825091839821648e-06,
+      "loss": 0.5816,
+      "step": 902
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0242936924161343,
+      "learning_rate": 9.82456453736692e-06,
+      "loss": 0.5167,
+      "step": 903
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.434184689682166,
+      "learning_rate": 9.8240364554613e-06,
+      "loss": 0.4836,
+      "step": 904
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6775533855530116,
+      "learning_rate": 9.823507594190103e-06,
+      "loss": 0.4603,
+      "step": 905
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6827627319527105,
+      "learning_rate": 9.822977953638772e-06,
+      "loss": 0.5396,
+      "step": 906
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3676057637250456,
+      "learning_rate": 9.822447533892877e-06,
+      "loss": 0.5059,
+      "step": 907
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3258876354387583,
+      "learning_rate": 9.821916335038111e-06,
+      "loss": 0.5057,
+      "step": 908
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 134.66857492128472,
+      "learning_rate": 9.821384357160292e-06,
+      "loss": 0.5451,
+      "step": 909
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2585093599662378,
+      "learning_rate": 9.820851600345368e-06,
+      "loss": 0.5558,
+      "step": 910
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5964531920502647,
+      "learning_rate": 9.820318064679408e-06,
+      "loss": 0.505,
+      "step": 911
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.969482196985344,
+      "learning_rate": 9.819783750248612e-06,
+      "loss": 0.5582,
+      "step": 912
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5323419980885993,
+      "learning_rate": 9.819248657139304e-06,
+      "loss": 0.5278,
+      "step": 913
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.449348072753027,
+      "learning_rate": 9.818712785437929e-06,
+      "loss": 0.5646,
+      "step": 914
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.113064908635709,
+      "learning_rate": 9.818176135231066e-06,
+      "loss": 0.5043,
+      "step": 915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.46539588571356,
+      "learning_rate": 9.817638706605413e-06,
+      "loss": 0.5927,
+      "step": 916
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9591832216721343,
+      "learning_rate": 9.817100499647796e-06,
+      "loss": 0.4954,
+      "step": 917
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8904706076981188,
+      "learning_rate": 9.81656151444517e-06,
+      "loss": 0.5074,
+      "step": 918
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.603397229383552,
+      "learning_rate": 9.81602175108461e-06,
+      "loss": 0.5213,
+      "step": 919
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4407858273805725,
+      "learning_rate": 9.815481209653318e-06,
+      "loss": 0.5073,
+      "step": 920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6780133192986095,
+      "learning_rate": 9.814939890238629e-06,
+      "loss": 0.5749,
+      "step": 921
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3890417822540029,
+      "learning_rate": 9.814397792927993e-06,
+      "loss": 0.5103,
+      "step": 922
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4123428104356073,
+      "learning_rate": 9.813854917808993e-06,
+      "loss": 0.511,
+      "step": 923
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.3239410306747974,
+      "learning_rate": 9.813311264969333e-06,
+      "loss": 0.5895,
+      "step": 924
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6481482416750444,
+      "learning_rate": 9.812766834496847e-06,
+      "loss": 0.5619,
+      "step": 925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5682167709528458,
+      "learning_rate": 9.812221626479493e-06,
+      "loss": 0.5106,
+      "step": 926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.108483178988802,
+      "learning_rate": 9.811675641005352e-06,
+      "loss": 0.4574,
+      "step": 927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.272404284404087,
+      "learning_rate": 9.811128878162633e-06,
+      "loss": 0.5017,
+      "step": 928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.229753751249011,
+      "learning_rate": 9.81058133803967e-06,
+      "loss": 0.4652,
+      "step": 929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8807442456609524,
+      "learning_rate": 9.810033020724927e-06,
+      "loss": 0.5045,
+      "step": 930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5325223466990254,
+      "learning_rate": 9.809483926306984e-06,
+      "loss": 0.5153,
+      "step": 931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.293134868842404,
+      "learning_rate": 9.808934054874557e-06,
+      "loss": 0.4899,
+      "step": 932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.398989709606525,
+      "learning_rate": 9.808383406516478e-06,
+      "loss": 0.5187,
+      "step": 933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.071140988213683,
+      "learning_rate": 9.807831981321711e-06,
+      "loss": 0.5403,
+      "step": 934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6607542531288317,
+      "learning_rate": 9.807279779379346e-06,
+      "loss": 0.4914,
+      "step": 935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3850092138390073,
+      "learning_rate": 9.806726800778592e-06,
+      "loss": 0.4681,
+      "step": 936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5181764228685495,
+      "learning_rate": 9.806173045608791e-06,
+      "loss": 0.4813,
+      "step": 937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.71479008444105,
+      "learning_rate": 9.805618513959405e-06,
+      "loss": 0.5298,
+      "step": 938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7241969898445435,
+      "learning_rate": 9.805063205920024e-06,
+      "loss": 0.5251,
+      "step": 939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4856505135312046,
+      "learning_rate": 9.804507121580363e-06,
+      "loss": 0.5439,
+      "step": 940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5218457627879594,
+      "learning_rate": 9.80395026103026e-06,
+      "loss": 0.538,
+      "step": 941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9835057866029318,
+      "learning_rate": 9.803392624359686e-06,
+      "loss": 0.5694,
+      "step": 942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7890943425296235,
+      "learning_rate": 9.802834211658727e-06,
+      "loss": 0.556,
+      "step": 943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.823356520693511,
+      "learning_rate": 9.802275023017603e-06,
+      "loss": 0.4954,
+      "step": 944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.778959746033984,
+      "learning_rate": 9.801715058526654e-06,
+      "loss": 0.5047,
+      "step": 945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5026611567465495,
+      "learning_rate": 9.801154318276346e-06,
+      "loss": 0.5611,
+      "step": 946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3920378820144184,
+      "learning_rate": 9.800592802357274e-06,
+      "loss": 0.5048,
+      "step": 947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5561386982115961,
+      "learning_rate": 9.800030510860154e-06,
+      "loss": 0.5821,
+      "step": 948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.0569047792119814,
+      "learning_rate": 9.799467443875832e-06,
+      "loss": 0.4784,
+      "step": 949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5279500400553616,
+      "learning_rate": 9.798903601495274e-06,
+      "loss": 0.5358,
+      "step": 950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4977870060684855,
+      "learning_rate": 9.798338983809574e-06,
+      "loss": 0.5229,
+      "step": 951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7866661445205454,
+      "learning_rate": 9.797773590909951e-06,
+      "loss": 0.5753,
+      "step": 952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6653644954339666,
+      "learning_rate": 9.797207422887749e-06,
+      "loss": 0.5222,
+      "step": 953
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4942644071101057,
+      "learning_rate": 9.796640479834439e-06,
+      "loss": 0.5648,
+      "step": 954
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3212009783951033,
+      "learning_rate": 9.796072761841614e-06,
+      "loss": 0.4975,
+      "step": 955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3157046474966476,
+      "learning_rate": 9.795504269000993e-06,
+      "loss": 0.5508,
+      "step": 956
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5525174218441946,
+      "learning_rate": 9.794935001404424e-06,
+      "loss": 0.5046,
+      "step": 957
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5409578454278048,
+      "learning_rate": 9.794364959143876e-06,
+      "loss": 0.5008,
+      "step": 958
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0286529491143623,
+      "learning_rate": 9.793794142311442e-06,
+      "loss": 0.5295,
+      "step": 959
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.735399312413163,
+      "learning_rate": 9.793222550999343e-06,
+      "loss": 0.4717,
+      "step": 960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.364301097630389,
+      "learning_rate": 9.79265018529993e-06,
+      "loss": 0.5434,
+      "step": 961
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6307910155590302,
+      "learning_rate": 9.792077045305667e-06,
+      "loss": 0.5009,
+      "step": 962
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0261251017328084,
+      "learning_rate": 9.791503131109153e-06,
+      "loss": 0.5054,
+      "step": 963
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8934827285210059,
+      "learning_rate": 9.790928442803109e-06,
+      "loss": 0.5408,
+      "step": 964
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7390284854415334,
+      "learning_rate": 9.79035298048038e-06,
+      "loss": 0.5017,
+      "step": 965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5967519421160552,
+      "learning_rate": 9.789776744233937e-06,
+      "loss": 0.5105,
+      "step": 966
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4326649427114473,
+      "learning_rate": 9.789199734156876e-06,
+      "loss": 0.5143,
+      "step": 967
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0047895387472607,
+      "learning_rate": 9.788621950342419e-06,
+      "loss": 0.5886,
+      "step": 968
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.059058151775754,
+      "learning_rate": 9.788043392883913e-06,
+      "loss": 0.5324,
+      "step": 969
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7462739610027802,
+      "learning_rate": 9.787464061874826e-06,
+      "loss": 0.5006,
+      "step": 970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.440421328798765,
+      "learning_rate": 9.786883957408755e-06,
+      "loss": 0.5404,
+      "step": 971
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4410596054166127,
+      "learning_rate": 9.786303079579423e-06,
+      "loss": 0.562,
+      "step": 972
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3845526423586314,
+      "learning_rate": 9.785721428480672e-06,
+      "loss": 0.5482,
+      "step": 973
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6455669709278569,
+      "learning_rate": 9.785139004206478e-06,
+      "loss": 0.5413,
+      "step": 974
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.965987983372974,
+      "learning_rate": 9.784555806850932e-06,
+      "loss": 0.546,
+      "step": 975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3477974890308504,
+      "learning_rate": 9.783971836508258e-06,
+      "loss": 0.5016,
+      "step": 976
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4996179118803765,
+      "learning_rate": 9.783387093272801e-06,
+      "loss": 0.603,
+      "step": 977
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6326168420110307,
+      "learning_rate": 9.78280157723903e-06,
+      "loss": 0.4822,
+      "step": 978
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1691513988958606,
+      "learning_rate": 9.782215288501541e-06,
+      "loss": 0.5877,
+      "step": 979
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3980271196527394,
+      "learning_rate": 9.781628227155056e-06,
+      "loss": 0.4766,
+      "step": 980
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5816788788667602,
+      "learning_rate": 9.781040393294415e-06,
+      "loss": 0.5204,
+      "step": 981
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4866512555478568,
+      "learning_rate": 9.780451787014593e-06,
+      "loss": 0.5519,
+      "step": 982
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.500045693403318,
+      "learning_rate": 9.779862408410682e-06,
+      "loss": 0.5559,
+      "step": 983
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.539271350449334,
+      "learning_rate": 9.779272257577901e-06,
+      "loss": 0.5247,
+      "step": 984
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.452860261604936,
+      "learning_rate": 9.778681334611595e-06,
+      "loss": 0.5641,
+      "step": 985
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.598342601973268,
+      "learning_rate": 9.778089639607232e-06,
+      "loss": 0.5046,
+      "step": 986
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5798505955932483,
+      "learning_rate": 9.777497172660406e-06,
+      "loss": 0.4981,
+      "step": 987
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4824775493039155,
+      "learning_rate": 9.776903933866838e-06,
+      "loss": 0.5306,
+      "step": 988
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7456312083003678,
+      "learning_rate": 9.776309923322365e-06,
+      "loss": 0.5612,
+      "step": 989
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0909503716589932,
+      "learning_rate": 9.77571514112296e-06,
+      "loss": 0.5093,
+      "step": 990
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.043487520816119,
+      "learning_rate": 9.77511958736471e-06,
+      "loss": 0.5899,
+      "step": 991
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5009302169192755,
+      "learning_rate": 9.774523262143838e-06,
+      "loss": 0.5234,
+      "step": 992
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.693164828745711,
+      "learning_rate": 9.773926165556682e-06,
+      "loss": 0.5293,
+      "step": 993
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.320473825716359,
+      "learning_rate": 9.773328297699709e-06,
+      "loss": 0.4707,
+      "step": 994
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.493561778782454,
+      "learning_rate": 9.772729658669511e-06,
+      "loss": 0.5299,
+      "step": 995
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3194409305794026,
+      "learning_rate": 9.7721302485628e-06,
+      "loss": 0.5391,
+      "step": 996
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5548800287869815,
+      "learning_rate": 9.77153006747642e-06,
+      "loss": 0.5221,
+      "step": 997
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3753921622169898,
+      "learning_rate": 9.770929115507333e-06,
+      "loss": 0.4613,
+      "step": 998
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4793123657960134,
+      "learning_rate": 9.770327392752627e-06,
+      "loss": 0.5219,
+      "step": 999
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6888355213538224,
+      "learning_rate": 9.76972489930952e-06,
+      "loss": 0.5256,
+      "step": 1000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.510342478308107,
+      "learning_rate": 9.769121635275348e-06,
+      "loss": 0.506,
+      "step": 1001
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3832178583611607,
+      "learning_rate": 9.768517600747572e-06,
+      "loss": 0.5308,
+      "step": 1002
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3804901712551811,
+      "learning_rate": 9.76791279582378e-06,
+      "loss": 0.5085,
+      "step": 1003
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5874172918864036,
+      "learning_rate": 9.767307220601682e-06,
+      "loss": 0.564,
+      "step": 1004
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.709225786964121,
+      "learning_rate": 9.766700875179118e-06,
+      "loss": 0.5527,
+      "step": 1005
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6674120712422085,
+      "learning_rate": 9.766093759654047e-06,
+      "loss": 0.549,
+      "step": 1006
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5771344976496824,
+      "learning_rate": 9.765485874124551e-06,
+      "loss": 0.541,
+      "step": 1007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0461275407031114,
+      "learning_rate": 9.764877218688844e-06,
+      "loss": 0.5979,
+      "step": 1008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5508383621666664,
+      "learning_rate": 9.764267793445254e-06,
+      "loss": 0.5072,
+      "step": 1009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5389952214975169,
+      "learning_rate": 9.763657598492244e-06,
+      "loss": 0.5267,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.19076555655587,
+      "learning_rate": 9.763046633928394e-06,
+      "loss": 0.5695,
+      "step": 1011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1357995975740542,
+      "learning_rate": 9.762434899852412e-06,
+      "loss": 0.5243,
+      "step": 1012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.30912398608652,
+      "learning_rate": 9.761822396363127e-06,
+      "loss": 0.5061,
+      "step": 1013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7211057851844471,
+      "learning_rate": 9.761209123559497e-06,
+      "loss": 0.5614,
+      "step": 1014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6430505979593055,
+      "learning_rate": 9.760595081540598e-06,
+      "loss": 0.5224,
+      "step": 1015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7293605186403183,
+      "learning_rate": 9.759980270405636e-06,
+      "loss": 0.5426,
+      "step": 1016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6819165374186842,
+      "learning_rate": 9.75936469025394e-06,
+      "loss": 0.5263,
+      "step": 1017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0238097603809133,
+      "learning_rate": 9.758748341184963e-06,
+      "loss": 0.5028,
+      "step": 1018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.498056742579276,
+      "learning_rate": 9.758131223298277e-06,
+      "loss": 0.4839,
+      "step": 1019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9442501356371493,
+      "learning_rate": 9.757513336693589e-06,
+      "loss": 0.4814,
+      "step": 1020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5512458163562854,
+      "learning_rate": 9.756894681470718e-06,
+      "loss": 0.4852,
+      "step": 1021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1008113097679795,
+      "learning_rate": 9.756275257729619e-06,
+      "loss": 0.5667,
+      "step": 1022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8076424691587238,
+      "learning_rate": 9.755655065570363e-06,
+      "loss": 0.5121,
+      "step": 1023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.799319617461757,
+      "learning_rate": 9.755034105093143e-06,
+      "loss": 0.5469,
+      "step": 1024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.364799714460202,
+      "learning_rate": 9.754412376398288e-06,
+      "loss": 0.5173,
+      "step": 1025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5452171924103493,
+      "learning_rate": 9.75378987958624e-06,
+      "loss": 0.5226,
+      "step": 1026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.548293795499255,
+      "learning_rate": 9.75316661475757e-06,
+      "loss": 0.5025,
+      "step": 1027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5622088483551202,
+      "learning_rate": 9.752542582012969e-06,
+      "loss": 0.5667,
+      "step": 1028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2172447624890994,
+      "learning_rate": 9.75191778145326e-06,
+      "loss": 0.5161,
+      "step": 1029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5164628494052323,
+      "learning_rate": 9.75129221317938e-06,
+      "loss": 0.5504,
+      "step": 1030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.803976492652808,
+      "learning_rate": 9.750665877292399e-06,
+      "loss": 0.5444,
+      "step": 1031
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6078363814371985,
+      "learning_rate": 9.750038773893504e-06,
+      "loss": 0.4997,
+      "step": 1032
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5506457096195898,
+      "learning_rate": 9.749410903084012e-06,
+      "loss": 0.5513,
+      "step": 1033
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9688749331319362,
+      "learning_rate": 9.748782264965358e-06,
+      "loss": 0.5313,
+      "step": 1034
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.383066967658014,
+      "learning_rate": 9.748152859639106e-06,
+      "loss": 0.5261,
+      "step": 1035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.39514071147194,
+      "learning_rate": 9.74752268720694e-06,
+      "loss": 0.4936,
+      "step": 1036
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8765304882173792,
+      "learning_rate": 9.746891747770674e-06,
+      "loss": 0.5348,
+      "step": 1037
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6546071258012456,
+      "learning_rate": 9.746260041432238e-06,
+      "loss": 0.5669,
+      "step": 1038
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8137595354780747,
+      "learning_rate": 9.745627568293692e-06,
+      "loss": 0.5026,
+      "step": 1039
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7045762249053962,
+      "learning_rate": 9.744994328457216e-06,
+      "loss": 0.5531,
+      "step": 1040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.227903087362228,
+      "learning_rate": 9.744360322025116e-06,
+      "loss": 0.5391,
+      "step": 1041
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.685844631800516,
+      "learning_rate": 9.743725549099821e-06,
+      "loss": 0.4648,
+      "step": 1042
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.569940061174605,
+      "learning_rate": 9.743090009783884e-06,
+      "loss": 0.4944,
+      "step": 1043
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.723142959449064,
+      "learning_rate": 9.742453704179984e-06,
+      "loss": 0.5125,
+      "step": 1044
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4912047218885114,
+      "learning_rate": 9.74181663239092e-06,
+      "loss": 0.4813,
+      "step": 1045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5235008521313596,
+      "learning_rate": 9.741178794519617e-06,
+      "loss": 0.4845,
+      "step": 1046
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.138918688090799,
+      "learning_rate": 9.740540190669123e-06,
+      "loss": 0.5212,
+      "step": 1047
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6402131120648562,
+      "learning_rate": 9.73990082094261e-06,
+      "loss": 0.5325,
+      "step": 1048
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3317651053778317,
+      "learning_rate": 9.739260685443373e-06,
+      "loss": 0.6036,
+      "step": 1049
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.870170924854555,
+      "learning_rate": 9.738619784274833e-06,
+      "loss": 0.547,
+      "step": 1050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.645815534813316,
+      "learning_rate": 9.737978117540535e-06,
+      "loss": 0.5063,
+      "step": 1051
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.392607147245962,
+      "learning_rate": 9.73733568534414e-06,
+      "loss": 0.5087,
+      "step": 1052
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5143165627174358,
+      "learning_rate": 9.736692487789445e-06,
+      "loss": 0.5406,
+      "step": 1053
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8633855334581346,
+      "learning_rate": 9.736048524980361e-06,
+      "loss": 0.5339,
+      "step": 1054
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4131426572227155,
+      "learning_rate": 9.735403797020927e-06,
+      "loss": 0.5163,
+      "step": 1055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7726441057919677,
+      "learning_rate": 9.734758304015304e-06,
+      "loss": 0.5178,
+      "step": 1056
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4664637428741596,
+      "learning_rate": 9.734112046067776e-06,
+      "loss": 0.5636,
+      "step": 1057
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.250829141697953,
+      "learning_rate": 9.733465023282752e-06,
+      "loss": 0.5415,
+      "step": 1058
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.034243227211706,
+      "learning_rate": 9.732817235764766e-06,
+      "loss": 0.5273,
+      "step": 1059
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4069863342768443,
+      "learning_rate": 9.732168683618473e-06,
+      "loss": 0.5037,
+      "step": 1060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9783336033315735,
+      "learning_rate": 9.731519366948649e-06,
+      "loss": 0.4826,
+      "step": 1061
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.083472842664245,
+      "learning_rate": 9.730869285860203e-06,
+      "loss": 0.539,
+      "step": 1062
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1097580935012274,
+      "learning_rate": 9.730218440458157e-06,
+      "loss": 0.5468,
+      "step": 1063
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.532998375527931,
+      "learning_rate": 9.729566830847662e-06,
+      "loss": 0.5339,
+      "step": 1064
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.236358533233571,
+      "learning_rate": 9.72891445713399e-06,
+      "loss": 0.526,
+      "step": 1065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6568477345833922,
+      "learning_rate": 9.728261319422541e-06,
+      "loss": 0.5168,
+      "step": 1066
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.524033186735707,
+      "learning_rate": 9.727607417818831e-06,
+      "loss": 0.5217,
+      "step": 1067
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.71213752658569,
+      "learning_rate": 9.726952752428509e-06,
+      "loss": 0.5378,
+      "step": 1068
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.898260673307282,
+      "learning_rate": 9.726297323357335e-06,
+      "loss": 0.5476,
+      "step": 1069
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4246224608948932,
+      "learning_rate": 9.725641130711205e-06,
+      "loss": 0.4839,
+      "step": 1070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7794987519402379,
+      "learning_rate": 9.724984174596129e-06,
+      "loss": 0.5076,
+      "step": 1071
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9061616591589945,
+      "learning_rate": 9.724326455118247e-06,
+      "loss": 0.4956,
+      "step": 1072
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.173745750118815,
+      "learning_rate": 9.723667972383816e-06,
+      "loss": 0.6017,
+      "step": 1073
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9698620736533579,
+      "learning_rate": 9.723008726499224e-06,
+      "loss": 0.5057,
+      "step": 1074
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6466460434953547,
+      "learning_rate": 9.722348717570974e-06,
+      "loss": 0.4908,
+      "step": 1075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8438028852879214,
+      "learning_rate": 9.7216879457057e-06,
+      "loss": 0.5895,
+      "step": 1076
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7790940990355575,
+      "learning_rate": 9.721026411010152e-06,
+      "loss": 0.5293,
+      "step": 1077
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5031660420904411,
+      "learning_rate": 9.720364113591209e-06,
+      "loss": 0.5265,
+      "step": 1078
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.536355883265839,
+      "learning_rate": 9.71970105355587e-06,
+      "loss": 0.5356,
+      "step": 1079
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7276800472528657,
+      "learning_rate": 9.719037231011258e-06,
+      "loss": 0.5216,
+      "step": 1080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5881462730538876,
+      "learning_rate": 9.71837264606462e-06,
+      "loss": 0.5356,
+      "step": 1081
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6318467478851473,
+      "learning_rate": 9.717707298823325e-06,
+      "loss": 0.5273,
+      "step": 1082
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9848151538178227,
+      "learning_rate": 9.717041189394865e-06,
+      "loss": 0.5267,
+      "step": 1083
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6819293422314283,
+      "learning_rate": 9.716374317886858e-06,
+      "loss": 0.5276,
+      "step": 1084
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8559724047291661,
+      "learning_rate": 9.715706684407042e-06,
+      "loss": 0.5191,
+      "step": 1085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5339688193120449,
+      "learning_rate": 9.715038289063278e-06,
+      "loss": 0.5334,
+      "step": 1086
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5166992172797564,
+      "learning_rate": 9.714369131963554e-06,
+      "loss": 0.4737,
+      "step": 1087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.134209639993174,
+      "learning_rate": 9.713699213215974e-06,
+      "loss": 0.5355,
+      "step": 1088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.96269880054436,
+      "learning_rate": 9.713028532928771e-06,
+      "loss": 0.5655,
+      "step": 1089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8661433241028522,
+      "learning_rate": 9.712357091210303e-06,
+      "loss": 0.5408,
+      "step": 1090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6968495452565269,
+      "learning_rate": 9.711684888169043e-06,
+      "loss": 0.5566,
+      "step": 1091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7839624583219784,
+      "learning_rate": 9.711011923913592e-06,
+      "loss": 0.4723,
+      "step": 1092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9629701859367892,
+      "learning_rate": 9.710338198552673e-06,
+      "loss": 0.5142,
+      "step": 1093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.614758849948944,
+      "learning_rate": 9.709663712195134e-06,
+      "loss": 0.5048,
+      "step": 1094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.667464215660333,
+      "learning_rate": 9.708988464949944e-06,
+      "loss": 0.5172,
+      "step": 1095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7918874492859533,
+      "learning_rate": 9.708312456926195e-06,
+      "loss": 0.5401,
+      "step": 1096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7004882321052495,
+      "learning_rate": 9.707635688233098e-06,
+      "loss": 0.5092,
+      "step": 1097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.701452526614193,
+      "learning_rate": 9.706958158979997e-06,
+      "loss": 0.5389,
+      "step": 1098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7636567128885436,
+      "learning_rate": 9.70627986927635e-06,
+      "loss": 0.6421,
+      "step": 1099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8276796485736944,
+      "learning_rate": 9.705600819231743e-06,
+      "loss": 0.5333,
+      "step": 1100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.246268565950378,
+      "learning_rate": 9.704921008955876e-06,
+      "loss": 0.5561,
+      "step": 1101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.438082199230945,
+      "learning_rate": 9.704240438558585e-06,
+      "loss": 0.569,
+      "step": 1102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9144248893253415,
+      "learning_rate": 9.70355910814982e-06,
+      "loss": 0.4746,
+      "step": 1103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9469849927167315,
+      "learning_rate": 9.702877017839656e-06,
+      "loss": 0.5467,
+      "step": 1104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.645061673320711,
+      "learning_rate": 9.70219416773829e-06,
+      "loss": 0.5401,
+      "step": 1105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7672589402175505,
+      "learning_rate": 9.701510557956041e-06,
+      "loss": 0.5233,
+      "step": 1106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6562526037884509,
+      "learning_rate": 9.700826188603358e-06,
+      "loss": 0.573,
+      "step": 1107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.1079545454951174,
+      "learning_rate": 9.700141059790801e-06,
+      "loss": 0.5971,
+      "step": 1108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.484603231624597,
+      "learning_rate": 9.699455171629063e-06,
+      "loss": 0.5241,
+      "step": 1109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5334514844206923,
+      "learning_rate": 9.69876852422895e-06,
+      "loss": 0.4631,
+      "step": 1110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.456284075939431,
+      "learning_rate": 9.698081117701399e-06,
+      "loss": 0.5309,
+      "step": 1111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1560014571082715,
+      "learning_rate": 9.697392952157467e-06,
+      "loss": 0.4962,
+      "step": 1112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6442324739481327,
+      "learning_rate": 9.696704027708332e-06,
+      "loss": 0.5049,
+      "step": 1113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7439283088142261,
+      "learning_rate": 9.6960143444653e-06,
+      "loss": 0.5787,
+      "step": 1114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8398447633635922,
+      "learning_rate": 9.695323902539787e-06,
+      "loss": 0.5191,
+      "step": 1115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6251227622842839,
+      "learning_rate": 9.694632702043347e-06,
+      "loss": 0.5029,
+      "step": 1116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.79714755581544,
+      "learning_rate": 9.693940743087647e-06,
+      "loss": 0.5115,
+      "step": 1117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.064944948011489,
+      "learning_rate": 9.693248025784481e-06,
+      "loss": 0.5948,
+      "step": 1118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.179435891042551,
+      "learning_rate": 9.692554550245759e-06,
+      "loss": 0.55,
+      "step": 1119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.815622421678713,
+      "learning_rate": 9.691860316583523e-06,
+      "loss": 0.4634,
+      "step": 1120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8473286275242924,
+      "learning_rate": 9.69116532490993e-06,
+      "loss": 0.4928,
+      "step": 1121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6666329256151406,
+      "learning_rate": 9.690469575337265e-06,
+      "loss": 0.5003,
+      "step": 1122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.5464761651711085,
+      "learning_rate": 9.689773067977927e-06,
+      "loss": 0.4924,
+      "step": 1123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.378718901137373,
+      "learning_rate": 9.689075802944447e-06,
+      "loss": 0.5337,
+      "step": 1124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8700837581036818,
+      "learning_rate": 9.688377780349475e-06,
+      "loss": 0.5231,
+      "step": 1125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7010701035393139,
+      "learning_rate": 9.687679000305779e-06,
+      "loss": 0.5754,
+      "step": 1126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.6688092521075686,
+      "learning_rate": 9.686979462926255e-06,
+      "loss": 0.5792,
+      "step": 1127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8813021532141243,
+      "learning_rate": 9.68627916832392e-06,
+      "loss": 0.6095,
+      "step": 1128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6581996476541911,
+      "learning_rate": 9.685578116611913e-06,
+      "loss": 0.5042,
+      "step": 1129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6582970311756564,
+      "learning_rate": 9.684876307903495e-06,
+      "loss": 0.4988,
+      "step": 1130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.683884028197148,
+      "learning_rate": 9.684173742312047e-06,
+      "loss": 0.4801,
+      "step": 1131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.588948092366968,
+      "learning_rate": 9.683470419951076e-06,
+      "loss": 0.5665,
+      "step": 1132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.310523817864352,
+      "learning_rate": 9.682766340934212e-06,
+      "loss": 0.5613,
+      "step": 1133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.195517805689883,
+      "learning_rate": 9.682061505375203e-06,
+      "loss": 0.5476,
+      "step": 1134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6244317571837932,
+      "learning_rate": 9.681355913387921e-06,
+      "loss": 0.5026,
+      "step": 1135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8925554979075394,
+      "learning_rate": 9.680649565086363e-06,
+      "loss": 0.5229,
+      "step": 1136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.632376881931433,
+      "learning_rate": 9.679942460584643e-06,
+      "loss": 0.5724,
+      "step": 1137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7020674525959631,
+      "learning_rate": 9.679234599997003e-06,
+      "loss": 0.4889,
+      "step": 1138
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5508086290219265,
+      "learning_rate": 9.6785259834378e-06,
+      "loss": 0.5535,
+      "step": 1139
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3720437102762606,
+      "learning_rate": 9.677816611021522e-06,
+      "loss": 0.5221,
+      "step": 1140
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6902927516019997,
+      "learning_rate": 9.67710648286277e-06,
+      "loss": 0.5167,
+      "step": 1141
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.430557957008761,
+      "learning_rate": 9.676395599076274e-06,
+      "loss": 0.4993,
+      "step": 1142
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8064802832707425,
+      "learning_rate": 9.675683959776883e-06,
+      "loss": 0.4976,
+      "step": 1143
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7692889961512195,
+      "learning_rate": 9.67497156507957e-06,
+      "loss": 0.6078,
+      "step": 1144
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7279284571335646,
+      "learning_rate": 9.674258415099424e-06,
+      "loss": 0.4917,
+      "step": 1145
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5671149456296096,
+      "learning_rate": 9.673544509951666e-06,
+      "loss": 0.5279,
+      "step": 1146
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3614047258344826,
+      "learning_rate": 9.672829849751633e-06,
+      "loss": 0.5778,
+      "step": 1147
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6605177543197422,
+      "learning_rate": 9.67211443461478e-06,
+      "loss": 0.5539,
+      "step": 1148
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.745652249143334,
+      "learning_rate": 9.671398264656693e-06,
+      "loss": 0.5044,
+      "step": 1149
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6403940468245404,
+      "learning_rate": 9.670681339993076e-06,
+      "loss": 0.5688,
+      "step": 1150
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3705504539528528,
+      "learning_rate": 9.669963660739749e-06,
+      "loss": 0.4915,
+      "step": 1151
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0052403405897046,
+      "learning_rate": 9.669245227012667e-06,
+      "loss": 0.5709,
+      "step": 1152
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.555499863113405,
+      "learning_rate": 9.668526038927895e-06,
+      "loss": 0.4766,
+      "step": 1153
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.531025831393635,
+      "learning_rate": 9.667806096601621e-06,
+      "loss": 0.4946,
+      "step": 1154
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0443285220787972,
+      "learning_rate": 9.667085400150167e-06,
+      "loss": 0.4962,
+      "step": 1155
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4424983450814588,
+      "learning_rate": 9.666363949689959e-06,
+      "loss": 0.4927,
+      "step": 1156
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.059446725914697,
+      "learning_rate": 9.665641745337558e-06,
+      "loss": 0.5343,
+      "step": 1157
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.0802914443470293,
+      "learning_rate": 9.664918787209643e-06,
+      "loss": 0.532,
+      "step": 1158
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6472887207129542,
+      "learning_rate": 9.664195075423011e-06,
+      "loss": 0.4588,
+      "step": 1159
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7292441767378381,
+      "learning_rate": 9.663470610094587e-06,
+      "loss": 0.5606,
+      "step": 1160
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.788898862305559,
+      "learning_rate": 9.662745391341415e-06,
+      "loss": 0.513,
+      "step": 1161
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5309024684748662,
+      "learning_rate": 9.662019419280659e-06,
+      "loss": 0.5694,
+      "step": 1162
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4995683508223077,
+      "learning_rate": 9.661292694029605e-06,
+      "loss": 0.5268,
+      "step": 1163
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7653072066574216,
+      "learning_rate": 9.660565215705664e-06,
+      "loss": 0.5659,
+      "step": 1164
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.963524493317451,
+      "learning_rate": 9.659836984426366e-06,
+      "loss": 0.4975,
+      "step": 1165
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6558752769707101,
+      "learning_rate": 9.65910800030936e-06,
+      "loss": 0.4612,
+      "step": 1166
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.569688065552612,
+      "learning_rate": 9.658378263472428e-06,
+      "loss": 0.5306,
+      "step": 1167
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.600377545926572,
+      "learning_rate": 9.657647774033456e-06,
+      "loss": 0.5281,
+      "step": 1168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.094413769681627,
+      "learning_rate": 9.656916532110468e-06,
+      "loss": 0.4827,
+      "step": 1169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4195357712238534,
+      "learning_rate": 9.656184537821598e-06,
+      "loss": 0.4979,
+      "step": 1170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5100123439164768,
+      "learning_rate": 9.655451791285108e-06,
+      "loss": 0.5261,
+      "step": 1171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.688940664076167,
+      "learning_rate": 9.65471829261938e-06,
+      "loss": 0.5062,
+      "step": 1172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5548426778604,
+      "learning_rate": 9.653984041942917e-06,
+      "loss": 0.5516,
+      "step": 1173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7100786513962827,
+      "learning_rate": 9.653249039374344e-06,
+      "loss": 0.5542,
+      "step": 1174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6241572009144818,
+      "learning_rate": 9.652513285032406e-06,
+      "loss": 0.4834,
+      "step": 1175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6386928401709666,
+      "learning_rate": 9.65177677903597e-06,
+      "loss": 0.5102,
+      "step": 1176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6867664454182212,
+      "learning_rate": 9.651039521504026e-06,
+      "loss": 0.5255,
+      "step": 1177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4139399172397975,
+      "learning_rate": 9.650301512555687e-06,
+      "loss": 0.5622,
+      "step": 1178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.094308688561519,
+      "learning_rate": 9.64956275231018e-06,
+      "loss": 0.4882,
+      "step": 1179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.865048965284112,
+      "learning_rate": 9.648823240886862e-06,
+      "loss": 0.5096,
+      "step": 1180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.494701863132983,
+      "learning_rate": 9.648082978405207e-06,
+      "loss": 0.5809,
+      "step": 1181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6793676392905069,
+      "learning_rate": 9.647341964984808e-06,
+      "loss": 0.5012,
+      "step": 1182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.213005790580897,
+      "learning_rate": 9.646600200745386e-06,
+      "loss": 0.5641,
+      "step": 1183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.695252467817391,
+      "learning_rate": 9.645857685806776e-06,
+      "loss": 0.4946,
+      "step": 1184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6872888667684667,
+      "learning_rate": 9.645114420288943e-06,
+      "loss": 0.5504,
+      "step": 1185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1474607546107864,
+      "learning_rate": 9.644370404311962e-06,
+      "loss": 0.5554,
+      "step": 1186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.65613332157712,
+      "learning_rate": 9.64362563799604e-06,
+      "loss": 0.5397,
+      "step": 1187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1984130027521513,
+      "learning_rate": 9.642880121461498e-06,
+      "loss": 0.5274,
+      "step": 1188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9276657533703743,
+      "learning_rate": 9.642133854828782e-06,
+      "loss": 0.5502,
+      "step": 1189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6981372675873604,
+      "learning_rate": 9.641386838218457e-06,
+      "loss": 0.471,
+      "step": 1190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.117793484644697,
+      "learning_rate": 9.640639071751211e-06,
+      "loss": 0.5068,
+      "step": 1191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.737088935356651,
+      "learning_rate": 9.639890555547851e-06,
+      "loss": 0.5434,
+      "step": 1192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.215032646390322,
+      "learning_rate": 9.639141289729308e-06,
+      "loss": 0.5475,
+      "step": 1193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.586242568006778,
+      "learning_rate": 9.638391274416631e-06,
+      "loss": 0.5123,
+      "step": 1194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7004319954959887,
+      "learning_rate": 9.637640509730994e-06,
+      "loss": 0.5152,
+      "step": 1195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.300052946091436,
+      "learning_rate": 9.636888995793688e-06,
+      "loss": 0.516,
+      "step": 1196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3913417939583823,
+      "learning_rate": 9.636136732726125e-06,
+      "loss": 0.5321,
+      "step": 1197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6905813420549474,
+      "learning_rate": 9.635383720649842e-06,
+      "loss": 0.5162,
+      "step": 1198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1986911451055486,
+      "learning_rate": 9.634629959686495e-06,
+      "loss": 0.5013,
+      "step": 1199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.310535302411778,
+      "learning_rate": 9.633875449957858e-06,
+      "loss": 0.535,
+      "step": 1200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8547606195531325,
+      "learning_rate": 9.633120191585831e-06,
+      "loss": 0.51,
+      "step": 1201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.613164494224369,
+      "learning_rate": 9.632364184692433e-06,
+      "loss": 0.5094,
+      "step": 1202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.85274118986579,
+      "learning_rate": 9.631607429399804e-06,
+      "loss": 0.5153,
+      "step": 1203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5773498083449433,
+      "learning_rate": 9.6308499258302e-06,
+      "loss": 0.5393,
+      "step": 1204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.010815599427557,
+      "learning_rate": 9.630091674106007e-06,
+      "loss": 0.5093,
+      "step": 1205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9770650666755865,
+      "learning_rate": 9.629332674349726e-06,
+      "loss": 0.4783,
+      "step": 1206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.437068958047755,
+      "learning_rate": 9.62857292668398e-06,
+      "loss": 0.5322,
+      "step": 1207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.74042908492775,
+      "learning_rate": 9.627812431231513e-06,
+      "loss": 0.5359,
+      "step": 1208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.01476278074651,
+      "learning_rate": 9.627051188115188e-06,
+      "loss": 0.5211,
+      "step": 1209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.8731844210119135,
+      "learning_rate": 9.626289197457994e-06,
+      "loss": 0.5703,
+      "step": 1210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3127041969062683,
+      "learning_rate": 9.625526459383036e-06,
+      "loss": 0.5378,
+      "step": 1211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2867488219449634,
+      "learning_rate": 9.62476297401354e-06,
+      "loss": 0.5546,
+      "step": 1212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5360675967832114,
+      "learning_rate": 9.623998741472853e-06,
+      "loss": 0.4797,
+      "step": 1213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0113735199718814,
+      "learning_rate": 9.623233761884445e-06,
+      "loss": 0.5743,
+      "step": 1214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.367006214996671,
+      "learning_rate": 9.622468035371905e-06,
+      "loss": 0.5069,
+      "step": 1215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.009616179978305,
+      "learning_rate": 9.621701562058945e-06,
+      "loss": 0.5268,
+      "step": 1216
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.7213470140986056,
+      "learning_rate": 9.620934342069391e-06,
+      "loss": 0.5008,
+      "step": 1217
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8416928579180722,
+      "learning_rate": 9.620166375527199e-06,
+      "loss": 0.5389,
+      "step": 1218
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5360768675242755,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.5154,
+      "step": 1219
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.666612843631181,
+      "learning_rate": 9.618628203281295e-06,
+      "loss": 0.5828,
+      "step": 1220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2125291127077475,
+      "learning_rate": 9.617857997826093e-06,
+      "loss": 0.5387,
+      "step": 1221
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.7546061712011785,
+      "learning_rate": 9.617087046315261e-06,
+      "loss": 0.4793,
+      "step": 1222
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.464835676400265,
+      "learning_rate": 9.616315348873351e-06,
+      "loss": 0.5424,
+      "step": 1223
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.164795916642207,
+      "learning_rate": 9.615542905625041e-06,
+      "loss": 0.5119,
+      "step": 1224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6484271639304557,
+      "learning_rate": 9.614769716695124e-06,
+      "loss": 0.5667,
+      "step": 1225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.150313277541799,
+      "learning_rate": 9.613995782208519e-06,
+      "loss": 0.5262,
+      "step": 1226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.321639794941608,
+      "learning_rate": 9.613221102290256e-06,
+      "loss": 0.4829,
+      "step": 1227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6141589182441907,
+      "learning_rate": 9.612445677065494e-06,
+      "loss": 0.5041,
+      "step": 1228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9570658209637812,
+      "learning_rate": 9.611669506659512e-06,
+      "loss": 0.5886,
+      "step": 1229
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8923861340904082,
+      "learning_rate": 9.610892591197702e-06,
+      "loss": 0.5248,
+      "step": 1230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.8984705127022097,
+      "learning_rate": 9.610114930805588e-06,
+      "loss": 0.5049,
+      "step": 1231
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4040263275304556,
+      "learning_rate": 9.609336525608804e-06,
+      "loss": 0.5476,
+      "step": 1232
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5329760895843587,
+      "learning_rate": 9.608557375733108e-06,
+      "loss": 0.5361,
+      "step": 1233
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4091340353572606,
+      "learning_rate": 9.607777481304378e-06,
+      "loss": 0.5241,
+      "step": 1234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.083298586043196,
+      "learning_rate": 9.606996842448617e-06,
+      "loss": 0.5274,
+      "step": 1235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.131902615128241,
+      "learning_rate": 9.60621545929194e-06,
+      "loss": 0.5581,
+      "step": 1236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7455567595837368,
+      "learning_rate": 9.605433331960589e-06,
+      "loss": 0.4963,
+      "step": 1237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2914875837784163,
+      "learning_rate": 9.60465046058092e-06,
+      "loss": 0.4893,
+      "step": 1238
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2680055610154803,
+      "learning_rate": 9.603866845279416e-06,
+      "loss": 0.547,
+      "step": 1239
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3527127987284717,
+      "learning_rate": 9.603082486182677e-06,
+      "loss": 0.5201,
+      "step": 1240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8932230576608493,
+      "learning_rate": 9.60229738341742e-06,
+      "loss": 0.5067,
+      "step": 1241
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.521453010399267,
+      "learning_rate": 9.601511537110488e-06,
+      "loss": 0.5039,
+      "step": 1242
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.902197545583407,
+      "learning_rate": 9.600724947388842e-06,
+      "loss": 0.5102,
+      "step": 1243
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8936990939314446,
+      "learning_rate": 9.59993761437956e-06,
+      "loss": 0.4766,
+      "step": 1244
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.818935057937169,
+      "learning_rate": 9.599149538209844e-06,
+      "loss": 0.4479,
+      "step": 1245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4841266281922247,
+      "learning_rate": 9.598360719007014e-06,
+      "loss": 0.5502,
+      "step": 1246
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6573299803407252,
+      "learning_rate": 9.597571156898512e-06,
+      "loss": 0.4948,
+      "step": 1247
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9694935567001643,
+      "learning_rate": 9.596780852011898e-06,
+      "loss": 0.5235,
+      "step": 1248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7258316103612048,
+      "learning_rate": 9.59598980447485e-06,
+      "loss": 0.5048,
+      "step": 1249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4025233679193496,
+      "learning_rate": 9.595198014415175e-06,
+      "loss": 0.4647,
+      "step": 1250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9544947136772115,
+      "learning_rate": 9.594405481960788e-06,
+      "loss": 0.5649,
+      "step": 1251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9777545383227084,
+      "learning_rate": 9.593612207239731e-06,
+      "loss": 0.5665,
+      "step": 1252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.117499822077094,
+      "learning_rate": 9.592818190380164e-06,
+      "loss": 0.5135,
+      "step": 1253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.076146244025416,
+      "learning_rate": 9.59202343151037e-06,
+      "loss": 0.5269,
+      "step": 1254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.223126647867816,
+      "learning_rate": 9.591227930758747e-06,
+      "loss": 0.5216,
+      "step": 1255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0328596241189216,
+      "learning_rate": 9.590431688253816e-06,
+      "loss": 0.5427,
+      "step": 1256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6556957352883486,
+      "learning_rate": 9.589634704124218e-06,
+      "loss": 0.5067,
+      "step": 1257
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.804364284260751,
+      "learning_rate": 9.58883697849871e-06,
+      "loss": 0.5287,
+      "step": 1258
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5480925200101965,
+      "learning_rate": 9.588038511506174e-06,
+      "loss": 0.45,
+      "step": 1259
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.730298843291153,
+      "learning_rate": 9.587239303275609e-06,
+      "loss": 0.5133,
+      "step": 1260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.908091031732794,
+      "learning_rate": 9.586439353936134e-06,
+      "loss": 0.4956,
+      "step": 1261
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.708475563781002,
+      "learning_rate": 9.585638663616988e-06,
+      "loss": 0.5518,
+      "step": 1262
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9115489553355347,
+      "learning_rate": 9.584837232447528e-06,
+      "loss": 0.484,
+      "step": 1263
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6690112469223255,
+      "learning_rate": 9.584035060557232e-06,
+      "loss": 0.558,
+      "step": 1264
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8645618150965557,
+      "learning_rate": 9.583232148075704e-06,
+      "loss": 0.5433,
+      "step": 1265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6006910031055772,
+      "learning_rate": 9.582428495132652e-06,
+      "loss": 0.5044,
+      "step": 1266
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.643665291170957,
+      "learning_rate": 9.58162410185792e-06,
+      "loss": 0.5421,
+      "step": 1267
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0755796639475856,
+      "learning_rate": 9.580818968381465e-06,
+      "loss": 0.5329,
+      "step": 1268
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4024091543968857,
+      "learning_rate": 9.580013094833358e-06,
+      "loss": 0.5128,
+      "step": 1269
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4212575894723896,
+      "learning_rate": 9.579206481343802e-06,
+      "loss": 0.4955,
+      "step": 1270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5678792635077088,
+      "learning_rate": 9.578399128043106e-06,
+      "loss": 0.4909,
+      "step": 1271
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.2412182086972967,
+      "learning_rate": 9.577591035061709e-06,
+      "loss": 0.4752,
+      "step": 1272
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0582740833905353,
+      "learning_rate": 9.576782202530164e-06,
+      "loss": 0.5094,
+      "step": 1273
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1595951744380097,
+      "learning_rate": 9.575972630579147e-06,
+      "loss": 0.4988,
+      "step": 1274
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7575760712452722,
+      "learning_rate": 9.575162319339448e-06,
+      "loss": 0.5597,
+      "step": 1275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1055777822631323,
+      "learning_rate": 9.574351268941982e-06,
+      "loss": 0.5209,
+      "step": 1276
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8901163453768444,
+      "learning_rate": 9.573539479517782e-06,
+      "loss": 0.5642,
+      "step": 1277
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.781553074512771,
+      "learning_rate": 9.572726951198e-06,
+      "loss": 0.5262,
+      "step": 1278
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9014220500004342,
+      "learning_rate": 9.571913684113905e-06,
+      "loss": 0.5311,
+      "step": 1279
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5394281191849384,
+      "learning_rate": 9.571099678396886e-06,
+      "loss": 0.5251,
+      "step": 1280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.749150285162678,
+      "learning_rate": 9.57028493417846e-06,
+      "loss": 0.558,
+      "step": 1281
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.194358414181551,
+      "learning_rate": 9.569469451590248e-06,
+      "loss": 0.5159,
+      "step": 1282
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3419010510041824,
+      "learning_rate": 9.568653230764003e-06,
+      "loss": 0.4754,
+      "step": 1283
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.60353026025585,
+      "learning_rate": 9.567836271831592e-06,
+      "loss": 0.5184,
+      "step": 1284
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.08991136852093,
+      "learning_rate": 9.567018574925e-06,
+      "loss": 0.5461,
+      "step": 1285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3027449836301668,
+      "learning_rate": 9.566200140176336e-06,
+      "loss": 0.5126,
+      "step": 1286
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.675273115727022,
+      "learning_rate": 9.565380967717824e-06,
+      "loss": 0.5403,
+      "step": 1287
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5043081964292935,
+      "learning_rate": 9.564561057681805e-06,
+      "loss": 0.4648,
+      "step": 1288
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3349823760180684,
+      "learning_rate": 9.56374041020075e-06,
+      "loss": 0.5427,
+      "step": 1289
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.211619642804284,
+      "learning_rate": 9.562919025407236e-06,
+      "loss": 0.5283,
+      "step": 1290
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.45557876575767,
+      "learning_rate": 9.562096903433968e-06,
+      "loss": 0.5412,
+      "step": 1291
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9953449011412303,
+      "learning_rate": 9.561274044413764e-06,
+      "loss": 0.5083,
+      "step": 1292
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.668342208345377,
+      "learning_rate": 9.560450448479567e-06,
+      "loss": 0.5149,
+      "step": 1293
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.170669371423883,
+      "learning_rate": 9.559626115764437e-06,
+      "loss": 0.5424,
+      "step": 1294
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.794732304031168,
+      "learning_rate": 9.558801046401547e-06,
+      "loss": 0.5584,
+      "step": 1295
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8864192744126171,
+      "learning_rate": 9.5579752405242e-06,
+      "loss": 0.5109,
+      "step": 1296
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9524520373710526,
+      "learning_rate": 9.55714869826581e-06,
+      "loss": 0.5622,
+      "step": 1297
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6924629447195305,
+      "learning_rate": 9.55632141975991e-06,
+      "loss": 0.5265,
+      "step": 1298
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5101536120787722,
+      "learning_rate": 9.555493405140158e-06,
+      "loss": 0.5426,
+      "step": 1299
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4822593254375265,
+      "learning_rate": 9.554664654540324e-06,
+      "loss": 0.5498,
+      "step": 1300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.1051335057769935,
+      "learning_rate": 9.553835168094302e-06,
+      "loss": 0.5347,
+      "step": 1301
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7821949833507682,
+      "learning_rate": 9.553004945936101e-06,
+      "loss": 0.5628,
+      "step": 1302
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7230312544660243,
+      "learning_rate": 9.552173988199854e-06,
+      "loss": 0.5047,
+      "step": 1303
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.525803056649358,
+      "learning_rate": 9.551342295019805e-06,
+      "loss": 0.5135,
+      "step": 1304
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6930037398549997,
+      "learning_rate": 9.550509866530323e-06,
+      "loss": 0.4996,
+      "step": 1305
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5104572978125144,
+      "learning_rate": 9.549676702865897e-06,
+      "loss": 0.5371,
+      "step": 1306
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8562563426485768,
+      "learning_rate": 9.54884280416113e-06,
+      "loss": 0.5612,
+      "step": 1307
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.926794889658595,
+      "learning_rate": 9.548008170550744e-06,
+      "loss": 0.4897,
+      "step": 1308
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.639247567223645,
+      "learning_rate": 9.547172802169582e-06,
+      "loss": 0.5497,
+      "step": 1309
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8091300839987694,
+      "learning_rate": 9.546336699152608e-06,
+      "loss": 0.5176,
+      "step": 1310
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9440447348284993,
+      "learning_rate": 9.545499861634897e-06,
+      "loss": 0.5783,
+      "step": 1311
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8076948729636986,
+      "learning_rate": 9.544662289751651e-06,
+      "loss": 0.5092,
+      "step": 1312
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.715210485185207,
+      "learning_rate": 9.543823983638187e-06,
+      "loss": 0.4853,
+      "step": 1313
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.513563901986961,
+      "learning_rate": 9.54298494342994e-06,
+      "loss": 0.5144,
+      "step": 1314
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3014768871227727,
+      "learning_rate": 9.542145169262465e-06,
+      "loss": 0.5047,
+      "step": 1315
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8557169997552467,
+      "learning_rate": 9.541304661271433e-06,
+      "loss": 0.5442,
+      "step": 1316
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.638193691189023,
+      "learning_rate": 9.540463419592638e-06,
+      "loss": 0.5737,
+      "step": 1317
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4331612514067824,
+      "learning_rate": 9.539621444361988e-06,
+      "loss": 0.4952,
+      "step": 1318
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.814975821494965,
+      "learning_rate": 9.538778735715512e-06,
+      "loss": 0.5214,
+      "step": 1319
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6358206258929424,
+      "learning_rate": 9.537935293789357e-06,
+      "loss": 0.5852,
+      "step": 1320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.785035433589641,
+      "learning_rate": 9.53709111871979e-06,
+      "loss": 0.4918,
+      "step": 1321
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7894871619403805,
+      "learning_rate": 9.536246210643192e-06,
+      "loss": 0.4917,
+      "step": 1322
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6960296087976123,
+      "learning_rate": 9.535400569696068e-06,
+      "loss": 0.4808,
+      "step": 1323
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6707368064014259,
+      "learning_rate": 9.534554196015038e-06,
+      "loss": 0.5208,
+      "step": 1324
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.418989617577566,
+      "learning_rate": 9.53370708973684e-06,
+      "loss": 0.5419,
+      "step": 1325
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9542552706872764,
+      "learning_rate": 9.532859250998332e-06,
+      "loss": 0.5268,
+      "step": 1326
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.109108500431418,
+      "learning_rate": 9.532010679936491e-06,
+      "loss": 0.5321,
+      "step": 1327
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.40541389320721,
+      "learning_rate": 9.53116137668841e-06,
+      "loss": 0.5123,
+      "step": 1328
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.120180365123593,
+      "learning_rate": 9.530311341391303e-06,
+      "loss": 0.5523,
+      "step": 1329
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.837416018100702,
+      "learning_rate": 9.529460574182498e-06,
+      "loss": 0.5293,
+      "step": 1330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8130170198172144,
+      "learning_rate": 9.528609075199445e-06,
+      "loss": 0.5078,
+      "step": 1331
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5671538001377203,
+      "learning_rate": 9.527756844579711e-06,
+      "loss": 0.5376,
+      "step": 1332
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7851923640009364,
+      "learning_rate": 9.526903882460983e-06,
+      "loss": 0.5495,
+      "step": 1333
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.058190778408725,
+      "learning_rate": 9.526050188981064e-06,
+      "loss": 0.488,
+      "step": 1334
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9634306303556182,
+      "learning_rate": 9.525195764277874e-06,
+      "loss": 0.5185,
+      "step": 1335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.032739483474685,
+      "learning_rate": 9.524340608489454e-06,
+      "loss": 0.5238,
+      "step": 1336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.412786853743459,
+      "learning_rate": 9.523484721753961e-06,
+      "loss": 0.4834,
+      "step": 1337
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7243055388004805,
+      "learning_rate": 9.522628104209675e-06,
+      "loss": 0.5075,
+      "step": 1338
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4597665228729926,
+      "learning_rate": 9.521770755994983e-06,
+      "loss": 0.5515,
+      "step": 1339
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.616644720048126,
+      "learning_rate": 9.520912677248403e-06,
+      "loss": 0.5475,
+      "step": 1340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5366952523712116,
+      "learning_rate": 9.520053868108566e-06,
+      "loss": 0.4744,
+      "step": 1341
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.09729427720111,
+      "learning_rate": 9.519194328714214e-06,
+      "loss": 0.5455,
+      "step": 1342
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.556155394623956,
+      "learning_rate": 9.518334059204218e-06,
+      "loss": 0.5228,
+      "step": 1343
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2820341211515673,
+      "learning_rate": 9.517473059717559e-06,
+      "loss": 0.5088,
+      "step": 1344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7841649432175868,
+      "learning_rate": 9.516611330393343e-06,
+      "loss": 0.499,
+      "step": 1345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3170294309282338,
+      "learning_rate": 9.515748871370786e-06,
+      "loss": 0.5223,
+      "step": 1346
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7675365545954447,
+      "learning_rate": 9.51488568278923e-06,
+      "loss": 0.5589,
+      "step": 1347
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.876530940821112,
+      "learning_rate": 9.514021764788127e-06,
+      "loss": 0.5056,
+      "step": 1348
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8763300559518927,
+      "learning_rate": 9.513157117507053e-06,
+      "loss": 0.5009,
+      "step": 1349
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.299660676483901,
+      "learning_rate": 9.512291741085696e-06,
+      "loss": 0.4943,
+      "step": 1350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8836219482182772,
+      "learning_rate": 9.51142563566387e-06,
+      "loss": 0.5369,
+      "step": 1351
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7009943354735515,
+      "learning_rate": 9.510558801381497e-06,
+      "loss": 0.5177,
+      "step": 1352
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.007229156877643,
+      "learning_rate": 9.509691238378626e-06,
+      "loss": 0.5089,
+      "step": 1353
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4387259770576408,
+      "learning_rate": 9.508822946795417e-06,
+      "loss": 0.4878,
+      "step": 1354
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1516983884115723,
+      "learning_rate": 9.507953926772152e-06,
+      "loss": 0.5281,
+      "step": 1355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8865060786578052,
+      "learning_rate": 9.507084178449226e-06,
+      "loss": 0.5245,
+      "step": 1356
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8621271424152859,
+      "learning_rate": 9.506213701967157e-06,
+      "loss": 0.5745,
+      "step": 1357
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6162835398610804,
+      "learning_rate": 9.505342497466577e-06,
+      "loss": 0.5083,
+      "step": 1358
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9539800242472292,
+      "learning_rate": 9.504470565088237e-06,
+      "loss": 0.5211,
+      "step": 1359
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8163097041032972,
+      "learning_rate": 9.503597904973006e-06,
+      "loss": 0.5399,
+      "step": 1360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6481196635818798,
+      "learning_rate": 9.502724517261867e-06,
+      "loss": 0.4951,
+      "step": 1361
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7434840971002281,
+      "learning_rate": 9.50185040209593e-06,
+      "loss": 0.4897,
+      "step": 1362
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.938999967638018,
+      "learning_rate": 9.500975559616407e-06,
+      "loss": 0.5322,
+      "step": 1363
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5166896975682522,
+      "learning_rate": 9.500099989964644e-06,
+      "loss": 0.4906,
+      "step": 1364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6493103354228393,
+      "learning_rate": 9.499223693282095e-06,
+      "loss": 0.4824,
+      "step": 1365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5580997421435645,
+      "learning_rate": 9.498346669710331e-06,
+      "loss": 0.5484,
+      "step": 1366
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5722533449608065,
+      "learning_rate": 9.497468919391046e-06,
+      "loss": 0.5286,
+      "step": 1367
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1379581735023523,
+      "learning_rate": 9.496590442466045e-06,
+      "loss": 0.5285,
+      "step": 1368
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.004542983588922,
+      "learning_rate": 9.495711239077256e-06,
+      "loss": 0.5383,
+      "step": 1369
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.723423923699892,
+      "learning_rate": 9.494831309366723e-06,
+      "loss": 0.5023,
+      "step": 1370
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.294621546390402,
+      "learning_rate": 9.493950653476604e-06,
+      "loss": 0.4739,
+      "step": 1371
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8962630752434209,
+      "learning_rate": 9.493069271549179e-06,
+      "loss": 0.5278,
+      "step": 1372
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2888590584106818,
+      "learning_rate": 9.49218716372684e-06,
+      "loss": 0.4836,
+      "step": 1373
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7787949561950795,
+      "learning_rate": 9.491304330152102e-06,
+      "loss": 0.5112,
+      "step": 1374
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1784795200831315,
+      "learning_rate": 9.490420770967594e-06,
+      "loss": 0.5267,
+      "step": 1375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6375490466954763,
+      "learning_rate": 9.489536486316062e-06,
+      "loss": 0.5242,
+      "step": 1376
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.477956447346823,
+      "learning_rate": 9.48865147634037e-06,
+      "loss": 0.5442,
+      "step": 1377
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.409881186281856,
+      "learning_rate": 9.487765741183499e-06,
+      "loss": 0.4991,
+      "step": 1378
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6800854486941126,
+      "learning_rate": 9.486879280988548e-06,
+      "loss": 0.5077,
+      "step": 1379
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9714553284132872,
+      "learning_rate": 9.485992095898734e-06,
+      "loss": 0.4876,
+      "step": 1380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5844507509035346,
+      "learning_rate": 9.485104186057386e-06,
+      "loss": 0.4705,
+      "step": 1381
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4796161279871252,
+      "learning_rate": 9.484215551607956e-06,
+      "loss": 0.4945,
+      "step": 1382
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3891117518032612,
+      "learning_rate": 9.48332619269401e-06,
+      "loss": 0.5054,
+      "step": 1383
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7929379067706905,
+      "learning_rate": 9.482436109459231e-06,
+      "loss": 0.5061,
+      "step": 1384
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6640724080747038,
+      "learning_rate": 9.481545302047423e-06,
+      "loss": 0.502,
+      "step": 1385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5469445671131699,
+      "learning_rate": 9.480653770602502e-06,
+      "loss": 0.5031,
+      "step": 1386
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.356349412795799,
+      "learning_rate": 9.479761515268499e-06,
+      "loss": 0.5271,
+      "step": 1387
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.44417765309099,
+      "learning_rate": 9.478868536189571e-06,
+      "loss": 0.4856,
+      "step": 1388
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.332946182241321,
+      "learning_rate": 9.477974833509984e-06,
+      "loss": 0.5198,
+      "step": 1389
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7290837012935454,
+      "learning_rate": 9.477080407374124e-06,
+      "loss": 0.4916,
+      "step": 1390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.4563147780501335,
+      "learning_rate": 9.476185257926496e-06,
+      "loss": 0.5489,
+      "step": 1391
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5845936119723913,
+      "learning_rate": 9.475289385311714e-06,
+      "loss": 0.5404,
+      "step": 1392
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6851537462201612,
+      "learning_rate": 9.474392789674517e-06,
+      "loss": 0.4969,
+      "step": 1393
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2924924777343387,
+      "learning_rate": 9.473495471159759e-06,
+      "loss": 0.4973,
+      "step": 1394
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8189207932754312,
+      "learning_rate": 9.472597429912409e-06,
+      "loss": 0.5301,
+      "step": 1395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9522121006208017,
+      "learning_rate": 9.471698666077554e-06,
+      "loss": 0.5599,
+      "step": 1396
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.654001619422945,
+      "learning_rate": 9.470799179800393e-06,
+      "loss": 0.513,
+      "step": 1397
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4693946073677817,
+      "learning_rate": 9.469898971226251e-06,
+      "loss": 0.4943,
+      "step": 1398
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.352280469067633,
+      "learning_rate": 9.468998040500563e-06,
+      "loss": 0.4851,
+      "step": 1399
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.713410517258502,
+      "learning_rate": 9.468096387768882e-06,
+      "loss": 0.4895,
+      "step": 1400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6359837268312648,
+      "learning_rate": 9.467194013176878e-06,
+      "loss": 0.4622,
+      "step": 1401
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9106064409196937,
+      "learning_rate": 9.466290916870338e-06,
+      "loss": 0.5316,
+      "step": 1402
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5652596983399456,
+      "learning_rate": 9.465387098995165e-06,
+      "loss": 0.524,
+      "step": 1403
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4641608430759605,
+      "learning_rate": 9.464482559697377e-06,
+      "loss": 0.5306,
+      "step": 1404
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.365277486999446,
+      "learning_rate": 9.463577299123113e-06,
+      "loss": 0.4959,
+      "step": 1405
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3531053244732796,
+      "learning_rate": 9.462671317418625e-06,
+      "loss": 0.4777,
+      "step": 1406
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.529796907855802,
+      "learning_rate": 9.461764614730282e-06,
+      "loss": 0.5073,
+      "step": 1407
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6037050306107454,
+      "learning_rate": 9.46085719120457e-06,
+      "loss": 0.5345,
+      "step": 1408
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.184384273671617,
+      "learning_rate": 9.459949046988089e-06,
+      "loss": 0.5283,
+      "step": 1409
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.971143226559445,
+      "learning_rate": 9.459040182227561e-06,
+      "loss": 0.4918,
+      "step": 1410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6162454601784557,
+      "learning_rate": 9.458130597069818e-06,
+      "loss": 0.5123,
+      "step": 1411
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.187548515299508,
+      "learning_rate": 9.457220291661817e-06,
+      "loss": 0.6018,
+      "step": 1412
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.417545035108629,
+      "learning_rate": 9.456309266150621e-06,
+      "loss": 0.5043,
+      "step": 1413
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6683483655573892,
+      "learning_rate": 9.455397520683414e-06,
+      "loss": 0.5066,
+      "step": 1414
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8267071607296352,
+      "learning_rate": 9.454485055407498e-06,
+      "loss": 0.5395,
+      "step": 1415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2769052311188753,
+      "learning_rate": 9.45357187047029e-06,
+      "loss": 0.5016,
+      "step": 1416
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5160384864380638,
+      "learning_rate": 9.452657966019324e-06,
+      "loss": 0.5127,
+      "step": 1417
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5904869300889308,
+      "learning_rate": 9.451743342202248e-06,
+      "loss": 0.5364,
+      "step": 1418
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6504050121743716,
+      "learning_rate": 9.450827999166825e-06,
+      "loss": 0.5016,
+      "step": 1419
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5705800991617183,
+      "learning_rate": 9.449911937060943e-06,
+      "loss": 0.5539,
+      "step": 1420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4645945472912039,
+      "learning_rate": 9.448995156032595e-06,
+      "loss": 0.5034,
+      "step": 1421
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4074577541628757,
+      "learning_rate": 9.448077656229895e-06,
+      "loss": 0.5257,
+      "step": 1422
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6932730921411916,
+      "learning_rate": 9.447159437801074e-06,
+      "loss": 0.525,
+      "step": 1423
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.508238112135693,
+      "learning_rate": 9.44624050089448e-06,
+      "loss": 0.5297,
+      "step": 1424
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.386080118480711,
+      "learning_rate": 9.445320845658574e-06,
+      "loss": 0.4918,
+      "step": 1425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7120372086794187,
+      "learning_rate": 9.444400472241934e-06,
+      "loss": 0.4962,
+      "step": 1426
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2643237638504257,
+      "learning_rate": 9.443479380793256e-06,
+      "loss": 0.5,
+      "step": 1427
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6849918060542648,
+      "learning_rate": 9.44255757146135e-06,
+      "loss": 0.5596,
+      "step": 1428
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7220988937001047,
+      "learning_rate": 9.44163504439514e-06,
+      "loss": 0.5398,
+      "step": 1429
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7402435053337213,
+      "learning_rate": 9.44071179974367e-06,
+      "loss": 0.5265,
+      "step": 1430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7330985961158227,
+      "learning_rate": 9.4397878376561e-06,
+      "loss": 0.4891,
+      "step": 1431
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9559886571410015,
+      "learning_rate": 9.438863158281702e-06,
+      "loss": 0.5233,
+      "step": 1432
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6463205429913006,
+      "learning_rate": 9.437937761769867e-06,
+      "loss": 0.5349,
+      "step": 1433
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6663696669907717,
+      "learning_rate": 9.4370116482701e-06,
+      "loss": 0.5131,
+      "step": 1434
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7039126134585667,
+      "learning_rate": 9.436084817932023e-06,
+      "loss": 0.537,
+      "step": 1435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5876792100322659,
+      "learning_rate": 9.435157270905375e-06,
+      "loss": 0.5268,
+      "step": 1436
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4258461786668768,
+      "learning_rate": 9.434229007340008e-06,
+      "loss": 0.4963,
+      "step": 1437
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.899112215769487,
+      "learning_rate": 9.433300027385891e-06,
+      "loss": 0.5093,
+      "step": 1438
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5966334451115977,
+      "learning_rate": 9.432370331193112e-06,
+      "loss": 0.506,
+      "step": 1439
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2001297668171955,
+      "learning_rate": 9.43143991891187e-06,
+      "loss": 0.5542,
+      "step": 1440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5979829933620182,
+      "learning_rate": 9.430508790692477e-06,
+      "loss": 0.5353,
+      "step": 1441
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3938257090441009,
+      "learning_rate": 9.429576946685369e-06,
+      "loss": 0.5508,
+      "step": 1442
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.887500531029625,
+      "learning_rate": 9.428644387041094e-06,
+      "loss": 0.5216,
+      "step": 1443
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.830904789433637,
+      "learning_rate": 9.427711111910314e-06,
+      "loss": 0.559,
+      "step": 1444
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4963688236773471,
+      "learning_rate": 9.42677712144381e-06,
+      "loss": 0.5282,
+      "step": 1445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4798735024063236,
+      "learning_rate": 9.42584241579247e-06,
+      "loss": 0.5549,
+      "step": 1446
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2502644942892267,
+      "learning_rate": 9.424906995107312e-06,
+      "loss": 0.541,
+      "step": 1447
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3223247871109725,
+      "learning_rate": 9.423970859539456e-06,
+      "loss": 0.4921,
+      "step": 1448
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.891800694340005,
+      "learning_rate": 9.423034009240146e-06,
+      "loss": 0.4867,
+      "step": 1449
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5131892594414695,
+      "learning_rate": 9.422096444360736e-06,
+      "loss": 0.5101,
+      "step": 1450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4027176765940803,
+      "learning_rate": 9.4211581650527e-06,
+      "loss": 0.4721,
+      "step": 1451
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.05923787696074,
+      "learning_rate": 9.420219171467624e-06,
+      "loss": 0.5484,
+      "step": 1452
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6191945144015785,
+      "learning_rate": 9.419279463757212e-06,
+      "loss": 0.5314,
+      "step": 1453
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5430874657099605,
+      "learning_rate": 9.41833904207328e-06,
+      "loss": 0.5428,
+      "step": 1454
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.544150626032831,
+      "learning_rate": 9.417397906567762e-06,
+      "loss": 0.5192,
+      "step": 1455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.828887008965527,
+      "learning_rate": 9.41645605739271e-06,
+      "loss": 0.5674,
+      "step": 1456
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.446962550807306,
+      "learning_rate": 9.415513494700281e-06,
+      "loss": 0.5534,
+      "step": 1457
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.409505263960533,
+      "learning_rate": 9.414570218642762e-06,
+      "loss": 0.5118,
+      "step": 1458
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9473100463811759,
+      "learning_rate": 9.413626229372543e-06,
+      "loss": 0.5302,
+      "step": 1459
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0115230441559584,
+      "learning_rate": 9.412681527042135e-06,
+      "loss": 0.5832,
+      "step": 1460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.504650850157876,
+      "learning_rate": 9.411736111804161e-06,
+      "loss": 0.5084,
+      "step": 1461
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5976807919316702,
+      "learning_rate": 9.410789983811366e-06,
+      "loss": 0.5124,
+      "step": 1462
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7259305630841915,
+      "learning_rate": 9.409843143216602e-06,
+      "loss": 0.5163,
+      "step": 1463
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3109812844968418,
+      "learning_rate": 9.408895590172837e-06,
+      "loss": 0.5135,
+      "step": 1464
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.501449859949129,
+      "learning_rate": 9.407947324833161e-06,
+      "loss": 0.5308,
+      "step": 1465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8953543710540945,
+      "learning_rate": 9.406998347350774e-06,
+      "loss": 0.4843,
+      "step": 1466
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7027642277487309,
+      "learning_rate": 9.40604865787899e-06,
+      "loss": 0.4771,
+      "step": 1467
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7233924413310757,
+      "learning_rate": 9.40509825657124e-06,
+      "loss": 0.5098,
+      "step": 1468
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7107652884518568,
+      "learning_rate": 9.40414714358107e-06,
+      "loss": 0.5001,
+      "step": 1469
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.582631591668211,
+      "learning_rate": 9.403195319062142e-06,
+      "loss": 0.5248,
+      "step": 1470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.387342543665346,
+      "learning_rate": 9.402242783168228e-06,
+      "loss": 0.5538,
+      "step": 1471
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.672928018604899,
+      "learning_rate": 9.401289536053223e-06,
+      "loss": 0.5351,
+      "step": 1472
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9915864472791334,
+      "learning_rate": 9.400335577871128e-06,
+      "loss": 0.5635,
+      "step": 1473
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6520622595751824,
+      "learning_rate": 9.399380908776068e-06,
+      "loss": 0.5298,
+      "step": 1474
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.931210566761977,
+      "learning_rate": 9.398425528922275e-06,
+      "loss": 0.558,
+      "step": 1475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.7686227437021595,
+      "learning_rate": 9.3974694384641e-06,
+      "loss": 0.5278,
+      "step": 1476
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5142014550507987,
+      "learning_rate": 9.396512637556007e-06,
+      "loss": 0.4852,
+      "step": 1477
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.121450635344311,
+      "learning_rate": 9.395555126352576e-06,
+      "loss": 0.5849,
+      "step": 1478
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9456554463436584,
+      "learning_rate": 9.394596905008504e-06,
+      "loss": 0.5376,
+      "step": 1479
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.697792582575253,
+      "learning_rate": 9.393637973678595e-06,
+      "loss": 0.5283,
+      "step": 1480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.311532952026007,
+      "learning_rate": 9.392678332517778e-06,
+      "loss": 0.5285,
+      "step": 1481
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6887015599370656,
+      "learning_rate": 9.391717981681089e-06,
+      "loss": 0.5606,
+      "step": 1482
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8749079141161675,
+      "learning_rate": 9.39075692132368e-06,
+      "loss": 0.5518,
+      "step": 1483
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5186622420467013,
+      "learning_rate": 9.38979515160082e-06,
+      "loss": 0.5646,
+      "step": 1484
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5302094559262753,
+      "learning_rate": 9.388832672667893e-06,
+      "loss": 0.5038,
+      "step": 1485
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7591373713642573,
+      "learning_rate": 9.387869484680395e-06,
+      "loss": 0.5247,
+      "step": 1486
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9377967760499746,
+      "learning_rate": 9.386905587793933e-06,
+      "loss": 0.5463,
+      "step": 1487
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3403669028777074,
+      "learning_rate": 9.385940982164239e-06,
+      "loss": 0.5146,
+      "step": 1488
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.811749023747539,
+      "learning_rate": 9.384975667947152e-06,
+      "loss": 0.5511,
+      "step": 1489
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.577574886181875,
+      "learning_rate": 9.384009645298627e-06,
+      "loss": 0.5508,
+      "step": 1490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.301454623132674,
+      "learning_rate": 9.383042914374731e-06,
+      "loss": 0.5493,
+      "step": 1491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4533170252188423,
+      "learning_rate": 9.382075475331652e-06,
+      "loss": 0.5466,
+      "step": 1492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5421032500320355,
+      "learning_rate": 9.381107328325683e-06,
+      "loss": 0.52,
+      "step": 1493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.352474522527251,
+      "learning_rate": 9.380138473513241e-06,
+      "loss": 0.4819,
+      "step": 1494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3336804024823565,
+      "learning_rate": 9.379168911050853e-06,
+      "loss": 0.5604,
+      "step": 1495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.450935676102443,
+      "learning_rate": 9.378198641095159e-06,
+      "loss": 0.5465,
+      "step": 1496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8670887608442013,
+      "learning_rate": 9.377227663802913e-06,
+      "loss": 0.5355,
+      "step": 1497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8171098999218793,
+      "learning_rate": 9.376255979330988e-06,
+      "loss": 0.5302,
+      "step": 1498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2358050289181253,
+      "learning_rate": 9.375283587836368e-06,
+      "loss": 0.5323,
+      "step": 1499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.586997073057225,
+      "learning_rate": 9.374310489476149e-06,
+      "loss": 0.5512,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.55579447130064,
+      "learning_rate": 9.373336684407545e-06,
+      "loss": 0.5447,
+      "step": 1501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6129422186439102,
+      "learning_rate": 9.372362172787882e-06,
+      "loss": 0.4549,
+      "step": 1502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.055117530737087,
+      "learning_rate": 9.371386954774603e-06,
+      "loss": 0.5521,
+      "step": 1503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6137472317864316,
+      "learning_rate": 9.370411030525261e-06,
+      "loss": 0.4941,
+      "step": 1504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6825693446935227,
+      "learning_rate": 9.369434400197526e-06,
+      "loss": 0.4765,
+      "step": 1505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0047785918869723,
+      "learning_rate": 9.36845706394918e-06,
+      "loss": 0.4718,
+      "step": 1506
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.184404353395833,
+      "learning_rate": 9.367479021938123e-06,
+      "loss": 0.5639,
+      "step": 1507
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.128493573070691,
+      "learning_rate": 9.366500274322365e-06,
+      "loss": 0.5322,
+      "step": 1508
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2077003304423544,
+      "learning_rate": 9.36552082126003e-06,
+      "loss": 0.5479,
+      "step": 1509
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5001203755032226,
+      "learning_rate": 9.364540662909358e-06,
+      "loss": 0.5705,
+      "step": 1510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.706994629835939,
+      "learning_rate": 9.363559799428704e-06,
+      "loss": 0.5147,
+      "step": 1511
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9071940375807923,
+      "learning_rate": 9.362578230976532e-06,
+      "loss": 0.5358,
+      "step": 1512
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.199072784364784,
+      "learning_rate": 9.361595957711425e-06,
+      "loss": 0.5274,
+      "step": 1513
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0735830354933347,
+      "learning_rate": 9.360612979792078e-06,
+      "loss": 0.5311,
+      "step": 1514
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2066166573545036,
+      "learning_rate": 9.3596292973773e-06,
+      "loss": 0.4923,
+      "step": 1515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4159872038811874,
+      "learning_rate": 9.358644910626012e-06,
+      "loss": 0.54,
+      "step": 1516
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.479018280513656,
+      "learning_rate": 9.35765981969725e-06,
+      "loss": 0.5287,
+      "step": 1517
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.089308304624241,
+      "learning_rate": 9.356674024750166e-06,
+      "loss": 0.5309,
+      "step": 1518
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2037636350371397,
+      "learning_rate": 9.355687525944025e-06,
+      "loss": 0.5754,
+      "step": 1519
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3408652508775556,
+      "learning_rate": 9.3547003234382e-06,
+      "loss": 0.4907,
+      "step": 1520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.913428552260077,
+      "learning_rate": 9.353712417392186e-06,
+      "loss": 0.5455,
+      "step": 1521
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6113559679228633,
+      "learning_rate": 9.352723807965586e-06,
+      "loss": 0.5563,
+      "step": 1522
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8314547370094012,
+      "learning_rate": 9.35173449531812e-06,
+      "loss": 0.4734,
+      "step": 1523
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0034022703098637,
+      "learning_rate": 9.350744479609622e-06,
+      "loss": 0.4954,
+      "step": 1524
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9027907885598594,
+      "learning_rate": 9.349753761000034e-06,
+      "loss": 0.5279,
+      "step": 1525
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.928394262689405,
+      "learning_rate": 9.348762339649416e-06,
+      "loss": 0.5482,
+      "step": 1526
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.638848020410885,
+      "learning_rate": 9.347770215717941e-06,
+      "loss": 0.6168,
+      "step": 1527
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9437404774526785,
+      "learning_rate": 9.346777389365896e-06,
+      "loss": 0.5453,
+      "step": 1528
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.581529142975484,
+      "learning_rate": 9.345783860753681e-06,
+      "loss": 0.5647,
+      "step": 1529
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 32.45158123886623,
+      "learning_rate": 9.344789630041811e-06,
+      "loss": 0.5425,
+      "step": 1530
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4210339557850866,
+      "learning_rate": 9.343794697390908e-06,
+      "loss": 0.5446,
+      "step": 1531
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0194607726636917,
+      "learning_rate": 9.342799062961716e-06,
+      "loss": 0.5179,
+      "step": 1532
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.666606781187195,
+      "learning_rate": 9.341802726915088e-06,
+      "loss": 0.5297,
+      "step": 1533
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.596514276782259,
+      "learning_rate": 9.340805689411989e-06,
+      "loss": 0.4748,
+      "step": 1534
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.127376711598774,
+      "learning_rate": 9.339807950613502e-06,
+      "loss": 0.5213,
+      "step": 1535
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.213578919065974,
+      "learning_rate": 9.338809510680818e-06,
+      "loss": 0.4784,
+      "step": 1536
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.6406784786202318,
+      "learning_rate": 9.337810369775245e-06,
+      "loss": 0.5167,
+      "step": 1537
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.710331055507111,
+      "learning_rate": 9.336810528058202e-06,
+      "loss": 0.513,
+      "step": 1538
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1709666221921617,
+      "learning_rate": 9.335809985691224e-06,
+      "loss": 0.5563,
+      "step": 1539
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.128357963800714,
+      "learning_rate": 9.334808742835956e-06,
+      "loss": 0.5162,
+      "step": 1540
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.18402549553364,
+      "learning_rate": 9.33380679965416e-06,
+      "loss": 0.5271,
+      "step": 1541
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.855067283290758,
+      "learning_rate": 9.332804156307705e-06,
+      "loss": 0.4588,
+      "step": 1542
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7396959013303646,
+      "learning_rate": 9.331800812958576e-06,
+      "loss": 0.4747,
+      "step": 1543
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2383869045867417,
+      "learning_rate": 9.330796769768876e-06,
+      "loss": 0.4842,
+      "step": 1544
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.501740378958404,
+      "learning_rate": 9.32979202690082e-06,
+      "loss": 0.5378,
+      "step": 1545
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5750775887331003,
+      "learning_rate": 9.328786584516725e-06,
+      "loss": 0.5284,
+      "step": 1546
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8052053950979472,
+      "learning_rate": 9.327780442779032e-06,
+      "loss": 0.4851,
+      "step": 1547
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.419443490992145,
+      "learning_rate": 9.326773601850294e-06,
+      "loss": 0.4742,
+      "step": 1548
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.79993293703367,
+      "learning_rate": 9.325766061893174e-06,
+      "loss": 0.5099,
+      "step": 1549
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9208065957405682,
+      "learning_rate": 9.324757823070448e-06,
+      "loss": 0.4399,
+      "step": 1550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.868678691846867,
+      "learning_rate": 9.323748885545006e-06,
+      "loss": 0.5128,
+      "step": 1551
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.714265342211101,
+      "learning_rate": 9.322739249479853e-06,
+      "loss": 0.5015,
+      "step": 1552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9555294131012604,
+      "learning_rate": 9.321728915038101e-06,
+      "loss": 0.4642,
+      "step": 1553
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.956602369828446,
+      "learning_rate": 9.320717882382983e-06,
+      "loss": 0.5373,
+      "step": 1554
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0660493606510353,
+      "learning_rate": 9.319706151677837e-06,
+      "loss": 0.5311,
+      "step": 1555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6164649314832973,
+      "learning_rate": 9.318693723086117e-06,
+      "loss": 0.5194,
+      "step": 1556
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8465365990890934,
+      "learning_rate": 9.317680596771389e-06,
+      "loss": 0.5243,
+      "step": 1557
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.587440187024195,
+      "learning_rate": 9.316666772897336e-06,
+      "loss": 0.5472,
+      "step": 1558
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9684250433411536,
+      "learning_rate": 9.315652251627747e-06,
+      "loss": 0.5507,
+      "step": 1559
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.7791357535604546,
+      "learning_rate": 9.314637033126529e-06,
+      "loss": 0.4917,
+      "step": 1560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.948195593969024,
+      "learning_rate": 9.313621117557696e-06,
+      "loss": 0.5052,
+      "step": 1561
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9947405314980446,
+      "learning_rate": 9.312604505085383e-06,
+      "loss": 0.4566,
+      "step": 1562
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2847089946324814,
+      "learning_rate": 9.311587195873828e-06,
+      "loss": 0.5639,
+      "step": 1563
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6519078478731801,
+      "learning_rate": 9.310569190087389e-06,
+      "loss": 0.5206,
+      "step": 1564
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.990246449649651,
+      "learning_rate": 9.309550487890533e-06,
+      "loss": 0.5136,
+      "step": 1565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.839093204640025,
+      "learning_rate": 9.308531089447842e-06,
+      "loss": 0.5123,
+      "step": 1566
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.90895040912295,
+      "learning_rate": 9.307510994924008e-06,
+      "loss": 0.5208,
+      "step": 1567
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6585239573059181,
+      "learning_rate": 9.306490204483834e-06,
+      "loss": 0.4448,
+      "step": 1568
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.499030323961249,
+      "learning_rate": 9.305468718292239e-06,
+      "loss": 0.5165,
+      "step": 1569
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.685184213028577,
+      "learning_rate": 9.304446536514253e-06,
+      "loss": 0.5815,
+      "step": 1570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.653346665510227,
+      "learning_rate": 9.303423659315021e-06,
+      "loss": 0.5341,
+      "step": 1571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.487347787640413,
+      "learning_rate": 9.302400086859792e-06,
+      "loss": 0.4511,
+      "step": 1572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.835863837095015,
+      "learning_rate": 9.30137581931394e-06,
+      "loss": 0.5304,
+      "step": 1573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2039324524357453,
+      "learning_rate": 9.30035085684294e-06,
+      "loss": 0.4979,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9724914467420795,
+      "learning_rate": 9.299325199612387e-06,
+      "loss": 0.5541,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7016534015009618,
+      "learning_rate": 9.29829884778798e-06,
+      "loss": 0.501,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6458296648826666,
+      "learning_rate": 9.29727180153554e-06,
+      "loss": 0.4692,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2079147099479624,
+      "learning_rate": 9.296244061020993e-06,
+      "loss": 0.5151,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.636960136634182,
+      "learning_rate": 9.295215626410382e-06,
+      "loss": 0.5485,
+      "step": 1579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.709804496192575,
+      "learning_rate": 9.294186497869854e-06,
+      "loss": 0.4918,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4096354356444096,
+      "learning_rate": 9.29315667556568e-06,
+      "loss": 0.4803,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0354798485566814,
+      "learning_rate": 9.292126159664231e-06,
+      "loss": 0.509,
+      "step": 1582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2221754554853494,
+      "learning_rate": 9.291094950332002e-06,
+      "loss": 0.5475,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9423280254716395,
+      "learning_rate": 9.290063047735592e-06,
+      "loss": 0.5052,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1675876477187406,
+      "learning_rate": 9.289030452041712e-06,
+      "loss": 0.5406,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0501627862867737,
+      "learning_rate": 9.287997163417189e-06,
+      "loss": 0.5008,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9004854423947835,
+      "learning_rate": 9.286963182028956e-06,
+      "loss": 0.4894,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1099537386796596,
+      "learning_rate": 9.285928508044067e-06,
+      "loss": 0.5329,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8551813231815677,
+      "learning_rate": 9.284893141629681e-06,
+      "loss": 0.5531,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.793130285133734,
+      "learning_rate": 9.283857082953069e-06,
+      "loss": 0.4937,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0523223005721443,
+      "learning_rate": 9.282820332181617e-06,
+      "loss": 0.5133,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8939020501115116,
+      "learning_rate": 9.281782889482819e-06,
+      "loss": 0.4925,
+      "step": 1592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0010390620354106,
+      "learning_rate": 9.280744755024286e-06,
+      "loss": 0.5246,
+      "step": 1593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0795143229091786,
+      "learning_rate": 9.279705928973736e-06,
+      "loss": 0.5206,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.924256171448095,
+      "learning_rate": 9.278666411499e-06,
+      "loss": 0.507,
+      "step": 1595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.186584015434262,
+      "learning_rate": 9.277626202768024e-06,
+      "loss": 0.5503,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.910497080223889,
+      "learning_rate": 9.276585302948861e-06,
+      "loss": 0.4957,
+      "step": 1597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.177423463231272,
+      "learning_rate": 9.275543712209675e-06,
+      "loss": 0.546,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7135803240589209,
+      "learning_rate": 9.27450143071875e-06,
+      "loss": 0.4871,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.6993256970327915,
+      "learning_rate": 9.27345845864447e-06,
+      "loss": 0.5413,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.760557263160197,
+      "learning_rate": 9.27241479615534e-06,
+      "loss": 0.5103,
+      "step": 1601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.91286078054797,
+      "learning_rate": 9.271370443419971e-06,
+      "loss": 0.5083,
+      "step": 1602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2604935256819743,
+      "learning_rate": 9.270325400607087e-06,
+      "loss": 0.5342,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.506226488554768,
+      "learning_rate": 9.269279667885527e-06,
+      "loss": 0.5538,
+      "step": 1604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2697591097698377,
+      "learning_rate": 9.268233245424235e-06,
+      "loss": 0.5124,
+      "step": 1605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5951640188619323,
+      "learning_rate": 9.267186133392272e-06,
+      "loss": 0.5352,
+      "step": 1606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.932511232716658,
+      "learning_rate": 9.266138331958805e-06,
+      "loss": 0.4822,
+      "step": 1607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9386140715583693,
+      "learning_rate": 9.265089841293119e-06,
+      "loss": 0.4821,
+      "step": 1608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6560129362849357,
+      "learning_rate": 9.264040661564606e-06,
+      "loss": 0.4677,
+      "step": 1609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8727163082095566,
+      "learning_rate": 9.262990792942769e-06,
+      "loss": 0.5077,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6686106337025884,
+      "learning_rate": 9.261940235597225e-06,
+      "loss": 0.4837,
+      "step": 1611
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7054558852668906,
+      "learning_rate": 9.260888989697699e-06,
+      "loss": 0.5034,
+      "step": 1612
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5885063961864796,
+      "learning_rate": 9.259837055414032e-06,
+      "loss": 0.5449,
+      "step": 1613
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5075623137868153,
+      "learning_rate": 9.258784432916169e-06,
+      "loss": 0.5099,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4840673530181774,
+      "learning_rate": 9.257731122374175e-06,
+      "loss": 0.5326,
+      "step": 1615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5251406015657,
+      "learning_rate": 9.256677123958218e-06,
+      "loss": 0.4976,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6575980627911382,
+      "learning_rate": 9.255622437838583e-06,
+      "loss": 0.4913,
+      "step": 1617
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5594753292416925,
+      "learning_rate": 9.254567064185662e-06,
+      "loss": 0.5068,
+      "step": 1618
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8252392196648803,
+      "learning_rate": 9.253511003169962e-06,
+      "loss": 0.5559,
+      "step": 1619
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.602915905442299,
+      "learning_rate": 9.252454254962098e-06,
+      "loss": 0.5077,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3477427111270988,
+      "learning_rate": 9.251396819732796e-06,
+      "loss": 0.5271,
+      "step": 1621
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.08775411330249,
+      "learning_rate": 9.250338697652894e-06,
+      "loss": 0.5081,
+      "step": 1622
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6607209226617103,
+      "learning_rate": 9.249279888893343e-06,
+      "loss": 0.5438,
+      "step": 1623
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6352177119328937,
+      "learning_rate": 9.248220393625203e-06,
+      "loss": 0.5325,
+      "step": 1624
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1768861690048777,
+      "learning_rate": 9.247160212019642e-06,
+      "loss": 0.5183,
+      "step": 1625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.385573126299658,
+      "learning_rate": 9.246099344247942e-06,
+      "loss": 0.5403,
+      "step": 1626
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.489702228564718,
+      "learning_rate": 9.245037790481497e-06,
+      "loss": 0.5016,
+      "step": 1627
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.341294944395603,
+      "learning_rate": 9.243975550891811e-06,
+      "loss": 0.4733,
+      "step": 1628
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3586603475734966,
+      "learning_rate": 9.242912625650497e-06,
+      "loss": 0.499,
+      "step": 1629
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.2481656074495446,
+      "learning_rate": 9.24184901492928e-06,
+      "loss": 0.4805,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3536741952704943,
+      "learning_rate": 9.240784718899996e-06,
+      "loss": 0.5595,
+      "step": 1631
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.099961793335952,
+      "learning_rate": 9.23971973773459e-06,
+      "loss": 0.463,
+      "step": 1632
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.036367273374173,
+      "learning_rate": 9.238654071605122e-06,
+      "loss": 0.5815,
+      "step": 1633
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6846706618374232,
+      "learning_rate": 9.237587720683757e-06,
+      "loss": 0.5025,
+      "step": 1634
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7727570659719454,
+      "learning_rate": 9.236520685142774e-06,
+      "loss": 0.4953,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.125186372058996,
+      "learning_rate": 9.235452965154563e-06,
+      "loss": 0.4817,
+      "step": 1636
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.119531132765183,
+      "learning_rate": 9.234384560891623e-06,
+      "loss": 0.5453,
+      "step": 1637
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7431348160011935,
+      "learning_rate": 9.233315472526564e-06,
+      "loss": 0.5565,
+      "step": 1638
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0585035517080037,
+      "learning_rate": 9.232245700232106e-06,
+      "loss": 0.5027,
+      "step": 1639
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.890438150417577,
+      "learning_rate": 9.231175244181081e-06,
+      "loss": 0.5089,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7057878522990162,
+      "learning_rate": 9.230104104546432e-06,
+      "loss": 0.4724,
+      "step": 1641
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.412450440165766,
+      "learning_rate": 9.229032281501209e-06,
+      "loss": 0.4912,
+      "step": 1642
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.886096384576177,
+      "learning_rate": 9.227959775218573e-06,
+      "loss": 0.5774,
+      "step": 1643
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 46.04729752018606,
+      "learning_rate": 9.2268865858718e-06,
+      "loss": 0.5116,
+      "step": 1644
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.643269640799219,
+      "learning_rate": 9.225812713634272e-06,
+      "loss": 0.5138,
+      "step": 1645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3246059542850226,
+      "learning_rate": 9.224738158679482e-06,
+      "loss": 0.4874,
+      "step": 1646
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9035809681676414,
+      "learning_rate": 9.223662921181036e-06,
+      "loss": 0.5225,
+      "step": 1647
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.151399192600683,
+      "learning_rate": 9.222587001312643e-06,
+      "loss": 0.532,
+      "step": 1648
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.844265666331204,
+      "learning_rate": 9.221510399248135e-06,
+      "loss": 0.5036,
+      "step": 1649
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4196112703467962,
+      "learning_rate": 9.220433115161438e-06,
+      "loss": 0.4798,
+      "step": 1650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.252588897696059,
+      "learning_rate": 9.219355149226604e-06,
+      "loss": 0.5083,
+      "step": 1651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6791610636974595,
+      "learning_rate": 9.218276501617784e-06,
+      "loss": 0.5234,
+      "step": 1652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6705201817878617,
+      "learning_rate": 9.217197172509245e-06,
+      "loss": 0.4772,
+      "step": 1653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5807569647746058,
+      "learning_rate": 9.216117162075358e-06,
+      "loss": 0.537,
+      "step": 1654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9277385315437,
+      "learning_rate": 9.215036470490614e-06,
+      "loss": 0.4648,
+      "step": 1655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.4803467893276085,
+      "learning_rate": 9.213955097929605e-06,
+      "loss": 0.5672,
+      "step": 1656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.188819558803046,
+      "learning_rate": 9.212873044567037e-06,
+      "loss": 0.5753,
+      "step": 1657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.865195647290685,
+      "learning_rate": 9.211790310577723e-06,
+      "loss": 0.5022,
+      "step": 1658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4354281045268755,
+      "learning_rate": 9.210706896136592e-06,
+      "loss": 0.4858,
+      "step": 1659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2891107036246554,
+      "learning_rate": 9.209622801418676e-06,
+      "loss": 0.4664,
+      "step": 1660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.4323787584802106,
+      "learning_rate": 9.208538026599124e-06,
+      "loss": 0.4623,
+      "step": 1661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.08512998477563,
+      "learning_rate": 9.207452571853184e-06,
+      "loss": 0.5376,
+      "step": 1662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5491397801481632,
+      "learning_rate": 9.206366437356229e-06,
+      "loss": 0.4774,
+      "step": 1663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.6062166151520283,
+      "learning_rate": 9.205279623283726e-06,
+      "loss": 0.5563,
+      "step": 1664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.764291375497078,
+      "learning_rate": 9.204192129811267e-06,
+      "loss": 0.5121,
+      "step": 1665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7158970499491477,
+      "learning_rate": 9.203103957114542e-06,
+      "loss": 0.5255,
+      "step": 1666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4593279660480194,
+      "learning_rate": 9.202015105369352e-06,
+      "loss": 0.4987,
+      "step": 1667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8706544392705933,
+      "learning_rate": 9.200925574751616e-06,
+      "loss": 0.5272,
+      "step": 1668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5665826039814967,
+      "learning_rate": 9.199835365437353e-06,
+      "loss": 0.4803,
+      "step": 1669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5876414430314243,
+      "learning_rate": 9.1987444776027e-06,
+      "loss": 0.4969,
+      "step": 1670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8052599557337923,
+      "learning_rate": 9.197652911423896e-06,
+      "loss": 0.4989,
+      "step": 1671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1687653891029224,
+      "learning_rate": 9.196560667077294e-06,
+      "loss": 0.5182,
+      "step": 1672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9959404100536227,
+      "learning_rate": 9.19546774473936e-06,
+      "loss": 0.5439,
+      "step": 1673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6202007061973254,
+      "learning_rate": 9.194374144586657e-06,
+      "loss": 0.5115,
+      "step": 1674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9857764472674335,
+      "learning_rate": 9.193279866795872e-06,
+      "loss": 0.5182,
+      "step": 1675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9178192223929929,
+      "learning_rate": 9.192184911543794e-06,
+      "loss": 0.4937,
+      "step": 1676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.602441531524967,
+      "learning_rate": 9.19108927900732e-06,
+      "loss": 0.5622,
+      "step": 1677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7741676838270075,
+      "learning_rate": 9.189992969363463e-06,
+      "loss": 0.4866,
+      "step": 1678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.48544166609113,
+      "learning_rate": 9.18889598278934e-06,
+      "loss": 0.5614,
+      "step": 1679
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.533303252704799,
+      "learning_rate": 9.187798319462177e-06,
+      "loss": 0.5184,
+      "step": 1680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6051159343776416,
+      "learning_rate": 9.186699979559314e-06,
+      "loss": 0.5394,
+      "step": 1681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.432811234549796,
+      "learning_rate": 9.185600963258194e-06,
+      "loss": 0.4813,
+      "step": 1682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.408143189466315,
+      "learning_rate": 9.184501270736378e-06,
+      "loss": 0.471,
+      "step": 1683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.659302765312688,
+      "learning_rate": 9.183400902171527e-06,
+      "loss": 0.5021,
+      "step": 1684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3610849090522068,
+      "learning_rate": 9.182299857741415e-06,
+      "loss": 0.4949,
+      "step": 1685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7850082372585498,
+      "learning_rate": 9.181198137623925e-06,
+      "loss": 0.5339,
+      "step": 1686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7702658610128066,
+      "learning_rate": 9.180095741997055e-06,
+      "loss": 0.4974,
+      "step": 1687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6362295093390105,
+      "learning_rate": 9.1789926710389e-06,
+      "loss": 0.6164,
+      "step": 1688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.651506207170075,
+      "learning_rate": 9.177888924927675e-06,
+      "loss": 0.5517,
+      "step": 1689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7460872196879362,
+      "learning_rate": 9.176784503841699e-06,
+      "loss": 0.5577,
+      "step": 1690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4073768477156903,
+      "learning_rate": 9.175679407959399e-06,
+      "loss": 0.5033,
+      "step": 1691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9985932651418432,
+      "learning_rate": 9.174573637459317e-06,
+      "loss": 0.5495,
+      "step": 1692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.634411342635036,
+      "learning_rate": 9.173467192520095e-06,
+      "loss": 0.5501,
+      "step": 1693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.768504397733744,
+      "learning_rate": 9.172360073320493e-06,
+      "loss": 0.496,
+      "step": 1694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5573104075423723,
+      "learning_rate": 9.171252280039374e-06,
+      "loss": 0.5204,
+      "step": 1695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3393545628964552,
+      "learning_rate": 9.170143812855711e-06,
+      "loss": 0.5045,
+      "step": 1696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.975789021549798,
+      "learning_rate": 9.169034671948589e-06,
+      "loss": 0.5396,
+      "step": 1697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6801972267122265,
+      "learning_rate": 9.167924857497197e-06,
+      "loss": 0.5435,
+      "step": 1698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7246085283194033,
+      "learning_rate": 9.166814369680837e-06,
+      "loss": 0.5395,
+      "step": 1699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7107966860149284,
+      "learning_rate": 9.165703208678917e-06,
+      "loss": 0.5077,
+      "step": 1700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9115099585645674,
+      "learning_rate": 9.164591374670957e-06,
+      "loss": 0.5343,
+      "step": 1701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.588858967214631,
+      "learning_rate": 9.163478867836582e-06,
+      "loss": 0.58,
+      "step": 1702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.554126927044856,
+      "learning_rate": 9.162365688355526e-06,
+      "loss": 0.4872,
+      "step": 1703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3298965104985374,
+      "learning_rate": 9.161251836407635e-06,
+      "loss": 0.5012,
+      "step": 1704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4103330693756335,
+      "learning_rate": 9.16013731217286e-06,
+      "loss": 0.5279,
+      "step": 1705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2852666812092637,
+      "learning_rate": 9.159022115831263e-06,
+      "loss": 0.5313,
+      "step": 1706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3941519860589664,
+      "learning_rate": 9.157906247563013e-06,
+      "loss": 0.5133,
+      "step": 1707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.413329364028389,
+      "learning_rate": 9.15678970754839e-06,
+      "loss": 0.5126,
+      "step": 1708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.42572581665144,
+      "learning_rate": 9.15567249596778e-06,
+      "loss": 0.5391,
+      "step": 1709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.739416463421691,
+      "learning_rate": 9.154554613001679e-06,
+      "loss": 0.5283,
+      "step": 1710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.014337410080953,
+      "learning_rate": 9.15343605883069e-06,
+      "loss": 0.5153,
+      "step": 1711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9909225909292738,
+      "learning_rate": 9.152316833635526e-06,
+      "loss": 0.5594,
+      "step": 1712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7323915263313325,
+      "learning_rate": 9.151196937597006e-06,
+      "loss": 0.5514,
+      "step": 1713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5930929939562986,
+      "learning_rate": 9.150076370896061e-06,
+      "loss": 0.5265,
+      "step": 1714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.561376133480682,
+      "learning_rate": 9.148955133713728e-06,
+      "loss": 0.4775,
+      "step": 1715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.357765216075096,
+      "learning_rate": 9.147833226231151e-06,
+      "loss": 0.4957,
+      "step": 1716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.316501717867558,
+      "learning_rate": 9.146710648629587e-06,
+      "loss": 0.5084,
+      "step": 1717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.622111052086209,
+      "learning_rate": 9.145587401090394e-06,
+      "loss": 0.5024,
+      "step": 1718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6842905276591803,
+      "learning_rate": 9.144463483795049e-06,
+      "loss": 0.5024,
+      "step": 1719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7523349635715433,
+      "learning_rate": 9.143338896925124e-06,
+      "loss": 0.5035,
+      "step": 1720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9129601709051165,
+      "learning_rate": 9.142213640662312e-06,
+      "loss": 0.6002,
+      "step": 1721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7903992740470418,
+      "learning_rate": 9.141087715188402e-06,
+      "loss": 0.5401,
+      "step": 1722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9102915119427542,
+      "learning_rate": 9.1399611206853e-06,
+      "loss": 0.5402,
+      "step": 1723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5467085531898992,
+      "learning_rate": 9.138833857335021e-06,
+      "loss": 0.5115,
+      "step": 1724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1841285847142387,
+      "learning_rate": 9.137705925319677e-06,
+      "loss": 0.5619,
+      "step": 1725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6965895219325391,
+      "learning_rate": 9.136577324821501e-06,
+      "loss": 0.4995,
+      "step": 1726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6859448097443563,
+      "learning_rate": 9.135448056022827e-06,
+      "loss": 0.4778,
+      "step": 1727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2927993289518513,
+      "learning_rate": 9.134318119106098e-06,
+      "loss": 0.4931,
+      "step": 1728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2036250081564543,
+      "learning_rate": 9.133187514253862e-06,
+      "loss": 0.4606,
+      "step": 1729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4265361266647725,
+      "learning_rate": 9.132056241648784e-06,
+      "loss": 0.5319,
+      "step": 1730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3683153281725227,
+      "learning_rate": 9.130924301473627e-06,
+      "loss": 0.512,
+      "step": 1731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3247604798264128,
+      "learning_rate": 9.129791693911268e-06,
+      "loss": 0.5108,
+      "step": 1732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.446197135116374,
+      "learning_rate": 9.128658419144689e-06,
+      "loss": 0.5134,
+      "step": 1733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4740690486820653,
+      "learning_rate": 9.127524477356979e-06,
+      "loss": 0.4845,
+      "step": 1734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5251618719712636,
+      "learning_rate": 9.12638986873134e-06,
+      "loss": 0.5262,
+      "step": 1735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3988676213731237,
+      "learning_rate": 9.125254593451074e-06,
+      "loss": 0.5271,
+      "step": 1736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4890342420503115,
+      "learning_rate": 9.124118651699596e-06,
+      "loss": 0.4984,
+      "step": 1737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7355545059416543,
+      "learning_rate": 9.122982043660428e-06,
+      "loss": 0.5047,
+      "step": 1738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4378053013014622,
+      "learning_rate": 9.121844769517201e-06,
+      "loss": 0.5143,
+      "step": 1739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0339115280934568,
+      "learning_rate": 9.120706829453648e-06,
+      "loss": 0.5517,
+      "step": 1740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1660031363051604,
+      "learning_rate": 9.119568223653614e-06,
+      "loss": 0.5358,
+      "step": 1741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.646667354488079,
+      "learning_rate": 9.118428952301052e-06,
+      "loss": 0.5022,
+      "step": 1742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6285272562157975,
+      "learning_rate": 9.117289015580022e-06,
+      "loss": 0.5378,
+      "step": 1743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3783401691815536,
+      "learning_rate": 9.116148413674688e-06,
+      "loss": 0.5544,
+      "step": 1744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8822723672322517,
+      "learning_rate": 9.115007146769326e-06,
+      "loss": 0.5124,
+      "step": 1745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.791138047925575,
+      "learning_rate": 9.11386521504832e-06,
+      "loss": 0.5198,
+      "step": 1746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7874423277192169,
+      "learning_rate": 9.112722618696155e-06,
+      "loss": 0.5609,
+      "step": 1747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5767870564083402,
+      "learning_rate": 9.11157935789743e-06,
+      "loss": 0.5131,
+      "step": 1748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7726451129948841,
+      "learning_rate": 9.110435432836847e-06,
+      "loss": 0.6044,
+      "step": 1749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6653491459083822,
+      "learning_rate": 9.109290843699219e-06,
+      "loss": 0.531,
+      "step": 1750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3072485709613986,
+      "learning_rate": 9.108145590669464e-06,
+      "loss": 0.5101,
+      "step": 1751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.094158973402322,
+      "learning_rate": 9.106999673932608e-06,
+      "loss": 0.5323,
+      "step": 1752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.620502440367058,
+      "learning_rate": 9.105853093673782e-06,
+      "loss": 0.5151,
+      "step": 1753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.028921567090642,
+      "learning_rate": 9.104705850078229e-06,
+      "loss": 0.5269,
+      "step": 1754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8064643452958555,
+      "learning_rate": 9.103557943331294e-06,
+      "loss": 0.5825,
+      "step": 1755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6589087033035828,
+      "learning_rate": 9.102409373618433e-06,
+      "loss": 0.5637,
+      "step": 1756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5052128611635847,
+      "learning_rate": 9.101260141125207e-06,
+      "loss": 0.5489,
+      "step": 1757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.444200630168782,
+      "learning_rate": 9.100110246037284e-06,
+      "loss": 0.523,
+      "step": 1758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3240725750282878,
+      "learning_rate": 9.09895968854044e-06,
+      "loss": 0.5259,
+      "step": 1759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4390570656456079,
+      "learning_rate": 9.097808468820559e-06,
+      "loss": 0.5009,
+      "step": 1760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7482633852588438,
+      "learning_rate": 9.096656587063628e-06,
+      "loss": 0.5182,
+      "step": 1761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4406976180426263,
+      "learning_rate": 9.095504043455747e-06,
+      "loss": 0.5115,
+      "step": 1762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3518685475810814,
+      "learning_rate": 9.094350838183117e-06,
+      "loss": 0.5111,
+      "step": 1763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2916236319235064,
+      "learning_rate": 9.093196971432048e-06,
+      "loss": 0.5317,
+      "step": 1764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8835108269892396,
+      "learning_rate": 9.09204244338896e-06,
+      "loss": 0.549,
+      "step": 1765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6435708630601518,
+      "learning_rate": 9.090887254240375e-06,
+      "loss": 0.5201,
+      "step": 1766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8321184478689867,
+      "learning_rate": 9.089731404172926e-06,
+      "loss": 0.5411,
+      "step": 1767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2943460426664266,
+      "learning_rate": 9.08857489337335e-06,
+      "loss": 0.5587,
+      "step": 1768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6125378503682504,
+      "learning_rate": 9.08741772202849e-06,
+      "loss": 0.4937,
+      "step": 1769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2238112852588428,
+      "learning_rate": 9.086259890325297e-06,
+      "loss": 0.5269,
+      "step": 1770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6489029783453386,
+      "learning_rate": 9.08510139845083e-06,
+      "loss": 0.5192,
+      "step": 1771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.552669094441354,
+      "learning_rate": 9.083942246592256e-06,
+      "loss": 0.5442,
+      "step": 1772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3550158439915052,
+      "learning_rate": 9.082782434936844e-06,
+      "loss": 0.4781,
+      "step": 1773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3058746893566382,
+      "learning_rate": 9.08162196367197e-06,
+      "loss": 0.5051,
+      "step": 1774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5481492836508022,
+      "learning_rate": 9.08046083298512e-06,
+      "loss": 0.5488,
+      "step": 1775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4310769766264753,
+      "learning_rate": 9.079299043063885e-06,
+      "loss": 0.5832,
+      "step": 1776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6531261378265354,
+      "learning_rate": 9.078136594095964e-06,
+      "loss": 0.5043,
+      "step": 1777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0159329090312497,
+      "learning_rate": 9.076973486269158e-06,
+      "loss": 0.517,
+      "step": 1778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9511807427420154,
+      "learning_rate": 9.075809719771378e-06,
+      "loss": 0.5206,
+      "step": 1779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5978232229037173,
+      "learning_rate": 9.074645294790643e-06,
+      "loss": 0.5324,
+      "step": 1780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4798428851101637,
+      "learning_rate": 9.073480211515071e-06,
+      "loss": 0.5067,
+      "step": 1781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0306639014145014,
+      "learning_rate": 9.0723144701329e-06,
+      "loss": 0.5573,
+      "step": 1782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4728696179541005,
+      "learning_rate": 9.071148070832456e-06,
+      "loss": 0.5273,
+      "step": 1783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.627559526949772,
+      "learning_rate": 9.069981013802188e-06,
+      "loss": 0.5133,
+      "step": 1784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4834810512434706,
+      "learning_rate": 9.06881329923064e-06,
+      "loss": 0.4869,
+      "step": 1785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.372365155180784,
+      "learning_rate": 9.067644927306471e-06,
+      "loss": 0.5169,
+      "step": 1786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6951708523849913,
+      "learning_rate": 9.066475898218439e-06,
+      "loss": 0.4823,
+      "step": 1787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3878873806746732,
+      "learning_rate": 9.06530621215541e-06,
+      "loss": 0.55,
+      "step": 1788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4695670506341068,
+      "learning_rate": 9.064135869306359e-06,
+      "loss": 0.4979,
+      "step": 1789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3469285321385347,
+      "learning_rate": 9.062964869860364e-06,
+      "loss": 0.5153,
+      "step": 1790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3684711900202686,
+      "learning_rate": 9.06179321400661e-06,
+      "loss": 0.5329,
+      "step": 1791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4754969731161858,
+      "learning_rate": 9.060620901934393e-06,
+      "loss": 0.5619,
+      "step": 1792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3439809939316543,
+      "learning_rate": 9.059447933833103e-06,
+      "loss": 0.4887,
+      "step": 1793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4620094045772167,
+      "learning_rate": 9.058274309892248e-06,
+      "loss": 0.5352,
+      "step": 1794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.336998172657835,
+      "learning_rate": 9.057100030301438e-06,
+      "loss": 0.5195,
+      "step": 1795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.740380989435937,
+      "learning_rate": 9.055925095250384e-06,
+      "loss": 0.5416,
+      "step": 1796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.654630803377701,
+      "learning_rate": 9.05474950492891e-06,
+      "loss": 0.4519,
+      "step": 1797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9937748629540164,
+      "learning_rate": 9.053573259526941e-06,
+      "loss": 0.5766,
+      "step": 1798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7901930378208728,
+      "learning_rate": 9.052396359234514e-06,
+      "loss": 0.4942,
+      "step": 1799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6924941301747856,
+      "learning_rate": 9.051218804241764e-06,
+      "loss": 0.5251,
+      "step": 1800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5309637058525534,
+      "learning_rate": 9.050040594738937e-06,
+      "loss": 0.5792,
+      "step": 1801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.355722641980439,
+      "learning_rate": 9.048861730916381e-06,
+      "loss": 0.4687,
+      "step": 1802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.845814720364885,
+      "learning_rate": 9.047682212964553e-06,
+      "loss": 0.5346,
+      "step": 1803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3425542240340513,
+      "learning_rate": 9.046502041074014e-06,
+      "loss": 0.481,
+      "step": 1804
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.617695978622734,
+      "learning_rate": 9.045321215435433e-06,
+      "loss": 0.4759,
+      "step": 1805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5334485516373337,
+      "learning_rate": 9.044139736239581e-06,
+      "loss": 0.562,
+      "step": 1806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5226781612458151,
+      "learning_rate": 9.042957603677338e-06,
+      "loss": 0.4891,
+      "step": 1807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3686279475665037,
+      "learning_rate": 9.041774817939686e-06,
+      "loss": 0.5542,
+      "step": 1808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5436064750306893,
+      "learning_rate": 9.040591379217718e-06,
+      "loss": 0.4889,
+      "step": 1809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.501701685513641,
+      "learning_rate": 9.039407287702622e-06,
+      "loss": 0.5308,
+      "step": 1810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8016377081845487,
+      "learning_rate": 9.038222543585706e-06,
+      "loss": 0.5534,
+      "step": 1811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4173818191469483,
+      "learning_rate": 9.037037147058372e-06,
+      "loss": 0.499,
+      "step": 1812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4346008752805965,
+      "learning_rate": 9.035851098312131e-06,
+      "loss": 0.4631,
+      "step": 1813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4053591169104527,
+      "learning_rate": 9.0346643975386e-06,
+      "loss": 0.5564,
+      "step": 1814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.275060750971617,
+      "learning_rate": 9.033477044929504e-06,
+      "loss": 0.5159,
+      "step": 1815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.77859597684764,
+      "learning_rate": 9.032289040676665e-06,
+      "loss": 0.5705,
+      "step": 1816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7519356417482121,
+      "learning_rate": 9.03110038497202e-06,
+      "loss": 0.5405,
+      "step": 1817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4251065097842237,
+      "learning_rate": 9.029911078007604e-06,
+      "loss": 0.5315,
+      "step": 1818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2492462894322915,
+      "learning_rate": 9.02872111997556e-06,
+      "loss": 0.5089,
+      "step": 1819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4463132073460823,
+      "learning_rate": 9.027530511068139e-06,
+      "loss": 0.4827,
+      "step": 1820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3032911990407365,
+      "learning_rate": 9.026339251477692e-06,
+      "loss": 0.5078,
+      "step": 1821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.626706774190053,
+      "learning_rate": 9.025147341396678e-06,
+      "loss": 0.5205,
+      "step": 1822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3509087885099418,
+      "learning_rate": 9.023954781017662e-06,
+      "loss": 0.4761,
+      "step": 1823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8946587915817568,
+      "learning_rate": 9.02276157053331e-06,
+      "loss": 0.5425,
+      "step": 1824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4134687843412066,
+      "learning_rate": 9.021567710136397e-06,
+      "loss": 0.4827,
+      "step": 1825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5361484675395476,
+      "learning_rate": 9.020373200019802e-06,
+      "loss": 0.5378,
+      "step": 1826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.513940693690181,
+      "learning_rate": 9.019178040376509e-06,
+      "loss": 0.5161,
+      "step": 1827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.507508231829813,
+      "learning_rate": 9.017982231399604e-06,
+      "loss": 0.5442,
+      "step": 1828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5256370685141893,
+      "learning_rate": 9.016785773282284e-06,
+      "loss": 0.5371,
+      "step": 1829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.497173577544496,
+      "learning_rate": 9.015588666217845e-06,
+      "loss": 0.5134,
+      "step": 1830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.0854993405691715,
+      "learning_rate": 9.014390910399691e-06,
+      "loss": 0.5161,
+      "step": 1831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.518221248548298,
+      "learning_rate": 9.013192506021328e-06,
+      "loss": 0.5122,
+      "step": 1832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6377551021684842,
+      "learning_rate": 9.011993453276373e-06,
+      "loss": 0.5408,
+      "step": 1833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5068880542901153,
+      "learning_rate": 9.01079375235854e-06,
+      "loss": 0.4908,
+      "step": 1834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4292090316250703,
+      "learning_rate": 9.009593403461652e-06,
+      "loss": 0.5954,
+      "step": 1835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.733599321570736,
+      "learning_rate": 9.008392406779638e-06,
+      "loss": 0.5509,
+      "step": 1836
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8927517677665815,
+      "learning_rate": 9.007190762506527e-06,
+      "loss": 0.5052,
+      "step": 1837
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.541189839125486,
+      "learning_rate": 9.005988470836456e-06,
+      "loss": 0.5135,
+      "step": 1838
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9180698533113243,
+      "learning_rate": 9.004785531963665e-06,
+      "loss": 0.5047,
+      "step": 1839
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3942542286939754,
+      "learning_rate": 9.003581946082503e-06,
+      "loss": 0.5786,
+      "step": 1840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0924217443966975,
+      "learning_rate": 9.002377713387415e-06,
+      "loss": 0.4938,
+      "step": 1841
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8068366112672747,
+      "learning_rate": 9.001172834072958e-06,
+      "loss": 0.4915,
+      "step": 1842
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4742349524709262,
+      "learning_rate": 8.999967308333791e-06,
+      "loss": 0.4864,
+      "step": 1843
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9117290532257252,
+      "learning_rate": 8.998761136364675e-06,
+      "loss": 0.5857,
+      "step": 1844
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8116074153470343,
+      "learning_rate": 8.997554318360482e-06,
+      "loss": 0.489,
+      "step": 1845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0987590232716706,
+      "learning_rate": 8.99634685451618e-06,
+      "loss": 0.5531,
+      "step": 1846
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5604499423857239,
+      "learning_rate": 8.995138745026847e-06,
+      "loss": 0.4864,
+      "step": 1847
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5388505102375079,
+      "learning_rate": 8.993929990087664e-06,
+      "loss": 0.5363,
+      "step": 1848
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.29222554323978,
+      "learning_rate": 8.992720589893915e-06,
+      "loss": 0.5159,
+      "step": 1849
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9916577839753073,
+      "learning_rate": 8.991510544640992e-06,
+      "loss": 0.5397,
+      "step": 1850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6948489030363554,
+      "learning_rate": 8.990299854524384e-06,
+      "loss": 0.5693,
+      "step": 1851
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9513734930911082,
+      "learning_rate": 8.989088519739693e-06,
+      "loss": 0.5161,
+      "step": 1852
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4587444301512456,
+      "learning_rate": 8.987876540482618e-06,
+      "loss": 0.4824,
+      "step": 1853
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6710499041249678,
+      "learning_rate": 8.986663916948965e-06,
+      "loss": 0.5454,
+      "step": 1854
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9152211306936873,
+      "learning_rate": 8.985450649334646e-06,
+      "loss": 0.5987,
+      "step": 1855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6560317393676094,
+      "learning_rate": 8.984236737835673e-06,
+      "loss": 0.5399,
+      "step": 1856
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5195714881992344,
+      "learning_rate": 8.983022182648166e-06,
+      "loss": 0.597,
+      "step": 1857
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.355222695904536,
+      "learning_rate": 8.981806983968346e-06,
+      "loss": 0.5864,
+      "step": 1858
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2761619323097473,
+      "learning_rate": 8.980591141992538e-06,
+      "loss": 0.5001,
+      "step": 1859
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4872401648051745,
+      "learning_rate": 8.979374656917174e-06,
+      "loss": 0.4997,
+      "step": 1860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.588819005755008,
+      "learning_rate": 8.978157528938786e-06,
+      "loss": 0.5829,
+      "step": 1861
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6542912440317863,
+      "learning_rate": 8.976939758254015e-06,
+      "loss": 0.5478,
+      "step": 1862
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 26.682459025536943,
+      "learning_rate": 8.975721345059598e-06,
+      "loss": 0.5039,
+      "step": 1863
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0430820852157336,
+      "learning_rate": 8.974502289552384e-06,
+      "loss": 0.5205,
+      "step": 1864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.43661731754798,
+      "learning_rate": 8.973282591929319e-06,
+      "loss": 0.5473,
+      "step": 1865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.688736566388985,
+      "learning_rate": 8.97206225238746e-06,
+      "loss": 0.5611,
+      "step": 1866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4425314258470237,
+      "learning_rate": 8.970841271123962e-06,
+      "loss": 0.5586,
+      "step": 1867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8776702201654907,
+      "learning_rate": 8.969619648336082e-06,
+      "loss": 0.5739,
+      "step": 1868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7012564447126629,
+      "learning_rate": 8.968397384221188e-06,
+      "loss": 0.5036,
+      "step": 1869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3566032268734702,
+      "learning_rate": 8.967174478976745e-06,
+      "loss": 0.5345,
+      "step": 1870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8068390422598124,
+      "learning_rate": 8.965950932800326e-06,
+      "loss": 0.5652,
+      "step": 1871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.435850750817673,
+      "learning_rate": 8.964726745889606e-06,
+      "loss": 0.5618,
+      "step": 1872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2816893239440885,
+      "learning_rate": 8.963501918442359e-06,
+      "loss": 0.4831,
+      "step": 1873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3104746693874945,
+      "learning_rate": 8.962276450656471e-06,
+      "loss": 0.4523,
+      "step": 1874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4357443614868366,
+      "learning_rate": 8.961050342729927e-06,
+      "loss": 0.4957,
+      "step": 1875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.568259871694868,
+      "learning_rate": 8.959823594860813e-06,
+      "loss": 0.4929,
+      "step": 1876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.372551882835372,
+      "learning_rate": 8.958596207247322e-06,
+      "loss": 0.585,
+      "step": 1877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3909216826293513,
+      "learning_rate": 8.95736818008775e-06,
+      "loss": 0.4942,
+      "step": 1878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6154288710912335,
+      "learning_rate": 8.956139513580495e-06,
+      "loss": 0.4675,
+      "step": 1879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4455950323321096,
+      "learning_rate": 8.95491020792406e-06,
+      "loss": 0.5411,
+      "step": 1880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.259868701492697,
+      "learning_rate": 8.953680263317048e-06,
+      "loss": 0.4533,
+      "step": 1881
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.522655048402381,
+      "learning_rate": 8.952449679958168e-06,
+      "loss": 0.5309,
+      "step": 1882
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6655362560273492,
+      "learning_rate": 8.951218458046233e-06,
+      "loss": 0.5407,
+      "step": 1883
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3048223977489455,
+      "learning_rate": 8.949986597780157e-06,
+      "loss": 0.5039,
+      "step": 1884
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5562776861902379,
+      "learning_rate": 8.94875409935896e-06,
+      "loss": 0.5763,
+      "step": 1885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2745792479455942,
+      "learning_rate": 8.947520962981758e-06,
+      "loss": 0.4908,
+      "step": 1886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6672757070096937,
+      "learning_rate": 8.946287188847778e-06,
+      "loss": 0.5517,
+      "step": 1887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3056030322769647,
+      "learning_rate": 8.945052777156346e-06,
+      "loss": 0.5022,
+      "step": 1888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8962485527766033,
+      "learning_rate": 8.943817728106894e-06,
+      "loss": 0.506,
+      "step": 1889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4549494573125639,
+      "learning_rate": 8.942582041898954e-06,
+      "loss": 0.4836,
+      "step": 1890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7727010742859703,
+      "learning_rate": 8.941345718732162e-06,
+      "loss": 0.5042,
+      "step": 1891
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.383471940931024,
+      "learning_rate": 8.940108758806258e-06,
+      "loss": 0.5198,
+      "step": 1892
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6303903494517051,
+      "learning_rate": 8.938871162321082e-06,
+      "loss": 0.5107,
+      "step": 1893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.783742959636969,
+      "learning_rate": 8.937632929476578e-06,
+      "loss": 0.5629,
+      "step": 1894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4634858088918852,
+      "learning_rate": 8.936394060472796e-06,
+      "loss": 0.4866,
+      "step": 1895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.339140151238824,
+      "learning_rate": 8.935154555509883e-06,
+      "loss": 0.5651,
+      "step": 1896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4343314700039496,
+      "learning_rate": 8.933914414788095e-06,
+      "loss": 0.5304,
+      "step": 1897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.447195399814874,
+      "learning_rate": 8.932673638507787e-06,
+      "loss": 0.5352,
+      "step": 1898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5450789392126534,
+      "learning_rate": 8.931432226869416e-06,
+      "loss": 0.4946,
+      "step": 1899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1174940010910475,
+      "learning_rate": 8.930190180073544e-06,
+      "loss": 0.5236,
+      "step": 1900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6545234531448152,
+      "learning_rate": 8.928947498320835e-06,
+      "loss": 0.5477,
+      "step": 1901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3292775696499728,
+      "learning_rate": 8.927704181812053e-06,
+      "loss": 0.4817,
+      "step": 1902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5071985105494015,
+      "learning_rate": 8.92646023074807e-06,
+      "loss": 0.5697,
+      "step": 1903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9818677569842058,
+      "learning_rate": 8.925215645329854e-06,
+      "loss": 0.4559,
+      "step": 1904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.580179304566661,
+      "learning_rate": 8.923970425758481e-06,
+      "loss": 0.5159,
+      "step": 1905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.458276371266938,
+      "learning_rate": 8.922724572235128e-06,
+      "loss": 0.5959,
+      "step": 1906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7012565794755887,
+      "learning_rate": 8.921478084961071e-06,
+      "loss": 0.5144,
+      "step": 1907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4290181998565596,
+      "learning_rate": 8.92023096413769e-06,
+      "loss": 0.5189,
+      "step": 1908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4312359811126811,
+      "learning_rate": 8.918983209966475e-06,
+      "loss": 0.5151,
+      "step": 1909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.336001691916545,
+      "learning_rate": 8.917734822649002e-06,
+      "loss": 0.4681,
+      "step": 1910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9200723965237108,
+      "learning_rate": 8.916485802386968e-06,
+      "loss": 0.5125,
+      "step": 1911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5797716719857873,
+      "learning_rate": 8.915236149382155e-06,
+      "loss": 0.5416,
+      "step": 1912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.783287209569103,
+      "learning_rate": 8.913985863836465e-06,
+      "loss": 0.5443,
+      "step": 1913
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8272401889966197,
+      "learning_rate": 8.912734945951884e-06,
+      "loss": 0.5295,
+      "step": 1914
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6138889353965085,
+      "learning_rate": 8.911483395930514e-06,
+      "loss": 0.5037,
+      "step": 1915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.495452942442373,
+      "learning_rate": 8.910231213974549e-06,
+      "loss": 0.5173,
+      "step": 1916
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.146077073992904,
+      "learning_rate": 8.908978400286297e-06,
+      "loss": 0.5296,
+      "step": 1917
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7738406140487761,
+      "learning_rate": 8.907724955068156e-06,
+      "loss": 0.5089,
+      "step": 1918
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4445001960204054,
+      "learning_rate": 8.90647087852263e-06,
+      "loss": 0.4729,
+      "step": 1919
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.064579524062797,
+      "learning_rate": 8.905216170852332e-06,
+      "loss": 0.4936,
+      "step": 1920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5744305050856235,
+      "learning_rate": 8.903960832259966e-06,
+      "loss": 0.4603,
+      "step": 1921
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6956951304877073,
+      "learning_rate": 8.902704862948344e-06,
+      "loss": 0.496,
+      "step": 1922
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1043226523378875,
+      "learning_rate": 8.901448263120379e-06,
+      "loss": 0.5216,
+      "step": 1923
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2288602774625417,
+      "learning_rate": 8.900191032979088e-06,
+      "loss": 0.4907,
+      "step": 1924
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6974651681887711,
+      "learning_rate": 8.898933172727584e-06,
+      "loss": 0.5437,
+      "step": 1925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1277108296392484,
+      "learning_rate": 8.897674682569088e-06,
+      "loss": 0.4111,
+      "step": 1926
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3434009357471774,
+      "learning_rate": 8.896415562706919e-06,
+      "loss": 0.5192,
+      "step": 1927
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5168120489469041,
+      "learning_rate": 8.8951558133445e-06,
+      "loss": 0.5363,
+      "step": 1928
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.372964230231597,
+      "learning_rate": 8.893895434685353e-06,
+      "loss": 0.5477,
+      "step": 1929
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3341145832969576,
+      "learning_rate": 8.892634426933106e-06,
+      "loss": 0.5334,
+      "step": 1930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.381671572338022,
+      "learning_rate": 8.891372790291482e-06,
+      "loss": 0.4387,
+      "step": 1931
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3889571726074057,
+      "learning_rate": 8.890110524964313e-06,
+      "loss": 0.5352,
+      "step": 1932
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6346808498868517,
+      "learning_rate": 8.888847631155525e-06,
+      "loss": 0.5435,
+      "step": 1933
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.389766354005252,
+      "learning_rate": 8.887584109069157e-06,
+      "loss": 0.4923,
+      "step": 1934
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.506276037797442,
+      "learning_rate": 8.886319958909334e-06,
+      "loss": 0.5358,
+      "step": 1935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8617829100438767,
+      "learning_rate": 8.885055180880294e-06,
+      "loss": 0.5455,
+      "step": 1936
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4403503924568222,
+      "learning_rate": 8.883789775186374e-06,
+      "loss": 0.4609,
+      "step": 1937
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5025770716659845,
+      "learning_rate": 8.88252374203201e-06,
+      "loss": 0.5037,
+      "step": 1938
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.713090061036998,
+      "learning_rate": 8.881257081621741e-06,
+      "loss": 0.4901,
+      "step": 1939
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7078369186879294,
+      "learning_rate": 8.879989794160208e-06,
+      "loss": 0.4839,
+      "step": 1940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5524521922610566,
+      "learning_rate": 8.878721879852153e-06,
+      "loss": 0.5419,
+      "step": 1941
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7800261176459817,
+      "learning_rate": 8.877453338902415e-06,
+      "loss": 0.5274,
+      "step": 1942
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.438442558784284,
+      "learning_rate": 8.876184171515943e-06,
+      "loss": 0.5588,
+      "step": 1943
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.579247057716277,
+      "learning_rate": 8.874914377897778e-06,
+      "loss": 0.4872,
+      "step": 1944
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7752495014824574,
+      "learning_rate": 8.87364395825307e-06,
+      "loss": 0.5349,
+      "step": 1945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5952965241983228,
+      "learning_rate": 8.872372912787061e-06,
+      "loss": 0.5209,
+      "step": 1946
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4540745806687716,
+      "learning_rate": 8.871101241705105e-06,
+      "loss": 0.5252,
+      "step": 1947
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5604738622505752,
+      "learning_rate": 8.86982894521265e-06,
+      "loss": 0.4982,
+      "step": 1948
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7187200851975826,
+      "learning_rate": 8.868556023515247e-06,
+      "loss": 0.5191,
+      "step": 1949
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3690982491354926,
+      "learning_rate": 8.867282476818546e-06,
+      "loss": 0.5194,
+      "step": 1950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.872490573096125,
+      "learning_rate": 8.866008305328303e-06,
+      "loss": 0.4693,
+      "step": 1951
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.653900492109943,
+      "learning_rate": 8.864733509250367e-06,
+      "loss": 0.4991,
+      "step": 1952
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8198036332691097,
+      "learning_rate": 8.863458088790695e-06,
+      "loss": 0.5657,
+      "step": 1953
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7071267036238204,
+      "learning_rate": 8.862182044155345e-06,
+      "loss": 0.5261,
+      "step": 1954
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4544658968334625,
+      "learning_rate": 8.860905375550469e-06,
+      "loss": 0.5482,
+      "step": 1955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5844352202023428,
+      "learning_rate": 8.859628083182326e-06,
+      "loss": 0.5331,
+      "step": 1956
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8270985290546589,
+      "learning_rate": 8.858350167257275e-06,
+      "loss": 0.5229,
+      "step": 1957
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2358524365219887,
+      "learning_rate": 8.85707162798177e-06,
+      "loss": 0.5315,
+      "step": 1958
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6698059723203507,
+      "learning_rate": 8.855792465562377e-06,
+      "loss": 0.5325,
+      "step": 1959
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7400217029843602,
+      "learning_rate": 8.854512680205748e-06,
+      "loss": 0.5117,
+      "step": 1960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4293314877485304,
+      "learning_rate": 8.853232272118653e-06,
+      "loss": 0.5194,
+      "step": 1961
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3617601801455612,
+      "learning_rate": 8.851951241507945e-06,
+      "loss": 0.48,
+      "step": 1962
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5593848472649214,
+      "learning_rate": 8.850669588580591e-06,
+      "loss": 0.5879,
+      "step": 1963
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.329342422082542,
+      "learning_rate": 8.84938731354365e-06,
+      "loss": 0.5234,
+      "step": 1964
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6158031021826076,
+      "learning_rate": 8.848104416604287e-06,
+      "loss": 0.5032,
+      "step": 1965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4835217258908457,
+      "learning_rate": 8.846820897969763e-06,
+      "loss": 0.53,
+      "step": 1966
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6760264757911146,
+      "learning_rate": 8.845536757847444e-06,
+      "loss": 0.502,
+      "step": 1967
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8477646844223596,
+      "learning_rate": 8.844251996444792e-06,
+      "loss": 0.5484,
+      "step": 1968
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7376392792536866,
+      "learning_rate": 8.842966613969376e-06,
+      "loss": 0.5252,
+      "step": 1969
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.165308432040499,
+      "learning_rate": 8.841680610628853e-06,
+      "loss": 0.4646,
+      "step": 1970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4974071983421144,
+      "learning_rate": 8.840393986630996e-06,
+      "loss": 0.4945,
+      "step": 1971
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.845611195832243,
+      "learning_rate": 8.839106742183668e-06,
+      "loss": 0.5535,
+      "step": 1972
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5394864096122054,
+      "learning_rate": 8.837818877494833e-06,
+      "loss": 0.5139,
+      "step": 1973
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4206296295423873,
+      "learning_rate": 8.836530392772555e-06,
+      "loss": 0.5274,
+      "step": 1974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6776243792818417,
+      "learning_rate": 8.835241288225007e-06,
+      "loss": 0.5352,
+      "step": 1975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3565471277974326,
+      "learning_rate": 8.83395156406045e-06,
+      "loss": 0.5327,
+      "step": 1976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4313562691928976,
+      "learning_rate": 8.832661220487251e-06,
+      "loss": 0.5132,
+      "step": 1977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7967161826065228,
+      "learning_rate": 8.831370257713877e-06,
+      "loss": 0.4947,
+      "step": 1978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.467448215573434,
+      "learning_rate": 8.830078675948894e-06,
+      "loss": 0.5552,
+      "step": 1979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5835629523108656,
+      "learning_rate": 8.828786475400968e-06,
+      "loss": 0.5141,
+      "step": 1980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0099353012114616,
+      "learning_rate": 8.827493656278867e-06,
+      "loss": 0.4699,
+      "step": 1981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5473553695010045,
+      "learning_rate": 8.826200218791455e-06,
+      "loss": 0.5123,
+      "step": 1982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5795956022466797,
+      "learning_rate": 8.8249061631477e-06,
+      "loss": 0.522,
+      "step": 1983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.222064880909354,
+      "learning_rate": 8.823611489556668e-06,
+      "loss": 0.5207,
+      "step": 1984
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7684840911887174,
+      "learning_rate": 8.822316198227525e-06,
+      "loss": 0.5149,
+      "step": 1985
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.476508776650612,
+      "learning_rate": 8.821020289369535e-06,
+      "loss": 0.5619,
+      "step": 1986
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3676529888008777,
+      "learning_rate": 8.819723763192065e-06,
+      "loss": 0.4907,
+      "step": 1987
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8100143598963343,
+      "learning_rate": 8.81842661990458e-06,
+      "loss": 0.4757,
+      "step": 1988
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6704306315147124,
+      "learning_rate": 8.817128859716646e-06,
+      "loss": 0.5198,
+      "step": 1989
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6174741992271353,
+      "learning_rate": 8.815830482837925e-06,
+      "loss": 0.4942,
+      "step": 1990
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.1235866635095344,
+      "learning_rate": 8.814531489478183e-06,
+      "loss": 0.5388,
+      "step": 1991
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.749572638929686,
+      "learning_rate": 8.813231879847284e-06,
+      "loss": 0.4947,
+      "step": 1992
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8765404375963373,
+      "learning_rate": 8.811931654155191e-06,
+      "loss": 0.4857,
+      "step": 1993
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7127206859801521,
+      "learning_rate": 8.810630812611965e-06,
+      "loss": 0.5374,
+      "step": 1994
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3765373181685499,
+      "learning_rate": 8.80932935542777e-06,
+      "loss": 0.5574,
+      "step": 1995
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4220313237505227,
+      "learning_rate": 8.808027282812871e-06,
+      "loss": 0.551,
+      "step": 1996
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.390487340074321,
+      "learning_rate": 8.806724594977625e-06,
+      "loss": 0.5951,
+      "step": 1997
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6304586844315077,
+      "learning_rate": 8.805421292132495e-06,
+      "loss": 0.4819,
+      "step": 1998
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.548784626397272,
+      "learning_rate": 8.804117374488037e-06,
+      "loss": 0.4743,
+      "step": 1999
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7079415571467975,
+      "learning_rate": 8.802812842254917e-06,
+      "loss": 0.5411,
+      "step": 2000
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7494791245756824,
+      "learning_rate": 8.801507695643886e-06,
+      "loss": 0.552,
+      "step": 2001
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5625148747215771,
+      "learning_rate": 8.80020193486581e-06,
+      "loss": 0.5551,
+      "step": 2002
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5943239427434999,
+      "learning_rate": 8.798895560131642e-06,
+      "loss": 0.46,
+      "step": 2003
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8847274473426747,
+      "learning_rate": 8.797588571652439e-06,
+      "loss": 0.5404,
+      "step": 2004
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7756131902868117,
+      "learning_rate": 8.796280969639353e-06,
+      "loss": 0.4842,
+      "step": 2005
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.852938704264008,
+      "learning_rate": 8.794972754303644e-06,
+      "loss": 0.5657,
+      "step": 2006
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.412632757606416,
+      "learning_rate": 8.793663925856662e-06,
+      "loss": 0.5468,
+      "step": 2007
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6268459566919793,
+      "learning_rate": 8.792354484509863e-06,
+      "loss": 0.5665,
+      "step": 2008
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.318500406755192,
+      "learning_rate": 8.791044430474795e-06,
+      "loss": 0.5019,
+      "step": 2009
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4252737383395815,
+      "learning_rate": 8.789733763963112e-06,
+      "loss": 0.5352,
+      "step": 2010
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7028911675339917,
+      "learning_rate": 8.788422485186561e-06,
+      "loss": 0.4947,
+      "step": 2011
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.295162872112619,
+      "learning_rate": 8.787110594356993e-06,
+      "loss": 0.5005,
+      "step": 2012
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2920138918598154,
+      "learning_rate": 8.785798091686356e-06,
+      "loss": 0.4892,
+      "step": 2013
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5507368136823545,
+      "learning_rate": 8.784484977386691e-06,
+      "loss": 0.5525,
+      "step": 2014
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4858272518440248,
+      "learning_rate": 8.783171251670151e-06,
+      "loss": 0.5463,
+      "step": 2015
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6740133113942713,
+      "learning_rate": 8.781856914748974e-06,
+      "loss": 0.5366,
+      "step": 2016
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6130432936138699,
+      "learning_rate": 8.780541966835506e-06,
+      "loss": 0.4703,
+      "step": 2017
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4865849447276784,
+      "learning_rate": 8.779226408142187e-06,
+      "loss": 0.5175,
+      "step": 2018
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.477496458338817,
+      "learning_rate": 8.777910238881557e-06,
+      "loss": 0.5737,
+      "step": 2019
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.590972661806151,
+      "learning_rate": 8.776593459266256e-06,
+      "loss": 0.5884,
+      "step": 2020
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5083726906854342,
+      "learning_rate": 8.77527606950902e-06,
+      "loss": 0.5236,
+      "step": 2021
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6928660682592702,
+      "learning_rate": 8.773958069822684e-06,
+      "loss": 0.5336,
+      "step": 2022
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8285529561145588,
+      "learning_rate": 8.772639460420184e-06,
+      "loss": 0.5104,
+      "step": 2023
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4625362649817097,
+      "learning_rate": 8.771320241514553e-06,
+      "loss": 0.5071,
+      "step": 2024
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.209422273835458,
+      "learning_rate": 8.770000413318923e-06,
+      "loss": 0.5072,
+      "step": 2025
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3313275590933291,
+      "learning_rate": 8.768679976046523e-06,
+      "loss": 0.4998,
+      "step": 2026
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2482005079526335,
+      "learning_rate": 8.767358929910681e-06,
+      "loss": 0.469,
+      "step": 2027
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4071157891730426,
+      "learning_rate": 8.766037275124824e-06,
+      "loss": 0.5088,
+      "step": 2028
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6073781895786337,
+      "learning_rate": 8.764715011902477e-06,
+      "loss": 0.5158,
+      "step": 2029
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.084224486603882,
+      "learning_rate": 8.763392140457261e-06,
+      "loss": 0.526,
+      "step": 2030
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7510253233417739,
+      "learning_rate": 8.762068661002902e-06,
+      "loss": 0.4855,
+      "step": 2031
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.3038810606284432,
+      "learning_rate": 8.760744573753218e-06,
+      "loss": 0.5514,
+      "step": 2032
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6022421653238001,
+      "learning_rate": 8.759419878922125e-06,
+      "loss": 0.5673,
+      "step": 2033
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.20301441399954,
+      "learning_rate": 8.758094576723641e-06,
+      "loss": 0.539,
+      "step": 2034
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6577453684894792,
+      "learning_rate": 8.75676866737188e-06,
+      "loss": 0.5521,
+      "step": 2035
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6839437161456582,
+      "learning_rate": 8.755442151081054e-06,
+      "loss": 0.4958,
+      "step": 2036
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5834842804236429,
+      "learning_rate": 8.754115028065474e-06,
+      "loss": 0.4872,
+      "step": 2037
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2421241088167467,
+      "learning_rate": 8.752787298539547e-06,
+      "loss": 0.5161,
+      "step": 2038
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.161941749490066,
+      "learning_rate": 8.75145896271778e-06,
+      "loss": 0.4987,
+      "step": 2039
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4927151108479508,
+      "learning_rate": 8.750130020814779e-06,
+      "loss": 0.4894,
+      "step": 2040
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.502074237777443,
+      "learning_rate": 8.748800473045245e-06,
+      "loss": 0.5471,
+      "step": 2041
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5446193105812245,
+      "learning_rate": 8.747470319623976e-06,
+      "loss": 0.5269,
+      "step": 2042
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8144058798330323,
+      "learning_rate": 8.746139560765873e-06,
+      "loss": 0.5298,
+      "step": 2043
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3392301561466349,
+      "learning_rate": 8.744808196685933e-06,
+      "loss": 0.5376,
+      "step": 2044
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3212574679321898,
+      "learning_rate": 8.743476227599245e-06,
+      "loss": 0.5288,
+      "step": 2045
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.953631424817136,
+      "learning_rate": 8.742143653721004e-06,
+      "loss": 0.5045,
+      "step": 2046
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3254459527990614,
+      "learning_rate": 8.740810475266497e-06,
+      "loss": 0.5295,
+      "step": 2047
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2471258040754365,
+      "learning_rate": 8.739476692451112e-06,
+      "loss": 0.5237,
+      "step": 2048
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.2406500377543024,
+      "learning_rate": 8.738142305490335e-06,
+      "loss": 0.5653,
+      "step": 2049
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3120908521833168,
+      "learning_rate": 8.736807314599744e-06,
+      "loss": 0.4988,
+      "step": 2050
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.56240581885028,
+      "learning_rate": 8.73547171999502e-06,
+      "loss": 0.5052,
+      "step": 2051
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3269004649671916,
+      "learning_rate": 8.734135521891941e-06,
+      "loss": 0.5342,
+      "step": 2052
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.472749561684422,
+      "learning_rate": 8.732798720506381e-06,
+      "loss": 0.4932,
+      "step": 2053
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5456774928408388,
+      "learning_rate": 8.731461316054313e-06,
+      "loss": 0.5342,
+      "step": 2054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2554204562914544,
+      "learning_rate": 8.730123308751806e-06,
+      "loss": 0.4592,
+      "step": 2055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3112293500770817,
+      "learning_rate": 8.728784698815026e-06,
+      "loss": 0.5272,
+      "step": 2056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.1667461673415906,
+      "learning_rate": 8.727445486460236e-06,
+      "loss": 0.4595,
+      "step": 2057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6367370511644874,
+      "learning_rate": 8.7261056719038e-06,
+      "loss": 0.5088,
+      "step": 2058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.50527731481556,
+      "learning_rate": 8.724765255362176e-06,
+      "loss": 0.5197,
+      "step": 2059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5322878593937856,
+      "learning_rate": 8.72342423705192e-06,
+      "loss": 0.5218,
+      "step": 2060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4793378737196778,
+      "learning_rate": 8.722082617189688e-06,
+      "loss": 0.5401,
+      "step": 2061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4610037837615666,
+      "learning_rate": 8.720740395992225e-06,
+      "loss": 0.5336,
+      "step": 2062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5622701747115364,
+      "learning_rate": 8.719397573676384e-06,
+      "loss": 0.5043,
+      "step": 2063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4367898452648369,
+      "learning_rate": 8.718054150459106e-06,
+      "loss": 0.5192,
+      "step": 2064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5175871163956165,
+      "learning_rate": 8.716710126557435e-06,
+      "loss": 0.5194,
+      "step": 2065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3458435739714716,
+      "learning_rate": 8.715365502188508e-06,
+      "loss": 0.5238,
+      "step": 2066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7803905820852695,
+      "learning_rate": 8.714020277569564e-06,
+      "loss": 0.5646,
+      "step": 2067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7385985543427787,
+      "learning_rate": 8.712674452917934e-06,
+      "loss": 0.5495,
+      "step": 2068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4759265210333785,
+      "learning_rate": 8.711328028451045e-06,
+      "loss": 0.4545,
+      "step": 2069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3288442314806546,
+      "learning_rate": 8.709981004386429e-06,
+      "loss": 0.5034,
+      "step": 2070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4758448460788367,
+      "learning_rate": 8.708633380941706e-06,
+      "loss": 0.5045,
+      "step": 2071
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3920330580189069,
+      "learning_rate": 8.707285158334598e-06,
+      "loss": 0.4962,
+      "step": 2072
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2930994944296241,
+      "learning_rate": 8.705936336782921e-06,
+      "loss": 0.5308,
+      "step": 2073
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8773196012080102,
+      "learning_rate": 8.704586916504592e-06,
+      "loss": 0.4816,
+      "step": 2074
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3165810492564216,
+      "learning_rate": 8.703236897717617e-06,
+      "loss": 0.4903,
+      "step": 2075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5163855156336135,
+      "learning_rate": 8.701886280640109e-06,
+      "loss": 0.4882,
+      "step": 2076
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6838193810705397,
+      "learning_rate": 8.700535065490266e-06,
+      "loss": 0.5349,
+      "step": 2077
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3303115176381106,
+      "learning_rate": 8.699183252486395e-06,
+      "loss": 0.5408,
+      "step": 2078
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3360266668930008,
+      "learning_rate": 8.697830841846887e-06,
+      "loss": 0.525,
+      "step": 2079
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.439059769559954,
+      "learning_rate": 8.69647783379024e-06,
+      "loss": 0.5354,
+      "step": 2080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3942462512315814,
+      "learning_rate": 8.695124228535044e-06,
+      "loss": 0.5391,
+      "step": 2081
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.626043540508294,
+      "learning_rate": 8.693770026299984e-06,
+      "loss": 0.475,
+      "step": 2082
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3662165330675327,
+      "learning_rate": 8.692415227303844e-06,
+      "loss": 0.5206,
+      "step": 2083
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4217799982549042,
+      "learning_rate": 8.691059831765505e-06,
+      "loss": 0.5648,
+      "step": 2084
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7365946002808297,
+      "learning_rate": 8.689703839903943e-06,
+      "loss": 0.5146,
+      "step": 2085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6943207518116408,
+      "learning_rate": 8.688347251938229e-06,
+      "loss": 0.5379,
+      "step": 2086
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5989235306650624,
+      "learning_rate": 8.686990068087532e-06,
+      "loss": 0.5212,
+      "step": 2087
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5245090919991986,
+      "learning_rate": 8.685632288571118e-06,
+      "loss": 0.5261,
+      "step": 2088
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3285545188358303,
+      "learning_rate": 8.684273913608346e-06,
+      "loss": 0.4977,
+      "step": 2089
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.488096499314304,
+      "learning_rate": 8.682914943418677e-06,
+      "loss": 0.5072,
+      "step": 2090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4838765109056207,
+      "learning_rate": 8.681555378221661e-06,
+      "loss": 0.5111,
+      "step": 2091
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6550785967751374,
+      "learning_rate": 8.680195218236951e-06,
+      "loss": 0.4896,
+      "step": 2092
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4018756217200874,
+      "learning_rate": 8.67883446368429e-06,
+      "loss": 0.547,
+      "step": 2093
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3465600335196564,
+      "learning_rate": 8.677473114783524e-06,
+      "loss": 0.5501,
+      "step": 2094
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.296107978657512,
+      "learning_rate": 8.676111171754585e-06,
+      "loss": 0.5285,
+      "step": 2095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3967902124449163,
+      "learning_rate": 8.67474863481751e-06,
+      "loss": 0.5699,
+      "step": 2096
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5257350142040416,
+      "learning_rate": 8.673385504192428e-06,
+      "loss": 0.5171,
+      "step": 2097
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3449457010675756,
+      "learning_rate": 8.672021780099569e-06,
+      "loss": 0.524,
+      "step": 2098
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.519683542426732,
+      "learning_rate": 8.670657462759248e-06,
+      "loss": 0.5065,
+      "step": 2099
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4787471182390501,
+      "learning_rate": 8.669292552391888e-06,
+      "loss": 0.5024,
+      "step": 2100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6463547942760857,
+      "learning_rate": 8.667927049217997e-06,
+      "loss": 0.5639,
+      "step": 2101
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8358536422213843,
+      "learning_rate": 8.66656095345819e-06,
+      "loss": 0.4962,
+      "step": 2102
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7067416368535604,
+      "learning_rate": 8.665194265333167e-06,
+      "loss": 0.5317,
+      "step": 2103
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3435245189908687,
+      "learning_rate": 8.66382698506373e-06,
+      "loss": 0.5432,
+      "step": 2104
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5260755140956461,
+      "learning_rate": 8.662459112870777e-06,
+      "loss": 0.4877,
+      "step": 2105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8155532123302065,
+      "learning_rate": 8.661090648975297e-06,
+      "loss": 0.5655,
+      "step": 2106
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4194970138329932,
+      "learning_rate": 8.659721593598379e-06,
+      "loss": 0.5591,
+      "step": 2107
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2344258030946587,
+      "learning_rate": 8.658351946961206e-06,
+      "loss": 0.4829,
+      "step": 2108
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5723751549266327,
+      "learning_rate": 8.656981709285054e-06,
+      "loss": 0.5254,
+      "step": 2109
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.204612264343951,
+      "learning_rate": 8.655610880791303e-06,
+      "loss": 0.4789,
+      "step": 2110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3796945954800486,
+      "learning_rate": 8.654239461701414e-06,
+      "loss": 0.587,
+      "step": 2111
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.526936464451939,
+      "learning_rate": 8.65286745223696e-06,
+      "loss": 0.5601,
+      "step": 2112
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5491201410612896,
+      "learning_rate": 8.651494852619596e-06,
+      "loss": 0.5303,
+      "step": 2113
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3635548901742522,
+      "learning_rate": 8.650121663071079e-06,
+      "loss": 0.4809,
+      "step": 2114
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6497764251763823,
+      "learning_rate": 8.64874788381326e-06,
+      "loss": 0.4995,
+      "step": 2115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4728552317424948,
+      "learning_rate": 8.647373515068085e-06,
+      "loss": 0.4796,
+      "step": 2116
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3439837298838473,
+      "learning_rate": 8.645998557057595e-06,
+      "loss": 0.5647,
+      "step": 2117
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6876273724296037,
+      "learning_rate": 8.644623010003928e-06,
+      "loss": 0.5131,
+      "step": 2118
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5868033275463422,
+      "learning_rate": 8.643246874129316e-06,
+      "loss": 0.5644,
+      "step": 2119
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.511917733457017,
+      "learning_rate": 8.641870149656082e-06,
+      "loss": 0.5396,
+      "step": 2120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3439071166591285,
+      "learning_rate": 8.64049283680665e-06,
+      "loss": 0.5074,
+      "step": 2121
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2768752235566887,
+      "learning_rate": 8.639114935803542e-06,
+      "loss": 0.5101,
+      "step": 2122
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3351149989408924,
+      "learning_rate": 8.637736446869362e-06,
+      "loss": 0.5379,
+      "step": 2123
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7684832724906134,
+      "learning_rate": 8.63635737022682e-06,
+      "loss": 0.506,
+      "step": 2124
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5063907422372096,
+      "learning_rate": 8.634977706098721e-06,
+      "loss": 0.5006,
+      "step": 2125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3146592975273879,
+      "learning_rate": 8.633597454707959e-06,
+      "loss": 0.5306,
+      "step": 2126
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3187236273362868,
+      "learning_rate": 8.632216616277527e-06,
+      "loss": 0.4807,
+      "step": 2127
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.077537638693262,
+      "learning_rate": 8.630835191030508e-06,
+      "loss": 0.5125,
+      "step": 2128
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4717538690131045,
+      "learning_rate": 8.62945317919009e-06,
+      "loss": 0.5009,
+      "step": 2129
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5072010277133576,
+      "learning_rate": 8.628070580979544e-06,
+      "loss": 0.5108,
+      "step": 2130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7909802743135568,
+      "learning_rate": 8.626687396622242e-06,
+      "loss": 0.4817,
+      "step": 2131
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.522209883270738,
+      "learning_rate": 8.62530362634165e-06,
+      "loss": 0.5689,
+      "step": 2132
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6072768318168984,
+      "learning_rate": 8.623919270361329e-06,
+      "loss": 0.4947,
+      "step": 2133
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.479690164963758,
+      "learning_rate": 8.622534328904932e-06,
+      "loss": 0.5179,
+      "step": 2134
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2871577638299292,
+      "learning_rate": 8.621148802196211e-06,
+      "loss": 0.4856,
+      "step": 2135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6382205250968358,
+      "learning_rate": 8.619762690459008e-06,
+      "loss": 0.5136,
+      "step": 2136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6979597886359754,
+      "learning_rate": 8.618375993917263e-06,
+      "loss": 0.4986,
+      "step": 2137
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3555980447677438,
+      "learning_rate": 8.616988712795008e-06,
+      "loss": 0.4785,
+      "step": 2138
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.748291939308693,
+      "learning_rate": 8.615600847316372e-06,
+      "loss": 0.5331,
+      "step": 2139
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.00088888375592,
+      "learning_rate": 8.614212397705575e-06,
+      "loss": 0.5701,
+      "step": 2140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.573457827272612,
+      "learning_rate": 8.612823364186933e-06,
+      "loss": 0.4721,
+      "step": 2141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.551083989047537,
+      "learning_rate": 8.611433746984858e-06,
+      "loss": 0.5552,
+      "step": 2142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4912064054984209,
+      "learning_rate": 8.610043546323855e-06,
+      "loss": 0.5488,
+      "step": 2143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2760824885449171,
+      "learning_rate": 8.608652762428521e-06,
+      "loss": 0.5275,
+      "step": 2144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5981452550912965,
+      "learning_rate": 8.607261395523554e-06,
+      "loss": 0.4953,
+      "step": 2145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4329804648625786,
+      "learning_rate": 8.605869445833737e-06,
+      "loss": 0.5042,
+      "step": 2146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3602541838086386,
+      "learning_rate": 8.604476913583955e-06,
+      "loss": 0.5388,
+      "step": 2147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7925939709059873,
+      "learning_rate": 8.60308379899918e-06,
+      "loss": 0.5228,
+      "step": 2148
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2888107219051665,
+      "learning_rate": 8.601690102304486e-06,
+      "loss": 0.5206,
+      "step": 2149
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3557563294974224,
+      "learning_rate": 8.600295823725033e-06,
+      "loss": 0.5455,
+      "step": 2150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3427580417032645,
+      "learning_rate": 8.598900963486083e-06,
+      "loss": 0.5269,
+      "step": 2151
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4120668579056157,
+      "learning_rate": 8.597505521812984e-06,
+      "loss": 0.5046,
+      "step": 2152
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3108192996952082,
+      "learning_rate": 8.596109498931185e-06,
+      "loss": 0.5103,
+      "step": 2153
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5495240706755715,
+      "learning_rate": 8.594712895066226e-06,
+      "loss": 0.4984,
+      "step": 2154
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3005502095759358,
+      "learning_rate": 8.593315710443739e-06,
+      "loss": 0.549,
+      "step": 2155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4965928633369276,
+      "learning_rate": 8.59191794528945e-06,
+      "loss": 0.4768,
+      "step": 2156
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3650726161121465,
+      "learning_rate": 8.590519599829184e-06,
+      "loss": 0.5329,
+      "step": 2157
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4564801871656392,
+      "learning_rate": 8.589120674288853e-06,
+      "loss": 0.4863,
+      "step": 2158
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.632300836612375,
+      "learning_rate": 8.587721168894466e-06,
+      "loss": 0.5334,
+      "step": 2159
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1623408678050393,
+      "learning_rate": 8.58632108387213e-06,
+      "loss": 0.4864,
+      "step": 2160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1603521892398447,
+      "learning_rate": 8.584920419448035e-06,
+      "loss": 0.4918,
+      "step": 2161
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2579644869776925,
+      "learning_rate": 8.583519175848474e-06,
+      "loss": 0.4942,
+      "step": 2162
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.129015625363322,
+      "learning_rate": 8.582117353299828e-06,
+      "loss": 0.469,
+      "step": 2163
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.924953776540199,
+      "learning_rate": 8.580714952028575e-06,
+      "loss": 0.5748,
+      "step": 2164
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.507743860015434,
+      "learning_rate": 8.579311972261286e-06,
+      "loss": 0.5841,
+      "step": 2165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.480253101467622,
+      "learning_rate": 8.577908414224625e-06,
+      "loss": 0.5915,
+      "step": 2166
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3203271426351713,
+      "learning_rate": 8.576504278145348e-06,
+      "loss": 0.4924,
+      "step": 2167
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2872361086433888,
+      "learning_rate": 8.575099564250304e-06,
+      "loss": 0.5593,
+      "step": 2168
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4261526727740197,
+      "learning_rate": 8.57369427276644e-06,
+      "loss": 0.5528,
+      "step": 2169
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.435250926712347,
+      "learning_rate": 8.572288403920791e-06,
+      "loss": 0.5494,
+      "step": 2170
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3595312099940866,
+      "learning_rate": 8.570881957940491e-06,
+      "loss": 0.5357,
+      "step": 2171
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7425534944356988,
+      "learning_rate": 8.56947493505276e-06,
+      "loss": 0.5608,
+      "step": 2172
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5923240614789778,
+      "learning_rate": 8.568067335484915e-06,
+      "loss": 0.5255,
+      "step": 2173
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3081107286282319,
+      "learning_rate": 8.566659159464367e-06,
+      "loss": 0.5069,
+      "step": 2174
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2351708937483399,
+      "learning_rate": 8.565250407218622e-06,
+      "loss": 0.506,
+      "step": 2175
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5097611716714756,
+      "learning_rate": 8.563841078975273e-06,
+      "loss": 0.5172,
+      "step": 2176
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.71048705718627,
+      "learning_rate": 8.562431174962009e-06,
+      "loss": 0.5111,
+      "step": 2177
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5677375671906304,
+      "learning_rate": 8.561020695406614e-06,
+      "loss": 0.5155,
+      "step": 2178
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5195255872079538,
+      "learning_rate": 8.559609640536962e-06,
+      "loss": 0.5085,
+      "step": 2179
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.541472808973938,
+      "learning_rate": 8.558198010581022e-06,
+      "loss": 0.5067,
+      "step": 2180
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5948220379252283,
+      "learning_rate": 8.556785805766859e-06,
+      "loss": 0.4828,
+      "step": 2181
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3828488243455312,
+      "learning_rate": 8.55537302632262e-06,
+      "loss": 0.5429,
+      "step": 2182
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6313382434068164,
+      "learning_rate": 8.553959672476558e-06,
+      "loss": 0.4702,
+      "step": 2183
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9515593617267215,
+      "learning_rate": 8.55254574445701e-06,
+      "loss": 0.5574,
+      "step": 2184
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6291863643827857,
+      "learning_rate": 8.551131242492407e-06,
+      "loss": 0.5196,
+      "step": 2185
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4610097753460443,
+      "learning_rate": 8.54971616681128e-06,
+      "loss": 0.486,
+      "step": 2186
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.243411626461023,
+      "learning_rate": 8.54830051764224e-06,
+      "loss": 0.4869,
+      "step": 2187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5254988279151824,
+      "learning_rate": 8.546884295214002e-06,
+      "loss": 0.5065,
+      "step": 2188
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.348325328290899,
+      "learning_rate": 8.545467499755366e-06,
+      "loss": 0.5287,
+      "step": 2189
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4484265724296916,
+      "learning_rate": 8.544050131495233e-06,
+      "loss": 0.4964,
+      "step": 2190
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3966660469184176,
+      "learning_rate": 8.542632190662586e-06,
+      "loss": 0.5599,
+      "step": 2191
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7947891690402593,
+      "learning_rate": 8.541213677486509e-06,
+      "loss": 0.5735,
+      "step": 2192
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.306427889420237,
+      "learning_rate": 8.539794592196173e-06,
+      "loss": 0.5341,
+      "step": 2193
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0480279716104333,
+      "learning_rate": 8.538374935020846e-06,
+      "loss": 0.5067,
+      "step": 2194
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3681137444792715,
+      "learning_rate": 8.536954706189883e-06,
+      "loss": 0.5132,
+      "step": 2195
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7656680141879872,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.5815,
+      "step": 2196
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2767603247891768,
+      "learning_rate": 8.534112534478953e-06,
+      "loss": 0.4554,
+      "step": 2197
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5866330370451214,
+      "learning_rate": 8.532690592058161e-06,
+      "loss": 0.5033,
+      "step": 2198
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9089032305015174,
+      "learning_rate": 8.53126807890009e-06,
+      "loss": 0.5001,
+      "step": 2199
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.474155149864476,
+      "learning_rate": 8.529844995234563e-06,
+      "loss": 0.5069,
+      "step": 2200
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4499060008784328,
+      "learning_rate": 8.528421341291488e-06,
+      "loss": 0.5084,
+      "step": 2201
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3897034097569219,
+      "learning_rate": 8.526997117300868e-06,
+      "loss": 0.5091,
+      "step": 2202
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3914446118557149,
+      "learning_rate": 8.525572323492803e-06,
+      "loss": 0.5164,
+      "step": 2203
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0644901750892553,
+      "learning_rate": 8.524146960097476e-06,
+      "loss": 0.4661,
+      "step": 2204
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5115707604600233,
+      "learning_rate": 8.522721027345173e-06,
+      "loss": 0.5254,
+      "step": 2205
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.529006318561144,
+      "learning_rate": 8.52129452546626e-06,
+      "loss": 0.5786,
+      "step": 2206
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6774863769516049,
+      "learning_rate": 8.519867454691204e-06,
+      "loss": 0.505,
+      "step": 2207
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7550570033418513,
+      "learning_rate": 8.518439815250561e-06,
+      "loss": 0.5145,
+      "step": 2208
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8311153441470829,
+      "learning_rate": 8.517011607374978e-06,
+      "loss": 0.494,
+      "step": 2209
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.508065531312494,
+      "learning_rate": 8.515582831295195e-06,
+      "loss": 0.5514,
+      "step": 2210
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8391102633980023,
+      "learning_rate": 8.514153487242042e-06,
+      "loss": 0.5094,
+      "step": 2211
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6333778269326167,
+      "learning_rate": 8.512723575446446e-06,
+      "loss": 0.5497,
+      "step": 2212
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5528173792460196,
+      "learning_rate": 8.511293096139417e-06,
+      "loss": 0.551,
+      "step": 2213
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5720104471169536,
+      "learning_rate": 8.509862049552065e-06,
+      "loss": 0.4421,
+      "step": 2214
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3400011066040898,
+      "learning_rate": 8.508430435915584e-06,
+      "loss": 0.4903,
+      "step": 2215
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5794600856104057,
+      "learning_rate": 8.506998255461269e-06,
+      "loss": 0.494,
+      "step": 2216
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5033180990212154,
+      "learning_rate": 8.505565508420498e-06,
+      "loss": 0.5592,
+      "step": 2217
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.359279658821798,
+      "learning_rate": 8.504132195024747e-06,
+      "loss": 0.546,
+      "step": 2218
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4952159845406447,
+      "learning_rate": 8.502698315505573e-06,
+      "loss": 0.5359,
+      "step": 2219
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5247955931103605,
+      "learning_rate": 8.501263870094642e-06,
+      "loss": 0.5179,
+      "step": 2220
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7782806600137773,
+      "learning_rate": 8.499828859023696e-06,
+      "loss": 0.5474,
+      "step": 2221
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5995731206263903,
+      "learning_rate": 8.498393282524572e-06,
+      "loss": 0.4935,
+      "step": 2222
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2088287114101004,
+      "learning_rate": 8.496957140829203e-06,
+      "loss": 0.4073,
+      "step": 2223
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0638784444279548,
+      "learning_rate": 8.495520434169609e-06,
+      "loss": 0.5159,
+      "step": 2224
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.497881780594768,
+      "learning_rate": 8.494083162777903e-06,
+      "loss": 0.5605,
+      "step": 2225
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.413788779749229,
+      "learning_rate": 8.492645326886291e-06,
+      "loss": 0.5486,
+      "step": 2226
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.654873395286464,
+      "learning_rate": 8.491206926727064e-06,
+      "loss": 0.5209,
+      "step": 2227
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6926087836331076,
+      "learning_rate": 8.489767962532611e-06,
+      "loss": 0.5036,
+      "step": 2228
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.422679228906343,
+      "learning_rate": 8.488328434535408e-06,
+      "loss": 0.5435,
+      "step": 2229
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1664800474007877,
+      "learning_rate": 8.486888342968023e-06,
+      "loss": 0.4997,
+      "step": 2230
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4461426110457904,
+      "learning_rate": 8.485447688063117e-06,
+      "loss": 0.522,
+      "step": 2231
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1007331833215424,
+      "learning_rate": 8.484006470053441e-06,
+      "loss": 0.498,
+      "step": 2232
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3641751224103698,
+      "learning_rate": 8.482564689171834e-06,
+      "loss": 0.4979,
+      "step": 2233
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5884561346617245,
+      "learning_rate": 8.481122345651233e-06,
+      "loss": 0.5074,
+      "step": 2234
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4192325095037206,
+      "learning_rate": 8.479679439724654e-06,
+      "loss": 0.5471,
+      "step": 2235
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5989179484159823,
+      "learning_rate": 8.478235971625218e-06,
+      "loss": 0.5239,
+      "step": 2236
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8668321080947397,
+      "learning_rate": 8.476791941586126e-06,
+      "loss": 0.5037,
+      "step": 2237
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5158066815377742,
+      "learning_rate": 8.475347349840674e-06,
+      "loss": 0.5369,
+      "step": 2238
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3645982218539816,
+      "learning_rate": 8.473902196622252e-06,
+      "loss": 0.5108,
+      "step": 2239
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.265074006713083,
+      "learning_rate": 8.472456482164332e-06,
+      "loss": 0.498,
+      "step": 2240
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2999038326538637,
+      "learning_rate": 8.471010206700488e-06,
+      "loss": 0.5254,
+      "step": 2241
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6060369052945243,
+      "learning_rate": 8.469563370464372e-06,
+      "loss": 0.5791,
+      "step": 2242
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.426987421889657,
+      "learning_rate": 8.468115973689739e-06,
+      "loss": 0.5261,
+      "step": 2243
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3635787901340615,
+      "learning_rate": 8.466668016610423e-06,
+      "loss": 0.5393,
+      "step": 2244
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.369209352696687,
+      "learning_rate": 8.46521949946036e-06,
+      "loss": 0.508,
+      "step": 2245
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8226736763186273,
+      "learning_rate": 8.463770422473566e-06,
+      "loss": 0.5892,
+      "step": 2246
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5394075544765462,
+      "learning_rate": 8.462320785884155e-06,
+      "loss": 0.5151,
+      "step": 2247
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.815742457072043,
+      "learning_rate": 8.460870589926327e-06,
+      "loss": 0.5484,
+      "step": 2248
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4875209217946046,
+      "learning_rate": 8.459419834834374e-06,
+      "loss": 0.5414,
+      "step": 2249
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.274126755614334,
+      "learning_rate": 8.45796852084268e-06,
+      "loss": 0.4801,
+      "step": 2250
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4106176029586979,
+      "learning_rate": 8.456516648185717e-06,
+      "loss": 0.4854,
+      "step": 2251
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.621250166058811,
+      "learning_rate": 8.455064217098046e-06,
+      "loss": 0.5174,
+      "step": 2252
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6601874583519945,
+      "learning_rate": 8.453611227814322e-06,
+      "loss": 0.496,
+      "step": 2253
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1827510625394257,
+      "learning_rate": 8.452157680569287e-06,
+      "loss": 0.5282,
+      "step": 2254
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4471288796504076,
+      "learning_rate": 8.450703575597775e-06,
+      "loss": 0.4927,
+      "step": 2255
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.569274387714447,
+      "learning_rate": 8.449248913134709e-06,
+      "loss": 0.5473,
+      "step": 2256
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.221832571991313,
+      "learning_rate": 8.447793693415103e-06,
+      "loss": 0.5164,
+      "step": 2257
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4929632157250627,
+      "learning_rate": 8.446337916674062e-06,
+      "loss": 0.5397,
+      "step": 2258
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4075589123438812,
+      "learning_rate": 8.444881583146776e-06,
+      "loss": 0.5151,
+      "step": 2259
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.2567410524008205,
+      "learning_rate": 8.44342469306853e-06,
+      "loss": 0.5171,
+      "step": 2260
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.27510092229832,
+      "learning_rate": 8.441967246674698e-06,
+      "loss": 0.5706,
+      "step": 2261
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7994978265127097,
+      "learning_rate": 8.440509244200743e-06,
+      "loss": 0.5459,
+      "step": 2262
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6112199363869086,
+      "learning_rate": 8.43905068588222e-06,
+      "loss": 0.5131,
+      "step": 2263
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9579492063060595,
+      "learning_rate": 8.437591571954768e-06,
+      "loss": 0.4655,
+      "step": 2264
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6073308539208877,
+      "learning_rate": 8.436131902654123e-06,
+      "loss": 0.5265,
+      "step": 2265
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7586176288861322,
+      "learning_rate": 8.43467167821611e-06,
+      "loss": 0.5683,
+      "step": 2266
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3915881330613307,
+      "learning_rate": 8.433210898876632e-06,
+      "loss": 0.4483,
+      "step": 2267
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8019170267996496,
+      "learning_rate": 8.431749564871698e-06,
+      "loss": 0.5564,
+      "step": 2268
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1320048897516775,
+      "learning_rate": 8.430287676437399e-06,
+      "loss": 0.5397,
+      "step": 2269
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.967697613735462,
+      "learning_rate": 8.428825233809914e-06,
+      "loss": 0.5517,
+      "step": 2270
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.446804864822726,
+      "learning_rate": 8.427362237225513e-06,
+      "loss": 0.5354,
+      "step": 2271
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5128368460256831,
+      "learning_rate": 8.425898686920557e-06,
+      "loss": 0.5594,
+      "step": 2272
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3255339105369666,
+      "learning_rate": 8.424434583131496e-06,
+      "loss": 0.4988,
+      "step": 2273
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.1360665417876894,
+      "learning_rate": 8.42296992609487e-06,
+      "loss": 0.5444,
+      "step": 2274
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4927309323485651,
+      "learning_rate": 8.421504716047305e-06,
+      "loss": 0.5291,
+      "step": 2275
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4284228996787336,
+      "learning_rate": 8.420038953225518e-06,
+      "loss": 0.4888,
+      "step": 2276
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.98282046193224,
+      "learning_rate": 8.418572637866316e-06,
+      "loss": 0.4727,
+      "step": 2277
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3807713542293747,
+      "learning_rate": 8.417105770206598e-06,
+      "loss": 0.523,
+      "step": 2278
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8528045536909103,
+      "learning_rate": 8.415638350483348e-06,
+      "loss": 0.5091,
+      "step": 2279
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.632035367705425,
+      "learning_rate": 8.41417037893364e-06,
+      "loss": 0.5403,
+      "step": 2280
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7208618427001102,
+      "learning_rate": 8.412701855794637e-06,
+      "loss": 0.5022,
+      "step": 2281
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4717548446224873,
+      "learning_rate": 8.411232781303593e-06,
+      "loss": 0.4975,
+      "step": 2282
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4415037838550437,
+      "learning_rate": 8.409763155697852e-06,
+      "loss": 0.553,
+      "step": 2283
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6165367478786659,
+      "learning_rate": 8.40829297921484e-06,
+      "loss": 0.4942,
+      "step": 2284
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4188139758484897,
+      "learning_rate": 8.40682225209208e-06,
+      "loss": 0.4914,
+      "step": 2285
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.0225226153497595,
+      "learning_rate": 8.405350974567182e-06,
+      "loss": 0.4985,
+      "step": 2286
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6167432111333073,
+      "learning_rate": 8.403879146877841e-06,
+      "loss": 0.4977,
+      "step": 2287
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5411985035248754,
+      "learning_rate": 8.402406769261846e-06,
+      "loss": 0.563,
+      "step": 2288
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4103723311543415,
+      "learning_rate": 8.400933841957072e-06,
+      "loss": 0.4985,
+      "step": 2289
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4687767595342092,
+      "learning_rate": 8.399460365201481e-06,
+      "loss": 0.4976,
+      "step": 2290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9856889826237913,
+      "learning_rate": 8.397986339233128e-06,
+      "loss": 0.5085,
+      "step": 2291
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7220986166950227,
+      "learning_rate": 8.396511764290158e-06,
+      "loss": 0.5138,
+      "step": 2292
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.421132299453895,
+      "learning_rate": 8.395036640610796e-06,
+      "loss": 0.5148,
+      "step": 2293
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8056847578745472,
+      "learning_rate": 8.393560968433366e-06,
+      "loss": 0.5342,
+      "step": 2294
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5731443883103524,
+      "learning_rate": 8.392084747996275e-06,
+      "loss": 0.4951,
+      "step": 2295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.265804124392798,
+      "learning_rate": 8.390607979538014e-06,
+      "loss": 0.4872,
+      "step": 2296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.297878688851731,
+      "learning_rate": 8.389130663297175e-06,
+      "loss": 0.5321,
+      "step": 2297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8342068119319483,
+      "learning_rate": 8.387652799512427e-06,
+      "loss": 0.5113,
+      "step": 2298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.34842173900626,
+      "learning_rate": 8.386174388422535e-06,
+      "loss": 0.531,
+      "step": 2299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.59218032004979,
+      "learning_rate": 8.384695430266348e-06,
+      "loss": 0.5145,
+      "step": 2300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5341020416331983,
+      "learning_rate": 8.383215925282802e-06,
+      "loss": 0.5395,
+      "step": 2301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7895960441730228,
+      "learning_rate": 8.381735873710928e-06,
+      "loss": 0.4936,
+      "step": 2302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5468634833190629,
+      "learning_rate": 8.38025527578984e-06,
+      "loss": 0.4864,
+      "step": 2303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7169385720661268,
+      "learning_rate": 8.378774131758742e-06,
+      "loss": 0.5151,
+      "step": 2304
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7764810966446116,
+      "learning_rate": 8.377292441856926e-06,
+      "loss": 0.5585,
+      "step": 2305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4247792641176793,
+      "learning_rate": 8.37581020632377e-06,
+      "loss": 0.5499,
+      "step": 2306
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.24826140127667,
+      "learning_rate": 8.374327425398744e-06,
+      "loss": 0.5943,
+      "step": 2307
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5190074639788063,
+      "learning_rate": 8.372844099321404e-06,
+      "loss": 0.4983,
+      "step": 2308
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3870284274978932,
+      "learning_rate": 8.371360228331393e-06,
+      "loss": 0.4672,
+      "step": 2309
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8711754548899875,
+      "learning_rate": 8.369875812668449e-06,
+      "loss": 0.5546,
+      "step": 2310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.444759030629122,
+      "learning_rate": 8.368390852572384e-06,
+      "loss": 0.5414,
+      "step": 2311
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2826081853724405,
+      "learning_rate": 8.366905348283114e-06,
+      "loss": 0.4739,
+      "step": 2312
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6215351080572908,
+      "learning_rate": 8.365419300040628e-06,
+      "loss": 0.4838,
+      "step": 2313
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.368664685893942,
+      "learning_rate": 8.363932708085016e-06,
+      "loss": 0.4851,
+      "step": 2314
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.268709114460342,
+      "learning_rate": 8.362445572656451e-06,
+      "loss": 0.4537,
+      "step": 2315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2595580792264183,
+      "learning_rate": 8.360957893995187e-06,
+      "loss": 0.5065,
+      "step": 2316
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3180728188145547,
+      "learning_rate": 8.359469672341574e-06,
+      "loss": 0.449,
+      "step": 2317
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3748707275140444,
+      "learning_rate": 8.357980907936048e-06,
+      "loss": 0.5071,
+      "step": 2318
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2182849061073715,
+      "learning_rate": 8.356491601019135e-06,
+      "loss": 0.5262,
+      "step": 2319
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8497839329843278,
+      "learning_rate": 8.35500175183144e-06,
+      "loss": 0.5182,
+      "step": 2320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7064279805028324,
+      "learning_rate": 8.353511360613665e-06,
+      "loss": 0.488,
+      "step": 2321
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.696234532399231,
+      "learning_rate": 8.352020427606591e-06,
+      "loss": 0.5753,
+      "step": 2322
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6188147545472062,
+      "learning_rate": 8.350528953051098e-06,
+      "loss": 0.5407,
+      "step": 2323
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3692002533058387,
+      "learning_rate": 8.349036937188143e-06,
+      "loss": 0.5252,
+      "step": 2324
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.443875711526957,
+      "learning_rate": 8.347544380258777e-06,
+      "loss": 0.5334,
+      "step": 2325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4178971172449506,
+      "learning_rate": 8.34605128250413e-06,
+      "loss": 0.5319,
+      "step": 2326
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2259432902634666,
+      "learning_rate": 8.344557644165431e-06,
+      "loss": 0.5138,
+      "step": 2327
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6778873298316646,
+      "learning_rate": 8.34306346548399e-06,
+      "loss": 0.5238,
+      "step": 2328
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4288498248716177,
+      "learning_rate": 8.341568746701202e-06,
+      "loss": 0.5337,
+      "step": 2329
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5684048901647027,
+      "learning_rate": 8.340073488058552e-06,
+      "loss": 0.4903,
+      "step": 2330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2599485203250966,
+      "learning_rate": 8.338577689797615e-06,
+      "loss": 0.5134,
+      "step": 2331
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4253859816398988,
+      "learning_rate": 8.337081352160048e-06,
+      "loss": 0.505,
+      "step": 2332
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1339071401428553,
+      "learning_rate": 8.335584475387597e-06,
+      "loss": 0.5273,
+      "step": 2333
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4778341026992288,
+      "learning_rate": 8.334087059722097e-06,
+      "loss": 0.4823,
+      "step": 2334
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.63766176270152,
+      "learning_rate": 8.33258910540547e-06,
+      "loss": 0.4946,
+      "step": 2335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.888945724040837,
+      "learning_rate": 8.33109061267972e-06,
+      "loss": 0.5359,
+      "step": 2336
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6516161391520482,
+      "learning_rate": 8.329591581786946e-06,
+      "loss": 0.4986,
+      "step": 2337
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5261816214983497,
+      "learning_rate": 8.328092012969327e-06,
+      "loss": 0.5083,
+      "step": 2338
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4821802013711955,
+      "learning_rate": 8.326591906469132e-06,
+      "loss": 0.5237,
+      "step": 2339
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6683955615857764,
+      "learning_rate": 8.325091262528715e-06,
+      "loss": 0.5683,
+      "step": 2340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.357707619827874,
+      "learning_rate": 8.323590081390522e-06,
+      "loss": 0.5236,
+      "step": 2341
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.457783527864208,
+      "learning_rate": 8.322088363297078e-06,
+      "loss": 0.4967,
+      "step": 2342
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7460405208068128,
+      "learning_rate": 8.320586108491002e-06,
+      "loss": 0.5097,
+      "step": 2343
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7744582128639608,
+      "learning_rate": 8.319083317214996e-06,
+      "loss": 0.5355,
+      "step": 2344
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5390306253936152,
+      "learning_rate": 8.317579989711846e-06,
+      "loss": 0.5013,
+      "step": 2345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7215048565142089,
+      "learning_rate": 8.31607612622443e-06,
+      "loss": 0.4965,
+      "step": 2346
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4736527077899817,
+      "learning_rate": 8.314571726995711e-06,
+      "loss": 0.5626,
+      "step": 2347
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8277857445033219,
+      "learning_rate": 8.313066792268737e-06,
+      "loss": 0.5266,
+      "step": 2348
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6391679556371148,
+      "learning_rate": 8.311561322286645e-06,
+      "loss": 0.5405,
+      "step": 2349
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4251390271552782,
+      "learning_rate": 8.310055317292656e-06,
+      "loss": 0.4482,
+      "step": 2350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4518548043395432,
+      "learning_rate": 8.308548777530077e-06,
+      "loss": 0.5224,
+      "step": 2351
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.529537909676569,
+      "learning_rate": 8.307041703242305e-06,
+      "loss": 0.4757,
+      "step": 2352
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4641076631316765,
+      "learning_rate": 8.305534094672818e-06,
+      "loss": 0.5606,
+      "step": 2353
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6459680091281955,
+      "learning_rate": 8.304025952065187e-06,
+      "loss": 0.4939,
+      "step": 2354
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5502872604777798,
+      "learning_rate": 8.302517275663063e-06,
+      "loss": 0.5186,
+      "step": 2355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4767675987142823,
+      "learning_rate": 8.301008065710188e-06,
+      "loss": 0.5366,
+      "step": 2356
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4597094379451114,
+      "learning_rate": 8.299498322450388e-06,
+      "loss": 0.5207,
+      "step": 2357
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5145188084629189,
+      "learning_rate": 8.297988046127574e-06,
+      "loss": 0.468,
+      "step": 2358
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5879209005951993,
+      "learning_rate": 8.296477236985744e-06,
+      "loss": 0.468,
+      "step": 2359
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.389298414107315,
+      "learning_rate": 8.294965895268985e-06,
+      "loss": 0.5107,
+      "step": 2360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.367116410018487,
+      "learning_rate": 8.293454021221466e-06,
+      "loss": 0.4894,
+      "step": 2361
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4671157811272946,
+      "learning_rate": 8.291941615087442e-06,
+      "loss": 0.5283,
+      "step": 2362
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.500644862850445,
+      "learning_rate": 8.290428677111258e-06,
+      "loss": 0.4954,
+      "step": 2363
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4648161938444693,
+      "learning_rate": 8.288915207537343e-06,
+      "loss": 0.4996,
+      "step": 2364
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4305139064341361,
+      "learning_rate": 8.28740120661021e-06,
+      "loss": 0.4865,
+      "step": 2365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5603416367884653,
+      "learning_rate": 8.285886674574459e-06,
+      "loss": 0.4875,
+      "step": 2366
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2213046850748692,
+      "learning_rate": 8.284371611674776e-06,
+      "loss": 0.4688,
+      "step": 2367
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.212734451608669,
+      "learning_rate": 8.282856018155932e-06,
+      "loss": 0.4468,
+      "step": 2368
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4988856306143041,
+      "learning_rate": 8.281339894262786e-06,
+      "loss": 0.5517,
+      "step": 2369
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9316901312270092,
+      "learning_rate": 8.279823240240282e-06,
+      "loss": 0.4647,
+      "step": 2370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6550193444792216,
+      "learning_rate": 8.278306056333445e-06,
+      "loss": 0.5398,
+      "step": 2371
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6404194777907943,
+      "learning_rate": 8.276788342787394e-06,
+      "loss": 0.5123,
+      "step": 2372
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.315161270026293,
+      "learning_rate": 8.275270099847325e-06,
+      "loss": 0.5449,
+      "step": 2373
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.517240516199405,
+      "learning_rate": 8.273751327758526e-06,
+      "loss": 0.5153,
+      "step": 2374
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.2174133228862587,
+      "learning_rate": 8.272232026766368e-06,
+      "loss": 0.491,
+      "step": 2375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4053448725891333,
+      "learning_rate": 8.270712197116306e-06,
+      "loss": 0.5205,
+      "step": 2376
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9740569155576038,
+      "learning_rate": 8.269191839053884e-06,
+      "loss": 0.4661,
+      "step": 2377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.672953358964489,
+      "learning_rate": 8.267670952824726e-06,
+      "loss": 0.5081,
+      "step": 2378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.077306245332074,
+      "learning_rate": 8.266149538674548e-06,
+      "loss": 0.4973,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.409380624306681,
+      "learning_rate": 8.264627596849146e-06,
+      "loss": 0.5062,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8358787110603083,
+      "learning_rate": 8.263105127594405e-06,
+      "loss": 0.5327,
+      "step": 2381
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4579310672276335,
+      "learning_rate": 8.261582131156289e-06,
+      "loss": 0.4562,
+      "step": 2382
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4387234091141146,
+      "learning_rate": 8.260058607780857e-06,
+      "loss": 0.5368,
+      "step": 2383
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.285121771570786,
+      "learning_rate": 8.258534557714242e-06,
+      "loss": 0.5482,
+      "step": 2384
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4535365064196426,
+      "learning_rate": 8.257009981202673e-06,
+      "loss": 0.539,
+      "step": 2385
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4290733161456042,
+      "learning_rate": 8.255484878492454e-06,
+      "loss": 0.5029,
+      "step": 2386
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3993424374411836,
+      "learning_rate": 8.253959249829983e-06,
+      "loss": 0.5006,
+      "step": 2387
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6216777802783738,
+      "learning_rate": 8.252433095461736e-06,
+      "loss": 0.5374,
+      "step": 2388
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8538060165988877,
+      "learning_rate": 8.250906415634279e-06,
+      "loss": 0.493,
+      "step": 2389
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4053489404747652,
+      "learning_rate": 8.249379210594258e-06,
+      "loss": 0.5127,
+      "step": 2390
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.824381664090078,
+      "learning_rate": 8.247851480588407e-06,
+      "loss": 0.4867,
+      "step": 2391
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4129574000855118,
+      "learning_rate": 8.246323225863545e-06,
+      "loss": 0.5446,
+      "step": 2392
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4106399327103274,
+      "learning_rate": 8.244794446666575e-06,
+      "loss": 0.4639,
+      "step": 2393
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.449898253618827,
+      "learning_rate": 8.243265143244485e-06,
+      "loss": 0.5184,
+      "step": 2394
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4255742042291377,
+      "learning_rate": 8.241735315844348e-06,
+      "loss": 0.5316,
+      "step": 2395
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.495255075502373,
+      "learning_rate": 8.240204964713317e-06,
+      "loss": 0.4794,
+      "step": 2396
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.19925092790902,
+      "learning_rate": 8.238674090098639e-06,
+      "loss": 0.4861,
+      "step": 2397
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2584502352522227,
+      "learning_rate": 8.237142692247637e-06,
+      "loss": 0.5453,
+      "step": 2398
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7706569851686855,
+      "learning_rate": 8.235610771407725e-06,
+      "loss": 0.5128,
+      "step": 2399
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.393286411135955,
+      "learning_rate": 8.234078327826394e-06,
+      "loss": 0.5132,
+      "step": 2400
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5911983819549134,
+      "learning_rate": 8.232545361751227e-06,
+      "loss": 0.5423,
+      "step": 2401
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2768352893732755,
+      "learning_rate": 8.231011873429887e-06,
+      "loss": 0.5027,
+      "step": 2402
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6464670337009326,
+      "learning_rate": 8.22947786311012e-06,
+      "loss": 0.5297,
+      "step": 2403
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.277057973772232,
+      "learning_rate": 8.227943331039765e-06,
+      "loss": 0.53,
+      "step": 2404
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5165937159294598,
+      "learning_rate": 8.226408277466735e-06,
+      "loss": 0.5029,
+      "step": 2405
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6307182302388195,
+      "learning_rate": 8.22487270263903e-06,
+      "loss": 0.5076,
+      "step": 2406
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4016704206015314,
+      "learning_rate": 8.22333660680474e-06,
+      "loss": 0.4607,
+      "step": 2407
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5733249006556254,
+      "learning_rate": 8.221799990212031e-06,
+      "loss": 0.5086,
+      "step": 2408
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.428256429738577,
+      "learning_rate": 8.22026285310916e-06,
+      "loss": 0.515,
+      "step": 2409
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.974071549092424,
+      "learning_rate": 8.218725195744464e-06,
+      "loss": 0.5051,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3210469873226283,
+      "learning_rate": 8.217187018366364e-06,
+      "loss": 0.5219,
+      "step": 2411
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7354476740515565,
+      "learning_rate": 8.215648321223363e-06,
+      "loss": 0.5185,
+      "step": 2412
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3385225264897804,
+      "learning_rate": 8.21410910456406e-06,
+      "loss": 0.5344,
+      "step": 2413
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5655282839619544,
+      "learning_rate": 8.212569368637123e-06,
+      "loss": 0.5764,
+      "step": 2414
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.838377707828001,
+      "learning_rate": 8.21102911369131e-06,
+      "loss": 0.5015,
+      "step": 2415
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3334971659877586,
+      "learning_rate": 8.209488339975461e-06,
+      "loss": 0.5422,
+      "step": 2416
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.601034277173312,
+      "learning_rate": 8.207947047738508e-06,
+      "loss": 0.5079,
+      "step": 2417
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.732019949409373,
+      "learning_rate": 8.206405237229453e-06,
+      "loss": 0.5537,
+      "step": 2418
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.600774736304845,
+      "learning_rate": 8.204862908697396e-06,
+      "loss": 0.5462,
+      "step": 2419
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2915947660919527,
+      "learning_rate": 8.203320062391506e-06,
+      "loss": 0.4483,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.482367313516658,
+      "learning_rate": 8.201776698561049e-06,
+      "loss": 0.5089,
+      "step": 2421
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.379824868329647,
+      "learning_rate": 8.200232817455369e-06,
+      "loss": 0.522,
+      "step": 2422
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6160975125022232,
+      "learning_rate": 8.198688419323893e-06,
+      "loss": 0.5134,
+      "step": 2423
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4497050493829207,
+      "learning_rate": 8.197143504416127e-06,
+      "loss": 0.5213,
+      "step": 2424
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.370280107513277,
+      "learning_rate": 8.195598072981674e-06,
+      "loss": 0.4872,
+      "step": 2425
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5170579867779488,
+      "learning_rate": 8.194052125270207e-06,
+      "loss": 0.5569,
+      "step": 2426
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0868880854014726,
+      "learning_rate": 8.192505661531489e-06,
+      "loss": 0.5531,
+      "step": 2427
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.596934387560802,
+      "learning_rate": 8.190958682015362e-06,
+      "loss": 0.4936,
+      "step": 2428
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6234727453741495,
+      "learning_rate": 8.189411186971759e-06,
+      "loss": 0.4954,
+      "step": 2429
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3541932266704004,
+      "learning_rate": 8.187863176650688e-06,
+      "loss": 0.5169,
+      "step": 2430
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7951463915061205,
+      "learning_rate": 8.186314651302242e-06,
+      "loss": 0.4556,
+      "step": 2431
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3940159809171142,
+      "learning_rate": 8.184765611176605e-06,
+      "loss": 0.5454,
+      "step": 2432
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3019380098469693,
+      "learning_rate": 8.183216056524035e-06,
+      "loss": 0.522,
+      "step": 2433
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.233015420492335,
+      "learning_rate": 8.181665987594874e-06,
+      "loss": 0.478,
+      "step": 2434
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.347521272975239,
+      "learning_rate": 8.18011540463955e-06,
+      "loss": 0.5021,
+      "step": 2435
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4394524554290282,
+      "learning_rate": 8.178564307908577e-06,
+      "loss": 0.4458,
+      "step": 2436
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.705192311131923,
+      "learning_rate": 8.177012697652544e-06,
+      "loss": 0.4793,
+      "step": 2437
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5623035789488708,
+      "learning_rate": 8.17546057412213e-06,
+      "loss": 0.5614,
+      "step": 2438
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3216329542885294,
+      "learning_rate": 8.173907937568093e-06,
+      "loss": 0.4732,
+      "step": 2439
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4405423395533385,
+      "learning_rate": 8.172354788241277e-06,
+      "loss": 0.5659,
+      "step": 2440
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4266787372028973,
+      "learning_rate": 8.170801126392602e-06,
+      "loss": 0.533,
+      "step": 2441
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.544729767161987,
+      "learning_rate": 8.169246952273081e-06,
+      "loss": 0.4868,
+      "step": 2442
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7137674103325354,
+      "learning_rate": 8.167692266133804e-06,
+      "loss": 0.4834,
+      "step": 2443
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5556203466575353,
+      "learning_rate": 8.166137068225942e-06,
+      "loss": 0.5492,
+      "step": 2444
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7711514179158319,
+      "learning_rate": 8.164581358800749e-06,
+      "loss": 0.4767,
+      "step": 2445
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5301362267020098,
+      "learning_rate": 8.16302513810957e-06,
+      "loss": 0.5439,
+      "step": 2446
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6901022197817894,
+      "learning_rate": 8.16146840640382e-06,
+      "loss": 0.5311,
+      "step": 2447
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4387929530835506,
+      "learning_rate": 8.159911163935007e-06,
+      "loss": 0.5052,
+      "step": 2448
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4881162518562419,
+      "learning_rate": 8.158353410954715e-06,
+      "loss": 0.4841,
+      "step": 2449
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3232025249317436,
+      "learning_rate": 8.156795147714612e-06,
+      "loss": 0.5394,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1502311983887634,
+      "learning_rate": 8.155236374466452e-06,
+      "loss": 0.5267,
+      "step": 2451
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4184547623606276,
+      "learning_rate": 8.153677091462067e-06,
+      "loss": 0.4806,
+      "step": 2452
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3045598237461515,
+      "learning_rate": 8.15211729895337e-06,
+      "loss": 0.5313,
+      "step": 2453
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4262062024035553,
+      "learning_rate": 8.150556997192366e-06,
+      "loss": 0.5778,
+      "step": 2454
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9812787992571088,
+      "learning_rate": 8.148996186431129e-06,
+      "loss": 0.5107,
+      "step": 2455
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.49217716763415,
+      "learning_rate": 8.147434866921824e-06,
+      "loss": 0.5229,
+      "step": 2456
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5400517482926255,
+      "learning_rate": 8.145873038916696e-06,
+      "loss": 0.5363,
+      "step": 2457
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7288336436232086,
+      "learning_rate": 8.144310702668072e-06,
+      "loss": 0.4677,
+      "step": 2458
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.397832321796737,
+      "learning_rate": 8.142747858428364e-06,
+      "loss": 0.4786,
+      "step": 2459
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6129287143413962,
+      "learning_rate": 8.141184506450058e-06,
+      "loss": 0.5479,
+      "step": 2460
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4428208255741053,
+      "learning_rate": 8.13962064698573e-06,
+      "loss": 0.5064,
+      "step": 2461
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6383815783802438,
+      "learning_rate": 8.138056280288036e-06,
+      "loss": 0.5306,
+      "step": 2462
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7647730587391064,
+      "learning_rate": 8.136491406609712e-06,
+      "loss": 0.4727,
+      "step": 2463
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.424228033452532,
+      "learning_rate": 8.134926026203578e-06,
+      "loss": 0.5626,
+      "step": 2464
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.176560989777953,
+      "learning_rate": 8.133360139322533e-06,
+      "loss": 0.4763,
+      "step": 2465
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4747933472510153,
+      "learning_rate": 8.131793746219563e-06,
+      "loss": 0.5606,
+      "step": 2466
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5579184573495903,
+      "learning_rate": 8.13022684714773e-06,
+      "loss": 0.5483,
+      "step": 2467
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3212477231943376,
+      "learning_rate": 8.128659442360182e-06,
+      "loss": 0.5219,
+      "step": 2468
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4716718393118688,
+      "learning_rate": 8.127091532110147e-06,
+      "loss": 0.5485,
+      "step": 2469
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1543171145452837,
+      "learning_rate": 8.125523116650933e-06,
+      "loss": 0.5366,
+      "step": 2470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9532575568968857,
+      "learning_rate": 8.123954196235932e-06,
+      "loss": 0.4726,
+      "step": 2471
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3300282366699865,
+      "learning_rate": 8.122384771118619e-06,
+      "loss": 0.45,
+      "step": 2472
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6658818267429538,
+      "learning_rate": 8.120814841552544e-06,
+      "loss": 0.5226,
+      "step": 2473
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.015297622544853,
+      "learning_rate": 8.119244407791346e-06,
+      "loss": 0.5539,
+      "step": 2474
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9033582376604703,
+      "learning_rate": 8.117673470088745e-06,
+      "loss": 0.5468,
+      "step": 2475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8411435527307176,
+      "learning_rate": 8.116102028698536e-06,
+      "loss": 0.5181,
+      "step": 2476
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4750661354518269,
+      "learning_rate": 8.114530083874599e-06,
+      "loss": 0.5282,
+      "step": 2477
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5342175409546561,
+      "learning_rate": 8.112957635870895e-06,
+      "loss": 0.5107,
+      "step": 2478
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4270612045971265,
+      "learning_rate": 8.111384684941471e-06,
+      "loss": 0.4622,
+      "step": 2479
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.731355664199108,
+      "learning_rate": 8.109811231340448e-06,
+      "loss": 0.5026,
+      "step": 2480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4987372239820895,
+      "learning_rate": 8.108237275322031e-06,
+      "loss": 0.5075,
+      "step": 2481
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.195280425107586,
+      "learning_rate": 8.106662817140508e-06,
+      "loss": 0.5285,
+      "step": 2482
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.236829502099297,
+      "learning_rate": 8.105087857050246e-06,
+      "loss": 0.4448,
+      "step": 2483
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.552047281830834,
+      "learning_rate": 8.103512395305693e-06,
+      "loss": 0.5142,
+      "step": 2484
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6745640264545936,
+      "learning_rate": 8.10193643216138e-06,
+      "loss": 0.505,
+      "step": 2485
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3051593774586854,
+      "learning_rate": 8.100359967871915e-06,
+      "loss": 0.542,
+      "step": 2486
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6495030342810197,
+      "learning_rate": 8.098783002691994e-06,
+      "loss": 0.5699,
+      "step": 2487
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.166135329822737,
+      "learning_rate": 8.097205536876387e-06,
+      "loss": 0.5654,
+      "step": 2488
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7492440243837697,
+      "learning_rate": 8.095627570679947e-06,
+      "loss": 0.5052,
+      "step": 2489
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.550936281764954,
+      "learning_rate": 8.094049104357608e-06,
+      "loss": 0.553,
+      "step": 2490
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4772954354123726,
+      "learning_rate": 8.092470138164388e-06,
+      "loss": 0.5395,
+      "step": 2491
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.0886376026135105,
+      "learning_rate": 8.09089067235538e-06,
+      "loss": 0.5575,
+      "step": 2492
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.977419426615732,
+      "learning_rate": 8.089310707185763e-06,
+      "loss": 0.508,
+      "step": 2493
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8598515635774202,
+      "learning_rate": 8.087730242910792e-06,
+      "loss": 0.4959,
+      "step": 2494
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6130983505295127,
+      "learning_rate": 8.086149279785807e-06,
+      "loss": 0.5438,
+      "step": 2495
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6412643748883935,
+      "learning_rate": 8.084567818066225e-06,
+      "loss": 0.524,
+      "step": 2496
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.796679205539921,
+      "learning_rate": 8.082985858007544e-06,
+      "loss": 0.512,
+      "step": 2497
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.493724261833098,
+      "learning_rate": 8.081403399865347e-06,
+      "loss": 0.526,
+      "step": 2498
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2693373006893207,
+      "learning_rate": 8.079820443895292e-06,
+      "loss": 0.5074,
+      "step": 2499
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.523194363659194,
+      "learning_rate": 8.07823699035312e-06,
+      "loss": 0.4962,
+      "step": 2500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4165837605180032,
+      "learning_rate": 8.076653039494649e-06,
+      "loss": 0.4806,
+      "step": 2501
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5154178932893485,
+      "learning_rate": 8.075068591575783e-06,
+      "loss": 0.4951,
+      "step": 2502
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4374752979267316,
+      "learning_rate": 8.073483646852507e-06,
+      "loss": 0.4523,
+      "step": 2503
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4793590132895706,
+      "learning_rate": 8.071898205580877e-06,
+      "loss": 0.5124,
+      "step": 2504
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1797717404084764,
+      "learning_rate": 8.070312268017036e-06,
+      "loss": 0.5688,
+      "step": 2505
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3568393334677344,
+      "learning_rate": 8.068725834417208e-06,
+      "loss": 0.5011,
+      "step": 2506
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3726472919474497,
+      "learning_rate": 8.067138905037694e-06,
+      "loss": 0.549,
+      "step": 2507
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4125513596310788,
+      "learning_rate": 8.065551480134879e-06,
+      "loss": 0.4578,
+      "step": 2508
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3339617497863514,
+      "learning_rate": 8.063963559965221e-06,
+      "loss": 0.4802,
+      "step": 2509
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5081473397590144,
+      "learning_rate": 8.062375144785265e-06,
+      "loss": 0.4897,
+      "step": 2510
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.303668471143709,
+      "learning_rate": 8.060786234851634e-06,
+      "loss": 0.4694,
+      "step": 2511
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.573520854460718,
+      "learning_rate": 8.059196830421032e-06,
+      "loss": 0.5268,
+      "step": 2512
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3482665581028472,
+      "learning_rate": 8.057606931750235e-06,
+      "loss": 0.4916,
+      "step": 2513
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8330345165028457,
+      "learning_rate": 8.056016539096112e-06,
+      "loss": 0.4659,
+      "step": 2514
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.509864150941314,
+      "learning_rate": 8.0544256527156e-06,
+      "loss": 0.5633,
+      "step": 2515
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4224595451530193,
+      "learning_rate": 8.052834272865724e-06,
+      "loss": 0.5587,
+      "step": 2516
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9882292371596657,
+      "learning_rate": 8.051242399803586e-06,
+      "loss": 0.5115,
+      "step": 2517
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.443219180257739,
+      "learning_rate": 8.049650033786364e-06,
+      "loss": 0.506,
+      "step": 2518
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4703440636482108,
+      "learning_rate": 8.04805717507132e-06,
+      "loss": 0.5088,
+      "step": 2519
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.938646696914128,
+      "learning_rate": 8.046463823915794e-06,
+      "loss": 0.4723,
+      "step": 2520
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8214933656484336,
+      "learning_rate": 8.044869980577205e-06,
+      "loss": 0.4937,
+      "step": 2521
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5537002581659554,
+      "learning_rate": 8.043275645313058e-06,
+      "loss": 0.5428,
+      "step": 2522
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8339999808752838,
+      "learning_rate": 8.041680818380924e-06,
+      "loss": 0.5098,
+      "step": 2523
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3599642946469352,
+      "learning_rate": 8.040085500038465e-06,
+      "loss": 0.5299,
+      "step": 2524
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2462733428982753,
+      "learning_rate": 8.038489690543421e-06,
+      "loss": 0.4943,
+      "step": 2525
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4496657853471375,
+      "learning_rate": 8.036893390153606e-06,
+      "loss": 0.4942,
+      "step": 2526
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6281553227437193,
+      "learning_rate": 8.035296599126917e-06,
+      "loss": 0.5068,
+      "step": 2527
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3263402475379906,
+      "learning_rate": 8.033699317721331e-06,
+      "loss": 0.4942,
+      "step": 2528
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5396019747741265,
+      "learning_rate": 8.0321015461949e-06,
+      "loss": 0.4934,
+      "step": 2529
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2515443414601215,
+      "learning_rate": 8.030503284805762e-06,
+      "loss": 0.5595,
+      "step": 2530
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7028155212480076,
+      "learning_rate": 8.028904533812125e-06,
+      "loss": 0.5103,
+      "step": 2531
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3470718084717475,
+      "learning_rate": 8.027305293472287e-06,
+      "loss": 0.5419,
+      "step": 2532
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3517717148809503,
+      "learning_rate": 8.025705564044615e-06,
+      "loss": 0.542,
+      "step": 2533
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6400748484848002,
+      "learning_rate": 8.024105345787562e-06,
+      "loss": 0.5487,
+      "step": 2534
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4080101276322876,
+      "learning_rate": 8.022504638959657e-06,
+      "loss": 0.5375,
+      "step": 2535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4608135823018693,
+      "learning_rate": 8.020903443819507e-06,
+      "loss": 0.5419,
+      "step": 2536
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2070914576507898,
+      "learning_rate": 8.0193017606258e-06,
+      "loss": 0.5425,
+      "step": 2537
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2988911897831374,
+      "learning_rate": 8.017699589637302e-06,
+      "loss": 0.4764,
+      "step": 2538
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3755261690769904,
+      "learning_rate": 8.016096931112858e-06,
+      "loss": 0.5326,
+      "step": 2539
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7204618851485292,
+      "learning_rate": 8.014493785311391e-06,
+      "loss": 0.4886,
+      "step": 2540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3003456270794864,
+      "learning_rate": 8.012890152491904e-06,
+      "loss": 0.4928,
+      "step": 2541
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3400916153759963,
+      "learning_rate": 8.011286032913478e-06,
+      "loss": 0.5215,
+      "step": 2542
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4896133467098513,
+      "learning_rate": 8.009681426835273e-06,
+      "loss": 0.5422,
+      "step": 2543
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8446002305350264,
+      "learning_rate": 8.008076334516523e-06,
+      "loss": 0.4814,
+      "step": 2544
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4655503832221062,
+      "learning_rate": 8.006470756216551e-06,
+      "loss": 0.5761,
+      "step": 2545
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.332714497007832,
+      "learning_rate": 8.00486469219475e-06,
+      "loss": 0.4785,
+      "step": 2546
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.772899784431288,
+      "learning_rate": 8.003258142710593e-06,
+      "loss": 0.5127,
+      "step": 2547
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.18496507520994,
+      "learning_rate": 8.001651108023632e-06,
+      "loss": 0.5681,
+      "step": 2548
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5808393491809347,
+      "learning_rate": 8.0000435883935e-06,
+      "loss": 0.4923,
+      "step": 2549
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9652897069753408,
+      "learning_rate": 7.998435584079904e-06,
+      "loss": 0.5546,
+      "step": 2550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4685113192092831,
+      "learning_rate": 7.99682709534263e-06,
+      "loss": 0.5046,
+      "step": 2551
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3755259188398998,
+      "learning_rate": 7.995218122441545e-06,
+      "loss": 0.5402,
+      "step": 2552
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6402565725216716,
+      "learning_rate": 7.993608665636594e-06,
+      "loss": 0.4868,
+      "step": 2553
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.340748457030908,
+      "learning_rate": 7.991998725187797e-06,
+      "loss": 0.4743,
+      "step": 2554
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.509932753295099,
+      "learning_rate": 7.990388301355257e-06,
+      "loss": 0.5071,
+      "step": 2555
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4331341944162324,
+      "learning_rate": 7.988777394399146e-06,
+      "loss": 0.5645,
+      "step": 2556
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7525247336155012,
+      "learning_rate": 7.987166004579727e-06,
+      "loss": 0.4617,
+      "step": 2557
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4084454111962381,
+      "learning_rate": 7.98555413215733e-06,
+      "loss": 0.5474,
+      "step": 2558
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.303093075197306,
+      "learning_rate": 7.98394177739237e-06,
+      "loss": 0.4948,
+      "step": 2559
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2288203173720957,
+      "learning_rate": 7.982328940545334e-06,
+      "loss": 0.4921,
+      "step": 2560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2101474981944487,
+      "learning_rate": 7.980715621876793e-06,
+      "loss": 0.4831,
+      "step": 2561
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2970614349320075,
+      "learning_rate": 7.97910182164739e-06,
+      "loss": 0.495,
+      "step": 2562
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6276571869904369,
+      "learning_rate": 7.977487540117852e-06,
+      "loss": 0.4888,
+      "step": 2563
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.31307825281209,
+      "learning_rate": 7.975872777548977e-06,
+      "loss": 0.5072,
+      "step": 2564
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8758618093049964,
+      "learning_rate": 7.974257534201647e-06,
+      "loss": 0.5052,
+      "step": 2565
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4751921344767631,
+      "learning_rate": 7.972641810336816e-06,
+      "loss": 0.5237,
+      "step": 2566
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.354956242965793,
+      "learning_rate": 7.971025606215521e-06,
+      "loss": 0.4831,
+      "step": 2567
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8797787633285996,
+      "learning_rate": 7.969408922098871e-06,
+      "loss": 0.4967,
+      "step": 2568
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9374064384200749,
+      "learning_rate": 7.96779175824806e-06,
+      "loss": 0.5014,
+      "step": 2569
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.572450726443971,
+      "learning_rate": 7.966174114924352e-06,
+      "loss": 0.5579,
+      "step": 2570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.424117716618892,
+      "learning_rate": 7.964555992389092e-06,
+      "loss": 0.5364,
+      "step": 2571
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4695100357496795,
+      "learning_rate": 7.9629373909037e-06,
+      "loss": 0.5565,
+      "step": 2572
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5546732225028876,
+      "learning_rate": 7.961318310729678e-06,
+      "loss": 0.5246,
+      "step": 2573
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4686814158564432,
+      "learning_rate": 7.959698752128602e-06,
+      "loss": 0.5406,
+      "step": 2574
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5002993149698707,
+      "learning_rate": 7.958078715362127e-06,
+      "loss": 0.5398,
+      "step": 2575
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3502921398391599,
+      "learning_rate": 7.956458200691981e-06,
+      "loss": 0.529,
+      "step": 2576
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.138050239994702,
+      "learning_rate": 7.954837208379978e-06,
+      "loss": 0.5314,
+      "step": 2577
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2974781099214527,
+      "learning_rate": 7.953215738687997e-06,
+      "loss": 0.4856,
+      "step": 2578
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4912136233611968,
+      "learning_rate": 7.951593791878005e-06,
+      "loss": 0.5309,
+      "step": 2579
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.519571706645829,
+      "learning_rate": 7.94997136821204e-06,
+      "loss": 0.5126,
+      "step": 2580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5345595769482312,
+      "learning_rate": 7.948348467952221e-06,
+      "loss": 0.4793,
+      "step": 2581
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5034221535478678,
+      "learning_rate": 7.946725091360738e-06,
+      "loss": 0.5109,
+      "step": 2582
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6353221949559826,
+      "learning_rate": 7.945101238699865e-06,
+      "loss": 0.5091,
+      "step": 2583
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6441115764867356,
+      "learning_rate": 7.943476910231948e-06,
+      "loss": 0.5287,
+      "step": 2584
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.548879078222247,
+      "learning_rate": 7.941852106219414e-06,
+      "loss": 0.4787,
+      "step": 2585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5134010020200608,
+      "learning_rate": 7.940226826924761e-06,
+      "loss": 0.464,
+      "step": 2586
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3246410761168317,
+      "learning_rate": 7.938601072610573e-06,
+      "loss": 0.5143,
+      "step": 2587
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6072185208166537,
+      "learning_rate": 7.936974843539496e-06,
+      "loss": 0.4914,
+      "step": 2588
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5873833753998925,
+      "learning_rate": 7.935348139974268e-06,
+      "loss": 0.487,
+      "step": 2589
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3803477410016485,
+      "learning_rate": 7.933720962177696e-06,
+      "loss": 0.5445,
+      "step": 2590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.479728555323523,
+      "learning_rate": 7.932093310412665e-06,
+      "loss": 0.5432,
+      "step": 2591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.461786336220308,
+      "learning_rate": 7.930465184942135e-06,
+      "loss": 0.4729,
+      "step": 2592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.372907725327091,
+      "learning_rate": 7.928836586029146e-06,
+      "loss": 0.4982,
+      "step": 2593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3760889483532308,
+      "learning_rate": 7.927207513936812e-06,
+      "loss": 0.509,
+      "step": 2594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3209344770538574,
+      "learning_rate": 7.925577968928323e-06,
+      "loss": 0.5233,
+      "step": 2595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.263176280923,
+      "learning_rate": 7.923947951266947e-06,
+      "loss": 0.5248,
+      "step": 2596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.869151617512003,
+      "learning_rate": 7.922317461216027e-06,
+      "loss": 0.556,
+      "step": 2597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2799391376527303,
+      "learning_rate": 7.920686499038985e-06,
+      "loss": 0.4808,
+      "step": 2598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3579113165318353,
+      "learning_rate": 7.919055064999315e-06,
+      "loss": 0.4971,
+      "step": 2599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6106478321846343,
+      "learning_rate": 7.917423159360592e-06,
+      "loss": 0.5491,
+      "step": 2600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.060188368021533,
+      "learning_rate": 7.915790782386462e-06,
+      "loss": 0.4894,
+      "step": 2601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4593670616968968,
+      "learning_rate": 7.91415793434065e-06,
+      "loss": 0.5285,
+      "step": 2602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8946627434610617,
+      "learning_rate": 7.91252461548696e-06,
+      "loss": 0.5221,
+      "step": 2603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.486158593598876,
+      "learning_rate": 7.910890826089267e-06,
+      "loss": 0.5064,
+      "step": 2604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5781546113023983,
+      "learning_rate": 7.909256566411522e-06,
+      "loss": 0.5123,
+      "step": 2605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2732298072495625,
+      "learning_rate": 7.907621836717757e-06,
+      "loss": 0.4986,
+      "step": 2606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.469675681281866,
+      "learning_rate": 7.905986637272079e-06,
+      "loss": 0.5008,
+      "step": 2607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6806249896449872,
+      "learning_rate": 7.904350968338663e-06,
+      "loss": 0.5207,
+      "step": 2608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3999951993913762,
+      "learning_rate": 7.90271483018177e-06,
+      "loss": 0.5482,
+      "step": 2609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3996070047305675,
+      "learning_rate": 7.901078223065731e-06,
+      "loss": 0.528,
+      "step": 2610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3957940027695925,
+      "learning_rate": 7.899441147254956e-06,
+      "loss": 0.5362,
+      "step": 2611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4092902102297067,
+      "learning_rate": 7.897803603013927e-06,
+      "loss": 0.5231,
+      "step": 2612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.313813586992601,
+      "learning_rate": 7.896165590607204e-06,
+      "loss": 0.4838,
+      "step": 2613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.829202367501445,
+      "learning_rate": 7.894527110299422e-06,
+      "loss": 0.5226,
+      "step": 2614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4582804197443877,
+      "learning_rate": 7.892888162355293e-06,
+      "loss": 0.5125,
+      "step": 2615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4644426641480262,
+      "learning_rate": 7.891248747039605e-06,
+      "loss": 0.5256,
+      "step": 2616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5832117310103972,
+      "learning_rate": 7.889608864617216e-06,
+      "loss": 0.569,
+      "step": 2617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.349664804555716,
+      "learning_rate": 7.887968515353065e-06,
+      "loss": 0.508,
+      "step": 2618
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5097419964276189,
+      "learning_rate": 7.886327699512166e-06,
+      "loss": 0.5373,
+      "step": 2619
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.314238734393822,
+      "learning_rate": 7.884686417359609e-06,
+      "loss": 0.5263,
+      "step": 2620
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3193574606716998,
+      "learning_rate": 7.88304466916055e-06,
+      "loss": 0.4542,
+      "step": 2621
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3564255854315617,
+      "learning_rate": 7.881402455180238e-06,
+      "loss": 0.4647,
+      "step": 2622
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3643754065286928,
+      "learning_rate": 7.87975977568398e-06,
+      "loss": 0.4483,
+      "step": 2623
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6474917452755293,
+      "learning_rate": 7.878116630937169e-06,
+      "loss": 0.5312,
+      "step": 2624
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3898488341013995,
+      "learning_rate": 7.876473021205266e-06,
+      "loss": 0.4786,
+      "step": 2625
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4444599189015728,
+      "learning_rate": 7.874828946753814e-06,
+      "loss": 0.5075,
+      "step": 2626
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4016382588893839,
+      "learning_rate": 7.873184407848428e-06,
+      "loss": 0.4868,
+      "step": 2627
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3055135483450153,
+      "learning_rate": 7.871539404754793e-06,
+      "loss": 0.4448,
+      "step": 2628
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4260820129750083,
+      "learning_rate": 7.86989393773868e-06,
+      "loss": 0.5093,
+      "step": 2629
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4595683756814524,
+      "learning_rate": 7.868248007065923e-06,
+      "loss": 0.528,
+      "step": 2630
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.383304259086062,
+      "learning_rate": 7.86660161300244e-06,
+      "loss": 0.537,
+      "step": 2631
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7295511857203106,
+      "learning_rate": 7.86495475581422e-06,
+      "loss": 0.4833,
+      "step": 2632
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4553464257059818,
+      "learning_rate": 7.863307435767329e-06,
+      "loss": 0.5198,
+      "step": 2633
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2005384216193022,
+      "learning_rate": 7.861659653127899e-06,
+      "loss": 0.4692,
+      "step": 2634
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.300754657681368,
+      "learning_rate": 7.860011408162153e-06,
+      "loss": 0.5397,
+      "step": 2635
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3431649243225865,
+      "learning_rate": 7.858362701136374e-06,
+      "loss": 0.5268,
+      "step": 2636
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4874194965826724,
+      "learning_rate": 7.856713532316927e-06,
+      "loss": 0.5276,
+      "step": 2637
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6575599415205786,
+      "learning_rate": 7.855063901970248e-06,
+      "loss": 0.5225,
+      "step": 2638
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.41297561898523,
+      "learning_rate": 7.85341381036285e-06,
+      "loss": 0.5139,
+      "step": 2639
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.1716924968520417,
+      "learning_rate": 7.851763257761322e-06,
+      "loss": 0.5113,
+      "step": 2640
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.536281272372059,
+      "learning_rate": 7.850112244432322e-06,
+      "loss": 0.5021,
+      "step": 2641
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5685613187358625,
+      "learning_rate": 7.848460770642588e-06,
+      "loss": 0.5094,
+      "step": 2642
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5203119298235317,
+      "learning_rate": 7.846808836658931e-06,
+      "loss": 0.48,
+      "step": 2643
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5732793968340726,
+      "learning_rate": 7.845156442748232e-06,
+      "loss": 0.5238,
+      "step": 2644
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6797119267555378,
+      "learning_rate": 7.843503589177453e-06,
+      "loss": 0.5117,
+      "step": 2645
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3572741565536737,
+      "learning_rate": 7.841850276213626e-06,
+      "loss": 0.4669,
+      "step": 2646
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7178449432637385,
+      "learning_rate": 7.840196504123856e-06,
+      "loss": 0.5076,
+      "step": 2647
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6289373215406235,
+      "learning_rate": 7.838542273175328e-06,
+      "loss": 0.5302,
+      "step": 2648
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.306086112198206,
+      "learning_rate": 7.836887583635297e-06,
+      "loss": 0.5639,
+      "step": 2649
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4157064488153976,
+      "learning_rate": 7.835232435771089e-06,
+      "loss": 0.524,
+      "step": 2650
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4036284659590974,
+      "learning_rate": 7.833576829850113e-06,
+      "loss": 0.5574,
+      "step": 2651
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5176353764519612,
+      "learning_rate": 7.831920766139844e-06,
+      "loss": 0.5141,
+      "step": 2652
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6075456494439913,
+      "learning_rate": 7.830264244907834e-06,
+      "loss": 0.5325,
+      "step": 2653
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6382628147633554,
+      "learning_rate": 7.828607266421705e-06,
+      "loss": 0.4846,
+      "step": 2654
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6454288930497467,
+      "learning_rate": 7.826949830949164e-06,
+      "loss": 0.5323,
+      "step": 2655
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.994581273390472,
+      "learning_rate": 7.82529193875798e-06,
+      "loss": 0.4747,
+      "step": 2656
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.349328328927657,
+      "learning_rate": 7.823633590116e-06,
+      "loss": 0.5051,
+      "step": 2657
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8911900323622857,
+      "learning_rate": 7.821974785291145e-06,
+      "loss": 0.5159,
+      "step": 2658
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6299950870453979,
+      "learning_rate": 7.82031552455141e-06,
+      "loss": 0.5159,
+      "step": 2659
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7036980394919018,
+      "learning_rate": 7.81865580816486e-06,
+      "loss": 0.5079,
+      "step": 2660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2437606573047195,
+      "learning_rate": 7.816995636399644e-06,
+      "loss": 0.5237,
+      "step": 2661
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3132373861797184,
+      "learning_rate": 7.81533500952397e-06,
+      "loss": 0.5083,
+      "step": 2662
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.38307268654614,
+      "learning_rate": 7.813673927806132e-06,
+      "loss": 0.5469,
+      "step": 2663
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.649046553253519,
+      "learning_rate": 7.812012391514488e-06,
+      "loss": 0.528,
+      "step": 2664
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.381471144918611,
+      "learning_rate": 7.81035040091748e-06,
+      "loss": 0.5184,
+      "step": 2665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4352863734471388,
+      "learning_rate": 7.808687956283609e-06,
+      "loss": 0.5141,
+      "step": 2666
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3719121111184451,
+      "learning_rate": 7.807025057881463e-06,
+      "loss": 0.5035,
+      "step": 2667
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3920475091905369,
+      "learning_rate": 7.805361705979698e-06,
+      "loss": 0.5194,
+      "step": 2668
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5848959505401254,
+      "learning_rate": 7.803697900847042e-06,
+      "loss": 0.5866,
+      "step": 2669
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8520042447777643,
+      "learning_rate": 7.802033642752298e-06,
+      "loss": 0.5271,
+      "step": 2670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4521535881547414,
+      "learning_rate": 7.80036893196434e-06,
+      "loss": 0.5194,
+      "step": 2671
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5181509592818552,
+      "learning_rate": 7.798703768752116e-06,
+      "loss": 0.5228,
+      "step": 2672
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5555764528978064,
+      "learning_rate": 7.79703815338465e-06,
+      "loss": 0.5209,
+      "step": 2673
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8939780601693137,
+      "learning_rate": 7.795372086131038e-06,
+      "loss": 0.5103,
+      "step": 2674
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5966979452355454,
+      "learning_rate": 7.793705567260445e-06,
+      "loss": 0.4822,
+      "step": 2675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.574764738764695,
+      "learning_rate": 7.792038597042113e-06,
+      "loss": 0.5219,
+      "step": 2676
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3781605626328648,
+      "learning_rate": 7.790371175745355e-06,
+      "loss": 0.5177,
+      "step": 2677
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.31760703133334,
+      "learning_rate": 7.78870330363956e-06,
+      "loss": 0.5038,
+      "step": 2678
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4686543270678922,
+      "learning_rate": 7.787034980994184e-06,
+      "loss": 0.4935,
+      "step": 2679
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.898415093232563,
+      "learning_rate": 7.78536620807876e-06,
+      "loss": 0.5388,
+      "step": 2680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8871341002492472,
+      "learning_rate": 7.783696985162896e-06,
+      "loss": 0.5185,
+      "step": 2681
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.756160964701136,
+      "learning_rate": 7.782027312516267e-06,
+      "loss": 0.4977,
+      "step": 2682
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6106092870879871,
+      "learning_rate": 7.780357190408622e-06,
+      "loss": 0.5323,
+      "step": 2683
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2645675766050717,
+      "learning_rate": 7.778686619109787e-06,
+      "loss": 0.4845,
+      "step": 2684
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7084608637158094,
+      "learning_rate": 7.777015598889656e-06,
+      "loss": 0.5252,
+      "step": 2685
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.714675656281538,
+      "learning_rate": 7.775344130018196e-06,
+      "loss": 0.5275,
+      "step": 2686
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4409858609141166,
+      "learning_rate": 7.77367221276545e-06,
+      "loss": 0.4673,
+      "step": 2687
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4382057653735294,
+      "learning_rate": 7.77199984740153e-06,
+      "loss": 0.4813,
+      "step": 2688
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6903980772123022,
+      "learning_rate": 7.77032703419662e-06,
+      "loss": 0.4768,
+      "step": 2689
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7131248191772204,
+      "learning_rate": 7.76865377342098e-06,
+      "loss": 0.537,
+      "step": 2690
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5944381852953722,
+      "learning_rate": 7.766980065344938e-06,
+      "loss": 0.4751,
+      "step": 2691
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4666012126559613,
+      "learning_rate": 7.765305910238898e-06,
+      "loss": 0.5269,
+      "step": 2692
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.471755000511237,
+      "learning_rate": 7.763631308373333e-06,
+      "loss": 0.5487,
+      "step": 2693
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3624088593151427,
+      "learning_rate": 7.76195626001879e-06,
+      "loss": 0.5085,
+      "step": 2694
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.710395640324599,
+      "learning_rate": 7.760280765445888e-06,
+      "loss": 0.4905,
+      "step": 2695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4268137254771691,
+      "learning_rate": 7.758604824925318e-06,
+      "loss": 0.5352,
+      "step": 2696
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3876286121831773,
+      "learning_rate": 7.756928438727844e-06,
+      "loss": 0.5045,
+      "step": 2697
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.580257529292837,
+      "learning_rate": 7.755251607124298e-06,
+      "loss": 0.5337,
+      "step": 2698
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9168052951765553,
+      "learning_rate": 7.75357433038559e-06,
+      "loss": 0.5272,
+      "step": 2699
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6677726778026305,
+      "learning_rate": 7.751896608782696e-06,
+      "loss": 0.4915,
+      "step": 2700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5021980264138404,
+      "learning_rate": 7.75021844258667e-06,
+      "loss": 0.5036,
+      "step": 2701
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6929177369758843,
+      "learning_rate": 7.748539832068633e-06,
+      "loss": 0.5464,
+      "step": 2702
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5228255867041702,
+      "learning_rate": 7.746860777499778e-06,
+      "loss": 0.5678,
+      "step": 2703
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8989886501816657,
+      "learning_rate": 7.745181279151373e-06,
+      "loss": 0.5263,
+      "step": 2704
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2883835482583696,
+      "learning_rate": 7.743501337294754e-06,
+      "loss": 0.5216,
+      "step": 2705
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.960176675059009,
+      "learning_rate": 7.741820952201333e-06,
+      "loss": 0.5232,
+      "step": 2706
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7759698679308338,
+      "learning_rate": 7.740140124142587e-06,
+      "loss": 0.5104,
+      "step": 2707
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4527436454312945,
+      "learning_rate": 7.738458853390072e-06,
+      "loss": 0.5327,
+      "step": 2708
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3174291504094895,
+      "learning_rate": 7.736777140215412e-06,
+      "loss": 0.4753,
+      "step": 2709
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7018137791242018,
+      "learning_rate": 7.735094984890302e-06,
+      "loss": 0.5085,
+      "step": 2710
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5993123487899685,
+      "learning_rate": 7.733412387686508e-06,
+      "loss": 0.4816,
+      "step": 2711
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.904796254721156,
+      "learning_rate": 7.731729348875868e-06,
+      "loss": 0.5451,
+      "step": 2712
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.604283221653655,
+      "learning_rate": 7.730045868730294e-06,
+      "loss": 0.5185,
+      "step": 2713
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.93420487671135,
+      "learning_rate": 7.728361947521765e-06,
+      "loss": 0.5288,
+      "step": 2714
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6354612809092355,
+      "learning_rate": 7.726677585522335e-06,
+      "loss": 0.466,
+      "step": 2715
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5483837104096023,
+      "learning_rate": 7.724992783004125e-06,
+      "loss": 0.545,
+      "step": 2716
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5651476019378647,
+      "learning_rate": 7.723307540239332e-06,
+      "loss": 0.5436,
+      "step": 2717
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.886905512469594,
+      "learning_rate": 7.721621857500221e-06,
+      "loss": 0.4745,
+      "step": 2718
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1861427286963115,
+      "learning_rate": 7.719935735059131e-06,
+      "loss": 0.4465,
+      "step": 2719
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.866549648988967,
+      "learning_rate": 7.718249173188465e-06,
+      "loss": 0.5152,
+      "step": 2720
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.586552465840777,
+      "learning_rate": 7.716562172160706e-06,
+      "loss": 0.5207,
+      "step": 2721
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8513938439300908,
+      "learning_rate": 7.714874732248404e-06,
+      "loss": 0.4664,
+      "step": 2722
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4103660610251199,
+      "learning_rate": 7.713186853724176e-06,
+      "loss": 0.5427,
+      "step": 2723
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2258139433120145,
+      "learning_rate": 7.711498536860719e-06,
+      "loss": 0.5043,
+      "step": 2724
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3097524148833894,
+      "learning_rate": 7.709809781930791e-06,
+      "loss": 0.4709,
+      "step": 2725
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0590687856687837,
+      "learning_rate": 7.708120589207227e-06,
+      "loss": 0.5472,
+      "step": 2726
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0120462476542746,
+      "learning_rate": 7.706430958962932e-06,
+      "loss": 0.5511,
+      "step": 2727
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6327764346512121,
+      "learning_rate": 7.704740891470878e-06,
+      "loss": 0.4788,
+      "step": 2728
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4071899449611387,
+      "learning_rate": 7.703050387004111e-06,
+      "loss": 0.528,
+      "step": 2729
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4315602215673797,
+      "learning_rate": 7.70135944583575e-06,
+      "loss": 0.5531,
+      "step": 2730
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.755214304596285,
+      "learning_rate": 7.69966806823898e-06,
+      "loss": 0.4996,
+      "step": 2731
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2982408118906106,
+      "learning_rate": 7.697976254487054e-06,
+      "loss": 0.5133,
+      "step": 2732
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3617359856408846,
+      "learning_rate": 7.696284004853303e-06,
+      "loss": 0.4639,
+      "step": 2733
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.277470567398012,
+      "learning_rate": 7.694591319611124e-06,
+      "loss": 0.5406,
+      "step": 2734
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2375158989213473,
+      "learning_rate": 7.692898199033988e-06,
+      "loss": 0.4945,
+      "step": 2735
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6525320292638912,
+      "learning_rate": 7.691204643395426e-06,
+      "loss": 0.4814,
+      "step": 2736
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2916993899549458,
+      "learning_rate": 7.689510652969055e-06,
+      "loss": 0.4816,
+      "step": 2737
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.391677601694305,
+      "learning_rate": 7.687816228028552e-06,
+      "loss": 0.4606,
+      "step": 2738
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4519576798931901,
+      "learning_rate": 7.686121368847666e-06,
+      "loss": 0.5472,
+      "step": 2739
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5009721768272837,
+      "learning_rate": 7.68442607570021e-06,
+      "loss": 0.5325,
+      "step": 2740
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3792196069659568,
+      "learning_rate": 7.682730348860085e-06,
+      "loss": 0.5191,
+      "step": 2741
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5022991314730445,
+      "learning_rate": 7.681034188601242e-06,
+      "loss": 0.517,
+      "step": 2742
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5302237971630834,
+      "learning_rate": 7.679337595197715e-06,
+      "loss": 0.4969,
+      "step": 2743
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4982130605022177,
+      "learning_rate": 7.677640568923601e-06,
+      "loss": 0.5161,
+      "step": 2744
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3912316865151857,
+      "learning_rate": 7.67594311005307e-06,
+      "loss": 0.4482,
+      "step": 2745
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3729878686082877,
+      "learning_rate": 7.674245218860362e-06,
+      "loss": 0.4852,
+      "step": 2746
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.419155005976854,
+      "learning_rate": 7.672546895619786e-06,
+      "loss": 0.517,
+      "step": 2747
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4587619792042252,
+      "learning_rate": 7.670848140605723e-06,
+      "loss": 0.5279,
+      "step": 2748
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5269943732151012,
+      "learning_rate": 7.66914895409262e-06,
+      "loss": 0.5329,
+      "step": 2749
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.975878240194581,
+      "learning_rate": 7.667449336354996e-06,
+      "loss": 0.4831,
+      "step": 2750
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4672990700567374,
+      "learning_rate": 7.665749287667436e-06,
+      "loss": 0.4792,
+      "step": 2751
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.39519067576427,
+      "learning_rate": 7.664048808304603e-06,
+      "loss": 0.5704,
+      "step": 2752
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.81220504073997,
+      "learning_rate": 7.662347898541222e-06,
+      "loss": 0.474,
+      "step": 2753
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.454364363303493,
+      "learning_rate": 7.66064655865209e-06,
+      "loss": 0.4896,
+      "step": 2754
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.345457097405705,
+      "learning_rate": 7.658944788912073e-06,
+      "loss": 0.4907,
+      "step": 2755
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6197799716830366,
+      "learning_rate": 7.657242589596107e-06,
+      "loss": 0.5136,
+      "step": 2756
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4100260413117522,
+      "learning_rate": 7.655539960979199e-06,
+      "loss": 0.4799,
+      "step": 2757
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4355816404214223,
+      "learning_rate": 7.653836903336423e-06,
+      "loss": 0.5246,
+      "step": 2758
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4093389459141474,
+      "learning_rate": 7.652133416942921e-06,
+      "loss": 0.4405,
+      "step": 2759
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1045817688671162,
+      "learning_rate": 7.650429502073909e-06,
+      "loss": 0.484,
+      "step": 2760
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4695892914070552,
+      "learning_rate": 7.648725159004666e-06,
+      "loss": 0.542,
+      "step": 2761
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3991047999982942,
+      "learning_rate": 7.647020388010546e-06,
+      "loss": 0.5196,
+      "step": 2762
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4581867933362196,
+      "learning_rate": 7.64531518936697e-06,
+      "loss": 0.5035,
+      "step": 2763
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.505294682141353,
+      "learning_rate": 7.643609563349428e-06,
+      "loss": 0.5147,
+      "step": 2764
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6133177462796544,
+      "learning_rate": 7.641903510233478e-06,
+      "loss": 0.5426,
+      "step": 2765
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4806029734791937,
+      "learning_rate": 7.640197030294749e-06,
+      "loss": 0.5115,
+      "step": 2766
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.616248266794655,
+      "learning_rate": 7.638490123808935e-06,
+      "loss": 0.4804,
+      "step": 2767
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3394062449465627,
+      "learning_rate": 7.636782791051805e-06,
+      "loss": 0.516,
+      "step": 2768
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.177620453219728,
+      "learning_rate": 7.63507503229919e-06,
+      "loss": 0.5205,
+      "step": 2769
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.5007327575946827,
+      "learning_rate": 7.633366847826999e-06,
+      "loss": 0.5107,
+      "step": 2770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3918381642554443,
+      "learning_rate": 7.6316582379112e-06,
+      "loss": 0.5833,
+      "step": 2771
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8013497027412864,
+      "learning_rate": 7.629949202827835e-06,
+      "loss": 0.5097,
+      "step": 2772
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5598225631411284,
+      "learning_rate": 7.6282397428530135e-06,
+      "loss": 0.4993,
+      "step": 2773
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4219298971027583,
+      "learning_rate": 7.626529858262914e-06,
+      "loss": 0.5187,
+      "step": 2774
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.601086179767575,
+      "learning_rate": 7.624819549333784e-06,
+      "loss": 0.4492,
+      "step": 2775
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5134415486399717,
+      "learning_rate": 7.62310881634194e-06,
+      "loss": 0.5014,
+      "step": 2776
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3155320012758154,
+      "learning_rate": 7.621397659563761e-06,
+      "loss": 0.5282,
+      "step": 2777
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.026077379891773,
+      "learning_rate": 7.619686079275705e-06,
+      "loss": 0.516,
+      "step": 2778
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2985785718733407,
+      "learning_rate": 7.617974075754291e-06,
+      "loss": 0.487,
+      "step": 2779
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.487272549933582,
+      "learning_rate": 7.616261649276107e-06,
+      "loss": 0.5503,
+      "step": 2780
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4482877030737968,
+      "learning_rate": 7.614548800117812e-06,
+      "loss": 0.5212,
+      "step": 2781
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.264435244527073,
+      "learning_rate": 7.612835528556131e-06,
+      "loss": 0.4865,
+      "step": 2782
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.428380630654264,
+      "learning_rate": 7.611121834867858e-06,
+      "loss": 0.4424,
+      "step": 2783
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4593193495946697,
+      "learning_rate": 7.609407719329858e-06,
+      "loss": 0.5258,
+      "step": 2784
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.595849152307415,
+      "learning_rate": 7.607693182219058e-06,
+      "loss": 0.5197,
+      "step": 2785
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1367861478862946,
+      "learning_rate": 7.605978223812458e-06,
+      "loss": 0.5408,
+      "step": 2786
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.237539684863762,
+      "learning_rate": 7.604262844387125e-06,
+      "loss": 0.5326,
+      "step": 2787
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4710863852700993,
+      "learning_rate": 7.602547044220192e-06,
+      "loss": 0.481,
+      "step": 2788
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2748156376890598,
+      "learning_rate": 7.600830823588864e-06,
+      "loss": 0.465,
+      "step": 2789
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2474025526575772,
+      "learning_rate": 7.5991141827704084e-06,
+      "loss": 0.4957,
+      "step": 2790
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.361004388110375,
+      "learning_rate": 7.5973971220421685e-06,
+      "loss": 0.4817,
+      "step": 2791
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.727555590942853,
+      "learning_rate": 7.595679641681546e-06,
+      "loss": 0.5122,
+      "step": 2792
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.463675582477899,
+      "learning_rate": 7.593961741966019e-06,
+      "loss": 0.4723,
+      "step": 2793
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6756022980618925,
+      "learning_rate": 7.592243423173124e-06,
+      "loss": 0.5637,
+      "step": 2794
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.339013495875156,
+      "learning_rate": 7.5905246855804735e-06,
+      "loss": 0.4775,
+      "step": 2795
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9885912649008708,
+      "learning_rate": 7.588805529465747e-06,
+      "loss": 0.5408,
+      "step": 2796
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3473130209888207,
+      "learning_rate": 7.587085955106685e-06,
+      "loss": 0.4883,
+      "step": 2797
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.721228409362882,
+      "learning_rate": 7.585365962781103e-06,
+      "loss": 0.4898,
+      "step": 2798
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6689030328737668,
+      "learning_rate": 7.58364555276688e-06,
+      "loss": 0.5007,
+      "step": 2799
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.312323699854146,
+      "learning_rate": 7.5819247253419624e-06,
+      "loss": 0.4944,
+      "step": 2800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2682572784572763,
+      "learning_rate": 7.5802034807843675e-06,
+      "loss": 0.4607,
+      "step": 2801
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.536315066547835,
+      "learning_rate": 7.578481819372174e-06,
+      "loss": 0.5442,
+      "step": 2802
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6095888676927919,
+      "learning_rate": 7.576759741383534e-06,
+      "loss": 0.5528,
+      "step": 2803
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3160002037099694,
+      "learning_rate": 7.575037247096664e-06,
+      "loss": 0.5164,
+      "step": 2804
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8566511781629418,
+      "learning_rate": 7.5733143367898475e-06,
+      "loss": 0.4993,
+      "step": 2805
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5097008377535144,
+      "learning_rate": 7.571591010741436e-06,
+      "loss": 0.5012,
+      "step": 2806
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3973734343760345,
+      "learning_rate": 7.569867269229849e-06,
+      "loss": 0.5597,
+      "step": 2807
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3440934662874864,
+      "learning_rate": 7.56814311253357e-06,
+      "loss": 0.5049,
+      "step": 2808
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7619676156979018,
+      "learning_rate": 7.566418540931154e-06,
+      "loss": 0.5333,
+      "step": 2809
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4386815776036004,
+      "learning_rate": 7.56469355470122e-06,
+      "loss": 0.4954,
+      "step": 2810
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.347905562414754,
+      "learning_rate": 7.5629681541224544e-06,
+      "loss": 0.4376,
+      "step": 2811
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7566813074923715,
+      "learning_rate": 7.56124233947361e-06,
+      "loss": 0.5428,
+      "step": 2812
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4255239120541885,
+      "learning_rate": 7.559516111033509e-06,
+      "loss": 0.5485,
+      "step": 2813
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.169394255097597,
+      "learning_rate": 7.557789469081036e-06,
+      "loss": 0.4957,
+      "step": 2814
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.599542317523516,
+      "learning_rate": 7.55606241389515e-06,
+      "loss": 0.5238,
+      "step": 2815
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7719411350941612,
+      "learning_rate": 7.554334945754869e-06,
+      "loss": 0.5163,
+      "step": 2816
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.781268122054056,
+      "learning_rate": 7.552607064939279e-06,
+      "loss": 0.5377,
+      "step": 2817
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4987431909303914,
+      "learning_rate": 7.550878771727537e-06,
+      "loss": 0.5219,
+      "step": 2818
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2623810509286617,
+      "learning_rate": 7.549150066398865e-06,
+      "loss": 0.5444,
+      "step": 2819
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5077713293075448,
+      "learning_rate": 7.5474209492325466e-06,
+      "loss": 0.5356,
+      "step": 2820
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4323981568838295,
+      "learning_rate": 7.5456914205079405e-06,
+      "loss": 0.5173,
+      "step": 2821
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4514281182699293,
+      "learning_rate": 7.543961480504463e-06,
+      "loss": 0.4911,
+      "step": 2822
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3481986067972602,
+      "learning_rate": 7.542231129501603e-06,
+      "loss": 0.5285,
+      "step": 2823
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.539752312980218,
+      "learning_rate": 7.540500367778916e-06,
+      "loss": 0.4925,
+      "step": 2824
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3970147800199832,
+      "learning_rate": 7.538769195616018e-06,
+      "loss": 0.5636,
+      "step": 2825
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2279673122276384,
+      "learning_rate": 7.537037613292597e-06,
+      "loss": 0.4888,
+      "step": 2826
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7812718415678297,
+      "learning_rate": 7.535305621088407e-06,
+      "loss": 0.5011,
+      "step": 2827
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5110838443802594,
+      "learning_rate": 7.533573219283264e-06,
+      "loss": 0.5166,
+      "step": 2828
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3924394435460279,
+      "learning_rate": 7.531840408157054e-06,
+      "loss": 0.5189,
+      "step": 2829
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5926740455071986,
+      "learning_rate": 7.530107187989727e-06,
+      "loss": 0.5627,
+      "step": 2830
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.679223687426762,
+      "learning_rate": 7.528373559061299e-06,
+      "loss": 0.5035,
+      "step": 2831
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4755516250039995,
+      "learning_rate": 7.526639521651858e-06,
+      "loss": 0.4423,
+      "step": 2832
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.452773889809484,
+      "learning_rate": 7.524905076041548e-06,
+      "loss": 0.5392,
+      "step": 2833
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3163601597702044,
+      "learning_rate": 7.5231702225105855e-06,
+      "loss": 0.4859,
+      "step": 2834
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4811247611517149,
+      "learning_rate": 7.521434961339251e-06,
+      "loss": 0.5279,
+      "step": 2835
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3828663889486925,
+      "learning_rate": 7.519699292807891e-06,
+      "loss": 0.5661,
+      "step": 2836
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4661354083477,
+      "learning_rate": 7.517963217196922e-06,
+      "loss": 0.5295,
+      "step": 2837
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2459607024844084,
+      "learning_rate": 7.516226734786818e-06,
+      "loss": 0.4879,
+      "step": 2838
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6301408152941137,
+      "learning_rate": 7.514489845858122e-06,
+      "loss": 0.509,
+      "step": 2839
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.89784173623586,
+      "learning_rate": 7.512752550691447e-06,
+      "loss": 0.51,
+      "step": 2840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5483161180179505,
+      "learning_rate": 7.511014849567466e-06,
+      "loss": 0.4395,
+      "step": 2841
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.509431583014947,
+      "learning_rate": 7.509276742766922e-06,
+      "loss": 0.5203,
+      "step": 2842
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3967413235682298,
+      "learning_rate": 7.50753823057062e-06,
+      "loss": 0.5108,
+      "step": 2843
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.454115123765699,
+      "learning_rate": 7.505799313259433e-06,
+      "loss": 0.4783,
+      "step": 2844
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6608105137667641,
+      "learning_rate": 7.504059991114298e-06,
+      "loss": 0.5376,
+      "step": 2845
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3770781664755012,
+      "learning_rate": 7.502320264416217e-06,
+      "loss": 0.5124,
+      "step": 2846
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2886223833941657,
+      "learning_rate": 7.500580133446259e-06,
+      "loss": 0.4993,
+      "step": 2847
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3551384829891164,
+      "learning_rate": 7.498839598485557e-06,
+      "loss": 0.5414,
+      "step": 2848
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4519349854873853,
+      "learning_rate": 7.497098659815312e-06,
+      "loss": 0.4865,
+      "step": 2849
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5215147079444222,
+      "learning_rate": 7.495357317716784e-06,
+      "loss": 0.4933,
+      "step": 2850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6889627685059738,
+      "learning_rate": 7.493615572471303e-06,
+      "loss": 0.5437,
+      "step": 2851
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5426468176935286,
+      "learning_rate": 7.491873424360267e-06,
+      "loss": 0.5536,
+      "step": 2852
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2934185127919198,
+      "learning_rate": 7.490130873665131e-06,
+      "loss": 0.555,
+      "step": 2853
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8929268075082235,
+      "learning_rate": 7.488387920667423e-06,
+      "loss": 0.5058,
+      "step": 2854
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5327500386068296,
+      "learning_rate": 7.486644565648731e-06,
+      "loss": 0.4923,
+      "step": 2855
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7011582348896546,
+      "learning_rate": 7.484900808890707e-06,
+      "loss": 0.5602,
+      "step": 2856
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5780262628699275,
+      "learning_rate": 7.483156650675073e-06,
+      "loss": 0.5853,
+      "step": 2857
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9623520716665723,
+      "learning_rate": 7.481412091283613e-06,
+      "loss": 0.5624,
+      "step": 2858
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4578953352943917,
+      "learning_rate": 7.479667130998174e-06,
+      "loss": 0.5114,
+      "step": 2859
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5016057014942639,
+      "learning_rate": 7.477921770100672e-06,
+      "loss": 0.5319,
+      "step": 2860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6416102122722422,
+      "learning_rate": 7.476176008873084e-06,
+      "loss": 0.5037,
+      "step": 2861
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.179152985299466,
+      "learning_rate": 7.474429847597454e-06,
+      "loss": 0.5166,
+      "step": 2862
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2704451553872496,
+      "learning_rate": 7.472683286555889e-06,
+      "loss": 0.4983,
+      "step": 2863
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4942884809928834,
+      "learning_rate": 7.470936326030562e-06,
+      "loss": 0.5133,
+      "step": 2864
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.553793602713353,
+      "learning_rate": 7.46918896630371e-06,
+      "loss": 0.4992,
+      "step": 2865
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4876867607076067,
+      "learning_rate": 7.467441207657633e-06,
+      "loss": 0.5729,
+      "step": 2866
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3095012089466964,
+      "learning_rate": 7.465693050374698e-06,
+      "loss": 0.4689,
+      "step": 2867
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2044094333174193,
+      "learning_rate": 7.463944494737334e-06,
+      "loss": 0.5217,
+      "step": 2868
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8193312473869578,
+      "learning_rate": 7.462195541028037e-06,
+      "loss": 0.5285,
+      "step": 2869
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4909392607182927,
+      "learning_rate": 7.460446189529365e-06,
+      "loss": 0.5196,
+      "step": 2870
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.24481673162198,
+      "learning_rate": 7.458696440523942e-06,
+      "loss": 0.4666,
+      "step": 2871
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5376375614629765,
+      "learning_rate": 7.4569462942944525e-06,
+      "loss": 0.5228,
+      "step": 2872
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4790336815690879,
+      "learning_rate": 7.455195751123654e-06,
+      "loss": 0.539,
+      "step": 2873
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3890204848477787,
+      "learning_rate": 7.453444811294357e-06,
+      "loss": 0.493,
+      "step": 2874
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9199696185437813,
+      "learning_rate": 7.451693475089442e-06,
+      "loss": 0.5167,
+      "step": 2875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4822184521520818,
+      "learning_rate": 7.449941742791853e-06,
+      "loss": 0.5066,
+      "step": 2876
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3316945278591281,
+      "learning_rate": 7.448189614684599e-06,
+      "loss": 0.5055,
+      "step": 2877
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.5407533800955155,
+      "learning_rate": 7.446437091050751e-06,
+      "loss": 0.4829,
+      "step": 2878
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7652822301013407,
+      "learning_rate": 7.444684172173445e-06,
+      "loss": 0.4929,
+      "step": 2879
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4786722254819273,
+      "learning_rate": 7.442930858335879e-06,
+      "loss": 0.4984,
+      "step": 2880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5148769815243324,
+      "learning_rate": 7.44117714982132e-06,
+      "loss": 0.4862,
+      "step": 2881
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2571723121367273,
+      "learning_rate": 7.43942304691309e-06,
+      "loss": 0.5446,
+      "step": 2882
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4647084158513137,
+      "learning_rate": 7.437668549894583e-06,
+      "loss": 0.5334,
+      "step": 2883
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4714820556689951,
+      "learning_rate": 7.435913659049253e-06,
+      "loss": 0.4864,
+      "step": 2884
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2478116026485935,
+      "learning_rate": 7.434158374660617e-06,
+      "loss": 0.5445,
+      "step": 2885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.616471999591579,
+      "learning_rate": 7.432402697012258e-06,
+      "loss": 0.5104,
+      "step": 2886
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5126140439909073,
+      "learning_rate": 7.430646626387821e-06,
+      "loss": 0.5154,
+      "step": 2887
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4132347024921716,
+      "learning_rate": 7.428890163071013e-06,
+      "loss": 0.5495,
+      "step": 2888
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.301685970559841,
+      "learning_rate": 7.427133307345608e-06,
+      "loss": 0.4776,
+      "step": 2889
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.221108781930627,
+      "learning_rate": 7.425376059495442e-06,
+      "loss": 0.4889,
+      "step": 2890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.177713222796027,
+      "learning_rate": 7.4236184198044115e-06,
+      "loss": 0.4286,
+      "step": 2891
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.631555248333458,
+      "learning_rate": 7.421860388556481e-06,
+      "loss": 0.5464,
+      "step": 2892
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4870772812914097,
+      "learning_rate": 7.4201019660356745e-06,
+      "loss": 0.4909,
+      "step": 2893
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7172788153908546,
+      "learning_rate": 7.418343152526081e-06,
+      "loss": 0.5042,
+      "step": 2894
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8282438880838727,
+      "learning_rate": 7.416583948311852e-06,
+      "loss": 0.5513,
+      "step": 2895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7214347174465112,
+      "learning_rate": 7.414824353677202e-06,
+      "loss": 0.5221,
+      "step": 2896
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3148725718413026,
+      "learning_rate": 7.4130643689064105e-06,
+      "loss": 0.5182,
+      "step": 2897
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2618207897069937,
+      "learning_rate": 7.411303994283818e-06,
+      "loss": 0.4973,
+      "step": 2898
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5852620561225959,
+      "learning_rate": 7.4095432300938295e-06,
+      "loss": 0.5121,
+      "step": 2899
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3827972981621872,
+      "learning_rate": 7.407782076620909e-06,
+      "loss": 0.4795,
+      "step": 2900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3328404352084577,
+      "learning_rate": 7.4060205341495895e-06,
+      "loss": 0.5338,
+      "step": 2901
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5203616571569938,
+      "learning_rate": 7.404258602964462e-06,
+      "loss": 0.5179,
+      "step": 2902
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4862472301200262,
+      "learning_rate": 7.402496283350182e-06,
+      "loss": 0.582,
+      "step": 2903
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5447324937281806,
+      "learning_rate": 7.400733575591469e-06,
+      "loss": 0.5216,
+      "step": 2904
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.353338309003225,
+      "learning_rate": 7.398970479973101e-06,
+      "loss": 0.5196,
+      "step": 2905
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.390823770991522,
+      "learning_rate": 7.3972069967799255e-06,
+      "loss": 0.5827,
+      "step": 2906
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3209525144108847,
+      "learning_rate": 7.395443126296846e-06,
+      "loss": 0.5036,
+      "step": 2907
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5146042399906572,
+      "learning_rate": 7.3936788688088335e-06,
+      "loss": 0.5087,
+      "step": 2908
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7655408507349146,
+      "learning_rate": 7.391914224600918e-06,
+      "loss": 0.5265,
+      "step": 2909
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3637002400842049,
+      "learning_rate": 7.390149193958192e-06,
+      "loss": 0.5321,
+      "step": 2910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.559416043795649,
+      "learning_rate": 7.388383777165815e-06,
+      "loss": 0.5393,
+      "step": 2911
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2247524083410941,
+      "learning_rate": 7.386617974509002e-06,
+      "loss": 0.4754,
+      "step": 2912
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4028519626964275,
+      "learning_rate": 7.3848517862730364e-06,
+      "loss": 0.5671,
+      "step": 2913
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6481816227825374,
+      "learning_rate": 7.38308521274326e-06,
+      "loss": 0.4992,
+      "step": 2914
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5708813531444694,
+      "learning_rate": 7.381318254205081e-06,
+      "loss": 0.5116,
+      "step": 2915
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6203098339157995,
+      "learning_rate": 7.3795509109439645e-06,
+      "loss": 0.5183,
+      "step": 2916
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2587625270337148,
+      "learning_rate": 7.377783183245442e-06,
+      "loss": 0.4795,
+      "step": 2917
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4369800726343376,
+      "learning_rate": 7.376015071395103e-06,
+      "loss": 0.5057,
+      "step": 2918
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3720643090525502,
+      "learning_rate": 7.374246575678604e-06,
+      "loss": 0.4933,
+      "step": 2919
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4278576722593987,
+      "learning_rate": 7.372477696381659e-06,
+      "loss": 0.5417,
+      "step": 2920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8221467932962376,
+      "learning_rate": 7.370708433790048e-06,
+      "loss": 0.5501,
+      "step": 2921
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6611923403031745,
+      "learning_rate": 7.36893878818961e-06,
+      "loss": 0.545,
+      "step": 2922
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3367746208219013,
+      "learning_rate": 7.367168759866248e-06,
+      "loss": 0.5271,
+      "step": 2923
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3873043665769509,
+      "learning_rate": 7.3653983491059245e-06,
+      "loss": 0.5342,
+      "step": 2924
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.686022468418169,
+      "learning_rate": 7.363627556194663e-06,
+      "loss": 0.5073,
+      "step": 2925
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.864577609095847,
+      "learning_rate": 7.361856381418555e-06,
+      "loss": 0.5025,
+      "step": 2926
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5126834615642972,
+      "learning_rate": 7.360084825063748e-06,
+      "loss": 0.5628,
+      "step": 2927
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4677059654643878,
+      "learning_rate": 7.35831288741645e-06,
+      "loss": 0.493,
+      "step": 2928
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5816508458999734,
+      "learning_rate": 7.356540568762936e-06,
+      "loss": 0.5537,
+      "step": 2929
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5373700847214247,
+      "learning_rate": 7.354767869389537e-06,
+      "loss": 0.526,
+      "step": 2930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.407280590400603,
+      "learning_rate": 7.3529947895826505e-06,
+      "loss": 0.5235,
+      "step": 2931
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3570233205482056,
+      "learning_rate": 7.351221329628733e-06,
+      "loss": 0.485,
+      "step": 2932
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3560438717504448,
+      "learning_rate": 7.349447489814301e-06,
+      "loss": 0.539,
+      "step": 2933
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5107167346554402,
+      "learning_rate": 7.347673270425935e-06,
+      "loss": 0.5465,
+      "step": 2934
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5350279692878122,
+      "learning_rate": 7.345898671750277e-06,
+      "loss": 0.5205,
+      "step": 2935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9087051914119872,
+      "learning_rate": 7.344123694074028e-06,
+      "loss": 0.46,
+      "step": 2936
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.316312234164915,
+      "learning_rate": 7.342348337683949e-06,
+      "loss": 0.5125,
+      "step": 2937
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.237503490731533,
+      "learning_rate": 7.340572602866868e-06,
+      "loss": 0.4504,
+      "step": 2938
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4873898920144497,
+      "learning_rate": 7.338796489909668e-06,
+      "loss": 0.5169,
+      "step": 2939
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4055193178497787,
+      "learning_rate": 7.337019999099297e-06,
+      "loss": 0.5279,
+      "step": 2940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3350012468586496,
+      "learning_rate": 7.335243130722763e-06,
+      "loss": 0.5018,
+      "step": 2941
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3696508667171363,
+      "learning_rate": 7.333465885067133e-06,
+      "loss": 0.4933,
+      "step": 2942
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2081108750109073,
+      "learning_rate": 7.331688262419539e-06,
+      "loss": 0.4823,
+      "step": 2943
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.126683979625058,
+      "learning_rate": 7.329910263067172e-06,
+      "loss": 0.5157,
+      "step": 2944
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6644563292123145,
+      "learning_rate": 7.328131887297281e-06,
+      "loss": 0.4806,
+      "step": 2945
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.006276596512924,
+      "learning_rate": 7.326353135397177e-06,
+      "loss": 0.4804,
+      "step": 2946
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5446131874657179,
+      "learning_rate": 7.3245740076542385e-06,
+      "loss": 0.4882,
+      "step": 2947
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6923510062890936,
+      "learning_rate": 7.322794504355894e-06,
+      "loss": 0.5492,
+      "step": 2948
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3394289668212493,
+      "learning_rate": 7.321014625789641e-06,
+      "loss": 0.4962,
+      "step": 2949
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4749240895876494,
+      "learning_rate": 7.319234372243032e-06,
+      "loss": 0.5458,
+      "step": 2950
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6144924649110135,
+      "learning_rate": 7.317453744003686e-06,
+      "loss": 0.5033,
+      "step": 2951
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4906919511209462,
+      "learning_rate": 7.315672741359277e-06,
+      "loss": 0.5148,
+      "step": 2952
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4532325048454164,
+      "learning_rate": 7.313891364597541e-06,
+      "loss": 0.5407,
+      "step": 2953
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1811973839585446,
+      "learning_rate": 7.3121096140062776e-06,
+      "loss": 0.4752,
+      "step": 2954
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.461913376198192,
+      "learning_rate": 7.310327489873341e-06,
+      "loss": 0.5184,
+      "step": 2955
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4023156824094034,
+      "learning_rate": 7.308544992486653e-06,
+      "loss": 0.5481,
+      "step": 2956
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3419673390743536,
+      "learning_rate": 7.3067621221341875e-06,
+      "loss": 0.4797,
+      "step": 2957
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.821401888509523,
+      "learning_rate": 7.304978879103986e-06,
+      "loss": 0.5255,
+      "step": 2958
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7809055463536292,
+      "learning_rate": 7.303195263684146e-06,
+      "loss": 0.4903,
+      "step": 2959
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5198332285659275,
+      "learning_rate": 7.301411276162827e-06,
+      "loss": 0.4952,
+      "step": 2960
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4624158529664613,
+      "learning_rate": 7.299626916828246e-06,
+      "loss": 0.5132,
+      "step": 2961
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4693717947778802,
+      "learning_rate": 7.2978421859686845e-06,
+      "loss": 0.4981,
+      "step": 2962
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3892004234553528,
+      "learning_rate": 7.296057083872481e-06,
+      "loss": 0.5037,
+      "step": 2963
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6086533258064626,
+      "learning_rate": 7.294271610828032e-06,
+      "loss": 0.4843,
+      "step": 2964
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.438061788483176,
+      "learning_rate": 7.2924857671237996e-06,
+      "loss": 0.4706,
+      "step": 2965
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3813683589454293,
+      "learning_rate": 7.2906995530482986e-06,
+      "loss": 0.5731,
+      "step": 2966
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4969900098101476,
+      "learning_rate": 7.288912968890112e-06,
+      "loss": 0.4994,
+      "step": 2967
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.483277435038951,
+      "learning_rate": 7.287126014937876e-06,
+      "loss": 0.5449,
+      "step": 2968
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5714496524913182,
+      "learning_rate": 7.285338691480289e-06,
+      "loss": 0.5268,
+      "step": 2969
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.396912755938003,
+      "learning_rate": 7.283550998806108e-06,
+      "loss": 0.5263,
+      "step": 2970
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3184850804940975,
+      "learning_rate": 7.2817629372041544e-06,
+      "loss": 0.4417,
+      "step": 2971
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5471097162982606,
+      "learning_rate": 7.279974506963301e-06,
+      "loss": 0.5393,
+      "step": 2972
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.501741983849576,
+      "learning_rate": 7.278185708372485e-06,
+      "loss": 0.5418,
+      "step": 2973
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5656687536865674,
+      "learning_rate": 7.276396541720703e-06,
+      "loss": 0.5272,
+      "step": 2974
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3627469786622979,
+      "learning_rate": 7.274607007297011e-06,
+      "loss": 0.5223,
+      "step": 2975
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6736889928051844,
+      "learning_rate": 7.272817105390525e-06,
+      "loss": 0.5768,
+      "step": 2976
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.284650601665037,
+      "learning_rate": 7.271026836290418e-06,
+      "loss": 0.4861,
+      "step": 2977
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1384482986339133,
+      "learning_rate": 7.269236200285925e-06,
+      "loss": 0.4921,
+      "step": 2978
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3967900521778172,
+      "learning_rate": 7.267445197666336e-06,
+      "loss": 0.446,
+      "step": 2979
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4677001082213201,
+      "learning_rate": 7.265653828721007e-06,
+      "loss": 0.4801,
+      "step": 2980
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4026183591203205,
+      "learning_rate": 7.263862093739349e-06,
+      "loss": 0.5119,
+      "step": 2981
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.495942156017364,
+      "learning_rate": 7.2620699930108295e-06,
+      "loss": 0.5288,
+      "step": 2982
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5489605948938818,
+      "learning_rate": 7.26027752682498e-06,
+      "loss": 0.5008,
+      "step": 2983
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.70728934022771,
+      "learning_rate": 7.258484695471391e-06,
+      "loss": 0.4936,
+      "step": 2984
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4794145127355325,
+      "learning_rate": 7.256691499239708e-06,
+      "loss": 0.5802,
+      "step": 2985
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2814675673826028,
+      "learning_rate": 7.254897938419637e-06,
+      "loss": 0.5676,
+      "step": 2986
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.828637831299985,
+      "learning_rate": 7.253104013300944e-06,
+      "loss": 0.4818,
+      "step": 2987
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4575986554723903,
+      "learning_rate": 7.251309724173457e-06,
+      "loss": 0.5136,
+      "step": 2988
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4005442517007833,
+      "learning_rate": 7.249515071327054e-06,
+      "loss": 0.5347,
+      "step": 2989
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3491908553203378,
+      "learning_rate": 7.2477200550516805e-06,
+      "loss": 0.5472,
+      "step": 2990
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1381894261443333,
+      "learning_rate": 7.245924675637335e-06,
+      "loss": 0.4848,
+      "step": 2991
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3732018389100948,
+      "learning_rate": 7.244128933374078e-06,
+      "loss": 0.5319,
+      "step": 2992
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7890526708319623,
+      "learning_rate": 7.242332828552028e-06,
+      "loss": 0.4988,
+      "step": 2993
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0272958037206954,
+      "learning_rate": 7.240536361461361e-06,
+      "loss": 0.5384,
+      "step": 2994
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5727488158742704,
+      "learning_rate": 7.238739532392311e-06,
+      "loss": 0.5257,
+      "step": 2995
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6202822584631098,
+      "learning_rate": 7.236942341635172e-06,
+      "loss": 0.5044,
+      "step": 2996
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3387512396302006,
+      "learning_rate": 7.2351447894802975e-06,
+      "loss": 0.5279,
+      "step": 2997
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4411390442007235,
+      "learning_rate": 7.233346876218097e-06,
+      "loss": 0.5236,
+      "step": 2998
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5184640188144154,
+      "learning_rate": 7.231548602139038e-06,
+      "loss": 0.4996,
+      "step": 2999
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.94169479448034,
+      "learning_rate": 7.22974996753365e-06,
+      "loss": 0.5685,
+      "step": 3000
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6834086594487581,
+      "learning_rate": 7.227950972692517e-06,
+      "loss": 0.4798,
+      "step": 3001
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6671153007124514,
+      "learning_rate": 7.226151617906281e-06,
+      "loss": 0.5056,
+      "step": 3002
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.703830372372508,
+      "learning_rate": 7.224351903465644e-06,
+      "loss": 0.5304,
+      "step": 3003
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4897957844867715,
+      "learning_rate": 7.222551829661368e-06,
+      "loss": 0.5406,
+      "step": 3004
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2547913791026493,
+      "learning_rate": 7.22075139678427e-06,
+      "loss": 0.5353,
+      "step": 3005
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3197654635000653,
+      "learning_rate": 7.2189506051252255e-06,
+      "loss": 0.479,
+      "step": 3006
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7469509824577085,
+      "learning_rate": 7.217149454975168e-06,
+      "loss": 0.5134,
+      "step": 3007
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6336829981917924,
+      "learning_rate": 7.215347946625088e-06,
+      "loss": 0.5164,
+      "step": 3008
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3129226881612475,
+      "learning_rate": 7.213546080366036e-06,
+      "loss": 0.4935,
+      "step": 3009
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.476449022913655,
+      "learning_rate": 7.21174385648912e-06,
+      "loss": 0.5313,
+      "step": 3010
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.457575248183513,
+      "learning_rate": 7.209941275285504e-06,
+      "loss": 0.554,
+      "step": 3011
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4815888327448383,
+      "learning_rate": 7.208138337046413e-06,
+      "loss": 0.5028,
+      "step": 3012
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7516509103197164,
+      "learning_rate": 7.206335042063125e-06,
+      "loss": 0.5096,
+      "step": 3013
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.391886475393212,
+      "learning_rate": 7.20453139062698e-06,
+      "loss": 0.5031,
+      "step": 3014
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5367984607946585,
+      "learning_rate": 7.202727383029372e-06,
+      "loss": 0.5289,
+      "step": 3015
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7267274430184791,
+      "learning_rate": 7.200923019561756e-06,
+      "loss": 0.5325,
+      "step": 3016
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.538263403773552,
+      "learning_rate": 7.199118300515644e-06,
+      "loss": 0.4914,
+      "step": 3017
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5598289001903407,
+      "learning_rate": 7.197313226182601e-06,
+      "loss": 0.4706,
+      "step": 3018
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.39709071983512,
+      "learning_rate": 7.195507796854253e-06,
+      "loss": 0.519,
+      "step": 3019
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.97273352499986,
+      "learning_rate": 7.193702012822285e-06,
+      "loss": 0.521,
+      "step": 3020
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.484045494576766,
+      "learning_rate": 7.191895874378436e-06,
+      "loss": 0.5457,
+      "step": 3021
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4923881307280848,
+      "learning_rate": 7.190089381814505e-06,
+      "loss": 0.4928,
+      "step": 3022
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.044763977393002,
+      "learning_rate": 7.188282535422345e-06,
+      "loss": 0.5349,
+      "step": 3023
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6422923664857394,
+      "learning_rate": 7.186475335493867e-06,
+      "loss": 0.5185,
+      "step": 3024
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1337642672469794,
+      "learning_rate": 7.184667782321044e-06,
+      "loss": 0.568,
+      "step": 3025
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.285695554350569,
+      "learning_rate": 7.182859876195903e-06,
+      "loss": 0.5264,
+      "step": 3026
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7352072076106895,
+      "learning_rate": 7.1810516174105195e-06,
+      "loss": 0.5556,
+      "step": 3027
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.631523989032772,
+      "learning_rate": 7.179243006257038e-06,
+      "loss": 0.5464,
+      "step": 3028
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5491726163007693,
+      "learning_rate": 7.177434043027658e-06,
+      "loss": 0.4847,
+      "step": 3029
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7115960087940415,
+      "learning_rate": 7.175624728014631e-06,
+      "loss": 0.5297,
+      "step": 3030
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3545527937682438,
+      "learning_rate": 7.173815061510267e-06,
+      "loss": 0.45,
+      "step": 3031
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2777021474778372,
+      "learning_rate": 7.172005043806934e-06,
+      "loss": 0.5129,
+      "step": 3032
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5843455375733593,
+      "learning_rate": 7.170194675197059e-06,
+      "loss": 0.4926,
+      "step": 3033
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5214693147833276,
+      "learning_rate": 7.168383955973119e-06,
+      "loss": 0.5229,
+      "step": 3034
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.261454930205327,
+      "learning_rate": 7.166572886427655e-06,
+      "loss": 0.465,
+      "step": 3035
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.419593791340515,
+      "learning_rate": 7.164761466853258e-06,
+      "loss": 0.548,
+      "step": 3036
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.364144842925871,
+      "learning_rate": 7.162949697542583e-06,
+      "loss": 0.4924,
+      "step": 3037
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.417175570424057,
+      "learning_rate": 7.161137578788333e-06,
+      "loss": 0.5086,
+      "step": 3038
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6138840172605053,
+      "learning_rate": 7.159325110883274e-06,
+      "loss": 0.5614,
+      "step": 3039
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4548766722256878,
+      "learning_rate": 7.157512294120225e-06,
+      "loss": 0.5243,
+      "step": 3040
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6327820741598216,
+      "learning_rate": 7.155699128792063e-06,
+      "loss": 0.537,
+      "step": 3041
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6640787478353725,
+      "learning_rate": 7.153885615191723e-06,
+      "loss": 0.55,
+      "step": 3042
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.526062849121019,
+      "learning_rate": 7.15207175361219e-06,
+      "loss": 0.5163,
+      "step": 3043
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.488004814390481,
+      "learning_rate": 7.150257544346513e-06,
+      "loss": 0.5014,
+      "step": 3044
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.458619976622426,
+      "learning_rate": 7.148442987687792e-06,
+      "loss": 0.5324,
+      "step": 3045
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6743278152473341,
+      "learning_rate": 7.146628083929183e-06,
+      "loss": 0.5597,
+      "step": 3046
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4734133235019455,
+      "learning_rate": 7.144812833363902e-06,
+      "loss": 0.4929,
+      "step": 3047
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5452831563979532,
+      "learning_rate": 7.142997236285217e-06,
+      "loss": 0.5187,
+      "step": 3048
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2664513770058123,
+      "learning_rate": 7.141181292986457e-06,
+      "loss": 0.5586,
+      "step": 3049
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3615095997700022,
+      "learning_rate": 7.139365003760998e-06,
+      "loss": 0.5051,
+      "step": 3050
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.843213610546877,
+      "learning_rate": 7.137548368902284e-06,
+      "loss": 0.4999,
+      "step": 3051
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4640428774254095,
+      "learning_rate": 7.135731388703804e-06,
+      "loss": 0.4822,
+      "step": 3052
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5575020507385404,
+      "learning_rate": 7.133914063459108e-06,
+      "loss": 0.5176,
+      "step": 3053
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3698924554546537,
+      "learning_rate": 7.132096393461801e-06,
+      "loss": 0.5603,
+      "step": 3054
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3405839230187293,
+      "learning_rate": 7.130278379005545e-06,
+      "loss": 0.5195,
+      "step": 3055
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5140107590051919,
+      "learning_rate": 7.128460020384055e-06,
+      "loss": 0.4992,
+      "step": 3056
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1915114977633574,
+      "learning_rate": 7.126641317891101e-06,
+      "loss": 0.4718,
+      "step": 3057
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4336714376199327,
+      "learning_rate": 7.124822271820513e-06,
+      "loss": 0.4776,
+      "step": 3058
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.388417058377369,
+      "learning_rate": 7.123002882466174e-06,
+      "loss": 0.4925,
+      "step": 3059
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.538238246586762,
+      "learning_rate": 7.121183150122022e-06,
+      "loss": 0.5452,
+      "step": 3060
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7098794479468666,
+      "learning_rate": 7.1193630750820495e-06,
+      "loss": 0.5372,
+      "step": 3061
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6669472641299299,
+      "learning_rate": 7.117542657640307e-06,
+      "loss": 0.5566,
+      "step": 3062
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6543545849103167,
+      "learning_rate": 7.1157218980908984e-06,
+      "loss": 0.5145,
+      "step": 3063
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9544024341278428,
+      "learning_rate": 7.113900796727984e-06,
+      "loss": 0.5462,
+      "step": 3064
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.933868451967738,
+      "learning_rate": 7.112079353845775e-06,
+      "loss": 0.5273,
+      "step": 3065
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4942206033292202,
+      "learning_rate": 7.110257569738549e-06,
+      "loss": 0.4997,
+      "step": 3066
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9699941736244426,
+      "learning_rate": 7.108435444700626e-06,
+      "loss": 0.4467,
+      "step": 3067
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5284924845003787,
+      "learning_rate": 7.106612979026387e-06,
+      "loss": 0.4825,
+      "step": 3068
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6737801292785671,
+      "learning_rate": 7.104790173010268e-06,
+      "loss": 0.5245,
+      "step": 3069
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4078834972046943,
+      "learning_rate": 7.102967026946758e-06,
+      "loss": 0.5132,
+      "step": 3070
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3875591801648743,
+      "learning_rate": 7.101143541130407e-06,
+      "loss": 0.4935,
+      "step": 3071
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2158183608208777,
+      "learning_rate": 7.0993197158558095e-06,
+      "loss": 0.491,
+      "step": 3072
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7437431691806793,
+      "learning_rate": 7.097495551417621e-06,
+      "loss": 0.5167,
+      "step": 3073
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.429828259302202,
+      "learning_rate": 7.095671048110555e-06,
+      "loss": 0.5072,
+      "step": 3074
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4081079213848775,
+      "learning_rate": 7.093846206229373e-06,
+      "loss": 0.521,
+      "step": 3075
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.512805807502615,
+      "learning_rate": 7.092021026068897e-06,
+      "loss": 0.5068,
+      "step": 3076
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6821786251772541,
+      "learning_rate": 7.090195507923998e-06,
+      "loss": 0.5147,
+      "step": 3077
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5016684049091409,
+      "learning_rate": 7.088369652089607e-06,
+      "loss": 0.4811,
+      "step": 3078
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3965871276845612,
+      "learning_rate": 7.086543458860706e-06,
+      "loss": 0.517,
+      "step": 3079
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9053127055549555,
+      "learning_rate": 7.084716928532334e-06,
+      "loss": 0.4807,
+      "step": 3080
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5364628357558445,
+      "learning_rate": 7.0828900613995775e-06,
+      "loss": 0.5673,
+      "step": 3081
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5726149366761402,
+      "learning_rate": 7.08106285775759e-06,
+      "loss": 0.5595,
+      "step": 3082
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4407485616166227,
+      "learning_rate": 7.079235317901569e-06,
+      "loss": 0.5062,
+      "step": 3083
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9085919826528777,
+      "learning_rate": 7.07740744212677e-06,
+      "loss": 0.5158,
+      "step": 3084
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.486163869615989,
+      "learning_rate": 7.0755792307285024e-06,
+      "loss": 0.4841,
+      "step": 3085
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3083113514428966,
+      "learning_rate": 7.07375068400213e-06,
+      "loss": 0.4705,
+      "step": 3086
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.831123081797392,
+      "learning_rate": 7.0719218022430715e-06,
+      "loss": 0.5052,
+      "step": 3087
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6126664823754961,
+      "learning_rate": 7.070092585746798e-06,
+      "loss": 0.5042,
+      "step": 3088
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4079867138889446,
+      "learning_rate": 7.0682630348088336e-06,
+      "loss": 0.5271,
+      "step": 3089
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4242873896765287,
+      "learning_rate": 7.066433149724762e-06,
+      "loss": 0.4858,
+      "step": 3090
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4017570824767323,
+      "learning_rate": 7.064602930790215e-06,
+      "loss": 0.5615,
+      "step": 3091
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6840962622500598,
+      "learning_rate": 7.062772378300882e-06,
+      "loss": 0.4723,
+      "step": 3092
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3857989267211437,
+      "learning_rate": 7.060941492552502e-06,
+      "loss": 0.5404,
+      "step": 3093
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.490729207593928,
+      "learning_rate": 7.0591102738408735e-06,
+      "loss": 0.543,
+      "step": 3094
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5869525897166517,
+      "learning_rate": 7.057278722461845e-06,
+      "loss": 0.5254,
+      "step": 3095
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6003221119375428,
+      "learning_rate": 7.0554468387113214e-06,
+      "loss": 0.524,
+      "step": 3096
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.557506866511841,
+      "learning_rate": 7.053614622885258e-06,
+      "loss": 0.4996,
+      "step": 3097
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.973734265611302,
+      "learning_rate": 7.051782075279665e-06,
+      "loss": 0.4846,
+      "step": 3098
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.393020111046219,
+      "learning_rate": 7.049949196190607e-06,
+      "loss": 0.5014,
+      "step": 3099
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6516241639640925,
+      "learning_rate": 7.048115985914204e-06,
+      "loss": 0.4566,
+      "step": 3100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.314792255216263,
+      "learning_rate": 7.046282444746624e-06,
+      "loss": 0.4513,
+      "step": 3101
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5707224724922517,
+      "learning_rate": 7.044448572984091e-06,
+      "loss": 0.5088,
+      "step": 3102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4978671506491446,
+      "learning_rate": 7.042614370922887e-06,
+      "loss": 0.527,
+      "step": 3103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7301185957841059,
+      "learning_rate": 7.0407798388593415e-06,
+      "loss": 0.4978,
+      "step": 3104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2063768657721639,
+      "learning_rate": 7.03894497708984e-06,
+      "loss": 0.5352,
+      "step": 3105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5425265184295172,
+      "learning_rate": 7.03710978591082e-06,
+      "loss": 0.5108,
+      "step": 3106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.646854735973331,
+      "learning_rate": 7.035274265618772e-06,
+      "loss": 0.5169,
+      "step": 3107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5097235016060884,
+      "learning_rate": 7.033438416510241e-06,
+      "loss": 0.4808,
+      "step": 3108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6812662874754638,
+      "learning_rate": 7.031602238881826e-06,
+      "loss": 0.5268,
+      "step": 3109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.479023250144563,
+      "learning_rate": 7.029765733030175e-06,
+      "loss": 0.523,
+      "step": 3110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3145790277291225,
+      "learning_rate": 7.027928899251995e-06,
+      "loss": 0.5166,
+      "step": 3111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4619991861740071,
+      "learning_rate": 7.02609173784404e-06,
+      "loss": 0.5387,
+      "step": 3112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3539576115167475,
+      "learning_rate": 7.0242542491031205e-06,
+      "loss": 0.4792,
+      "step": 3113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.859711344496118,
+      "learning_rate": 7.022416433326099e-06,
+      "loss": 0.5065,
+      "step": 3114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6856514624428732,
+      "learning_rate": 7.020578290809892e-06,
+      "loss": 0.4905,
+      "step": 3115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.294158902627508,
+      "learning_rate": 7.018739821851466e-06,
+      "loss": 0.5295,
+      "step": 3116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5880413052145466,
+      "learning_rate": 7.016901026747842e-06,
+      "loss": 0.5467,
+      "step": 3117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3439530838922578,
+      "learning_rate": 7.0150619057960926e-06,
+      "loss": 0.4675,
+      "step": 3118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6216617436812264,
+      "learning_rate": 7.0132224592933464e-06,
+      "loss": 0.5298,
+      "step": 3119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5317615759809857,
+      "learning_rate": 7.011382687536781e-06,
+      "loss": 0.5099,
+      "step": 3120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8957008492341494,
+      "learning_rate": 7.009542590823628e-06,
+      "loss": 0.5106,
+      "step": 3121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3386803487504768,
+      "learning_rate": 7.007702169451169e-06,
+      "loss": 0.4911,
+      "step": 3122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.600767614933352,
+      "learning_rate": 7.0058614237167445e-06,
+      "loss": 0.4981,
+      "step": 3123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6010427135166088,
+      "learning_rate": 7.004020353917742e-06,
+      "loss": 0.4986,
+      "step": 3124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6392739027265255,
+      "learning_rate": 7.0021789603515995e-06,
+      "loss": 0.5125,
+      "step": 3125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7549088594860496,
+      "learning_rate": 7.000337243315812e-06,
+      "loss": 0.4807,
+      "step": 3126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7399474805452981,
+      "learning_rate": 6.998495203107927e-06,
+      "loss": 0.4826,
+      "step": 3127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0452652233489848,
+      "learning_rate": 6.996652840025539e-06,
+      "loss": 0.5083,
+      "step": 3128
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4516835604294887,
+      "learning_rate": 6.994810154366302e-06,
+      "loss": 0.481,
+      "step": 3129
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.229369461643623,
+      "learning_rate": 6.992967146427913e-06,
+      "loss": 0.5527,
+      "step": 3130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2383374837601475,
+      "learning_rate": 6.991123816508131e-06,
+      "loss": 0.4843,
+      "step": 3131
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4156075245677167,
+      "learning_rate": 6.989280164904759e-06,
+      "loss": 0.4891,
+      "step": 3132
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.711299286904344,
+      "learning_rate": 6.987436191915658e-06,
+      "loss": 0.5811,
+      "step": 3133
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4285050158864885,
+      "learning_rate": 6.985591897838736e-06,
+      "loss": 0.4793,
+      "step": 3134
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6946023949726925,
+      "learning_rate": 6.983747282971954e-06,
+      "loss": 0.5321,
+      "step": 3135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.468337455984439,
+      "learning_rate": 6.981902347613328e-06,
+      "loss": 0.4785,
+      "step": 3136
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5886861702690127,
+      "learning_rate": 6.980057092060924e-06,
+      "loss": 0.4854,
+      "step": 3137
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2264486795838963,
+      "learning_rate": 6.9782115166128565e-06,
+      "loss": 0.5179,
+      "step": 3138
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5917890050453414,
+      "learning_rate": 6.976365621567295e-06,
+      "loss": 0.5331,
+      "step": 3139
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.478122933457729,
+      "learning_rate": 6.974519407222462e-06,
+      "loss": 0.4715,
+      "step": 3140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7984096108555954,
+      "learning_rate": 6.9726728738766295e-06,
+      "loss": 0.5443,
+      "step": 3141
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6560769479857425,
+      "learning_rate": 6.97082602182812e-06,
+      "loss": 0.4792,
+      "step": 3142
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7874171722080707,
+      "learning_rate": 6.9689788513753094e-06,
+      "loss": 0.5356,
+      "step": 3143
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6950227531979957,
+      "learning_rate": 6.967131362816623e-06,
+      "loss": 0.4974,
+      "step": 3144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.539464285764076,
+      "learning_rate": 6.965283556450542e-06,
+      "loss": 0.5479,
+      "step": 3145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4146763723631428,
+      "learning_rate": 6.963435432575593e-06,
+      "loss": 0.5373,
+      "step": 3146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4136312113724916,
+      "learning_rate": 6.961586991490357e-06,
+      "loss": 0.529,
+      "step": 3147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9189851726240688,
+      "learning_rate": 6.959738233493466e-06,
+      "loss": 0.5202,
+      "step": 3148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.399927805798971,
+      "learning_rate": 6.957889158883604e-06,
+      "loss": 0.5258,
+      "step": 3149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6344488410503175,
+      "learning_rate": 6.9560397679595044e-06,
+      "loss": 0.5367,
+      "step": 3150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.441472487643674,
+      "learning_rate": 6.954190061019954e-06,
+      "loss": 0.5707,
+      "step": 3151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4277228917494604,
+      "learning_rate": 6.952340038363788e-06,
+      "loss": 0.4729,
+      "step": 3152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3571806069444707,
+      "learning_rate": 6.950489700289894e-06,
+      "loss": 0.4925,
+      "step": 3153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3716644997004555,
+      "learning_rate": 6.948639047097211e-06,
+      "loss": 0.433,
+      "step": 3154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6052244929492939,
+      "learning_rate": 6.946788079084727e-06,
+      "loss": 0.4785,
+      "step": 3155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4155172324818357,
+      "learning_rate": 6.944936796551482e-06,
+      "loss": 0.4885,
+      "step": 3156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.707369994295181,
+      "learning_rate": 6.943085199796571e-06,
+      "loss": 0.4817,
+      "step": 3157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4534014173887673,
+      "learning_rate": 6.9412332891191315e-06,
+      "loss": 0.5137,
+      "step": 3158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.103367237771373,
+      "learning_rate": 6.9393810648183566e-06,
+      "loss": 0.5128,
+      "step": 3159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6222332634783083,
+      "learning_rate": 6.937528527193491e-06,
+      "loss": 0.469,
+      "step": 3160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3360922385144436,
+      "learning_rate": 6.935675676543827e-06,
+      "loss": 0.5262,
+      "step": 3161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6389463231703143,
+      "learning_rate": 6.93382251316871e-06,
+      "loss": 0.5106,
+      "step": 3162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4765630400181307,
+      "learning_rate": 6.931969037367533e-06,
+      "loss": 0.5152,
+      "step": 3163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4176110067019618,
+      "learning_rate": 6.930115249439744e-06,
+      "loss": 0.4844,
+      "step": 3164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3708693622727413,
+      "learning_rate": 6.928261149684837e-06,
+      "loss": 0.5306,
+      "step": 3165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4359433735125964,
+      "learning_rate": 6.926406738402359e-06,
+      "loss": 0.4372,
+      "step": 3166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.567047899929472,
+      "learning_rate": 6.924552015891905e-06,
+      "loss": 0.5303,
+      "step": 3167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.490193781690818,
+      "learning_rate": 6.9226969824531254e-06,
+      "loss": 0.5776,
+      "step": 3168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.002612215630255,
+      "learning_rate": 6.920841638385715e-06,
+      "loss": 0.4715,
+      "step": 3169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8963193464549652,
+      "learning_rate": 6.918985983989418e-06,
+      "loss": 0.5875,
+      "step": 3170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.213321963170932,
+      "learning_rate": 6.917130019564034e-06,
+      "loss": 0.5225,
+      "step": 3171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4544021040788373,
+      "learning_rate": 6.915273745409413e-06,
+      "loss": 0.5195,
+      "step": 3172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9278156233429231,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 0.5123,
+      "step": 3173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6524752099528455,
+      "learning_rate": 6.911560269112092e-06,
+      "loss": 0.5542,
+      "step": 3174
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.541481926746516,
+      "learning_rate": 6.909703067569337e-06,
+      "loss": 0.5246,
+      "step": 3175
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4279355948714822,
+      "learning_rate": 6.907845557497231e-06,
+      "loss": 0.4483,
+      "step": 3176
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.743431193415441,
+      "learning_rate": 6.905987739195874e-06,
+      "loss": 0.5881,
+      "step": 3177
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.61440211095754,
+      "learning_rate": 6.9041296129654125e-06,
+      "loss": 0.5301,
+      "step": 3178
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6702754112326066,
+      "learning_rate": 6.902271179106041e-06,
+      "loss": 0.5233,
+      "step": 3179
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.920713679990864,
+      "learning_rate": 6.900412437918005e-06,
+      "loss": 0.4744,
+      "step": 3180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3697518409491982,
+      "learning_rate": 6.898553389701603e-06,
+      "loss": 0.5108,
+      "step": 3181
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1212805977850646,
+      "learning_rate": 6.896694034757181e-06,
+      "loss": 0.5294,
+      "step": 3182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5411551878710228,
+      "learning_rate": 6.894834373385132e-06,
+      "loss": 0.5106,
+      "step": 3183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4368851055143177,
+      "learning_rate": 6.892974405885902e-06,
+      "loss": 0.4735,
+      "step": 3184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9299020300383583,
+      "learning_rate": 6.891114132559985e-06,
+      "loss": 0.5244,
+      "step": 3185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.473232038624168,
+      "learning_rate": 6.8892535537079245e-06,
+      "loss": 0.5071,
+      "step": 3186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.588323591332897,
+      "learning_rate": 6.8873926696303135e-06,
+      "loss": 0.4919,
+      "step": 3187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7334748667952768,
+      "learning_rate": 6.885531480627794e-06,
+      "loss": 0.5087,
+      "step": 3188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4829244717106247,
+      "learning_rate": 6.883669987001058e-06,
+      "loss": 0.5095,
+      "step": 3189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6092064733528466,
+      "learning_rate": 6.8818081890508456e-06,
+      "loss": 0.5232,
+      "step": 3190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.341271386639204,
+      "learning_rate": 6.8799460870779465e-06,
+      "loss": 0.5029,
+      "step": 3191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4481930839888286,
+      "learning_rate": 6.878083681383198e-06,
+      "loss": 0.5057,
+      "step": 3192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3083656791753513,
+      "learning_rate": 6.876220972267494e-06,
+      "loss": 0.4886,
+      "step": 3193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.39296038386585,
+      "learning_rate": 6.874357960031765e-06,
+      "loss": 0.5084,
+      "step": 3194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.900143200573506,
+      "learning_rate": 6.872494644977e-06,
+      "loss": 0.5262,
+      "step": 3195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8050499601037346,
+      "learning_rate": 6.8706310274042345e-06,
+      "loss": 0.5068,
+      "step": 3196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.352345388329052,
+      "learning_rate": 6.868767107614552e-06,
+      "loss": 0.4604,
+      "step": 3197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6545175377347725,
+      "learning_rate": 6.866902885909083e-06,
+      "loss": 0.5316,
+      "step": 3198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2856771180257691,
+      "learning_rate": 6.865038362589012e-06,
+      "loss": 0.5097,
+      "step": 3199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6801585779785584,
+      "learning_rate": 6.863173537955566e-06,
+      "loss": 0.5037,
+      "step": 3200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6786852023536945,
+      "learning_rate": 6.861308412310026e-06,
+      "loss": 0.5134,
+      "step": 3201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3999297059160014,
+      "learning_rate": 6.85944298595372e-06,
+      "loss": 0.5327,
+      "step": 3202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3839457585180088,
+      "learning_rate": 6.857577259188022e-06,
+      "loss": 0.547,
+      "step": 3203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4739330800383263,
+      "learning_rate": 6.855711232314358e-06,
+      "loss": 0.5638,
+      "step": 3204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.569513748632115,
+      "learning_rate": 6.853844905634202e-06,
+      "loss": 0.4931,
+      "step": 3205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8421076393489355,
+      "learning_rate": 6.851978279449073e-06,
+      "loss": 0.5661,
+      "step": 3206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.956359431715477,
+      "learning_rate": 6.850111354060543e-06,
+      "loss": 0.4918,
+      "step": 3207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.577412127294876,
+      "learning_rate": 6.848244129770228e-06,
+      "loss": 0.5145,
+      "step": 3208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4958117513380969,
+      "learning_rate": 6.8463766068797964e-06,
+      "loss": 0.5455,
+      "step": 3209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4911133031146588,
+      "learning_rate": 6.844508785690964e-06,
+      "loss": 0.489,
+      "step": 3210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4088742719955414,
+      "learning_rate": 6.842640666505491e-06,
+      "loss": 0.5215,
+      "step": 3211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6629950094869201,
+      "learning_rate": 6.840772249625189e-06,
+      "loss": 0.5135,
+      "step": 3212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2384801027680652,
+      "learning_rate": 6.838903535351921e-06,
+      "loss": 0.5416,
+      "step": 3213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.460819899140803,
+      "learning_rate": 6.837034523987589e-06,
+      "loss": 0.5163,
+      "step": 3214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4184315724520054,
+      "learning_rate": 6.835165215834151e-06,
+      "loss": 0.4849,
+      "step": 3215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3412792084883807,
+      "learning_rate": 6.83329561119361e-06,
+      "loss": 0.5011,
+      "step": 3216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.877516849937572,
+      "learning_rate": 6.831425710368016e-06,
+      "loss": 0.5177,
+      "step": 3217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4397412895230428,
+      "learning_rate": 6.829555513659468e-06,
+      "loss": 0.4972,
+      "step": 3218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6384697735109914,
+      "learning_rate": 6.827685021370115e-06,
+      "loss": 0.5218,
+      "step": 3219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8141569051259778,
+      "learning_rate": 6.825814233802151e-06,
+      "loss": 0.4911,
+      "step": 3220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.306899734027489,
+      "learning_rate": 6.8239431512578135e-06,
+      "loss": 0.5356,
+      "step": 3221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3572460136657636,
+      "learning_rate": 6.822071774039399e-06,
+      "loss": 0.5072,
+      "step": 3222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5945829296918865,
+      "learning_rate": 6.820200102449243e-06,
+      "loss": 0.5119,
+      "step": 3223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2301820912603123,
+      "learning_rate": 6.818328136789727e-06,
+      "loss": 0.4695,
+      "step": 3224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.379457106022792,
+      "learning_rate": 6.816455877363286e-06,
+      "loss": 0.4784,
+      "step": 3225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3225665325357554,
+      "learning_rate": 6.814583324472401e-06,
+      "loss": 0.5207,
+      "step": 3226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2700246527823775,
+      "learning_rate": 6.8127104784196e-06,
+      "loss": 0.461,
+      "step": 3227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8472757730911826,
+      "learning_rate": 6.810837339507454e-06,
+      "loss": 0.5445,
+      "step": 3228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6707847338140217,
+      "learning_rate": 6.808963908038589e-06,
+      "loss": 0.4855,
+      "step": 3229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.691319968280161,
+      "learning_rate": 6.807090184315671e-06,
+      "loss": 0.5171,
+      "step": 3230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2302438266834328,
+      "learning_rate": 6.80521616864142e-06,
+      "loss": 0.4727,
+      "step": 3231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3207802784267022,
+      "learning_rate": 6.803341861318598e-06,
+      "loss": 0.5208,
+      "step": 3232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4476132703215128,
+      "learning_rate": 6.801467262650015e-06,
+      "loss": 0.476,
+      "step": 3233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6336738979678858,
+      "learning_rate": 6.799592372938529e-06,
+      "loss": 0.5643,
+      "step": 3234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.835620022589171,
+      "learning_rate": 6.797717192487046e-06,
+      "loss": 0.5108,
+      "step": 3235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5619037774003635,
+      "learning_rate": 6.79584172159852e-06,
+      "loss": 0.4783,
+      "step": 3236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7819998664226975,
+      "learning_rate": 6.793965960575944e-06,
+      "loss": 0.5288,
+      "step": 3237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4359135475085967,
+      "learning_rate": 6.79208990972237e-06,
+      "loss": 0.5227,
+      "step": 3238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6807354645349077,
+      "learning_rate": 6.790213569340887e-06,
+      "loss": 0.5521,
+      "step": 3239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3640185458322238,
+      "learning_rate": 6.788336939734634e-06,
+      "loss": 0.5437,
+      "step": 3240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5519889606097421,
+      "learning_rate": 6.7864600212068e-06,
+      "loss": 0.4967,
+      "step": 3241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3749868972314432,
+      "learning_rate": 6.784582814060615e-06,
+      "loss": 0.4629,
+      "step": 3242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7267913886036812,
+      "learning_rate": 6.78270531859936e-06,
+      "loss": 0.5539,
+      "step": 3243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.609565352487597,
+      "learning_rate": 6.7808275351263595e-06,
+      "loss": 0.4702,
+      "step": 3244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4380316093444492,
+      "learning_rate": 6.778949463944985e-06,
+      "loss": 0.4981,
+      "step": 3245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.515498940838077,
+      "learning_rate": 6.777071105358659e-06,
+      "loss": 0.5276,
+      "step": 3246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2043397648266816,
+      "learning_rate": 6.775192459670844e-06,
+      "loss": 0.5164,
+      "step": 3247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.364429422355196,
+      "learning_rate": 6.773313527185053e-06,
+      "loss": 0.5203,
+      "step": 3248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2666796513747796,
+      "learning_rate": 6.771434308204844e-06,
+      "loss": 0.48,
+      "step": 3249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6534994415436592,
+      "learning_rate": 6.769554803033821e-06,
+      "loss": 0.4969,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2553049497037567,
+      "learning_rate": 6.767675011975634e-06,
+      "loss": 0.531,
+      "step": 3251
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7674560829884376,
+      "learning_rate": 6.765794935333981e-06,
+      "loss": 0.5026,
+      "step": 3252
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.829940354533642,
+      "learning_rate": 6.763914573412604e-06,
+      "loss": 0.5008,
+      "step": 3253
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.78692529089743,
+      "learning_rate": 6.762033926515293e-06,
+      "loss": 0.4988,
+      "step": 3254
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6657136686058656,
+      "learning_rate": 6.760152994945882e-06,
+      "loss": 0.4766,
+      "step": 3255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3284019099181725,
+      "learning_rate": 6.758271779008254e-06,
+      "loss": 0.5209,
+      "step": 3256
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4389937674757496,
+      "learning_rate": 6.756390279006333e-06,
+      "loss": 0.5446,
+      "step": 3257
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4769530562233115,
+      "learning_rate": 6.754508495244096e-06,
+      "loss": 0.4907,
+      "step": 3258
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.562785102305474,
+      "learning_rate": 6.752626428025557e-06,
+      "loss": 0.4999,
+      "step": 3259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5520415303055872,
+      "learning_rate": 6.750744077654783e-06,
+      "loss": 0.547,
+      "step": 3260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2761812793650475,
+      "learning_rate": 6.748861444435885e-06,
+      "loss": 0.4909,
+      "step": 3261
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5786024125723916,
+      "learning_rate": 6.746978528673016e-06,
+      "loss": 0.5128,
+      "step": 3262
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5004031067813843,
+      "learning_rate": 6.7450953306703815e-06,
+      "loss": 0.4872,
+      "step": 3263
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7776122389127431,
+      "learning_rate": 6.743211850732227e-06,
+      "loss": 0.5313,
+      "step": 3264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5788272163994252,
+      "learning_rate": 6.7413280891628445e-06,
+      "loss": 0.5826,
+      "step": 3265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4900859967676385,
+      "learning_rate": 6.739444046266572e-06,
+      "loss": 0.5607,
+      "step": 3266
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5222895795374038,
+      "learning_rate": 6.7375597223477975e-06,
+      "loss": 0.4946,
+      "step": 3267
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3541859908879346,
+      "learning_rate": 6.7356751177109435e-06,
+      "loss": 0.5121,
+      "step": 3268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3777186829538355,
+      "learning_rate": 6.73379023266049e-06,
+      "loss": 0.56,
+      "step": 3269
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4314582012171,
+      "learning_rate": 6.731905067500952e-06,
+      "loss": 0.5078,
+      "step": 3270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.639823130130427,
+      "learning_rate": 6.730019622536899e-06,
+      "loss": 0.5102,
+      "step": 3271
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4987241384585417,
+      "learning_rate": 6.72813389807294e-06,
+      "loss": 0.5499,
+      "step": 3272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6573727062661225,
+      "learning_rate": 6.726247894413728e-06,
+      "loss": 0.5103,
+      "step": 3273
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.261235589777226,
+      "learning_rate": 6.724361611863964e-06,
+      "loss": 0.4597,
+      "step": 3274
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3445659210242793,
+      "learning_rate": 6.722475050728396e-06,
+      "loss": 0.55,
+      "step": 3275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5693155050154757,
+      "learning_rate": 6.720588211311815e-06,
+      "loss": 0.5569,
+      "step": 3276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6456907454871615,
+      "learning_rate": 6.7187010939190555e-06,
+      "loss": 0.5133,
+      "step": 3277
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6245573463874465,
+      "learning_rate": 6.7168136988549935e-06,
+      "loss": 0.4931,
+      "step": 3278
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8694239394978942,
+      "learning_rate": 6.714926026424561e-06,
+      "loss": 0.5084,
+      "step": 3279
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6244660101509918,
+      "learning_rate": 6.713038076932725e-06,
+      "loss": 0.5219,
+      "step": 3280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5859703131172727,
+      "learning_rate": 6.711149850684499e-06,
+      "loss": 0.511,
+      "step": 3281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6905258567661081,
+      "learning_rate": 6.709261347984946e-06,
+      "loss": 0.4687,
+      "step": 3282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3208341183094228,
+      "learning_rate": 6.707372569139167e-06,
+      "loss": 0.4472,
+      "step": 3283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7078454269081773,
+      "learning_rate": 6.705483514452314e-06,
+      "loss": 0.518,
+      "step": 3284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.776538177003102,
+      "learning_rate": 6.703594184229576e-06,
+      "loss": 0.5311,
+      "step": 3285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.429757999962165,
+      "learning_rate": 6.701704578776196e-06,
+      "loss": 0.4716,
+      "step": 3286
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2293398881643591,
+      "learning_rate": 6.699814698397454e-06,
+      "loss": 0.4841,
+      "step": 3287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.295501405232231,
+      "learning_rate": 6.697924543398675e-06,
+      "loss": 0.4988,
+      "step": 3288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4495497732771263,
+      "learning_rate": 6.696034114085233e-06,
+      "loss": 0.5466,
+      "step": 3289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.684673687979301,
+      "learning_rate": 6.694143410762543e-06,
+      "loss": 0.5172,
+      "step": 3290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4588467248806118,
+      "learning_rate": 6.692252433736063e-06,
+      "loss": 0.486,
+      "step": 3291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.454877955929744,
+      "learning_rate": 6.690361183311299e-06,
+      "loss": 0.5379,
+      "step": 3292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4608802875863345,
+      "learning_rate": 6.688469659793799e-06,
+      "loss": 0.5089,
+      "step": 3293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2498258415718937,
+      "learning_rate": 6.686577863489154e-06,
+      "loss": 0.5039,
+      "step": 3294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4267227972809884,
+      "learning_rate": 6.684685794703003e-06,
+      "loss": 0.5738,
+      "step": 3295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4259808443137425,
+      "learning_rate": 6.682793453741022e-06,
+      "loss": 0.5103,
+      "step": 3296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.506083836138939,
+      "learning_rate": 6.6809008409089396e-06,
+      "loss": 0.565,
+      "step": 3297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5930280829423225,
+      "learning_rate": 6.679007956512522e-06,
+      "loss": 0.5352,
+      "step": 3298
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4478525083131581,
+      "learning_rate": 6.6771148008575805e-06,
+      "loss": 0.4984,
+      "step": 3299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.562424146635213,
+      "learning_rate": 6.675221374249972e-06,
+      "loss": 0.5472,
+      "step": 3300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4015293961648203,
+      "learning_rate": 6.673327676995598e-06,
+      "loss": 0.492,
+      "step": 3301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5461992961632107,
+      "learning_rate": 6.671433709400399e-06,
+      "loss": 0.5391,
+      "step": 3302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0638014465505985,
+      "learning_rate": 6.6695394717703654e-06,
+      "loss": 0.5239,
+      "step": 3303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4292878955601038,
+      "learning_rate": 6.6676449644115246e-06,
+      "loss": 0.5406,
+      "step": 3304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8768562923306327,
+      "learning_rate": 6.665750187629953e-06,
+      "loss": 0.5434,
+      "step": 3305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.62994328696491,
+      "learning_rate": 6.6638551417317675e-06,
+      "loss": 0.547,
+      "step": 3306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5736630956087028,
+      "learning_rate": 6.66195982702313e-06,
+      "loss": 0.5195,
+      "step": 3307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4803308672254103,
+      "learning_rate": 6.6600642438102454e-06,
+      "loss": 0.527,
+      "step": 3308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4342128458844394,
+      "learning_rate": 6.658168392399362e-06,
+      "loss": 0.4953,
+      "step": 3309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.412313512785876,
+      "learning_rate": 6.656272273096771e-06,
+      "loss": 0.5431,
+      "step": 3310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6077253118836052,
+      "learning_rate": 6.654375886208806e-06,
+      "loss": 0.5168,
+      "step": 3311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7743537280984463,
+      "learning_rate": 6.652479232041849e-06,
+      "loss": 0.4673,
+      "step": 3312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3382429044577637,
+      "learning_rate": 6.650582310902316e-06,
+      "loss": 0.5023,
+      "step": 3313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.336504964303114,
+      "learning_rate": 6.648685123096674e-06,
+      "loss": 0.4931,
+      "step": 3314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.468379926370309,
+      "learning_rate": 6.646787668931429e-06,
+      "loss": 0.5196,
+      "step": 3315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4020497395917302,
+      "learning_rate": 6.644889948713135e-06,
+      "loss": 0.4779,
+      "step": 3316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.416022357706084,
+      "learning_rate": 6.642991962748381e-06,
+      "loss": 0.5308,
+      "step": 3317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4729653949046748,
+      "learning_rate": 6.641093711343806e-06,
+      "loss": 0.4772,
+      "step": 3318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.288982109275439,
+      "learning_rate": 6.639195194806087e-06,
+      "loss": 0.5182,
+      "step": 3319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2569075774740808,
+      "learning_rate": 6.637296413441949e-06,
+      "loss": 0.4978,
+      "step": 3320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8210600894090125,
+      "learning_rate": 6.635397367558156e-06,
+      "loss": 0.5356,
+      "step": 3321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4408394135319371,
+      "learning_rate": 6.633498057461514e-06,
+      "loss": 0.4808,
+      "step": 3322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3604884909269044,
+      "learning_rate": 6.631598483458874e-06,
+      "loss": 0.4612,
+      "step": 3323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6252149096540622,
+      "learning_rate": 6.629698645857129e-06,
+      "loss": 0.5598,
+      "step": 3324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.1562231472595759,
+      "learning_rate": 6.6277985449632155e-06,
+      "loss": 0.4699,
+      "step": 3325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9768297893227216,
+      "learning_rate": 6.625898181084111e-06,
+      "loss": 0.5483,
+      "step": 3326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5422501208070072,
+      "learning_rate": 6.623997554526833e-06,
+      "loss": 0.4897,
+      "step": 3327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.361915446466029,
+      "learning_rate": 6.62209666559845e-06,
+      "loss": 0.5565,
+      "step": 3328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4386159169197326,
+      "learning_rate": 6.620195514606063e-06,
+      "loss": 0.5225,
+      "step": 3329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4925820107560432,
+      "learning_rate": 6.6182941018568224e-06,
+      "loss": 0.5247,
+      "step": 3330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.414937074964594,
+      "learning_rate": 6.616392427657918e-06,
+      "loss": 0.5447,
+      "step": 3331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.419668460668176,
+      "learning_rate": 6.614490492316578e-06,
+      "loss": 0.5357,
+      "step": 3332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.781817186987209,
+      "learning_rate": 6.612588296140082e-06,
+      "loss": 0.5121,
+      "step": 3333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.48145844537759,
+      "learning_rate": 6.610685839435744e-06,
+      "loss": 0.5703,
+      "step": 3334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3152571504495545,
+      "learning_rate": 6.608783122510922e-06,
+      "loss": 0.515,
+      "step": 3335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.410781395131533,
+      "learning_rate": 6.606880145673018e-06,
+      "loss": 0.5163,
+      "step": 3336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5398437941017569,
+      "learning_rate": 6.604976909229475e-06,
+      "loss": 0.5226,
+      "step": 3337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8230133487038807,
+      "learning_rate": 6.603073413487777e-06,
+      "loss": 0.5104,
+      "step": 3338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.319234468105146,
+      "learning_rate": 6.6011696587554495e-06,
+      "loss": 0.5276,
+      "step": 3339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8885195195601567,
+      "learning_rate": 6.599265645340063e-06,
+      "loss": 0.5423,
+      "step": 3340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.341219235782096,
+      "learning_rate": 6.597361373549226e-06,
+      "loss": 0.4838,
+      "step": 3341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9725768133216375,
+      "learning_rate": 6.595456843690591e-06,
+      "loss": 0.5362,
+      "step": 3342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7764264535883638,
+      "learning_rate": 6.59355205607185e-06,
+      "loss": 0.4959,
+      "step": 3343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.657915233528162,
+      "learning_rate": 6.59164701100074e-06,
+      "loss": 0.5096,
+      "step": 3344
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.365255514796697,
+      "learning_rate": 6.589741708785038e-06,
+      "loss": 0.4849,
+      "step": 3345
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.385547860326123,
+      "learning_rate": 6.587836149732562e-06,
+      "loss": 0.5173,
+      "step": 3346
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6822192304164865,
+      "learning_rate": 6.585930334151172e-06,
+      "loss": 0.4999,
+      "step": 3347
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.401093568895059,
+      "learning_rate": 6.584024262348767e-06,
+      "loss": 0.5137,
+      "step": 3348
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5128507982800243,
+      "learning_rate": 6.582117934633293e-06,
+      "loss": 0.5186,
+      "step": 3349
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3496302725839122,
+      "learning_rate": 6.580211351312733e-06,
+      "loss": 0.5126,
+      "step": 3350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5700861571843783,
+      "learning_rate": 6.5783045126951104e-06,
+      "loss": 0.5038,
+      "step": 3351
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.332489990796265,
+      "learning_rate": 6.576397419088494e-06,
+      "loss": 0.5392,
+      "step": 3352
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.3029563614078903,
+      "learning_rate": 6.574490070800991e-06,
+      "loss": 0.508,
+      "step": 3353
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.434447606032874,
+      "learning_rate": 6.5725824681407505e-06,
+      "loss": 0.5029,
+      "step": 3354
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4351387854652031,
+      "learning_rate": 6.570674611415962e-06,
+      "loss": 0.5237,
+      "step": 3355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7332743393085002,
+      "learning_rate": 6.5687665009348564e-06,
+      "loss": 0.5281,
+      "step": 3356
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3254583143473562,
+      "learning_rate": 6.566858137005707e-06,
+      "loss": 0.5138,
+      "step": 3357
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6459229735641585,
+      "learning_rate": 6.564949519936825e-06,
+      "loss": 0.5463,
+      "step": 3358
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.798749927990085,
+      "learning_rate": 6.563040650036566e-06,
+      "loss": 0.5203,
+      "step": 3359
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8899694122457764,
+      "learning_rate": 6.5611315276133224e-06,
+      "loss": 0.488,
+      "step": 3360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2222412116863846,
+      "learning_rate": 6.559222152975533e-06,
+      "loss": 0.4919,
+      "step": 3361
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6101329963661517,
+      "learning_rate": 6.5573125264316715e-06,
+      "loss": 0.5285,
+      "step": 3362
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3733533959449282,
+      "learning_rate": 6.555402648290256e-06,
+      "loss": 0.4836,
+      "step": 3363
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3305592653903977,
+      "learning_rate": 6.553492518859843e-06,
+      "loss": 0.4739,
+      "step": 3364
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5819335222378026,
+      "learning_rate": 6.551582138449033e-06,
+      "loss": 0.5243,
+      "step": 3365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.68120273290888,
+      "learning_rate": 6.549671507366464e-06,
+      "loss": 0.5258,
+      "step": 3366
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6916725965006718,
+      "learning_rate": 6.547760625920814e-06,
+      "loss": 0.5715,
+      "step": 3367
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5603014124207755,
+      "learning_rate": 6.545849494420802e-06,
+      "loss": 0.5392,
+      "step": 3368
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.110977366460912,
+      "learning_rate": 6.543938113175191e-06,
+      "loss": 0.5224,
+      "step": 3369
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3684723813682567,
+      "learning_rate": 6.5420264824927796e-06,
+      "loss": 0.4657,
+      "step": 3370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4282419027256408,
+      "learning_rate": 6.540114602682409e-06,
+      "loss": 0.527,
+      "step": 3371
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4251475318763287,
+      "learning_rate": 6.5382024740529605e-06,
+      "loss": 0.4968,
+      "step": 3372
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3420448115949397,
+      "learning_rate": 6.536290096913354e-06,
+      "loss": 0.5144,
+      "step": 3373
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.295778067774979,
+      "learning_rate": 6.5343774715725525e-06,
+      "loss": 0.4714,
+      "step": 3374
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3489738242574505,
+      "learning_rate": 6.532464598339557e-06,
+      "loss": 0.5089,
+      "step": 3375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7563087403545115,
+      "learning_rate": 6.530551477523411e-06,
+      "loss": 0.535,
+      "step": 3376
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4205894446580483,
+      "learning_rate": 6.528638109433191e-06,
+      "loss": 0.5118,
+      "step": 3377
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5426548496710313,
+      "learning_rate": 6.526724494378023e-06,
+      "loss": 0.4733,
+      "step": 3378
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4987232712876852,
+      "learning_rate": 6.524810632667066e-06,
+      "loss": 0.4909,
+      "step": 3379
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.790263490504134,
+      "learning_rate": 6.522896524609521e-06,
+      "loss": 0.51,
+      "step": 3380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8194563671942883,
+      "learning_rate": 6.520982170514631e-06,
+      "loss": 0.4846,
+      "step": 3381
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2429209424748784,
+      "learning_rate": 6.519067570691675e-06,
+      "loss": 0.4647,
+      "step": 3382
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.342898948876468,
+      "learning_rate": 6.517152725449976e-06,
+      "loss": 0.5154,
+      "step": 3383
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3705724281814533,
+      "learning_rate": 6.515237635098891e-06,
+      "loss": 0.5706,
+      "step": 3384
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5642337602627236,
+      "learning_rate": 6.513322299947822e-06,
+      "loss": 0.5745,
+      "step": 3385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4367398118514414,
+      "learning_rate": 6.511406720306206e-06,
+      "loss": 0.527,
+      "step": 3386
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2768734239191017,
+      "learning_rate": 6.509490896483524e-06,
+      "loss": 0.4542,
+      "step": 3387
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4721952171363482,
+      "learning_rate": 6.507574828789292e-06,
+      "loss": 0.4273,
+      "step": 3388
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8717370896892942,
+      "learning_rate": 6.50565851753307e-06,
+      "loss": 0.5026,
+      "step": 3389
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4527504054398452,
+      "learning_rate": 6.503741963024454e-06,
+      "loss": 0.5421,
+      "step": 3390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.178249397843699,
+      "learning_rate": 6.5018251655730795e-06,
+      "loss": 0.5536,
+      "step": 3391
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2495856813403767,
+      "learning_rate": 6.499908125488623e-06,
+      "loss": 0.5353,
+      "step": 3392
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.096512316760758,
+      "learning_rate": 6.4979908430807995e-06,
+      "loss": 0.5541,
+      "step": 3393
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.408989150160295,
+      "learning_rate": 6.4960733186593604e-06,
+      "loss": 0.5071,
+      "step": 3394
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6106819885568957,
+      "learning_rate": 6.494155552534102e-06,
+      "loss": 0.4794,
+      "step": 3395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8647778316019838,
+      "learning_rate": 6.492237545014853e-06,
+      "loss": 0.5133,
+      "step": 3396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4265607978785886,
+      "learning_rate": 6.490319296411487e-06,
+      "loss": 0.5425,
+      "step": 3397
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.239661238066202,
+      "learning_rate": 6.488400807033913e-06,
+      "loss": 0.468,
+      "step": 3398
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5414791454347683,
+      "learning_rate": 6.486482077192081e-06,
+      "loss": 0.5226,
+      "step": 3399
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.610310909755451,
+      "learning_rate": 6.484563107195977e-06,
+      "loss": 0.5389,
+      "step": 3400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2558984430146658,
+      "learning_rate": 6.482643897355628e-06,
+      "loss": 0.5248,
+      "step": 3401
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6212989357041312,
+      "learning_rate": 6.4807244479810995e-06,
+      "loss": 0.5324,
+      "step": 3402
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1973737707139973,
+      "learning_rate": 6.478804759382495e-06,
+      "loss": 0.4898,
+      "step": 3403
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3683074620200943,
+      "learning_rate": 6.476884831869958e-06,
+      "loss": 0.4739,
+      "step": 3404
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4320958020667869,
+      "learning_rate": 6.4749646657536695e-06,
+      "loss": 0.4828,
+      "step": 3405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.118003030237448,
+      "learning_rate": 6.473044261343848e-06,
+      "loss": 0.5994,
+      "step": 3406
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5132374141721459,
+      "learning_rate": 6.4711236189507535e-06,
+      "loss": 0.5209,
+      "step": 3407
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9169913340489075,
+      "learning_rate": 6.469202738884681e-06,
+      "loss": 0.4791,
+      "step": 3408
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5541877617357418,
+      "learning_rate": 6.467281621455967e-06,
+      "loss": 0.5716,
+      "step": 3409
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3159596193980914,
+      "learning_rate": 6.465360266974984e-06,
+      "loss": 0.5009,
+      "step": 3410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5372368754643895,
+      "learning_rate": 6.463438675752145e-06,
+      "loss": 0.5137,
+      "step": 3411
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8482767490950647,
+      "learning_rate": 6.461516848097899e-06,
+      "loss": 0.5033,
+      "step": 3412
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3063262071097612,
+      "learning_rate": 6.459594784322734e-06,
+      "loss": 0.5729,
+      "step": 3413
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4749651475169578,
+      "learning_rate": 6.457672484737177e-06,
+      "loss": 0.5183,
+      "step": 3414
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.371908921907626,
+      "learning_rate": 6.455749949651791e-06,
+      "loss": 0.5636,
+      "step": 3415
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6074630823942877,
+      "learning_rate": 6.45382717937718e-06,
+      "loss": 0.5369,
+      "step": 3416
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4537404369016569,
+      "learning_rate": 6.4519041742239844e-06,
+      "loss": 0.4837,
+      "step": 3417
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.470038846548514,
+      "learning_rate": 6.449980934502881e-06,
+      "loss": 0.5673,
+      "step": 3418
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.411105238613095,
+      "learning_rate": 6.448057460524588e-06,
+      "loss": 0.5568,
+      "step": 3419
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.672003949426101,
+      "learning_rate": 6.44613375259986e-06,
+      "loss": 0.4877,
+      "step": 3420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7186118198150053,
+      "learning_rate": 6.444209811039488e-06,
+      "loss": 0.5366,
+      "step": 3421
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6707392924215835,
+      "learning_rate": 6.4422856361543e-06,
+      "loss": 0.549,
+      "step": 3422
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.479892625489135,
+      "learning_rate": 6.440361228255165e-06,
+      "loss": 0.5229,
+      "step": 3423
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6666535300936052,
+      "learning_rate": 6.438436587652989e-06,
+      "loss": 0.4785,
+      "step": 3424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6953266703542345,
+      "learning_rate": 6.436511714658713e-06,
+      "loss": 0.5738,
+      "step": 3425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6245088591621328,
+      "learning_rate": 6.434586609583316e-06,
+      "loss": 0.4554,
+      "step": 3426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5209450118584893,
+      "learning_rate": 6.43266127273782e-06,
+      "loss": 0.4935,
+      "step": 3427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3368355711738351,
+      "learning_rate": 6.430735704433278e-06,
+      "loss": 0.5396,
+      "step": 3428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.899006884541238,
+      "learning_rate": 6.428809904980782e-06,
+      "loss": 0.4872,
+      "step": 3429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5256633510601914,
+      "learning_rate": 6.426883874691461e-06,
+      "loss": 0.4879,
+      "step": 3430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7959945788719616,
+      "learning_rate": 6.424957613876483e-06,
+      "loss": 0.4867,
+      "step": 3431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.035467778275846,
+      "learning_rate": 6.4230311228470535e-06,
+      "loss": 0.5156,
+      "step": 3432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6409432369288743,
+      "learning_rate": 6.421104401914413e-06,
+      "loss": 0.4887,
+      "step": 3433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.316556873161966,
+      "learning_rate": 6.41917745138984e-06,
+      "loss": 0.4915,
+      "step": 3434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4845717808016468,
+      "learning_rate": 6.417250271584649e-06,
+      "loss": 0.482,
+      "step": 3435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3906828490733454,
+      "learning_rate": 6.415322862810198e-06,
+      "loss": 0.506,
+      "step": 3436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4789087457973646,
+      "learning_rate": 6.413395225377872e-06,
+      "loss": 0.5158,
+      "step": 3437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4001127768438166,
+      "learning_rate": 6.4114673595991e-06,
+      "loss": 0.5042,
+      "step": 3438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6352396847654843,
+      "learning_rate": 6.409539265785344e-06,
+      "loss": 0.4848,
+      "step": 3439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6458804502407448,
+      "learning_rate": 6.407610944248106e-06,
+      "loss": 0.4742,
+      "step": 3440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6893783720661152,
+      "learning_rate": 6.405682395298922e-06,
+      "loss": 0.5247,
+      "step": 3441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3762211902781725,
+      "learning_rate": 6.4037536192493665e-06,
+      "loss": 0.5144,
+      "step": 3442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8440755133843545,
+      "learning_rate": 6.401824616411052e-06,
+      "loss": 0.5605,
+      "step": 3443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.502330810149086,
+      "learning_rate": 6.399895387095624e-06,
+      "loss": 0.5145,
+      "step": 3444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.310208911902816,
+      "learning_rate": 6.397965931614767e-06,
+      "loss": 0.5021,
+      "step": 3445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4878071758351954,
+      "learning_rate": 6.396036250280202e-06,
+      "loss": 0.5111,
+      "step": 3446
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3499148482849967,
+      "learning_rate": 6.394106343403685e-06,
+      "loss": 0.4843,
+      "step": 3447
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6672272401015171,
+      "learning_rate": 6.392176211297011e-06,
+      "loss": 0.4892,
+      "step": 3448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1963725876544813,
+      "learning_rate": 6.3902458542720085e-06,
+      "loss": 0.5252,
+      "step": 3449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8342464367231288,
+      "learning_rate": 6.388315272640544e-06,
+      "loss": 0.4821,
+      "step": 3450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4035352327594603,
+      "learning_rate": 6.386384466714518e-06,
+      "loss": 0.4818,
+      "step": 3451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.342198015827931,
+      "learning_rate": 6.384453436805873e-06,
+      "loss": 0.4663,
+      "step": 3452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.704755890875126,
+      "learning_rate": 6.382522183226583e-06,
+      "loss": 0.5597,
+      "step": 3453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6219798429195122,
+      "learning_rate": 6.3805907062886564e-06,
+      "loss": 0.476,
+      "step": 3454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7164666803756226,
+      "learning_rate": 6.3786590063041434e-06,
+      "loss": 0.5154,
+      "step": 3455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3471849166172916,
+      "learning_rate": 6.376727083585126e-06,
+      "loss": 0.5078,
+      "step": 3456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6146640259005276,
+      "learning_rate": 6.374794938443722e-06,
+      "loss": 0.5718,
+      "step": 3457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6818007005864406,
+      "learning_rate": 6.372862571192088e-06,
+      "loss": 0.5091,
+      "step": 3458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.366665236720085,
+      "learning_rate": 6.370929982142413e-06,
+      "loss": 0.4912,
+      "step": 3459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0380886234440747,
+      "learning_rate": 6.368997171606927e-06,
+      "loss": 0.5472,
+      "step": 3460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6612318777400736,
+      "learning_rate": 6.367064139897891e-06,
+      "loss": 0.4455,
+      "step": 3461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.378939506184432,
+      "learning_rate": 6.365130887327603e-06,
+      "loss": 0.5667,
+      "step": 3462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3967605747616427,
+      "learning_rate": 6.363197414208396e-06,
+      "loss": 0.4876,
+      "step": 3463
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8320960845441028,
+      "learning_rate": 6.361263720852642e-06,
+      "loss": 0.4931,
+      "step": 3464
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6450256223949539,
+      "learning_rate": 6.359329807572746e-06,
+      "loss": 0.4911,
+      "step": 3465
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4463229538998164,
+      "learning_rate": 6.357395674681146e-06,
+      "loss": 0.455,
+      "step": 3466
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3119855418983852,
+      "learning_rate": 6.355461322490319e-06,
+      "loss": 0.5339,
+      "step": 3467
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0993905938665463,
+      "learning_rate": 6.35352675131278e-06,
+      "loss": 0.476,
+      "step": 3468
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0789495941589324,
+      "learning_rate": 6.3515919614610725e-06,
+      "loss": 0.5042,
+      "step": 3469
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.505001124357574,
+      "learning_rate": 6.3496569532477796e-06,
+      "loss": 0.5039,
+      "step": 3470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3543726155896296,
+      "learning_rate": 6.347721726985518e-06,
+      "loss": 0.4937,
+      "step": 3471
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.410733061613502,
+      "learning_rate": 6.345786282986944e-06,
+      "loss": 0.5036,
+      "step": 3472
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.288134741913164,
+      "learning_rate": 6.343850621564742e-06,
+      "loss": 0.5071,
+      "step": 3473
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3889345099741446,
+      "learning_rate": 6.3419147430316375e-06,
+      "loss": 0.5137,
+      "step": 3474
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.408902839713213,
+      "learning_rate": 6.3399786477003866e-06,
+      "loss": 0.5138,
+      "step": 3475
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4541724062815307,
+      "learning_rate": 6.338042335883784e-06,
+      "loss": 0.4752,
+      "step": 3476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0315020298988067,
+      "learning_rate": 6.336105807894658e-06,
+      "loss": 0.5071,
+      "step": 3477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.479139923284239,
+      "learning_rate": 6.334169064045871e-06,
+      "loss": 0.5207,
+      "step": 3478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6708097406058517,
+      "learning_rate": 6.332232104650321e-06,
+      "loss": 0.4893,
+      "step": 3479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7443151935211285,
+      "learning_rate": 6.330294930020941e-06,
+      "loss": 0.5204,
+      "step": 3480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4373772526834405,
+      "learning_rate": 6.3283575404706996e-06,
+      "loss": 0.5344,
+      "step": 3481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.2391986402563235,
+      "learning_rate": 6.326419936312599e-06,
+      "loss": 0.5516,
+      "step": 3482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4624592865974275,
+      "learning_rate": 6.324482117859676e-06,
+      "loss": 0.5251,
+      "step": 3483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.525858161534442,
+      "learning_rate": 6.322544085425001e-06,
+      "loss": 0.5243,
+      "step": 3484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6321469993075612,
+      "learning_rate": 6.320605839321681e-06,
+      "loss": 0.5121,
+      "step": 3485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6245468809709254,
+      "learning_rate": 6.318667379862856e-06,
+      "loss": 0.5002,
+      "step": 3486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3950302507184524,
+      "learning_rate": 6.3167287073617035e-06,
+      "loss": 0.4825,
+      "step": 3487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4184155641850023,
+      "learning_rate": 6.31478982213143e-06,
+      "loss": 0.5087,
+      "step": 3488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5052726070482894,
+      "learning_rate": 6.312850724485282e-06,
+      "loss": 0.4351,
+      "step": 3489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2968373517440517,
+      "learning_rate": 6.310911414736537e-06,
+      "loss": 0.5013,
+      "step": 3490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6496514532699211,
+      "learning_rate": 6.308971893198508e-06,
+      "loss": 0.5353,
+      "step": 3491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5839632150330647,
+      "learning_rate": 6.307032160184541e-06,
+      "loss": 0.5673,
+      "step": 3492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3248369734797494,
+      "learning_rate": 6.305092216008016e-06,
+      "loss": 0.5356,
+      "step": 3493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4246486211657718,
+      "learning_rate": 6.30315206098235e-06,
+      "loss": 0.5004,
+      "step": 3494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.829505588718067,
+      "learning_rate": 6.301211695420992e-06,
+      "loss": 0.4931,
+      "step": 3495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5619325285771204,
+      "learning_rate": 6.2992711196374236e-06,
+      "loss": 0.5172,
+      "step": 3496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3228374138272259,
+      "learning_rate": 6.297330333945164e-06,
+      "loss": 0.4938,
+      "step": 3497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3977534860361773,
+      "learning_rate": 6.2953893386577626e-06,
+      "loss": 0.4958,
+      "step": 3498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2944479809162501,
+      "learning_rate": 6.293448134088805e-06,
+      "loss": 0.5331,
+      "step": 3499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7182384782561086,
+      "learning_rate": 6.2915067205519085e-06,
+      "loss": 0.4769,
+      "step": 3500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2931322581828373,
+      "learning_rate": 6.289565098360728e-06,
+      "loss": 0.4586,
+      "step": 3501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.522939179168897,
+      "learning_rate": 6.287623267828948e-06,
+      "loss": 0.4692,
+      "step": 3502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.328118856112479,
+      "learning_rate": 6.2856812292702884e-06,
+      "loss": 0.4894,
+      "step": 3503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9947553241739746,
+      "learning_rate": 6.283738982998502e-06,
+      "loss": 0.4862,
+      "step": 3504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6311486849974827,
+      "learning_rate": 6.281796529327378e-06,
+      "loss": 0.5108,
+      "step": 3505
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6625257516752647,
+      "learning_rate": 6.279853868570736e-06,
+      "loss": 0.5143,
+      "step": 3506
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9066907850009018,
+      "learning_rate": 6.27791100104243e-06,
+      "loss": 0.5267,
+      "step": 3507
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4781973315845938,
+      "learning_rate": 6.2759679270563446e-06,
+      "loss": 0.5299,
+      "step": 3508
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.308884538474327,
+      "learning_rate": 6.274024646926405e-06,
+      "loss": 0.4653,
+      "step": 3509
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.278907688874593,
+      "learning_rate": 6.272081160966564e-06,
+      "loss": 0.5022,
+      "step": 3510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6621416069496444,
+      "learning_rate": 6.2701374694908045e-06,
+      "loss": 0.4967,
+      "step": 3511
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3440202170943447,
+      "learning_rate": 6.268193572813151e-06,
+      "loss": 0.4782,
+      "step": 3512
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7500464171050284,
+      "learning_rate": 6.266249471247659e-06,
+      "loss": 0.4943,
+      "step": 3513
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.411055795264252,
+      "learning_rate": 6.264305165108412e-06,
+      "loss": 0.4958,
+      "step": 3514
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2671764977020974,
+      "learning_rate": 6.262360654709529e-06,
+      "loss": 0.4663,
+      "step": 3515
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1726670323695136,
+      "learning_rate": 6.260415940365165e-06,
+      "loss": 0.4504,
+      "step": 3516
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5489519189048802,
+      "learning_rate": 6.258471022389506e-06,
+      "loss": 0.5345,
+      "step": 3517
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9116402906622896,
+      "learning_rate": 6.256525901096769e-06,
+      "loss": 0.4731,
+      "step": 3518
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4590621142943716,
+      "learning_rate": 6.254580576801208e-06,
+      "loss": 0.5038,
+      "step": 3519
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4887237061309944,
+      "learning_rate": 6.252635049817104e-06,
+      "loss": 0.5682,
+      "step": 3520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.554091148712146,
+      "learning_rate": 6.250689320458775e-06,
+      "loss": 0.4634,
+      "step": 3521
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5543738127421725,
+      "learning_rate": 6.248743389040573e-06,
+      "loss": 0.4984,
+      "step": 3522
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7629146754869744,
+      "learning_rate": 6.246797255876876e-06,
+      "loss": 0.5209,
+      "step": 3523
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9998417318041923,
+      "learning_rate": 6.244850921282102e-06,
+      "loss": 0.5541,
+      "step": 3524
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2551611920725345,
+      "learning_rate": 6.242904385570699e-06,
+      "loss": 0.5279,
+      "step": 3525
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5336429069292903,
+      "learning_rate": 6.240957649057145e-06,
+      "loss": 0.5148,
+      "step": 3526
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.044244210210011,
+      "learning_rate": 6.239010712055955e-06,
+      "loss": 0.5184,
+      "step": 3527
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4181219458895138,
+      "learning_rate": 6.2370635748816725e-06,
+      "loss": 0.4707,
+      "step": 3528
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6032922870244413,
+      "learning_rate": 6.235116237848872e-06,
+      "loss": 0.4305,
+      "step": 3529
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7304352877367216,
+      "learning_rate": 6.233168701272167e-06,
+      "loss": 0.5232,
+      "step": 3530
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3911196910792476,
+      "learning_rate": 6.231220965466197e-06,
+      "loss": 0.5021,
+      "step": 3531
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3165358719004459,
+      "learning_rate": 6.229273030745638e-06,
+      "loss": 0.5183,
+      "step": 3532
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7453145753076977,
+      "learning_rate": 6.227324897425191e-06,
+      "loss": 0.5097,
+      "step": 3533
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.5380079581301915,
+      "learning_rate": 6.2253765658195986e-06,
+      "loss": 0.5064,
+      "step": 3534
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0266057416813785,
+      "learning_rate": 6.223428036243631e-06,
+      "loss": 0.5195,
+      "step": 3535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3156969788819382,
+      "learning_rate": 6.2214793090120896e-06,
+      "loss": 0.4469,
+      "step": 3536
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5487418691974917,
+      "learning_rate": 6.219530384439807e-06,
+      "loss": 0.502,
+      "step": 3537
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4488340230656243,
+      "learning_rate": 6.217581262841651e-06,
+      "loss": 0.5008,
+      "step": 3538
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0032013299670752,
+      "learning_rate": 6.215631944532518e-06,
+      "loss": 0.5196,
+      "step": 3539
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6068169770942995,
+      "learning_rate": 6.213682429827338e-06,
+      "loss": 0.5189,
+      "step": 3540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1926932157629393,
+      "learning_rate": 6.21173271904107e-06,
+      "loss": 0.484,
+      "step": 3541
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6671431070107134,
+      "learning_rate": 6.209782812488713e-06,
+      "loss": 0.511,
+      "step": 3542
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6212237617207985,
+      "learning_rate": 6.207832710485285e-06,
+      "loss": 0.528,
+      "step": 3543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6305336046241468,
+      "learning_rate": 6.2058824133458476e-06,
+      "loss": 0.478,
+      "step": 3544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.822234471136306,
+      "learning_rate": 6.203931921385484e-06,
+      "loss": 0.5406,
+      "step": 3545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6679266912093794,
+      "learning_rate": 6.201981234919317e-06,
+      "loss": 0.4911,
+      "step": 3546
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.439282471232996,
+      "learning_rate": 6.200030354262493e-06,
+      "loss": 0.5135,
+      "step": 3547
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3884166014461918,
+      "learning_rate": 6.198079279730198e-06,
+      "loss": 0.4351,
+      "step": 3548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4047748956301707,
+      "learning_rate": 6.196128011637642e-06,
+      "loss": 0.5472,
+      "step": 3549
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4808855445369755,
+      "learning_rate": 6.194176550300071e-06,
+      "loss": 0.4965,
+      "step": 3550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0614228950515447,
+      "learning_rate": 6.1922248960327635e-06,
+      "loss": 0.5206,
+      "step": 3551
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7331701568695543,
+      "learning_rate": 6.190273049151022e-06,
+      "loss": 0.5328,
+      "step": 3552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8941941076976965,
+      "learning_rate": 6.1883210099701864e-06,
+      "loss": 0.5372,
+      "step": 3553
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.83335774832791,
+      "learning_rate": 6.186368778805628e-06,
+      "loss": 0.4962,
+      "step": 3554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5119225106231424,
+      "learning_rate": 6.184416355972743e-06,
+      "loss": 0.5088,
+      "step": 3555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2784469665914808,
+      "learning_rate": 6.182463741786965e-06,
+      "loss": 0.5033,
+      "step": 3556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.416495701529496,
+      "learning_rate": 6.180510936563754e-06,
+      "loss": 0.5359,
+      "step": 3557
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4934566071982733,
+      "learning_rate": 6.178557940618605e-06,
+      "loss": 0.486,
+      "step": 3558
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.258013858046018,
+      "learning_rate": 6.17660475426704e-06,
+      "loss": 0.5424,
+      "step": 3559
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4267716289763825,
+      "learning_rate": 6.174651377824615e-06,
+      "loss": 0.4984,
+      "step": 3560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5973749407245927,
+      "learning_rate": 6.172697811606914e-06,
+      "loss": 0.4614,
+      "step": 3561
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4895046489791153,
+      "learning_rate": 6.1707440559295554e-06,
+      "loss": 0.5066,
+      "step": 3562
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4827775512387729,
+      "learning_rate": 6.1687901111081826e-06,
+      "loss": 0.5524,
+      "step": 3563
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.715697107372859,
+      "learning_rate": 6.166835977458473e-06,
+      "loss": 0.5476,
+      "step": 3564
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4663246145687245,
+      "learning_rate": 6.1648816552961355e-06,
+      "loss": 0.5031,
+      "step": 3565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5067132761428363,
+      "learning_rate": 6.162927144936906e-06,
+      "loss": 0.465,
+      "step": 3566
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7021011233090093,
+      "learning_rate": 6.1609724466965535e-06,
+      "loss": 0.5044,
+      "step": 3567
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3699920347424173,
+      "learning_rate": 6.159017560890879e-06,
+      "loss": 0.4359,
+      "step": 3568
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.40510218511984,
+      "learning_rate": 6.15706248783571e-06,
+      "loss": 0.4913,
+      "step": 3569
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6211969399829059,
+      "learning_rate": 6.155107227846904e-06,
+      "loss": 0.4779,
+      "step": 3570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2953802588952739,
+      "learning_rate": 6.153151781240352e-06,
+      "loss": 0.4684,
+      "step": 3571
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3483986543487585,
+      "learning_rate": 6.151196148331975e-06,
+      "loss": 0.4908,
+      "step": 3572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5473054117537224,
+      "learning_rate": 6.1492403294377225e-06,
+      "loss": 0.4678,
+      "step": 3573
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2787212957409102,
+      "learning_rate": 6.14728432487357e-06,
+      "loss": 0.517,
+      "step": 3574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3290235549835527,
+      "learning_rate": 6.145328134955533e-06,
+      "loss": 0.504,
+      "step": 3575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4494007619059526,
+      "learning_rate": 6.143371759999648e-06,
+      "loss": 0.5333,
+      "step": 3576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5308204499267646,
+      "learning_rate": 6.1414152003219854e-06,
+      "loss": 0.5299,
+      "step": 3577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.481114838188324,
+      "learning_rate": 6.1394584562386425e-06,
+      "loss": 0.5139,
+      "step": 3578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.219964866020278,
+      "learning_rate": 6.137501528065752e-06,
+      "loss": 0.553,
+      "step": 3579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3323902271568275,
+      "learning_rate": 6.1355444161194724e-06,
+      "loss": 0.4944,
+      "step": 3580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.275208807005506,
+      "learning_rate": 6.13358712071599e-06,
+      "loss": 0.5012,
+      "step": 3581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7100296280459837,
+      "learning_rate": 6.131629642171526e-06,
+      "loss": 0.4876,
+      "step": 3582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3075143321738223,
+      "learning_rate": 6.129671980802325e-06,
+      "loss": 0.4824,
+      "step": 3583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8781822785139812,
+      "learning_rate": 6.127714136924667e-06,
+      "loss": 0.4955,
+      "step": 3584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8573339867952048,
+      "learning_rate": 6.125756110854859e-06,
+      "loss": 0.5335,
+      "step": 3585
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3234182294746977,
+      "learning_rate": 6.123797902909236e-06,
+      "loss": 0.5012,
+      "step": 3586
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5037302731280628,
+      "learning_rate": 6.121839513404163e-06,
+      "loss": 0.5497,
+      "step": 3587
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2704736827188403,
+      "learning_rate": 6.119880942656038e-06,
+      "loss": 0.5592,
+      "step": 3588
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.751140929108546,
+      "learning_rate": 6.117922190981282e-06,
+      "loss": 0.513,
+      "step": 3589
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.552318521249153,
+      "learning_rate": 6.1159632586963524e-06,
+      "loss": 0.5248,
+      "step": 3590
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4367054279952804,
+      "learning_rate": 6.114004146117729e-06,
+      "loss": 0.5626,
+      "step": 3591
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2107766082355556,
+      "learning_rate": 6.112044853561925e-06,
+      "loss": 0.4458,
+      "step": 3592
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4829866654121693,
+      "learning_rate": 6.11008538134548e-06,
+      "loss": 0.5067,
+      "step": 3593
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5846189280659801,
+      "learning_rate": 6.108125729784964e-06,
+      "loss": 0.5742,
+      "step": 3594
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5089263131720982,
+      "learning_rate": 6.106165899196978e-06,
+      "loss": 0.5726,
+      "step": 3595
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4150877058823383,
+      "learning_rate": 6.1042058898981484e-06,
+      "loss": 0.5297,
+      "step": 3596
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4547063878965534,
+      "learning_rate": 6.1022457022051325e-06,
+      "loss": 0.5352,
+      "step": 3597
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3973769343405766,
+      "learning_rate": 6.100285336434616e-06,
+      "loss": 0.5057,
+      "step": 3598
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.403887011962654,
+      "learning_rate": 6.098324792903313e-06,
+      "loss": 0.4837,
+      "step": 3599
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2463330225672213,
+      "learning_rate": 6.096364071927966e-06,
+      "loss": 0.4729,
+      "step": 3600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.050882947847235,
+      "learning_rate": 6.094403173825348e-06,
+      "loss": 0.4663,
+      "step": 3601
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.611584775857751,
+      "learning_rate": 6.0924420989122565e-06,
+      "loss": 0.5078,
+      "step": 3602
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5437249180480355,
+      "learning_rate": 6.0904808475055246e-06,
+      "loss": 0.5367,
+      "step": 3603
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.138459892149861,
+      "learning_rate": 6.088519419922008e-06,
+      "loss": 0.4932,
+      "step": 3604
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6913666855531088,
+      "learning_rate": 6.086557816478591e-06,
+      "loss": 0.5108,
+      "step": 3605
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5783740456870101,
+      "learning_rate": 6.08459603749219e-06,
+      "loss": 0.5018,
+      "step": 3606
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2359527048762666,
+      "learning_rate": 6.082634083279746e-06,
+      "loss": 0.4702,
+      "step": 3607
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4654275615242967,
+      "learning_rate": 6.080671954158232e-06,
+      "loss": 0.5437,
+      "step": 3608
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.272816051175789,
+      "learning_rate": 6.0787096504446465e-06,
+      "loss": 0.4884,
+      "step": 3609
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5154753911364525,
+      "learning_rate": 6.076747172456016e-06,
+      "loss": 0.5131,
+      "step": 3610
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.491605697034516,
+      "learning_rate": 6.074784520509395e-06,
+      "loss": 0.5344,
+      "step": 3611
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6704317411138851,
+      "learning_rate": 6.07282169492187e-06,
+      "loss": 0.517,
+      "step": 3612
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2193079275691656,
+      "learning_rate": 6.070858696010552e-06,
+      "loss": 0.5255,
+      "step": 3613
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.442941458805938,
+      "learning_rate": 6.068895524092581e-06,
+      "loss": 0.5055,
+      "step": 3614
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4500045510329642,
+      "learning_rate": 6.066932179485122e-06,
+      "loss": 0.544,
+      "step": 3615
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.243790000007731,
+      "learning_rate": 6.064968662505374e-06,
+      "loss": 0.501,
+      "step": 3616
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3412828250391715,
+      "learning_rate": 6.06300497347056e-06,
+      "loss": 0.5653,
+      "step": 3617
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2171866043917068,
+      "learning_rate": 6.06104111269793e-06,
+      "loss": 0.3964,
+      "step": 3618
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3164204240625614,
+      "learning_rate": 6.059077080504761e-06,
+      "loss": 0.4738,
+      "step": 3619
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.424702659928036,
+      "learning_rate": 6.057112877208364e-06,
+      "loss": 0.5737,
+      "step": 3620
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4070668142198544,
+      "learning_rate": 6.0551485031260725e-06,
+      "loss": 0.4949,
+      "step": 3621
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3408711147549046,
+      "learning_rate": 6.053183958575246e-06,
+      "loss": 0.4988,
+      "step": 3622
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8979002790542197,
+      "learning_rate": 6.051219243873275e-06,
+      "loss": 0.4505,
+      "step": 3623
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3990509882706048,
+      "learning_rate": 6.049254359337578e-06,
+      "loss": 0.5486,
+      "step": 3624
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3622160719714407,
+      "learning_rate": 6.0472893052855985e-06,
+      "loss": 0.4876,
+      "step": 3625
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3103141054320948,
+      "learning_rate": 6.0453240820348066e-06,
+      "loss": 0.4933,
+      "step": 3626
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5448586835372746,
+      "learning_rate": 6.043358689902704e-06,
+      "loss": 0.5565,
+      "step": 3627
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.740198154341534,
+      "learning_rate": 6.041393129206816e-06,
+      "loss": 0.5272,
+      "step": 3628
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4284041583081468,
+      "learning_rate": 6.0394274002646965e-06,
+      "loss": 0.4735,
+      "step": 3629
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.502198490514338,
+      "learning_rate": 6.037461503393925e-06,
+      "loss": 0.5462,
+      "step": 3630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2648890363953325,
+      "learning_rate": 6.03549543891211e-06,
+      "loss": 0.5275,
+      "step": 3631
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4609835118631087,
+      "learning_rate": 6.033529207136888e-06,
+      "loss": 0.5122,
+      "step": 3632
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5404894439925987,
+      "learning_rate": 6.03156280838592e-06,
+      "loss": 0.5405,
+      "step": 3633
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.681033615151205,
+      "learning_rate": 6.029596242976895e-06,
+      "loss": 0.5145,
+      "step": 3634
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2585368863622328,
+      "learning_rate": 6.02762951122753e-06,
+      "loss": 0.5044,
+      "step": 3635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6716525797681374,
+      "learning_rate": 6.025662613455566e-06,
+      "loss": 0.5353,
+      "step": 3636
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.690724018178695,
+      "learning_rate": 6.023695549978774e-06,
+      "loss": 0.5276,
+      "step": 3637
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6397336826076158,
+      "learning_rate": 6.021728321114949e-06,
+      "loss": 0.4837,
+      "step": 3638
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.333082952090204,
+      "learning_rate": 6.0197609271819145e-06,
+      "loss": 0.5052,
+      "step": 3639
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.638944018469146,
+      "learning_rate": 6.017793368497523e-06,
+      "loss": 0.548,
+      "step": 3640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3767723047356812,
+      "learning_rate": 6.0158256453796485e-06,
+      "loss": 0.492,
+      "step": 3641
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9370306650773428,
+      "learning_rate": 6.013857758146193e-06,
+      "loss": 0.538,
+      "step": 3642
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3766171654282706,
+      "learning_rate": 6.01188970711509e-06,
+      "loss": 0.47,
+      "step": 3643
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.555917116487695,
+      "learning_rate": 6.0099214926042905e-06,
+      "loss": 0.5003,
+      "step": 3644
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9513733336041112,
+      "learning_rate": 6.00795311493178e-06,
+      "loss": 0.5294,
+      "step": 3645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4435089295034635,
+      "learning_rate": 6.005984574415568e-06,
+      "loss": 0.5174,
+      "step": 3646
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3801760935816803,
+      "learning_rate": 6.0040158713736865e-06,
+      "loss": 0.5404,
+      "step": 3647
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5837482763871387,
+      "learning_rate": 6.002047006124198e-06,
+      "loss": 0.5395,
+      "step": 3648
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7192576032754558,
+      "learning_rate": 6.000077978985191e-06,
+      "loss": 0.4661,
+      "step": 3649
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3025945963336203,
+      "learning_rate": 5.9981087902747785e-06,
+      "loss": 0.5132,
+      "step": 3650
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.477148566837164,
+      "learning_rate": 5.996139440311099e-06,
+      "loss": 0.5258,
+      "step": 3651
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2595738614799532,
+      "learning_rate": 5.994169929412323e-06,
+      "loss": 0.4461,
+      "step": 3652
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2587053512253195,
+      "learning_rate": 5.9922002578966364e-06,
+      "loss": 0.5135,
+      "step": 3653
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4674379833778877,
+      "learning_rate": 5.990230426082261e-06,
+      "loss": 0.552,
+      "step": 3654
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5625476515475012,
+      "learning_rate": 5.988260434287438e-06,
+      "loss": 0.5393,
+      "step": 3655
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2436803191403425,
+      "learning_rate": 5.986290282830438e-06,
+      "loss": 0.5203,
+      "step": 3656
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4690916234746247,
+      "learning_rate": 5.984319972029558e-06,
+      "loss": 0.5051,
+      "step": 3657
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4158183487114746,
+      "learning_rate": 5.982349502203116e-06,
+      "loss": 0.537,
+      "step": 3658
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6914721822117508,
+      "learning_rate": 5.98037887366946e-06,
+      "loss": 0.5228,
+      "step": 3659
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3885647159507815,
+      "learning_rate": 5.978408086746962e-06,
+      "loss": 0.4751,
+      "step": 3660
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0146078697076937,
+      "learning_rate": 5.976437141754021e-06,
+      "loss": 0.4981,
+      "step": 3661
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.613300758549818,
+      "learning_rate": 5.974466039009063e-06,
+      "loss": 0.4567,
+      "step": 3662
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2454938124351362,
+      "learning_rate": 5.972494778830531e-06,
+      "loss": 0.4523,
+      "step": 3663
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.391952792088452,
+      "learning_rate": 5.970523361536903e-06,
+      "loss": 0.4386,
+      "step": 3664
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.470221657599158,
+      "learning_rate": 5.968551787446678e-06,
+      "loss": 0.5198,
+      "step": 3665
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5239604080432068,
+      "learning_rate": 5.966580056878382e-06,
+      "loss": 0.5177,
+      "step": 3666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3258176562134478,
+      "learning_rate": 5.9646081701505654e-06,
+      "loss": 0.4909,
+      "step": 3667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7142755784727002,
+      "learning_rate": 5.962636127581802e-06,
+      "loss": 0.4909,
+      "step": 3668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.484667936557781,
+      "learning_rate": 5.960663929490696e-06,
+      "loss": 0.5158,
+      "step": 3669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0552574632383447,
+      "learning_rate": 5.95869157619587e-06,
+      "loss": 0.4578,
+      "step": 3670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.592695773175072,
+      "learning_rate": 5.956719068015977e-06,
+      "loss": 0.5533,
+      "step": 3671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2148265940903142,
+      "learning_rate": 5.954746405269692e-06,
+      "loss": 0.493,
+      "step": 3672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.479906503336225,
+      "learning_rate": 5.952773588275718e-06,
+      "loss": 0.5627,
+      "step": 3673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4811317212396697,
+      "learning_rate": 5.950800617352778e-06,
+      "loss": 0.5211,
+      "step": 3674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4535882719954414,
+      "learning_rate": 5.9488274928196245e-06,
+      "loss": 0.4846,
+      "step": 3675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8321506613837863,
+      "learning_rate": 5.946854214995032e-06,
+      "loss": 0.4663,
+      "step": 3676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6928171693718244,
+      "learning_rate": 5.9448807841978044e-06,
+      "loss": 0.5501,
+      "step": 3677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3754667256719582,
+      "learning_rate": 5.942907200746762e-06,
+      "loss": 0.4814,
+      "step": 3678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6116269924251991,
+      "learning_rate": 5.940933464960759e-06,
+      "loss": 0.5573,
+      "step": 3679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3579375286635698,
+      "learning_rate": 5.9389595771586675e-06,
+      "loss": 0.4808,
+      "step": 3680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3432840262996693,
+      "learning_rate": 5.936985537659385e-06,
+      "loss": 0.49,
+      "step": 3681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9449076892590997,
+      "learning_rate": 5.93501134678184e-06,
+      "loss": 0.4858,
+      "step": 3682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.712462555171365,
+      "learning_rate": 5.9330370048449736e-06,
+      "loss": 0.5005,
+      "step": 3683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4118247291797974,
+      "learning_rate": 5.9310625121677636e-06,
+      "loss": 0.5104,
+      "step": 3684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.324781398126362,
+      "learning_rate": 5.929087869069202e-06,
+      "loss": 0.4799,
+      "step": 3685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2617013365478964,
+      "learning_rate": 5.927113075868315e-06,
+      "loss": 0.5406,
+      "step": 3686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5523795998609873,
+      "learning_rate": 5.925138132884145e-06,
+      "loss": 0.5128,
+      "step": 3687
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4891053513027392,
+      "learning_rate": 5.923163040435762e-06,
+      "loss": 0.5259,
+      "step": 3688
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.459970627139374,
+      "learning_rate": 5.921187798842258e-06,
+      "loss": 0.5568,
+      "step": 3689
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4923420303638777,
+      "learning_rate": 5.9192124084227534e-06,
+      "loss": 0.538,
+      "step": 3690
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4986361120402343,
+      "learning_rate": 5.917236869496388e-06,
+      "loss": 0.5481,
+      "step": 3691
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3799188836426888,
+      "learning_rate": 5.915261182382328e-06,
+      "loss": 0.5267,
+      "step": 3692
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4245940989096637,
+      "learning_rate": 5.913285347399762e-06,
+      "loss": 0.5152,
+      "step": 3693
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.752625008923162,
+      "learning_rate": 5.9113093648679065e-06,
+      "loss": 0.5141,
+      "step": 3694
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6379456852940069,
+      "learning_rate": 5.909333235105996e-06,
+      "loss": 0.5097,
+      "step": 3695
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4584429688128662,
+      "learning_rate": 5.907356958433292e-06,
+      "loss": 0.4911,
+      "step": 3696
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.441952787689109,
+      "learning_rate": 5.905380535169082e-06,
+      "loss": 0.5118,
+      "step": 3697
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5176438222118591,
+      "learning_rate": 5.903403965632672e-06,
+      "loss": 0.4929,
+      "step": 3698
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2751446770162105,
+      "learning_rate": 5.9014272501433945e-06,
+      "loss": 0.4591,
+      "step": 3699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.404506261471304,
+      "learning_rate": 5.899450389020605e-06,
+      "loss": 0.516,
+      "step": 3700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3889948229171092,
+      "learning_rate": 5.897473382583684e-06,
+      "loss": 0.4677,
+      "step": 3701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2631815193287492,
+      "learning_rate": 5.895496231152033e-06,
+      "loss": 0.5102,
+      "step": 3702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5626518136918286,
+      "learning_rate": 5.893518935045081e-06,
+      "loss": 0.5141,
+      "step": 3703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4104502177968783,
+      "learning_rate": 5.891541494582274e-06,
+      "loss": 0.5233,
+      "step": 3704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6254946366913998,
+      "learning_rate": 5.889563910083087e-06,
+      "loss": 0.5165,
+      "step": 3705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3610756542467268,
+      "learning_rate": 5.887586181867015e-06,
+      "loss": 0.5031,
+      "step": 3706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4098997632065402,
+      "learning_rate": 5.885608310253581e-06,
+      "loss": 0.5532,
+      "step": 3707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5357593910677314,
+      "learning_rate": 5.8836302955623225e-06,
+      "loss": 0.5059,
+      "step": 3708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.822301409849621,
+      "learning_rate": 5.881652138112808e-06,
+      "loss": 0.5389,
+      "step": 3709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4528299534532378,
+      "learning_rate": 5.879673838224625e-06,
+      "loss": 0.5259,
+      "step": 3710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6502222787439851,
+      "learning_rate": 5.877695396217386e-06,
+      "loss": 0.4961,
+      "step": 3711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5887911511047892,
+      "learning_rate": 5.875716812410727e-06,
+      "loss": 0.5187,
+      "step": 3712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6127291879447092,
+      "learning_rate": 5.873738087124302e-06,
+      "loss": 0.5522,
+      "step": 3713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3110231020898824,
+      "learning_rate": 5.871759220677795e-06,
+      "loss": 0.4711,
+      "step": 3714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2923517567560525,
+      "learning_rate": 5.8697802133909085e-06,
+      "loss": 0.494,
+      "step": 3715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0171625532417607,
+      "learning_rate": 5.867801065583369e-06,
+      "loss": 0.5045,
+      "step": 3716
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.459700446720111,
+      "learning_rate": 5.865821777574922e-06,
+      "loss": 0.4705,
+      "step": 3717
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.238435685941135,
+      "learning_rate": 5.863842349685344e-06,
+      "loss": 0.5557,
+      "step": 3718
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3170644542756293,
+      "learning_rate": 5.861862782234425e-06,
+      "loss": 0.5146,
+      "step": 3719
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.77112402196552,
+      "learning_rate": 5.8598830755419835e-06,
+      "loss": 0.5185,
+      "step": 3720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4227671148027017,
+      "learning_rate": 5.8579032299278585e-06,
+      "loss": 0.547,
+      "step": 3721
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.8600426263452965,
+      "learning_rate": 5.855923245711909e-06,
+      "loss": 0.4657,
+      "step": 3722
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8078423840950666,
+      "learning_rate": 5.853943123214022e-06,
+      "loss": 0.4963,
+      "step": 3723
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3140178472040205,
+      "learning_rate": 5.851962862754103e-06,
+      "loss": 0.4336,
+      "step": 3724
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3518973598901456,
+      "learning_rate": 5.84998246465208e-06,
+      "loss": 0.4447,
+      "step": 3725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6492949452638737,
+      "learning_rate": 5.848001929227902e-06,
+      "loss": 0.5199,
+      "step": 3726
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7079034101337343,
+      "learning_rate": 5.846021256801546e-06,
+      "loss": 0.4598,
+      "step": 3727
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3625776176116,
+      "learning_rate": 5.844040447693004e-06,
+      "loss": 0.4952,
+      "step": 3728
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5736507105134774,
+      "learning_rate": 5.842059502222295e-06,
+      "loss": 0.5335,
+      "step": 3729
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3900575889574893,
+      "learning_rate": 5.840078420709456e-06,
+      "loss": 0.4571,
+      "step": 3730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3081077443249847,
+      "learning_rate": 5.838097203474549e-06,
+      "loss": 0.5325,
+      "step": 3731
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4149913690397518,
+      "learning_rate": 5.8361158508376584e-06,
+      "loss": 0.5263,
+      "step": 3732
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4227367199239105,
+      "learning_rate": 5.834134363118889e-06,
+      "loss": 0.554,
+      "step": 3733
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3659653418941065,
+      "learning_rate": 5.832152740638366e-06,
+      "loss": 0.5594,
+      "step": 3734
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4498815236131395,
+      "learning_rate": 5.8301709837162375e-06,
+      "loss": 0.4839,
+      "step": 3735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6379826138444108,
+      "learning_rate": 5.828189092672677e-06,
+      "loss": 0.4814,
+      "step": 3736
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1874404120184479,
+      "learning_rate": 5.826207067827874e-06,
+      "loss": 0.4355,
+      "step": 3737
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1896968083972823,
+      "learning_rate": 5.824224909502042e-06,
+      "loss": 0.5023,
+      "step": 3738
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.28469351928837,
+      "learning_rate": 5.822242618015417e-06,
+      "loss": 0.4747,
+      "step": 3739
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3405004682609176,
+      "learning_rate": 5.8202601936882565e-06,
+      "loss": 0.4912,
+      "step": 3740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.521509508415757,
+      "learning_rate": 5.8182776368408365e-06,
+      "loss": 0.5042,
+      "step": 3741
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4304713086594372,
+      "learning_rate": 5.816294947793457e-06,
+      "loss": 0.4857,
+      "step": 3742
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.407516635409624,
+      "learning_rate": 5.814312126866441e-06,
+      "loss": 0.4815,
+      "step": 3743
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.374491776202101,
+      "learning_rate": 5.812329174380128e-06,
+      "loss": 0.4992,
+      "step": 3744
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2658042191199907,
+      "learning_rate": 5.810346090654883e-06,
+      "loss": 0.5197,
+      "step": 3745
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.589948736983883,
+      "learning_rate": 5.808362876011089e-06,
+      "loss": 0.5338,
+      "step": 3746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5472017582413746,
+      "learning_rate": 5.806379530769154e-06,
+      "loss": 0.5497,
+      "step": 3747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4941205729620681,
+      "learning_rate": 5.804396055249504e-06,
+      "loss": 0.4912,
+      "step": 3748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2873758547493115,
+      "learning_rate": 5.802412449772585e-06,
+      "loss": 0.4737,
+      "step": 3749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9862356925132492,
+      "learning_rate": 5.800428714658869e-06,
+      "loss": 0.5114,
+      "step": 3750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2015546976438813,
+      "learning_rate": 5.7984448502288455e-06,
+      "loss": 0.5073,
+      "step": 3751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2466761880916173,
+      "learning_rate": 5.796460856803024e-06,
+      "loss": 0.4497,
+      "step": 3752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4910622887566962,
+      "learning_rate": 5.794476734701936e-06,
+      "loss": 0.5127,
+      "step": 3753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.290204703858359,
+      "learning_rate": 5.792492484246134e-06,
+      "loss": 0.4756,
+      "step": 3754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4159973156842203,
+      "learning_rate": 5.790508105756192e-06,
+      "loss": 0.4534,
+      "step": 3755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4655772184973017,
+      "learning_rate": 5.788523599552704e-06,
+      "loss": 0.4617,
+      "step": 3756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.63522079548207,
+      "learning_rate": 5.7865389659562835e-06,
+      "loss": 0.5522,
+      "step": 3757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.342507328459723,
+      "learning_rate": 5.784554205287564e-06,
+      "loss": 0.5328,
+      "step": 3758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3710636892525927,
+      "learning_rate": 5.782569317867204e-06,
+      "loss": 0.4799,
+      "step": 3759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.238963629696016,
+      "learning_rate": 5.780584304015879e-06,
+      "loss": 0.4872,
+      "step": 3760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.516859888821994,
+      "learning_rate": 5.778599164054286e-06,
+      "loss": 0.5175,
+      "step": 3761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5242376957130266,
+      "learning_rate": 5.77661389830314e-06,
+      "loss": 0.5162,
+      "step": 3762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5959688209328888,
+      "learning_rate": 5.774628507083178e-06,
+      "loss": 0.478,
+      "step": 3763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5390937579591093,
+      "learning_rate": 5.772642990715158e-06,
+      "loss": 0.549,
+      "step": 3764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5459986321471937,
+      "learning_rate": 5.7706573495198595e-06,
+      "loss": 0.4824,
+      "step": 3765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7485615328310886,
+      "learning_rate": 5.768671583818077e-06,
+      "loss": 0.5419,
+      "step": 3766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6617800056061585,
+      "learning_rate": 5.7666856939306316e-06,
+      "loss": 0.5149,
+      "step": 3767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.035412095338443,
+      "learning_rate": 5.7646996801783595e-06,
+      "loss": 0.5039,
+      "step": 3768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2570451162662535,
+      "learning_rate": 5.762713542882119e-06,
+      "loss": 0.514,
+      "step": 3769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5096625637139065,
+      "learning_rate": 5.76072728236279e-06,
+      "loss": 0.4711,
+      "step": 3770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4769654769939338,
+      "learning_rate": 5.7587408989412666e-06,
+      "loss": 0.4925,
+      "step": 3771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7859026735626404,
+      "learning_rate": 5.7567543929384696e-06,
+      "loss": 0.4986,
+      "step": 3772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7966181544995699,
+      "learning_rate": 5.754767764675335e-06,
+      "loss": 0.4997,
+      "step": 3773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4071480749895844,
+      "learning_rate": 5.752781014472822e-06,
+      "loss": 0.482,
+      "step": 3774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3596223913000691,
+      "learning_rate": 5.750794142651904e-06,
+      "loss": 0.5223,
+      "step": 3775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4904250217919546,
+      "learning_rate": 5.74880714953358e-06,
+      "loss": 0.517,
+      "step": 3776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9235479357932799,
+      "learning_rate": 5.746820035438868e-06,
+      "loss": 0.5683,
+      "step": 3777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8101897413589039,
+      "learning_rate": 5.744832800688801e-06,
+      "loss": 0.4943,
+      "step": 3778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.255858221208562,
+      "learning_rate": 5.742845445604436e-06,
+      "loss": 0.5104,
+      "step": 3779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7438680744054342,
+      "learning_rate": 5.740857970506846e-06,
+      "loss": 0.5296,
+      "step": 3780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9525014070396836,
+      "learning_rate": 5.738870375717125e-06,
+      "loss": 0.5288,
+      "step": 3781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5688073783490026,
+      "learning_rate": 5.736882661556389e-06,
+      "loss": 0.5331,
+      "step": 3782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4595580168795548,
+      "learning_rate": 5.7348948283457675e-06,
+      "loss": 0.5149,
+      "step": 3783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3998571453059667,
+      "learning_rate": 5.7329068764064155e-06,
+      "loss": 0.4991,
+      "step": 3784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.538385560945485,
+      "learning_rate": 5.730918806059502e-06,
+      "loss": 0.5069,
+      "step": 3785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5397070432129656,
+      "learning_rate": 5.728930617626218e-06,
+      "loss": 0.5278,
+      "step": 3786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.64968775068539,
+      "learning_rate": 5.7269423114277745e-06,
+      "loss": 0.5205,
+      "step": 3787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.783297417690423,
+      "learning_rate": 5.724953887785396e-06,
+      "loss": 0.4983,
+      "step": 3788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8262445860846808,
+      "learning_rate": 5.722965347020334e-06,
+      "loss": 0.4443,
+      "step": 3789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.673647777908011,
+      "learning_rate": 5.7209766894538525e-06,
+      "loss": 0.5117,
+      "step": 3790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5446791654566836,
+      "learning_rate": 5.718987915407235e-06,
+      "loss": 0.4558,
+      "step": 3791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7108875475636238,
+      "learning_rate": 5.716999025201789e-06,
+      "loss": 0.518,
+      "step": 3792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.830804067213509,
+      "learning_rate": 5.715010019158835e-06,
+      "loss": 0.5068,
+      "step": 3793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4892383342831381,
+      "learning_rate": 5.713020897599717e-06,
+      "loss": 0.5711,
+      "step": 3794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6181834703601836,
+      "learning_rate": 5.7110316608457905e-06,
+      "loss": 0.492,
+      "step": 3795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.807414076229394,
+      "learning_rate": 5.70904230921844e-06,
+      "loss": 0.4895,
+      "step": 3796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3334993049574357,
+      "learning_rate": 5.7070528430390585e-06,
+      "loss": 0.4797,
+      "step": 3797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.481416961209663,
+      "learning_rate": 5.705063262629062e-06,
+      "loss": 0.521,
+      "step": 3798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5103932214495144,
+      "learning_rate": 5.703073568309888e-06,
+      "loss": 0.4994,
+      "step": 3799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.697328962339054,
+      "learning_rate": 5.701083760402984e-06,
+      "loss": 0.552,
+      "step": 3800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4739255860006288,
+      "learning_rate": 5.699093839229826e-06,
+      "loss": 0.4767,
+      "step": 3801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.634062539340076,
+      "learning_rate": 5.697103805111901e-06,
+      "loss": 0.4903,
+      "step": 3802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3965792045845735,
+      "learning_rate": 5.695113658370717e-06,
+      "loss": 0.5166,
+      "step": 3803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3024963936312466,
+      "learning_rate": 5.693123399327798e-06,
+      "loss": 0.5122,
+      "step": 3804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2964010800289418,
+      "learning_rate": 5.691133028304691e-06,
+      "loss": 0.4602,
+      "step": 3805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4052206110404424,
+      "learning_rate": 5.689142545622955e-06,
+      "loss": 0.5785,
+      "step": 3806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5247851003929445,
+      "learning_rate": 5.687151951604173e-06,
+      "loss": 0.5741,
+      "step": 3807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3450074532349305,
+      "learning_rate": 5.685161246569939e-06,
+      "loss": 0.4638,
+      "step": 3808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5339496052297188,
+      "learning_rate": 5.6831704308418726e-06,
+      "loss": 0.4924,
+      "step": 3809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3515665945195043,
+      "learning_rate": 5.681179504741606e-06,
+      "loss": 0.4642,
+      "step": 3810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5465504956664637,
+      "learning_rate": 5.679188468590792e-06,
+      "loss": 0.5482,
+      "step": 3811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4018583974536207,
+      "learning_rate": 5.6771973227110976e-06,
+      "loss": 0.526,
+      "step": 3812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2934414393005258,
+      "learning_rate": 5.675206067424212e-06,
+      "loss": 0.459,
+      "step": 3813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.315842288116888,
+      "learning_rate": 5.67321470305184e-06,
+      "loss": 0.4843,
+      "step": 3814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3248801202211313,
+      "learning_rate": 5.671223229915705e-06,
+      "loss": 0.5124,
+      "step": 3815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2614354452804377,
+      "learning_rate": 5.669231648337543e-06,
+      "loss": 0.511,
+      "step": 3816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4200889081262933,
+      "learning_rate": 5.667239958639114e-06,
+      "loss": 0.4547,
+      "step": 3817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4868397770474286,
+      "learning_rate": 5.665248161142196e-06,
+      "loss": 0.4711,
+      "step": 3818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4873342914262366,
+      "learning_rate": 5.663256256168577e-06,
+      "loss": 0.5622,
+      "step": 3819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5527944075539648,
+      "learning_rate": 5.661264244040067e-06,
+      "loss": 0.5585,
+      "step": 3820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.455195628107584,
+      "learning_rate": 5.659272125078495e-06,
+      "loss": 0.5281,
+      "step": 3821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4272424919821816,
+      "learning_rate": 5.6572798996057065e-06,
+      "loss": 0.4699,
+      "step": 3822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4537711024353266,
+      "learning_rate": 5.65528756794356e-06,
+      "loss": 0.526,
+      "step": 3823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7239883379658023,
+      "learning_rate": 5.653295130413937e-06,
+      "loss": 0.501,
+      "step": 3824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.388762025067527,
+      "learning_rate": 5.651302587338732e-06,
+      "loss": 0.465,
+      "step": 3825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.605357518243063,
+      "learning_rate": 5.649309939039856e-06,
+      "loss": 0.5448,
+      "step": 3826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3004710250305824,
+      "learning_rate": 5.647317185839243e-06,
+      "loss": 0.5147,
+      "step": 3827
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2453765452625825,
+      "learning_rate": 5.645324328058834e-06,
+      "loss": 0.4939,
+      "step": 3828
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6630221600508244,
+      "learning_rate": 5.643331366020599e-06,
+      "loss": 0.5313,
+      "step": 3829
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1403267643695534,
+      "learning_rate": 5.641338300046516e-06,
+      "loss": 0.5347,
+      "step": 3830
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6645397725384086,
+      "learning_rate": 5.639345130458582e-06,
+      "loss": 0.4903,
+      "step": 3831
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.30182281674491,
+      "learning_rate": 5.637351857578811e-06,
+      "loss": 0.4917,
+      "step": 3832
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4362963547064533,
+      "learning_rate": 5.635358481729234e-06,
+      "loss": 0.5219,
+      "step": 3833
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4382650532418015,
+      "learning_rate": 5.6333650032318985e-06,
+      "loss": 0.5203,
+      "step": 3834
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3442530910752974,
+      "learning_rate": 5.631371422408869e-06,
+      "loss": 0.4943,
+      "step": 3835
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4265060064893866,
+      "learning_rate": 5.629377739582225e-06,
+      "loss": 0.4498,
+      "step": 3836
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.216269417602553,
+      "learning_rate": 5.627383955074064e-06,
+      "loss": 0.477,
+      "step": 3837
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8816201018509997,
+      "learning_rate": 5.6253900692065e-06,
+      "loss": 0.5216,
+      "step": 3838
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5067962273330215,
+      "learning_rate": 5.623396082301662e-06,
+      "loss": 0.4927,
+      "step": 3839
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.553380864316869,
+      "learning_rate": 5.6214019946816975e-06,
+      "loss": 0.5067,
+      "step": 3840
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2467239685806337,
+      "learning_rate": 5.619407806668768e-06,
+      "loss": 0.5472,
+      "step": 3841
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3959033963800893,
+      "learning_rate": 5.617413518585051e-06,
+      "loss": 0.5428,
+      "step": 3842
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.833831320835665,
+      "learning_rate": 5.615419130752743e-06,
+      "loss": 0.5274,
+      "step": 3843
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.447765424375853,
+      "learning_rate": 5.613424643494054e-06,
+      "loss": 0.552,
+      "step": 3844
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.423418358339179,
+      "learning_rate": 5.611430057131211e-06,
+      "loss": 0.5416,
+      "step": 3845
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9508924456607224,
+      "learning_rate": 5.609435371986457e-06,
+      "loss": 0.5314,
+      "step": 3846
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1955084497773902,
+      "learning_rate": 5.607440588382052e-06,
+      "loss": 0.4636,
+      "step": 3847
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8439394899829424,
+      "learning_rate": 5.60544570664027e-06,
+      "loss": 0.5107,
+      "step": 3848
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.515882843463731,
+      "learning_rate": 5.6034507270834e-06,
+      "loss": 0.5218,
+      "step": 3849
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5241593853007185,
+      "learning_rate": 5.601455650033754e-06,
+      "loss": 0.5216,
+      "step": 3850
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 12.162683050587866,
+      "learning_rate": 5.599460475813648e-06,
+      "loss": 0.4684,
+      "step": 3851
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2701973747754316,
+      "learning_rate": 5.5974652047454235e-06,
+      "loss": 0.5149,
+      "step": 3852
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9791516248661194,
+      "learning_rate": 5.595469837151432e-06,
+      "loss": 0.531,
+      "step": 3853
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2696017553188235,
+      "learning_rate": 5.593474373354045e-06,
+      "loss": 0.5118,
+      "step": 3854
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3986543128280633,
+      "learning_rate": 5.591478813675646e-06,
+      "loss": 0.53,
+      "step": 3855
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2703344434000077,
+      "learning_rate": 5.589483158438636e-06,
+      "loss": 0.5266,
+      "step": 3856
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4011913209638538,
+      "learning_rate": 5.587487407965429e-06,
+      "loss": 0.5057,
+      "step": 3857
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.313243610761376,
+      "learning_rate": 5.585491562578456e-06,
+      "loss": 0.497,
+      "step": 3858
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4951327213403844,
+      "learning_rate": 5.583495622600168e-06,
+      "loss": 0.5133,
+      "step": 3859
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7132730355293824,
+      "learning_rate": 5.581499588353021e-06,
+      "loss": 0.4932,
+      "step": 3860
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.479779515901214,
+      "learning_rate": 5.579503460159493e-06,
+      "loss": 0.5218,
+      "step": 3861
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3103752607086714,
+      "learning_rate": 5.577507238342078e-06,
+      "loss": 0.5094,
+      "step": 3862
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5486202310935997,
+      "learning_rate": 5.575510923223284e-06,
+      "loss": 0.4926,
+      "step": 3863
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.033733047007289,
+      "learning_rate": 5.57351451512563e-06,
+      "loss": 0.4873,
+      "step": 3864
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3273892616667802,
+      "learning_rate": 5.5715180143716555e-06,
+      "loss": 0.5261,
+      "step": 3865
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2925855023656074,
+      "learning_rate": 5.569521421283912e-06,
+      "loss": 0.5083,
+      "step": 3866
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1539965994659505,
+      "learning_rate": 5.567524736184967e-06,
+      "loss": 0.4574,
+      "step": 3867
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5021009839892818,
+      "learning_rate": 5.565527959397403e-06,
+      "loss": 0.5179,
+      "step": 3868
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2228970177505332,
+      "learning_rate": 5.563531091243817e-06,
+      "loss": 0.5128,
+      "step": 3869
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7249298140042144,
+      "learning_rate": 5.5615341320468195e-06,
+      "loss": 0.5216,
+      "step": 3870
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.342039726006231,
+      "learning_rate": 5.559537082129037e-06,
+      "loss": 0.5276,
+      "step": 3871
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.660683024715625,
+      "learning_rate": 5.5575399418131115e-06,
+      "loss": 0.5323,
+      "step": 3872
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7599077618364973,
+      "learning_rate": 5.555542711421696e-06,
+      "loss": 0.5335,
+      "step": 3873
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4276717562059684,
+      "learning_rate": 5.553545391277465e-06,
+      "loss": 0.4936,
+      "step": 3874
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0651420932368425,
+      "learning_rate": 5.5515479817031005e-06,
+      "loss": 0.5023,
+      "step": 3875
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2854772402377153,
+      "learning_rate": 5.549550483021302e-06,
+      "loss": 0.4946,
+      "step": 3876
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.999653481852163,
+      "learning_rate": 5.547552895554783e-06,
+      "loss": 0.4761,
+      "step": 3877
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.355792789812382,
+      "learning_rate": 5.545555219626271e-06,
+      "loss": 0.5174,
+      "step": 3878
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4044057073752125,
+      "learning_rate": 5.543557455558509e-06,
+      "loss": 0.528,
+      "step": 3879
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4615035364180013,
+      "learning_rate": 5.541559603674252e-06,
+      "loss": 0.5201,
+      "step": 3880
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.052410206473402,
+      "learning_rate": 5.539561664296273e-06,
+      "loss": 0.519,
+      "step": 3881
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7101247179212373,
+      "learning_rate": 5.537563637747352e-06,
+      "loss": 0.4933,
+      "step": 3882
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4553625355001882,
+      "learning_rate": 5.535565524350293e-06,
+      "loss": 0.515,
+      "step": 3883
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5823929048265364,
+      "learning_rate": 5.533567324427906e-06,
+      "loss": 0.5251,
+      "step": 3884
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5125343261415436,
+      "learning_rate": 5.5315690383030195e-06,
+      "loss": 0.5091,
+      "step": 3885
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6701063803891216,
+      "learning_rate": 5.529570666298473e-06,
+      "loss": 0.4574,
+      "step": 3886
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7559956685205522,
+      "learning_rate": 5.5275722087371205e-06,
+      "loss": 0.5381,
+      "step": 3887
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9654385967049164,
+      "learning_rate": 5.525573665941833e-06,
+      "loss": 0.5253,
+      "step": 3888
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.92878214412045,
+      "learning_rate": 5.523575038235489e-06,
+      "loss": 0.4617,
+      "step": 3889
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9839273447811436,
+      "learning_rate": 5.521576325940986e-06,
+      "loss": 0.5056,
+      "step": 3890
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.448419304382558,
+      "learning_rate": 5.519577529381235e-06,
+      "loss": 0.509,
+      "step": 3891
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6441958788761752,
+      "learning_rate": 5.5175786488791575e-06,
+      "loss": 0.5632,
+      "step": 3892
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1335798112114635,
+      "learning_rate": 5.515579684757691e-06,
+      "loss": 0.5622,
+      "step": 3893
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.362047097744073,
+      "learning_rate": 5.513580637339786e-06,
+      "loss": 0.556,
+      "step": 3894
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4222148143798927,
+      "learning_rate": 5.511581506948407e-06,
+      "loss": 0.4875,
+      "step": 3895
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4232338129775202,
+      "learning_rate": 5.509582293906528e-06,
+      "loss": 0.4922,
+      "step": 3896
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.479210921980862,
+      "learning_rate": 5.507582998537142e-06,
+      "loss": 0.5159,
+      "step": 3897
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3022543152571098,
+      "learning_rate": 5.505583621163252e-06,
+      "loss": 0.5361,
+      "step": 3898
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3123499942506818,
+      "learning_rate": 5.503584162107876e-06,
+      "loss": 0.5051,
+      "step": 3899
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.434330700741348,
+      "learning_rate": 5.501584621694043e-06,
+      "loss": 0.5379,
+      "step": 3900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9198939466196938,
+      "learning_rate": 5.4995850002447955e-06,
+      "loss": 0.5165,
+      "step": 3901
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5452717097371367,
+      "learning_rate": 5.4975852980831925e-06,
+      "loss": 0.4853,
+      "step": 3902
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4825630920396455,
+      "learning_rate": 5.495585515532302e-06,
+      "loss": 0.526,
+      "step": 3903
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3945136410814463,
+      "learning_rate": 5.4935856529152075e-06,
+      "loss": 0.477,
+      "step": 3904
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8283627173286865,
+      "learning_rate": 5.491585710555004e-06,
+      "loss": 0.5037,
+      "step": 3905
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.863992368418988,
+      "learning_rate": 5.489585688774798e-06,
+      "loss": 0.5524,
+      "step": 3906
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5047987926389876,
+      "learning_rate": 5.487585587897713e-06,
+      "loss": 0.4737,
+      "step": 3907
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.051631341269757,
+      "learning_rate": 5.4855854082468805e-06,
+      "loss": 0.5492,
+      "step": 3908
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.411897797407055,
+      "learning_rate": 5.483585150145451e-06,
+      "loss": 0.4651,
+      "step": 3909
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5153953133448943,
+      "learning_rate": 5.481584813916579e-06,
+      "loss": 0.5339,
+      "step": 3910
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5773017112958216,
+      "learning_rate": 5.47958439988344e-06,
+      "loss": 0.5031,
+      "step": 3911
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4551035961762118,
+      "learning_rate": 5.477583908369219e-06,
+      "loss": 0.5042,
+      "step": 3912
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6593828724311739,
+      "learning_rate": 5.47558333969711e-06,
+      "loss": 0.5674,
+      "step": 3913
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.52905879375427,
+      "learning_rate": 5.473582694190323e-06,
+      "loss": 0.4978,
+      "step": 3914
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.068302465264425,
+      "learning_rate": 5.471581972172082e-06,
+      "loss": 0.4969,
+      "step": 3915
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9660457837978043,
+      "learning_rate": 5.4695811739656204e-06,
+      "loss": 0.5076,
+      "step": 3916
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.45853978070074,
+      "learning_rate": 5.467580299894183e-06,
+      "loss": 0.5606,
+      "step": 3917
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.664872886856364,
+      "learning_rate": 5.465579350281032e-06,
+      "loss": 0.513,
+      "step": 3918
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6934631719056283,
+      "learning_rate": 5.463578325449434e-06,
+      "loss": 0.5506,
+      "step": 3919
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0296482449817947,
+      "learning_rate": 5.461577225722676e-06,
+      "loss": 0.5128,
+      "step": 3920
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6661246497699034,
+      "learning_rate": 5.4595760514240525e-06,
+      "loss": 0.4678,
+      "step": 3921
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3550930215234538,
+      "learning_rate": 5.45757480287687e-06,
+      "loss": 0.5224,
+      "step": 3922
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4528187165438429,
+      "learning_rate": 5.455573480404448e-06,
+      "loss": 0.512,
+      "step": 3923
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4543519260418145,
+      "learning_rate": 5.453572084330117e-06,
+      "loss": 0.4873,
+      "step": 3924
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4530409679251879,
+      "learning_rate": 5.451570614977223e-06,
+      "loss": 0.5308,
+      "step": 3925
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4345990352945623,
+      "learning_rate": 5.449569072669119e-06,
+      "loss": 0.4699,
+      "step": 3926
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3333642766386664,
+      "learning_rate": 5.447567457729169e-06,
+      "loss": 0.4703,
+      "step": 3927
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9541630497777558,
+      "learning_rate": 5.4455657704807566e-06,
+      "loss": 0.5087,
+      "step": 3928
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5383476148156443,
+      "learning_rate": 5.44356401124727e-06,
+      "loss": 0.5101,
+      "step": 3929
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3848178623750136,
+      "learning_rate": 5.44156218035211e-06,
+      "loss": 0.5336,
+      "step": 3930
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2823093070822174,
+      "learning_rate": 5.439560278118692e-06,
+      "loss": 0.4547,
+      "step": 3931
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7596416413390763,
+      "learning_rate": 5.437558304870438e-06,
+      "loss": 0.5239,
+      "step": 3932
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6311320117875125,
+      "learning_rate": 5.435556260930788e-06,
+      "loss": 0.4999,
+      "step": 3933
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1424810969438024,
+      "learning_rate": 5.433554146623187e-06,
+      "loss": 0.4954,
+      "step": 3934
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2357186619718303,
+      "learning_rate": 5.431551962271094e-06,
+      "loss": 0.5075,
+      "step": 3935
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5599698797491244,
+      "learning_rate": 5.429549708197982e-06,
+      "loss": 0.4683,
+      "step": 3936
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.531008438280879,
+      "learning_rate": 5.427547384727332e-06,
+      "loss": 0.5502,
+      "step": 3937
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4741978722495628,
+      "learning_rate": 5.425544992182636e-06,
+      "loss": 0.5169,
+      "step": 3938
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8507114164050131,
+      "learning_rate": 5.423542530887399e-06,
+      "loss": 0.4601,
+      "step": 3939
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3637993319871706,
+      "learning_rate": 5.421540001165135e-06,
+      "loss": 0.5606,
+      "step": 3940
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7277813484271007,
+      "learning_rate": 5.419537403339372e-06,
+      "loss": 0.4859,
+      "step": 3941
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2992120001099763,
+      "learning_rate": 5.417534737733648e-06,
+      "loss": 0.5243,
+      "step": 3942
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3523696755445922,
+      "learning_rate": 5.415532004671506e-06,
+      "loss": 0.5216,
+      "step": 3943
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4589926049928805,
+      "learning_rate": 5.413529204476512e-06,
+      "loss": 0.4862,
+      "step": 3944
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.334457517749518,
+      "learning_rate": 5.411526337472232e-06,
+      "loss": 0.5187,
+      "step": 3945
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3357269722684226,
+      "learning_rate": 5.409523403982247e-06,
+      "loss": 0.5084,
+      "step": 3946
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.596655906918947,
+      "learning_rate": 5.40752040433015e-06,
+      "loss": 0.4783,
+      "step": 3947
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.386972195789598,
+      "learning_rate": 5.4055173388395445e-06,
+      "loss": 0.5405,
+      "step": 3948
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3719441891820827,
+      "learning_rate": 5.40351420783404e-06,
+      "loss": 0.4992,
+      "step": 3949
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5172485941704512,
+      "learning_rate": 5.40151101163726e-06,
+      "loss": 0.5282,
+      "step": 3950
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4264798256383389,
+      "learning_rate": 5.399507750572841e-06,
+      "loss": 0.5071,
+      "step": 3951
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.87510756362639,
+      "learning_rate": 5.397504424964426e-06,
+      "loss": 0.5177,
+      "step": 3952
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3567163748524484,
+      "learning_rate": 5.395501035135671e-06,
+      "loss": 0.4396,
+      "step": 3953
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3409907238416823,
+      "learning_rate": 5.39349758141024e-06,
+      "loss": 0.4629,
+      "step": 3954
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.982873474793292,
+      "learning_rate": 5.391494064111809e-06,
+      "loss": 0.4955,
+      "step": 3955
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.508085533195063,
+      "learning_rate": 5.389490483564064e-06,
+      "loss": 0.5198,
+      "step": 3956
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2250701542250673,
+      "learning_rate": 5.387486840090701e-06,
+      "loss": 0.4803,
+      "step": 3957
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4641385971210494,
+      "learning_rate": 5.3854831340154265e-06,
+      "loss": 0.5253,
+      "step": 3958
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3998754036190544,
+      "learning_rate": 5.383479365661958e-06,
+      "loss": 0.4961,
+      "step": 3959
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4065888899634549,
+      "learning_rate": 5.381475535354018e-06,
+      "loss": 0.5066,
+      "step": 3960
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4691889688853568,
+      "learning_rate": 5.379471643415347e-06,
+      "loss": 0.5421,
+      "step": 3961
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.39716223039857,
+      "learning_rate": 5.37746769016969e-06,
+      "loss": 0.5345,
+      "step": 3962
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2562632841531238,
+      "learning_rate": 5.375463675940803e-06,
+      "loss": 0.4499,
+      "step": 3963
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8008475312845498,
+      "learning_rate": 5.373459601052451e-06,
+      "loss": 0.5108,
+      "step": 3964
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.483453578025183,
+      "learning_rate": 5.3714554658284125e-06,
+      "loss": 0.565,
+      "step": 3965
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4975414533192573,
+      "learning_rate": 5.369451270592472e-06,
+      "loss": 0.5173,
+      "step": 3966
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.123985067004447,
+      "learning_rate": 5.3674470156684255e-06,
+      "loss": 0.522,
+      "step": 3967
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6500700336932811,
+      "learning_rate": 5.365442701380077e-06,
+      "loss": 0.4349,
+      "step": 3968
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6029152251625887,
+      "learning_rate": 5.3634383280512416e-06,
+      "loss": 0.5308,
+      "step": 3969
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3769653475610781,
+      "learning_rate": 5.361433896005743e-06,
+      "loss": 0.5098,
+      "step": 3970
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.020545630737324,
+      "learning_rate": 5.359429405567415e-06,
+      "loss": 0.5218,
+      "step": 3971
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4426438371728305,
+      "learning_rate": 5.357424857060102e-06,
+      "loss": 0.5149,
+      "step": 3972
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.498625679032598,
+      "learning_rate": 5.355420250807654e-06,
+      "loss": 0.5784,
+      "step": 3973
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.337563009204282,
+      "learning_rate": 5.353415587133936e-06,
+      "loss": 0.5322,
+      "step": 3974
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1258963333162684,
+      "learning_rate": 5.351410866362816e-06,
+      "loss": 0.5215,
+      "step": 3975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1845941268352767,
+      "learning_rate": 5.349406088818176e-06,
+      "loss": 0.5404,
+      "step": 3976
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5751923241999841,
+      "learning_rate": 5.347401254823906e-06,
+      "loss": 0.5373,
+      "step": 3977
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.646247402251166,
+      "learning_rate": 5.3453963647039035e-06,
+      "loss": 0.5345,
+      "step": 3978
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.58443216877823,
+      "learning_rate": 5.343391418782076e-06,
+      "loss": 0.4974,
+      "step": 3979
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.9367408343514425,
+      "learning_rate": 5.341386417382338e-06,
+      "loss": 0.4804,
+      "step": 3980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3965703455304954,
+      "learning_rate": 5.33938136082862e-06,
+      "loss": 0.5394,
+      "step": 3981
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.590559268954731,
+      "learning_rate": 5.3373762494448546e-06,
+      "loss": 0.5371,
+      "step": 3982
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.320429476244967,
+      "learning_rate": 5.335371083554984e-06,
+      "loss": 0.53,
+      "step": 3983
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.421899153361156,
+      "learning_rate": 5.333365863482961e-06,
+      "loss": 0.4955,
+      "step": 3984
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5542947739027118,
+      "learning_rate": 5.331360589552746e-06,
+      "loss": 0.4968,
+      "step": 3985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7457943000103398,
+      "learning_rate": 5.3293552620883115e-06,
+      "loss": 0.4566,
+      "step": 3986
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.2672149801902926,
+      "learning_rate": 5.327349881413632e-06,
+      "loss": 0.5328,
+      "step": 3987
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6192367900745737,
+      "learning_rate": 5.325344447852696e-06,
+      "loss": 0.508,
+      "step": 3988
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.478967517170313,
+      "learning_rate": 5.323338961729499e-06,
+      "loss": 0.5139,
+      "step": 3989
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5728731128199156,
+      "learning_rate": 5.321333423368047e-06,
+      "loss": 0.5522,
+      "step": 3990
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3371364747051662,
+      "learning_rate": 5.319327833092348e-06,
+      "loss": 0.4853,
+      "step": 3991
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6996233064142633,
+      "learning_rate": 5.317322191226426e-06,
+      "loss": 0.4729,
+      "step": 3992
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4012552294803955,
+      "learning_rate": 5.3153164980943104e-06,
+      "loss": 0.5021,
+      "step": 3993
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4430297337479316,
+      "learning_rate": 5.313310754020037e-06,
+      "loss": 0.4965,
+      "step": 3994
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2621347231021145,
+      "learning_rate": 5.311304959327651e-06,
+      "loss": 0.4459,
+      "step": 3995
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2817572185185382,
+      "learning_rate": 5.309299114341209e-06,
+      "loss": 0.5423,
+      "step": 3996
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5992598858188514,
+      "learning_rate": 5.307293219384768e-06,
+      "loss": 0.4952,
+      "step": 3997
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3738706690239122,
+      "learning_rate": 5.305287274782403e-06,
+      "loss": 0.5252,
+      "step": 3998
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.422877770604774,
+      "learning_rate": 5.303281280858189e-06,
+      "loss": 0.5104,
+      "step": 3999
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4768927119527948,
+      "learning_rate": 5.301275237936214e-06,
+      "loss": 0.5426,
+      "step": 4000
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3443790195812775,
+      "learning_rate": 5.29926914634057e-06,
+      "loss": 0.511,
+      "step": 4001
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7210003573811843,
+      "learning_rate": 5.29726300639536e-06,
+      "loss": 0.5112,
+      "step": 4002
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4934549647405275,
+      "learning_rate": 5.295256818424695e-06,
+      "loss": 0.5028,
+      "step": 4003
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5346083778331399,
+      "learning_rate": 5.293250582752689e-06,
+      "loss": 0.5074,
+      "step": 4004
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.51676023911406,
+      "learning_rate": 5.291244299703469e-06,
+      "loss": 0.4786,
+      "step": 4005
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2407408253081538,
+      "learning_rate": 5.2892379696011665e-06,
+      "loss": 0.549,
+      "step": 4006
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.482395526869878,
+      "learning_rate": 5.2872315927699235e-06,
+      "loss": 0.5327,
+      "step": 4007
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9826848283440506,
+      "learning_rate": 5.2852251695338865e-06,
+      "loss": 0.5153,
+      "step": 4008
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1060888524002794,
+      "learning_rate": 5.283218700217211e-06,
+      "loss": 0.5266,
+      "step": 4009
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4164346549674205,
+      "learning_rate": 5.28121218514406e-06,
+      "loss": 0.4965,
+      "step": 4010
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4669936305099687,
+      "learning_rate": 5.279205624638605e-06,
+      "loss": 0.4723,
+      "step": 4011
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4057306748118408,
+      "learning_rate": 5.277199019025022e-06,
+      "loss": 0.4843,
+      "step": 4012
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4158292097211684,
+      "learning_rate": 5.275192368627495e-06,
+      "loss": 0.5183,
+      "step": 4013
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.347733607048643,
+      "learning_rate": 5.2731856737702195e-06,
+      "loss": 0.5201,
+      "step": 4014
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4507919841534922,
+      "learning_rate": 5.27117893477739e-06,
+      "loss": 0.4904,
+      "step": 4015
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.351191256523609,
+      "learning_rate": 5.269172151973216e-06,
+      "loss": 0.5062,
+      "step": 4016
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8170541743140898,
+      "learning_rate": 5.26716532568191e-06,
+      "loss": 0.5185,
+      "step": 4017
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8948850199277016,
+      "learning_rate": 5.2651584562276935e-06,
+      "loss": 0.5088,
+      "step": 4018
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4895022409198124,
+      "learning_rate": 5.263151543934792e-06,
+      "loss": 0.4905,
+      "step": 4019
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2034136078269118,
+      "learning_rate": 5.261144589127441e-06,
+      "loss": 0.4706,
+      "step": 4020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5148654943676552,
+      "learning_rate": 5.259137592129883e-06,
+      "loss": 0.5068,
+      "step": 4021
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7539595585635952,
+      "learning_rate": 5.257130553266364e-06,
+      "loss": 0.4596,
+      "step": 4022
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1276164530176103,
+      "learning_rate": 5.25512347286114e-06,
+      "loss": 0.4614,
+      "step": 4023
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.293833851355961,
+      "learning_rate": 5.253116351238472e-06,
+      "loss": 0.4858,
+      "step": 4024
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.488875806713043,
+      "learning_rate": 5.251109188722626e-06,
+      "loss": 0.5054,
+      "step": 4025
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.436624795404598,
+      "learning_rate": 5.249101985637881e-06,
+      "loss": 0.5032,
+      "step": 4026
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.621443769321251,
+      "learning_rate": 5.247094742308516e-06,
+      "loss": 0.4775,
+      "step": 4027
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9501526582620479,
+      "learning_rate": 5.24508745905882e-06,
+      "loss": 0.5201,
+      "step": 4028
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.042610189384628,
+      "learning_rate": 5.243080136213085e-06,
+      "loss": 0.544,
+      "step": 4029
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5994079979309168,
+      "learning_rate": 5.241072774095615e-06,
+      "loss": 0.5709,
+      "step": 4030
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.251606242450616,
+      "learning_rate": 5.239065373030713e-06,
+      "loss": 0.4817,
+      "step": 4031
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4523263770339998,
+      "learning_rate": 5.237057933342696e-06,
+      "loss": 0.4666,
+      "step": 4032
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3870858370349122,
+      "learning_rate": 5.235050455355881e-06,
+      "loss": 0.5148,
+      "step": 4033
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2956128600540535,
+      "learning_rate": 5.233042939394595e-06,
+      "loss": 0.5353,
+      "step": 4034
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5130642043818001,
+      "learning_rate": 5.2310353857831695e-06,
+      "loss": 0.5045,
+      "step": 4035
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4965739798426347,
+      "learning_rate": 5.229027794845944e-06,
+      "loss": 0.5177,
+      "step": 4036
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3940798964222623,
+      "learning_rate": 5.227020166907259e-06,
+      "loss": 0.5138,
+      "step": 4037
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.216930430982062,
+      "learning_rate": 5.225012502291469e-06,
+      "loss": 0.4939,
+      "step": 4038
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4608752793636535,
+      "learning_rate": 5.223004801322926e-06,
+      "loss": 0.5409,
+      "step": 4039
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4213725178723116,
+      "learning_rate": 5.220997064325994e-06,
+      "loss": 0.5562,
+      "step": 4040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3870879937572156,
+      "learning_rate": 5.21898929162504e-06,
+      "loss": 0.5173,
+      "step": 4041
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.782760226940039,
+      "learning_rate": 5.2169814835444356e-06,
+      "loss": 0.4981,
+      "step": 4042
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6988563115846317,
+      "learning_rate": 5.214973640408563e-06,
+      "loss": 0.4882,
+      "step": 4043
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3902236234519256,
+      "learning_rate": 5.2129657625418055e-06,
+      "loss": 0.458,
+      "step": 4044
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3850148995382587,
+      "learning_rate": 5.2109578502685534e-06,
+      "loss": 0.5024,
+      "step": 4045
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3471538316223168,
+      "learning_rate": 5.208949903913201e-06,
+      "loss": 0.4689,
+      "step": 4046
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6714409613819254,
+      "learning_rate": 5.206941923800154e-06,
+      "loss": 0.5532,
+      "step": 4047
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2658631922999917,
+      "learning_rate": 5.2049339102538154e-06,
+      "loss": 0.4786,
+      "step": 4048
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.511130438261548,
+      "learning_rate": 5.202925863598599e-06,
+      "loss": 0.5137,
+      "step": 4049
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3184918589934416,
+      "learning_rate": 5.200917784158921e-06,
+      "loss": 0.4591,
+      "step": 4050
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.237884065088822,
+      "learning_rate": 5.1989096722592055e-06,
+      "loss": 0.4556,
+      "step": 4051
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2726858470097968,
+      "learning_rate": 5.1969015282238824e-06,
+      "loss": 0.4671,
+      "step": 4052
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4979136717841783,
+      "learning_rate": 5.1948933523773824e-06,
+      "loss": 0.5146,
+      "step": 4053
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.503751302350897,
+      "learning_rate": 5.192885145044143e-06,
+      "loss": 0.449,
+      "step": 4054
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3001077377892787,
+      "learning_rate": 5.190876906548612e-06,
+      "loss": 0.4992,
+      "step": 4055
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.35780055003118,
+      "learning_rate": 5.188868637215235e-06,
+      "loss": 0.5042,
+      "step": 4056
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8096393235457786,
+      "learning_rate": 5.186860337368468e-06,
+      "loss": 0.4997,
+      "step": 4057
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7976803429353938,
+      "learning_rate": 5.184852007332765e-06,
+      "loss": 0.5438,
+      "step": 4058
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4295847395483166,
+      "learning_rate": 5.182843647432593e-06,
+      "loss": 0.5195,
+      "step": 4059
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.329781113500144,
+      "learning_rate": 5.180835257992419e-06,
+      "loss": 0.4692,
+      "step": 4060
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4198247587720287,
+      "learning_rate": 5.178826839336718e-06,
+      "loss": 0.5149,
+      "step": 4061
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.45463475933709,
+      "learning_rate": 5.176818391789964e-06,
+      "loss": 0.4831,
+      "step": 4062
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.405498695171607,
+      "learning_rate": 5.174809915676643e-06,
+      "loss": 0.5151,
+      "step": 4063
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8636479793019347,
+      "learning_rate": 5.17280141132124e-06,
+      "loss": 0.5218,
+      "step": 4064
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3187889761676121,
+      "learning_rate": 5.170792879048248e-06,
+      "loss": 0.5031,
+      "step": 4065
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2851291534308047,
+      "learning_rate": 5.168784319182161e-06,
+      "loss": 0.4685,
+      "step": 4066
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.901671464065162,
+      "learning_rate": 5.166775732047481e-06,
+      "loss": 0.4776,
+      "step": 4067
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.577965546141581,
+      "learning_rate": 5.164767117968713e-06,
+      "loss": 0.504,
+      "step": 4068
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.611583306216855,
+      "learning_rate": 5.162758477270366e-06,
+      "loss": 0.5139,
+      "step": 4069
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4722771473445617,
+      "learning_rate": 5.160749810276952e-06,
+      "loss": 0.5049,
+      "step": 4070
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6841236873282883,
+      "learning_rate": 5.158741117312992e-06,
+      "loss": 0.4823,
+      "step": 4071
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7283913002786766,
+      "learning_rate": 5.156732398703007e-06,
+      "loss": 0.486,
+      "step": 4072
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3644169274208862,
+      "learning_rate": 5.154723654771522e-06,
+      "loss": 0.5316,
+      "step": 4073
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.412978580938532,
+      "learning_rate": 5.1527148858430675e-06,
+      "loss": 0.5205,
+      "step": 4074
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.304308989910714,
+      "learning_rate": 5.150706092242178e-06,
+      "loss": 0.5112,
+      "step": 4075
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5096581064631402,
+      "learning_rate": 5.148697274293392e-06,
+      "loss": 0.5137,
+      "step": 4076
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9507912798218543,
+      "learning_rate": 5.146688432321253e-06,
+      "loss": 0.5665,
+      "step": 4077
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3974637208424106,
+      "learning_rate": 5.144679566650306e-06,
+      "loss": 0.5266,
+      "step": 4078
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.381186236440111,
+      "learning_rate": 5.1426706776050985e-06,
+      "loss": 0.4926,
+      "step": 4079
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2908982111875704,
+      "learning_rate": 5.140661765510187e-06,
+      "loss": 0.5299,
+      "step": 4080
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.517278250782335,
+      "learning_rate": 5.138652830690129e-06,
+      "loss": 0.5097,
+      "step": 4081
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7679776693034879,
+      "learning_rate": 5.136643873469487e-06,
+      "loss": 0.536,
+      "step": 4082
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3584650382939063,
+      "learning_rate": 5.1346348941728215e-06,
+      "loss": 0.4941,
+      "step": 4083
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.260066066177889,
+      "learning_rate": 5.132625893124704e-06,
+      "loss": 0.5343,
+      "step": 4084
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2417882558433795,
+      "learning_rate": 5.130616870649705e-06,
+      "loss": 0.5519,
+      "step": 4085
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4348244112899506,
+      "learning_rate": 5.1286078270724e-06,
+      "loss": 0.5182,
+      "step": 4086
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3761732902076147,
+      "learning_rate": 5.126598762717367e-06,
+      "loss": 0.5643,
+      "step": 4087
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.093279684038697,
+      "learning_rate": 5.1245896779091905e-06,
+      "loss": 0.544,
+      "step": 4088
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1953773828090084,
+      "learning_rate": 5.122580572972453e-06,
+      "loss": 0.4644,
+      "step": 4089
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.335699984289526,
+      "learning_rate": 5.120571448231746e-06,
+      "loss": 0.481,
+      "step": 4090
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.539519553974143,
+      "learning_rate": 5.118562304011657e-06,
+      "loss": 0.5513,
+      "step": 4091
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3014018277586958,
+      "learning_rate": 5.116553140636788e-06,
+      "loss": 0.5649,
+      "step": 4092
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2963592808820954,
+      "learning_rate": 5.114543958431729e-06,
+      "loss": 0.5241,
+      "step": 4093
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3751884600265256,
+      "learning_rate": 5.112534757721086e-06,
+      "loss": 0.4718,
+      "step": 4094
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6805985594689253,
+      "learning_rate": 5.110525538829461e-06,
+      "loss": 0.5219,
+      "step": 4095
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7427122972421231,
+      "learning_rate": 5.108516302081461e-06,
+      "loss": 0.5227,
+      "step": 4096
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4777276883977413,
+      "learning_rate": 5.106507047801699e-06,
+      "loss": 0.4717,
+      "step": 4097
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4404767805534364,
+      "learning_rate": 5.104497776314784e-06,
+      "loss": 0.4985,
+      "step": 4098
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.32514183192607,
+      "learning_rate": 5.102488487945332e-06,
+      "loss": 0.4796,
+      "step": 4099
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6300526573029417,
+      "learning_rate": 5.100479183017963e-06,
+      "loss": 0.5484,
+      "step": 4100
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.506233040859698,
+      "learning_rate": 5.098469861857299e-06,
+      "loss": 0.4938,
+      "step": 4101
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5854909157924568,
+      "learning_rate": 5.09646052478796e-06,
+      "loss": 0.5012,
+      "step": 4102
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7334239667693274,
+      "learning_rate": 5.094451172134573e-06,
+      "loss": 0.5402,
+      "step": 4103
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.421782694911038,
+      "learning_rate": 5.092441804221767e-06,
+      "loss": 0.5402,
+      "step": 4104
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2380410240615574,
+      "learning_rate": 5.090432421374175e-06,
+      "loss": 0.4782,
+      "step": 4105
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8233363976081143,
+      "learning_rate": 5.0884230239164274e-06,
+      "loss": 0.5275,
+      "step": 4106
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2157977078482316,
+      "learning_rate": 5.0864136121731614e-06,
+      "loss": 0.5032,
+      "step": 4107
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0628077114535106,
+      "learning_rate": 5.084404186469016e-06,
+      "loss": 0.529,
+      "step": 4108
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4673681464236095,
+      "learning_rate": 5.082394747128632e-06,
+      "loss": 0.5267,
+      "step": 4109
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.405043504400975,
+      "learning_rate": 5.08038529447665e-06,
+      "loss": 0.4953,
+      "step": 4110
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.426360350391599,
+      "learning_rate": 5.078375828837716e-06,
+      "loss": 0.5208,
+      "step": 4111
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8072437193277209,
+      "learning_rate": 5.0763663505364754e-06,
+      "loss": 0.5057,
+      "step": 4112
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3660490562593417,
+      "learning_rate": 5.07435685989758e-06,
+      "loss": 0.462,
+      "step": 4113
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5486600415029512,
+      "learning_rate": 5.072347357245678e-06,
+      "loss": 0.5372,
+      "step": 4114
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.616974280163713,
+      "learning_rate": 5.0703378429054226e-06,
+      "loss": 0.5241,
+      "step": 4115
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4809366060080675,
+      "learning_rate": 5.06832831720147e-06,
+      "loss": 0.542,
+      "step": 4116
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8635109413540314,
+      "learning_rate": 5.066318780458476e-06,
+      "loss": 0.5301,
+      "step": 4117
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6926137434244817,
+      "learning_rate": 5.064309233001099e-06,
+      "loss": 0.4694,
+      "step": 4118
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.1567850084651718,
+      "learning_rate": 5.062299675153999e-06,
+      "loss": 0.4616,
+      "step": 4119
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.600980417226766,
+      "learning_rate": 5.0602901072418375e-06,
+      "loss": 0.4872,
+      "step": 4120
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6067968795720464,
+      "learning_rate": 5.058280529589279e-06,
+      "loss": 0.4697,
+      "step": 4121
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3858276131809577,
+      "learning_rate": 5.056270942520986e-06,
+      "loss": 0.5071,
+      "step": 4122
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5759322132167428,
+      "learning_rate": 5.054261346361628e-06,
+      "loss": 0.5134,
+      "step": 4123
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7641683615692467,
+      "learning_rate": 5.0522517414358705e-06,
+      "loss": 0.5418,
+      "step": 4124
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5814241502874156,
+      "learning_rate": 5.050242128068386e-06,
+      "loss": 0.453,
+      "step": 4125
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2912461088706648,
+      "learning_rate": 5.048232506583841e-06,
+      "loss": 0.4845,
+      "step": 4126
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2778937846886074,
+      "learning_rate": 5.046222877306911e-06,
+      "loss": 0.4403,
+      "step": 4127
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1456675490207737,
+      "learning_rate": 5.044213240562268e-06,
+      "loss": 0.5408,
+      "step": 4128
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6392616148549624,
+      "learning_rate": 5.042203596674586e-06,
+      "loss": 0.4769,
+      "step": 4129
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4748949968945793,
+      "learning_rate": 5.040193945968542e-06,
+      "loss": 0.507,
+      "step": 4130
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.385296056790979,
+      "learning_rate": 5.038184288768813e-06,
+      "loss": 0.5147,
+      "step": 4131
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.812264408884099,
+      "learning_rate": 5.036174625400073e-06,
+      "loss": 0.5232,
+      "step": 4132
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2807690983806912,
+      "learning_rate": 5.034164956187006e-06,
+      "loss": 0.53,
+      "step": 4133
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4308465971303546,
+      "learning_rate": 5.032155281454288e-06,
+      "loss": 0.5092,
+      "step": 4134
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2571579009281904,
+      "learning_rate": 5.030145601526603e-06,
+      "loss": 0.5141,
+      "step": 4135
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4038444686476588,
+      "learning_rate": 5.028135916728628e-06,
+      "loss": 0.4968,
+      "step": 4136
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5252270252095017,
+      "learning_rate": 5.02612622738505e-06,
+      "loss": 0.4903,
+      "step": 4137
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8263454315055303,
+      "learning_rate": 5.024116533820549e-06,
+      "loss": 0.5004,
+      "step": 4138
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1238635399425574,
+      "learning_rate": 5.02210683635981e-06,
+      "loss": 0.516,
+      "step": 4139
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.491279320066013,
+      "learning_rate": 5.020097135327515e-06,
+      "loss": 0.4862,
+      "step": 4140
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3561697321845458,
+      "learning_rate": 5.018087431048353e-06,
+      "loss": 0.4839,
+      "step": 4141
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.375828041356734,
+      "learning_rate": 5.016077723847006e-06,
+      "loss": 0.5209,
+      "step": 4142
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5082826103778804,
+      "learning_rate": 5.0140680140481625e-06,
+      "loss": 0.546,
+      "step": 4143
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4036655538213862,
+      "learning_rate": 5.012058301976505e-06,
+      "loss": 0.5226,
+      "step": 4144
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3933326710765965,
+      "learning_rate": 5.010048587956724e-06,
+      "loss": 0.4945,
+      "step": 4145
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.506118935176337,
+      "learning_rate": 5.008038872313506e-06,
+      "loss": 0.5511,
+      "step": 4146
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5381169525014873,
+      "learning_rate": 5.006029155371538e-06,
+      "loss": 0.5581,
+      "step": 4147
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2389131059981193,
+      "learning_rate": 5.004019437455504e-06,
+      "loss": 0.4614,
+      "step": 4148
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3076621463048856,
+      "learning_rate": 5.0020097188900965e-06,
+      "loss": 0.5189,
+      "step": 4149
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5725310274284423,
+      "learning_rate": 5e-06,
+      "loss": 0.469,
+      "step": 4150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4551892770544779,
+      "learning_rate": 4.997990281109905e-06,
+      "loss": 0.5506,
+      "step": 4151
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5266131165054042,
+      "learning_rate": 4.995980562544497e-06,
+      "loss": 0.5118,
+      "step": 4152
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4304922054207752,
+      "learning_rate": 4.993970844628464e-06,
+      "loss": 0.5216,
+      "step": 4153
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3661929147010907,
+      "learning_rate": 4.9919611276864956e-06,
+      "loss": 0.4755,
+      "step": 4154
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5063251153366095,
+      "learning_rate": 4.989951412043276e-06,
+      "loss": 0.4918,
+      "step": 4155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3266955956864537,
+      "learning_rate": 4.987941698023495e-06,
+      "loss": 0.5007,
+      "step": 4156
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.594738069327744,
+      "learning_rate": 4.985931985951839e-06,
+      "loss": 0.5322,
+      "step": 4157
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3599941087299574,
+      "learning_rate": 4.983922276152995e-06,
+      "loss": 0.5543,
+      "step": 4158
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6960652624546686,
+      "learning_rate": 4.981912568951649e-06,
+      "loss": 0.4587,
+      "step": 4159
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6547482077030211,
+      "learning_rate": 4.979902864672486e-06,
+      "loss": 0.5651,
+      "step": 4160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4706748716930576,
+      "learning_rate": 4.977893163640193e-06,
+      "loss": 0.4898,
+      "step": 4161
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4434440513618652,
+      "learning_rate": 4.975883466179453e-06,
+      "loss": 0.4831,
+      "step": 4162
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9645279976236119,
+      "learning_rate": 4.973873772614952e-06,
+      "loss": 0.4389,
+      "step": 4163
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7249626792549597,
+      "learning_rate": 4.9718640832713725e-06,
+      "loss": 0.5114,
+      "step": 4164
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7902153666586336,
+      "learning_rate": 4.9698543984733995e-06,
+      "loss": 0.5163,
+      "step": 4165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.784462551091202,
+      "learning_rate": 4.967844718545713e-06,
+      "loss": 0.4856,
+      "step": 4166
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.968567016437116,
+      "learning_rate": 4.965835043812996e-06,
+      "loss": 0.5301,
+      "step": 4167
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.415022168086972,
+      "learning_rate": 4.963825374599929e-06,
+      "loss": 0.495,
+      "step": 4168
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6551021929136456,
+      "learning_rate": 4.96181571123119e-06,
+      "loss": 0.4626,
+      "step": 4169
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2770436177440663,
+      "learning_rate": 4.959806054031459e-06,
+      "loss": 0.4402,
+      "step": 4170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4890483786874007,
+      "learning_rate": 4.957796403325415e-06,
+      "loss": 0.535,
+      "step": 4171
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6292700008320747,
+      "learning_rate": 4.955786759437733e-06,
+      "loss": 0.4561,
+      "step": 4172
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3794138507763596,
+      "learning_rate": 4.9537771226930895e-06,
+      "loss": 0.5358,
+      "step": 4173
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2242660546252222,
+      "learning_rate": 4.9517674934161595e-06,
+      "loss": 0.4633,
+      "step": 4174
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.644565449493744,
+      "learning_rate": 4.949757871931616e-06,
+      "loss": 0.4738,
+      "step": 4175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5253593098849336,
+      "learning_rate": 4.94774825856413e-06,
+      "loss": 0.5339,
+      "step": 4176
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.691103468736044,
+      "learning_rate": 4.945738653638374e-06,
+      "loss": 0.4447,
+      "step": 4177
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3440088062239994,
+      "learning_rate": 4.943729057479016e-06,
+      "loss": 0.5049,
+      "step": 4178
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.55358748969524,
+      "learning_rate": 4.941719470410722e-06,
+      "loss": 0.5487,
+      "step": 4179
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6662063178996205,
+      "learning_rate": 4.9397098927581625e-06,
+      "loss": 0.5347,
+      "step": 4180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1629719586070864,
+      "learning_rate": 4.937700324846002e-06,
+      "loss": 0.5606,
+      "step": 4181
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7344846885416572,
+      "learning_rate": 4.935690766998902e-06,
+      "loss": 0.5023,
+      "step": 4182
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9817220234093216,
+      "learning_rate": 4.9336812195415256e-06,
+      "loss": 0.452,
+      "step": 4183
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3397533477646433,
+      "learning_rate": 4.931671682798532e-06,
+      "loss": 0.4803,
+      "step": 4184
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7432222684048106,
+      "learning_rate": 4.929662157094579e-06,
+      "loss": 0.5058,
+      "step": 4185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7210622686065686,
+      "learning_rate": 4.9276526427543246e-06,
+      "loss": 0.506,
+      "step": 4186
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4401480942162428,
+      "learning_rate": 4.925643140102421e-06,
+      "loss": 0.4744,
+      "step": 4187
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3887530249461608,
+      "learning_rate": 4.9236336494635245e-06,
+      "loss": 0.4719,
+      "step": 4188
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6669286955765157,
+      "learning_rate": 4.921624171162285e-06,
+      "loss": 0.5595,
+      "step": 4189
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3289537902911408,
+      "learning_rate": 4.919614705523352e-06,
+      "loss": 0.5222,
+      "step": 4190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6561483477939754,
+      "learning_rate": 4.91760525287137e-06,
+      "loss": 0.4604,
+      "step": 4191
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2549264933188877,
+      "learning_rate": 4.915595813530985e-06,
+      "loss": 0.4916,
+      "step": 4192
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.337661073623808,
+      "learning_rate": 4.913586387826839e-06,
+      "loss": 0.5058,
+      "step": 4193
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7660930060464695,
+      "learning_rate": 4.911576976083574e-06,
+      "loss": 0.4937,
+      "step": 4194
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4433537947099384,
+      "learning_rate": 4.909567578625828e-06,
+      "loss": 0.5052,
+      "step": 4195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5505560786292858,
+      "learning_rate": 4.907558195778233e-06,
+      "loss": 0.4929,
+      "step": 4196
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0204960000285306,
+      "learning_rate": 4.905548827865428e-06,
+      "loss": 0.4991,
+      "step": 4197
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2185001831453082,
+      "learning_rate": 4.903539475212042e-06,
+      "loss": 0.4375,
+      "step": 4198
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3643582538915,
+      "learning_rate": 4.9015301381427024e-06,
+      "loss": 0.4703,
+      "step": 4199
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8839576911566032,
+      "learning_rate": 4.899520816982038e-06,
+      "loss": 0.5241,
+      "step": 4200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3994381417308601,
+      "learning_rate": 4.8975115120546696e-06,
+      "loss": 0.4646,
+      "step": 4201
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.495212091403681,
+      "learning_rate": 4.895502223685219e-06,
+      "loss": 0.5456,
+      "step": 4202
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6878448599306082,
+      "learning_rate": 4.8934929521983045e-06,
+      "loss": 0.5805,
+      "step": 4203
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4437728881272724,
+      "learning_rate": 4.891483697918539e-06,
+      "loss": 0.479,
+      "step": 4204
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7700415243388306,
+      "learning_rate": 4.88947446117054e-06,
+      "loss": 0.4935,
+      "step": 4205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5026598746843285,
+      "learning_rate": 4.887465242278915e-06,
+      "loss": 0.5104,
+      "step": 4206
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2791913508237092,
+      "learning_rate": 4.885456041568272e-06,
+      "loss": 0.4952,
+      "step": 4207
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3321160626968813,
+      "learning_rate": 4.883446859363215e-06,
+      "loss": 0.5013,
+      "step": 4208
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6531483540609091,
+      "learning_rate": 4.881437695988344e-06,
+      "loss": 0.5052,
+      "step": 4209
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8734086791695874,
+      "learning_rate": 4.8794285517682565e-06,
+      "loss": 0.5431,
+      "step": 4210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6352629459090195,
+      "learning_rate": 4.877419427027548e-06,
+      "loss": 0.4828,
+      "step": 4211
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.498449057211703,
+      "learning_rate": 4.87541032209081e-06,
+      "loss": 0.5139,
+      "step": 4212
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4920133056165215,
+      "learning_rate": 4.873401237282634e-06,
+      "loss": 0.5033,
+      "step": 4213
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4502112098248143,
+      "learning_rate": 4.8713921729276015e-06,
+      "loss": 0.5103,
+      "step": 4214
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6614611334638816,
+      "learning_rate": 4.869383129350297e-06,
+      "loss": 0.4647,
+      "step": 4215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7022236557980238,
+      "learning_rate": 4.867374106875298e-06,
+      "loss": 0.4935,
+      "step": 4216
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7900496633745702,
+      "learning_rate": 4.86536510582718e-06,
+      "loss": 0.5193,
+      "step": 4217
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3757970981568506,
+      "learning_rate": 4.8633561265305156e-06,
+      "loss": 0.4933,
+      "step": 4218
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.868217808906103,
+      "learning_rate": 4.8613471693098724e-06,
+      "loss": 0.5326,
+      "step": 4219
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2677455892833875,
+      "learning_rate": 4.859338234489813e-06,
+      "loss": 0.5134,
+      "step": 4220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3549139939818315,
+      "learning_rate": 4.857329322394902e-06,
+      "loss": 0.5593,
+      "step": 4221
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7574557695669282,
+      "learning_rate": 4.8553204333496965e-06,
+      "loss": 0.5168,
+      "step": 4222
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4457527589125292,
+      "learning_rate": 4.853311567678748e-06,
+      "loss": 0.5191,
+      "step": 4223
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4256473720212115,
+      "learning_rate": 4.8513027257066085e-06,
+      "loss": 0.4653,
+      "step": 4224
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6287755213141875,
+      "learning_rate": 4.849293907757823e-06,
+      "loss": 0.5237,
+      "step": 4225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.655830538154988,
+      "learning_rate": 4.847285114156934e-06,
+      "loss": 0.5446,
+      "step": 4226
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4664096866038583,
+      "learning_rate": 4.84527634522848e-06,
+      "loss": 0.5061,
+      "step": 4227
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6320403680583409,
+      "learning_rate": 4.843267601296994e-06,
+      "loss": 0.5454,
+      "step": 4228
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.450134674129634,
+      "learning_rate": 4.8412588826870075e-06,
+      "loss": 0.5739,
+      "step": 4229
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6707231135387525,
+      "learning_rate": 4.839250189723048e-06,
+      "loss": 0.5417,
+      "step": 4230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1014948435593066,
+      "learning_rate": 4.8372415227296355e-06,
+      "loss": 0.4824,
+      "step": 4231
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5850577118025315,
+      "learning_rate": 4.835232882031288e-06,
+      "loss": 0.5486,
+      "step": 4232
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.363920654716586,
+      "learning_rate": 4.83322426795252e-06,
+      "loss": 0.5245,
+      "step": 4233
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2817202214537684,
+      "learning_rate": 4.8312156808178405e-06,
+      "loss": 0.5268,
+      "step": 4234
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5467511661175979,
+      "learning_rate": 4.829207120951754e-06,
+      "loss": 0.4896,
+      "step": 4235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.147514907824803,
+      "learning_rate": 4.827198588678761e-06,
+      "loss": 0.4676,
+      "step": 4236
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.349527074343018,
+      "learning_rate": 4.825190084323358e-06,
+      "loss": 0.5278,
+      "step": 4237
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2386225759661265,
+      "learning_rate": 4.823181608210036e-06,
+      "loss": 0.4553,
+      "step": 4238
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3773870018897933,
+      "learning_rate": 4.821173160663284e-06,
+      "loss": 0.4915,
+      "step": 4239
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.337943712069457,
+      "learning_rate": 4.819164742007582e-06,
+      "loss": 0.4422,
+      "step": 4240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.50843833816639,
+      "learning_rate": 4.817156352567409e-06,
+      "loss": 0.4631,
+      "step": 4241
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.888360839786836,
+      "learning_rate": 4.815147992667237e-06,
+      "loss": 0.4285,
+      "step": 4242
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4492756322046154,
+      "learning_rate": 4.813139662631535e-06,
+      "loss": 0.5241,
+      "step": 4243
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.597220166897354,
+      "learning_rate": 4.811131362784766e-06,
+      "loss": 0.4924,
+      "step": 4244
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6341422036164526,
+      "learning_rate": 4.809123093451388e-06,
+      "loss": 0.4793,
+      "step": 4245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.381380178994785,
+      "learning_rate": 4.807114854955856e-06,
+      "loss": 0.5316,
+      "step": 4246
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6357430017650092,
+      "learning_rate": 4.805106647622619e-06,
+      "loss": 0.5142,
+      "step": 4247
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.323250740425993,
+      "learning_rate": 4.803098471776119e-06,
+      "loss": 0.5756,
+      "step": 4248
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2960946149258648,
+      "learning_rate": 4.801090327740795e-06,
+      "loss": 0.517,
+      "step": 4249
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3783769541175654,
+      "learning_rate": 4.799082215841081e-06,
+      "loss": 0.4731,
+      "step": 4250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4238922609555158,
+      "learning_rate": 4.797074136401403e-06,
+      "loss": 0.5305,
+      "step": 4251
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3187213899984527,
+      "learning_rate": 4.795066089746187e-06,
+      "loss": 0.5133,
+      "step": 4252
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5319986153504062,
+      "learning_rate": 4.793058076199847e-06,
+      "loss": 0.4949,
+      "step": 4253
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3313211380770655,
+      "learning_rate": 4.791050096086799e-06,
+      "loss": 0.5175,
+      "step": 4254
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.661121794113625,
+      "learning_rate": 4.789042149731448e-06,
+      "loss": 0.5314,
+      "step": 4255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8520098290199518,
+      "learning_rate": 4.787034237458195e-06,
+      "loss": 0.5665,
+      "step": 4256
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4383615798536171,
+      "learning_rate": 4.785026359591438e-06,
+      "loss": 0.4781,
+      "step": 4257
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.36015549004612,
+      "learning_rate": 4.783018516455565e-06,
+      "loss": 0.5016,
+      "step": 4258
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8240998590538453,
+      "learning_rate": 4.781010708374963e-06,
+      "loss": 0.5126,
+      "step": 4259
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4355718796301644,
+      "learning_rate": 4.779002935674008e-06,
+      "loss": 0.5586,
+      "step": 4260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7266807431996253,
+      "learning_rate": 4.776995198677075e-06,
+      "loss": 0.5162,
+      "step": 4261
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2775623472876219,
+      "learning_rate": 4.774987497708533e-06,
+      "loss": 0.4685,
+      "step": 4262
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.307364784532959,
+      "learning_rate": 4.7729798330927415e-06,
+      "loss": 0.4875,
+      "step": 4263
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.393824074495462,
+      "learning_rate": 4.770972205154058e-06,
+      "loss": 0.4782,
+      "step": 4264
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7554410138205054,
+      "learning_rate": 4.768964614216831e-06,
+      "loss": 0.4668,
+      "step": 4265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3652602730204615,
+      "learning_rate": 4.7669570606054066e-06,
+      "loss": 0.5051,
+      "step": 4266
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4637991891015811,
+      "learning_rate": 4.764949544644121e-06,
+      "loss": 0.5407,
+      "step": 4267
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2055380948074947,
+      "learning_rate": 4.7629420666573065e-06,
+      "loss": 0.4802,
+      "step": 4268
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3646970283412756,
+      "learning_rate": 4.760934626969289e-06,
+      "loss": 0.464,
+      "step": 4269
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2475512160334674,
+      "learning_rate": 4.7589272259043875e-06,
+      "loss": 0.4824,
+      "step": 4270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.382321088190782,
+      "learning_rate": 4.756919863786916e-06,
+      "loss": 0.4819,
+      "step": 4271
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2062445168984832,
+      "learning_rate": 4.754912540941182e-06,
+      "loss": 0.5103,
+      "step": 4272
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3685375269870788,
+      "learning_rate": 4.752905257691485e-06,
+      "loss": 0.4959,
+      "step": 4273
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6638309922683546,
+      "learning_rate": 4.7508980143621205e-06,
+      "loss": 0.4928,
+      "step": 4274
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4235012573773838,
+      "learning_rate": 4.7488908112773755e-06,
+      "loss": 0.4925,
+      "step": 4275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6268177268925723,
+      "learning_rate": 4.746883648761531e-06,
+      "loss": 0.4982,
+      "step": 4276
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4566520518954882,
+      "learning_rate": 4.744876527138863e-06,
+      "loss": 0.5318,
+      "step": 4277
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4686807392611485,
+      "learning_rate": 4.742869446733636e-06,
+      "loss": 0.4866,
+      "step": 4278
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3052234583127929,
+      "learning_rate": 4.740862407870118e-06,
+      "loss": 0.5233,
+      "step": 4279
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4928138119168188,
+      "learning_rate": 4.7388554108725594e-06,
+      "loss": 0.4965,
+      "step": 4280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3119229364651537,
+      "learning_rate": 4.7368484560652085e-06,
+      "loss": 0.4925,
+      "step": 4281
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8819516435959767,
+      "learning_rate": 4.734841543772308e-06,
+      "loss": 0.5072,
+      "step": 4282
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3552684872325136,
+      "learning_rate": 4.732834674318091e-06,
+      "loss": 0.4941,
+      "step": 4283
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.502677241539126,
+      "learning_rate": 4.7308278480267865e-06,
+      "loss": 0.4369,
+      "step": 4284
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.28365550736728,
+      "learning_rate": 4.728821065222612e-06,
+      "loss": 0.4956,
+      "step": 4285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4078836243131156,
+      "learning_rate": 4.726814326229781e-06,
+      "loss": 0.4885,
+      "step": 4286
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3401141295793377,
+      "learning_rate": 4.724807631372505e-06,
+      "loss": 0.4562,
+      "step": 4287
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6390815902227274,
+      "learning_rate": 4.722800980974979e-06,
+      "loss": 0.4953,
+      "step": 4288
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6719051916157315,
+      "learning_rate": 4.720794375361397e-06,
+      "loss": 0.4831,
+      "step": 4289
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3668822128904665,
+      "learning_rate": 4.718787814855942e-06,
+      "loss": 0.509,
+      "step": 4290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2587487927128795,
+      "learning_rate": 4.716781299782791e-06,
+      "loss": 0.4483,
+      "step": 4291
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4839484448291642,
+      "learning_rate": 4.714774830466116e-06,
+      "loss": 0.5056,
+      "step": 4292
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6584679793549012,
+      "learning_rate": 4.71276840723008e-06,
+      "loss": 0.5071,
+      "step": 4293
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4562233356371959,
+      "learning_rate": 4.7107620303988335e-06,
+      "loss": 0.5153,
+      "step": 4294
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2737569846521757,
+      "learning_rate": 4.708755700296532e-06,
+      "loss": 0.4427,
+      "step": 4295
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3727841747162295,
+      "learning_rate": 4.706749417247312e-06,
+      "loss": 0.4871,
+      "step": 4296
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5004984388484188,
+      "learning_rate": 4.704743181575306e-06,
+      "loss": 0.4623,
+      "step": 4297
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4457080763319423,
+      "learning_rate": 4.7027369936046415e-06,
+      "loss": 0.4888,
+      "step": 4298
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.476081348114596,
+      "learning_rate": 4.700730853659432e-06,
+      "loss": 0.4998,
+      "step": 4299
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3457600490722774,
+      "learning_rate": 4.698724762063789e-06,
+      "loss": 0.4999,
+      "step": 4300
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5710943313782562,
+      "learning_rate": 4.696718719141813e-06,
+      "loss": 0.5094,
+      "step": 4301
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.023508681986042,
+      "learning_rate": 4.694712725217598e-06,
+      "loss": 0.5668,
+      "step": 4302
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6785659078411073,
+      "learning_rate": 4.692706780615232e-06,
+      "loss": 0.5215,
+      "step": 4303
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3301372259571647,
+      "learning_rate": 4.690700885658793e-06,
+      "loss": 0.4892,
+      "step": 4304
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.942794004902333,
+      "learning_rate": 4.68869504067235e-06,
+      "loss": 0.5059,
+      "step": 4305
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3480045867939938,
+      "learning_rate": 4.686689245979965e-06,
+      "loss": 0.5091,
+      "step": 4306
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4420165452073062,
+      "learning_rate": 4.68468350190569e-06,
+      "loss": 0.48,
+      "step": 4307
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4199701993205618,
+      "learning_rate": 4.682677808773576e-06,
+      "loss": 0.5201,
+      "step": 4308
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2548107192351894,
+      "learning_rate": 4.680672166907654e-06,
+      "loss": 0.4661,
+      "step": 4309
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5109225918516362,
+      "learning_rate": 4.678666576631956e-06,
+      "loss": 0.5373,
+      "step": 4310
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2714281680666961,
+      "learning_rate": 4.676661038270501e-06,
+      "loss": 0.5001,
+      "step": 4311
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3277386522117836,
+      "learning_rate": 4.674655552147305e-06,
+      "loss": 0.5494,
+      "step": 4312
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4401191668265765,
+      "learning_rate": 4.6726501185863694e-06,
+      "loss": 0.4942,
+      "step": 4313
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3129932472587458,
+      "learning_rate": 4.67064473791169e-06,
+      "loss": 0.5189,
+      "step": 4314
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5058725829686856,
+      "learning_rate": 4.668639410447255e-06,
+      "loss": 0.4832,
+      "step": 4315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5243590945691707,
+      "learning_rate": 4.666634136517041e-06,
+      "loss": 0.4935,
+      "step": 4316
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.614027731511778,
+      "learning_rate": 4.664628916445018e-06,
+      "loss": 0.547,
+      "step": 4317
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7247810144381126,
+      "learning_rate": 4.662623750555149e-06,
+      "loss": 0.5399,
+      "step": 4318
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5402121533125563,
+      "learning_rate": 4.6606186391713805e-06,
+      "loss": 0.4774,
+      "step": 4319
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.282588310149704,
+      "learning_rate": 4.6586135826176625e-06,
+      "loss": 0.533,
+      "step": 4320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6803085872164045,
+      "learning_rate": 4.6566085812179265e-06,
+      "loss": 0.5164,
+      "step": 4321
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3154255919996705,
+      "learning_rate": 4.654603635296098e-06,
+      "loss": 0.5056,
+      "step": 4322
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.266041772416684,
+      "learning_rate": 4.652598745176095e-06,
+      "loss": 0.4811,
+      "step": 4323
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3636206626837686,
+      "learning_rate": 4.6505939111818246e-06,
+      "loss": 0.4925,
+      "step": 4324
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.478486601519347,
+      "learning_rate": 4.648589133637185e-06,
+      "loss": 0.4737,
+      "step": 4325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5515659815698486,
+      "learning_rate": 4.646584412866065e-06,
+      "loss": 0.4896,
+      "step": 4326
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2139131098187366,
+      "learning_rate": 4.644579749192346e-06,
+      "loss": 0.4533,
+      "step": 4327
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3647906625218897,
+      "learning_rate": 4.642575142939898e-06,
+      "loss": 0.5035,
+      "step": 4328
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2970981149774912,
+      "learning_rate": 4.640570594432586e-06,
+      "loss": 0.5153,
+      "step": 4329
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4040798471253462,
+      "learning_rate": 4.638566103994258e-06,
+      "loss": 0.552,
+      "step": 4330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6276214042065345,
+      "learning_rate": 4.63656167194876e-06,
+      "loss": 0.501,
+      "step": 4331
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5530490248697577,
+      "learning_rate": 4.634557298619924e-06,
+      "loss": 0.5374,
+      "step": 4332
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3407778756048003,
+      "learning_rate": 4.632552984331576e-06,
+      "loss": 0.4578,
+      "step": 4333
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4791583074513865,
+      "learning_rate": 4.630548729407529e-06,
+      "loss": 0.5266,
+      "step": 4334
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4249788214778287,
+      "learning_rate": 4.6285445341715875e-06,
+      "loss": 0.4824,
+      "step": 4335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2562567504953168,
+      "learning_rate": 4.626540398947549e-06,
+      "loss": 0.4949,
+      "step": 4336
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.321673506920769,
+      "learning_rate": 4.624536324059199e-06,
+      "loss": 0.51,
+      "step": 4337
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7009338899251323,
+      "learning_rate": 4.622532309830312e-06,
+      "loss": 0.5161,
+      "step": 4338
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3804748517386487,
+      "learning_rate": 4.620528356584655e-06,
+      "loss": 0.5386,
+      "step": 4339
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2710254221528257,
+      "learning_rate": 4.6185244646459835e-06,
+      "loss": 0.4611,
+      "step": 4340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8608849766501319,
+      "learning_rate": 4.616520634338045e-06,
+      "loss": 0.471,
+      "step": 4341
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6798704097529361,
+      "learning_rate": 4.614516865984575e-06,
+      "loss": 0.4846,
+      "step": 4342
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1922263872791192,
+      "learning_rate": 4.6125131599092995e-06,
+      "loss": 0.4922,
+      "step": 4343
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.409966446941975,
+      "learning_rate": 4.610509516435937e-06,
+      "loss": 0.542,
+      "step": 4344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2967390492106146,
+      "learning_rate": 4.608505935888192e-06,
+      "loss": 0.5277,
+      "step": 4345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4520281467958223,
+      "learning_rate": 4.606502418589762e-06,
+      "loss": 0.4888,
+      "step": 4346
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5444683940931594,
+      "learning_rate": 4.604498964864331e-06,
+      "loss": 0.5162,
+      "step": 4347
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4585220753140886,
+      "learning_rate": 4.6024955750355755e-06,
+      "loss": 0.4734,
+      "step": 4348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.054117529489168,
+      "learning_rate": 4.600492249427161e-06,
+      "loss": 0.5057,
+      "step": 4349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7080318860903294,
+      "learning_rate": 4.598488988362742e-06,
+      "loss": 0.4853,
+      "step": 4350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5706453071719417,
+      "learning_rate": 4.5964857921659635e-06,
+      "loss": 0.5186,
+      "step": 4351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3140820055921787,
+      "learning_rate": 4.594482661160458e-06,
+      "loss": 0.5169,
+      "step": 4352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2680956481661165,
+      "learning_rate": 4.59247959566985e-06,
+      "loss": 0.4936,
+      "step": 4353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3409682132801277,
+      "learning_rate": 4.5904765960177535e-06,
+      "loss": 0.5418,
+      "step": 4354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.310893595658717,
+      "learning_rate": 4.58847366252777e-06,
+      "loss": 0.495,
+      "step": 4355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5131214786662126,
+      "learning_rate": 4.58647079552349e-06,
+      "loss": 0.4563,
+      "step": 4356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.826228307832778,
+      "learning_rate": 4.5844679953284946e-06,
+      "loss": 0.5028,
+      "step": 4357
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.320033672212789,
+      "learning_rate": 4.582465262266355e-06,
+      "loss": 0.4299,
+      "step": 4358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5183627776360842,
+      "learning_rate": 4.58046259666063e-06,
+      "loss": 0.4833,
+      "step": 4359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4938905628289831,
+      "learning_rate": 4.5784599988348656e-06,
+      "loss": 0.5108,
+      "step": 4360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6296818643660147,
+      "learning_rate": 4.576457469112602e-06,
+      "loss": 0.5559,
+      "step": 4361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7693617181056263,
+      "learning_rate": 4.574455007817365e-06,
+      "loss": 0.5516,
+      "step": 4362
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3295911623522392,
+      "learning_rate": 4.57245261527267e-06,
+      "loss": 0.522,
+      "step": 4363
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5954330240640078,
+      "learning_rate": 4.570450291802019e-06,
+      "loss": 0.5147,
+      "step": 4364
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7853840557710536,
+      "learning_rate": 4.568448037728907e-06,
+      "loss": 0.5434,
+      "step": 4365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4879695205979435,
+      "learning_rate": 4.5664458533768155e-06,
+      "loss": 0.4781,
+      "step": 4366
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4936270775797755,
+      "learning_rate": 4.564443739069215e-06,
+      "loss": 0.5157,
+      "step": 4367
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.710118172072937,
+      "learning_rate": 4.562441695129563e-06,
+      "loss": 0.4956,
+      "step": 4368
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.479920114691241,
+      "learning_rate": 4.56043972188131e-06,
+      "loss": 0.5305,
+      "step": 4369
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4216813600328682,
+      "learning_rate": 4.558437819647892e-06,
+      "loss": 0.519,
+      "step": 4370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6207893049711012,
+      "learning_rate": 4.556435988752732e-06,
+      "loss": 0.4515,
+      "step": 4371
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7293017674348017,
+      "learning_rate": 4.554434229519244e-06,
+      "loss": 0.5257,
+      "step": 4372
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4846186532957242,
+      "learning_rate": 4.552432542270832e-06,
+      "loss": 0.4847,
+      "step": 4373
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4686377152136705,
+      "learning_rate": 4.550430927330885e-06,
+      "loss": 0.4727,
+      "step": 4374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0718575292030823,
+      "learning_rate": 4.54842938502278e-06,
+      "loss": 0.5046,
+      "step": 4375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5944676215700275,
+      "learning_rate": 4.546427915669882e-06,
+      "loss": 0.4735,
+      "step": 4376
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4081743207529693,
+      "learning_rate": 4.5444265195955525e-06,
+      "loss": 0.5041,
+      "step": 4377
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.484577775781484,
+      "learning_rate": 4.542425197123131e-06,
+      "loss": 0.539,
+      "step": 4378
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6249081818741813,
+      "learning_rate": 4.540423948575949e-06,
+      "loss": 0.4749,
+      "step": 4379
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2982887830474357,
+      "learning_rate": 4.538422774277325e-06,
+      "loss": 0.5252,
+      "step": 4380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3766822741656382,
+      "learning_rate": 4.536421674550567e-06,
+      "loss": 0.5047,
+      "step": 4381
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.406652765077386,
+      "learning_rate": 4.534420649718972e-06,
+      "loss": 0.4652,
+      "step": 4382
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6791428428602214,
+      "learning_rate": 4.532419700105819e-06,
+      "loss": 0.5047,
+      "step": 4383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9436755327984607,
+      "learning_rate": 4.53041882603438e-06,
+      "loss": 0.4829,
+      "step": 4384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5698938749165816,
+      "learning_rate": 4.528418027827918e-06,
+      "loss": 0.4655,
+      "step": 4385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7356508093213097,
+      "learning_rate": 4.526417305809677e-06,
+      "loss": 0.5052,
+      "step": 4386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.174600104406179,
+      "learning_rate": 4.5244166603028915e-06,
+      "loss": 0.4722,
+      "step": 4387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5507075320836625,
+      "learning_rate": 4.522416091630784e-06,
+      "loss": 0.4298,
+      "step": 4388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.771951957318884,
+      "learning_rate": 4.520415600116561e-06,
+      "loss": 0.4666,
+      "step": 4389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4399449863102074,
+      "learning_rate": 4.518415186083422e-06,
+      "loss": 0.4913,
+      "step": 4390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3278433017347138,
+      "learning_rate": 4.516414849854552e-06,
+      "loss": 0.5082,
+      "step": 4391
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4678291151658778,
+      "learning_rate": 4.514414591753121e-06,
+      "loss": 0.5698,
+      "step": 4392
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5435085395494517,
+      "learning_rate": 4.512414412102288e-06,
+      "loss": 0.5142,
+      "step": 4393
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.556647391808432,
+      "learning_rate": 4.510414311225203e-06,
+      "loss": 0.4913,
+      "step": 4394
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6916292073049486,
+      "learning_rate": 4.508414289444998e-06,
+      "loss": 0.4798,
+      "step": 4395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6692120312743866,
+      "learning_rate": 4.506414347084793e-06,
+      "loss": 0.5023,
+      "step": 4396
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.604330736759279,
+      "learning_rate": 4.5044144844676995e-06,
+      "loss": 0.4997,
+      "step": 4397
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.118799793257604,
+      "learning_rate": 4.50241470191681e-06,
+      "loss": 0.5001,
+      "step": 4398
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3528728363724072,
+      "learning_rate": 4.500414999755207e-06,
+      "loss": 0.4631,
+      "step": 4399
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5128477685762356,
+      "learning_rate": 4.498415378305961e-06,
+      "loss": 0.483,
+      "step": 4400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.454443965715015,
+      "learning_rate": 4.496415837892125e-06,
+      "loss": 0.4791,
+      "step": 4401
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3334002112817895,
+      "learning_rate": 4.494416378836749e-06,
+      "loss": 0.4398,
+      "step": 4402
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7095351293567198,
+      "learning_rate": 4.492417001462859e-06,
+      "loss": 0.5341,
+      "step": 4403
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.562988755978761,
+      "learning_rate": 4.490417706093473e-06,
+      "loss": 0.5465,
+      "step": 4404
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4290886758069221,
+      "learning_rate": 4.4884184930515955e-06,
+      "loss": 0.4642,
+      "step": 4405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.65549340328514,
+      "learning_rate": 4.486419362660214e-06,
+      "loss": 0.4921,
+      "step": 4406
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6983782171234239,
+      "learning_rate": 4.484420315242311e-06,
+      "loss": 0.48,
+      "step": 4407
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6234618539161976,
+      "learning_rate": 4.482421351120845e-06,
+      "loss": 0.5449,
+      "step": 4408
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6357978088500176,
+      "learning_rate": 4.480422470618766e-06,
+      "loss": 0.5332,
+      "step": 4409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9663873905647158,
+      "learning_rate": 4.478423674059015e-06,
+      "loss": 0.5126,
+      "step": 4410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4502008959206971,
+      "learning_rate": 4.476424961764513e-06,
+      "loss": 0.5099,
+      "step": 4411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4117212277671094,
+      "learning_rate": 4.47442633405817e-06,
+      "loss": 0.4995,
+      "step": 4412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.387171017364631,
+      "learning_rate": 4.472427791262881e-06,
+      "loss": 0.4811,
+      "step": 4413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.722836592836977,
+      "learning_rate": 4.470429333701529e-06,
+      "loss": 0.4993,
+      "step": 4414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.046809463793669,
+      "learning_rate": 4.468430961696982e-06,
+      "loss": 0.5041,
+      "step": 4415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4247780429245809,
+      "learning_rate": 4.466432675572096e-06,
+      "loss": 0.4771,
+      "step": 4416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4185856211803107,
+      "learning_rate": 4.464434475649708e-06,
+      "loss": 0.5237,
+      "step": 4417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.468863603545287,
+      "learning_rate": 4.462436362252648e-06,
+      "loss": 0.5028,
+      "step": 4418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5714487313977783,
+      "learning_rate": 4.46043833570373e-06,
+      "loss": 0.5409,
+      "step": 4419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5369611189212409,
+      "learning_rate": 4.4584403963257485e-06,
+      "loss": 0.4821,
+      "step": 4420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5062598221616934,
+      "learning_rate": 4.456442544441493e-06,
+      "loss": 0.4762,
+      "step": 4421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8761573257549888,
+      "learning_rate": 4.45444478037373e-06,
+      "loss": 0.5218,
+      "step": 4422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.780866183894427,
+      "learning_rate": 4.452447104445218e-06,
+      "loss": 0.4763,
+      "step": 4423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0530249281707036,
+      "learning_rate": 4.450449516978699e-06,
+      "loss": 0.507,
+      "step": 4424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.357116791630275,
+      "learning_rate": 4.4484520182969e-06,
+      "loss": 0.5084,
+      "step": 4425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4447708016997076,
+      "learning_rate": 4.4464546087225346e-06,
+      "loss": 0.4825,
+      "step": 4426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4769381978965075,
+      "learning_rate": 4.444457288578303e-06,
+      "loss": 0.4954,
+      "step": 4427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6761864909530553,
+      "learning_rate": 4.44246005818689e-06,
+      "loss": 0.4857,
+      "step": 4428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4748459109991388,
+      "learning_rate": 4.440462917870964e-06,
+      "loss": 0.5239,
+      "step": 4429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3350226162503687,
+      "learning_rate": 4.438465867953182e-06,
+      "loss": 0.5043,
+      "step": 4430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.80295600244093,
+      "learning_rate": 4.4364689087561845e-06,
+      "loss": 0.5169,
+      "step": 4431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.334114055897355,
+      "learning_rate": 4.434472040602599e-06,
+      "loss": 0.5331,
+      "step": 4432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5164473111659773,
+      "learning_rate": 4.432475263815035e-06,
+      "loss": 0.4775,
+      "step": 4433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3005993439552876,
+      "learning_rate": 4.430478578716089e-06,
+      "loss": 0.4822,
+      "step": 4434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.474955584683926,
+      "learning_rate": 4.428481985628345e-06,
+      "loss": 0.5167,
+      "step": 4435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4789894800140106,
+      "learning_rate": 4.426485484874371e-06,
+      "loss": 0.4987,
+      "step": 4436
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3903653813140513,
+      "learning_rate": 4.424489076776718e-06,
+      "loss": 0.5015,
+      "step": 4437
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3625637879056767,
+      "learning_rate": 4.422492761657923e-06,
+      "loss": 0.4851,
+      "step": 4438
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4422805656413178,
+      "learning_rate": 4.420496539840509e-06,
+      "loss": 0.4535,
+      "step": 4439
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.505004277895124,
+      "learning_rate": 4.4185004116469824e-06,
+      "loss": 0.4876,
+      "step": 4440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.293316550344579,
+      "learning_rate": 4.416504377399835e-06,
+      "loss": 0.4517,
+      "step": 4441
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4825354853263424,
+      "learning_rate": 4.414508437421544e-06,
+      "loss": 0.5174,
+      "step": 4442
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3531475040207626,
+      "learning_rate": 4.412512592034572e-06,
+      "loss": 0.5488,
+      "step": 4443
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3345359208586365,
+      "learning_rate": 4.410516841561366e-06,
+      "loss": 0.5341,
+      "step": 4444
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7683739150605549,
+      "learning_rate": 4.408521186324356e-06,
+      "loss": 0.4505,
+      "step": 4445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4109196527564847,
+      "learning_rate": 4.406525626645956e-06,
+      "loss": 0.4472,
+      "step": 4446
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4761501607577767,
+      "learning_rate": 4.404530162848569e-06,
+      "loss": 0.4795,
+      "step": 4447
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2703248406744538,
+      "learning_rate": 4.402534795254578e-06,
+      "loss": 0.4719,
+      "step": 4448
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1843683187084255,
+      "learning_rate": 4.4005395241863535e-06,
+      "loss": 0.502,
+      "step": 4449
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0136952093233424,
+      "learning_rate": 4.398544349966247e-06,
+      "loss": 0.5151,
+      "step": 4450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.106552861840512,
+      "learning_rate": 4.3965492729166e-06,
+      "loss": 0.4904,
+      "step": 4451
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6491909140414045,
+      "learning_rate": 4.394554293359731e-06,
+      "loss": 0.554,
+      "step": 4452
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4630506011879885,
+      "learning_rate": 4.392559411617949e-06,
+      "loss": 0.5344,
+      "step": 4453
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.55165485319063,
+      "learning_rate": 4.390564628013545e-06,
+      "loss": 0.5136,
+      "step": 4454
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.094936177696235,
+      "learning_rate": 4.388569942868791e-06,
+      "loss": 0.4817,
+      "step": 4455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9932684063707253,
+      "learning_rate": 4.3865753565059485e-06,
+      "loss": 0.5695,
+      "step": 4456
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8901580466559014,
+      "learning_rate": 4.384580869247259e-06,
+      "loss": 0.4978,
+      "step": 4457
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.505734659711322,
+      "learning_rate": 4.38258648141495e-06,
+      "loss": 0.4256,
+      "step": 4458
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6879853623833825,
+      "learning_rate": 4.380592193331234e-06,
+      "loss": 0.6263,
+      "step": 4459
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3953623635669,
+      "learning_rate": 4.378598005318304e-06,
+      "loss": 0.4724,
+      "step": 4460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7850285873847695,
+      "learning_rate": 4.376603917698339e-06,
+      "loss": 0.4961,
+      "step": 4461
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5945344825899566,
+      "learning_rate": 4.374609930793501e-06,
+      "loss": 0.4908,
+      "step": 4462
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5711748763666404,
+      "learning_rate": 4.372616044925938e-06,
+      "loss": 0.4778,
+      "step": 4463
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.362564324612832,
+      "learning_rate": 4.370622260417777e-06,
+      "loss": 0.4729,
+      "step": 4464
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5883608424991777,
+      "learning_rate": 4.368628577591134e-06,
+      "loss": 0.5014,
+      "step": 4465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4573401950173126,
+      "learning_rate": 4.366634996768104e-06,
+      "loss": 0.491,
+      "step": 4466
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3802325792037728,
+      "learning_rate": 4.364641518270767e-06,
+      "loss": 0.5029,
+      "step": 4467
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1344719427915315,
+      "learning_rate": 4.362648142421191e-06,
+      "loss": 0.4574,
+      "step": 4468
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.369050190285605,
+      "learning_rate": 4.360654869541419e-06,
+      "loss": 0.5084,
+      "step": 4469
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.909973556369964,
+      "learning_rate": 4.358661699953486e-06,
+      "loss": 0.4853,
+      "step": 4470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4003381183520367,
+      "learning_rate": 4.356668633979402e-06,
+      "loss": 0.4647,
+      "step": 4471
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.606974333963992,
+      "learning_rate": 4.354675671941167e-06,
+      "loss": 0.4842,
+      "step": 4472
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2865352907440146,
+      "learning_rate": 4.3526828141607605e-06,
+      "loss": 0.4676,
+      "step": 4473
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.239797259529583,
+      "learning_rate": 4.350690060960146e-06,
+      "loss": 0.4793,
+      "step": 4474
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4018255932426165,
+      "learning_rate": 4.348697412661269e-06,
+      "loss": 0.455,
+      "step": 4475
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2995263446919354,
+      "learning_rate": 4.346704869586064e-06,
+      "loss": 0.4692,
+      "step": 4476
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7508953414723027,
+      "learning_rate": 4.344712432056441e-06,
+      "loss": 0.5329,
+      "step": 4477
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4401002791730046,
+      "learning_rate": 4.342720100394295e-06,
+      "loss": 0.5505,
+      "step": 4478
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.581965730475437,
+      "learning_rate": 4.340727874921506e-06,
+      "loss": 0.534,
+      "step": 4479
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2937840348727583,
+      "learning_rate": 4.338735755959935e-06,
+      "loss": 0.4743,
+      "step": 4480
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4277893209041128,
+      "learning_rate": 4.336743743831426e-06,
+      "loss": 0.4859,
+      "step": 4481
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.727417677604167,
+      "learning_rate": 4.334751838857807e-06,
+      "loss": 0.4857,
+      "step": 4482
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3905935748983391,
+      "learning_rate": 4.332760041360885e-06,
+      "loss": 0.5166,
+      "step": 4483
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4155875374503135,
+      "learning_rate": 4.330768351662458e-06,
+      "loss": 0.505,
+      "step": 4484
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4320288646888266,
+      "learning_rate": 4.328776770084296e-06,
+      "loss": 0.5258,
+      "step": 4485
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2614713949066259,
+      "learning_rate": 4.326785296948162e-06,
+      "loss": 0.46,
+      "step": 4486
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.462031081338204,
+      "learning_rate": 4.324793932575789e-06,
+      "loss": 0.5246,
+      "step": 4487
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4480727488848604,
+      "learning_rate": 4.322802677288904e-06,
+      "loss": 0.5121,
+      "step": 4488
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5817956181131727,
+      "learning_rate": 4.320811531409211e-06,
+      "loss": 0.5003,
+      "step": 4489
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.362465636960399,
+      "learning_rate": 4.318820495258396e-06,
+      "loss": 0.4872,
+      "step": 4490
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4532442998980175,
+      "learning_rate": 4.316829569158127e-06,
+      "loss": 0.5363,
+      "step": 4491
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6472721593024459,
+      "learning_rate": 4.3148387534300615e-06,
+      "loss": 0.5004,
+      "step": 4492
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.560388920149118,
+      "learning_rate": 4.312848048395828e-06,
+      "loss": 0.5634,
+      "step": 4493
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6795898740165551,
+      "learning_rate": 4.310857454377045e-06,
+      "loss": 0.4687,
+      "step": 4494
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7042855561623385,
+      "learning_rate": 4.30886697169531e-06,
+      "loss": 0.4863,
+      "step": 4495
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6835008969281285,
+      "learning_rate": 4.306876600672204e-06,
+      "loss": 0.4789,
+      "step": 4496
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5439431277188411,
+      "learning_rate": 4.3048863416292866e-06,
+      "loss": 0.5109,
+      "step": 4497
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2984059818710183,
+      "learning_rate": 4.302896194888102e-06,
+      "loss": 0.4721,
+      "step": 4498
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9009395166991934,
+      "learning_rate": 4.300906160770174e-06,
+      "loss": 0.5337,
+      "step": 4499
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5366204918981408,
+      "learning_rate": 4.298916239597016e-06,
+      "loss": 0.5198,
+      "step": 4500
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4234969050909656,
+      "learning_rate": 4.2969264316901135e-06,
+      "loss": 0.5181,
+      "step": 4501
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8208805018296175,
+      "learning_rate": 4.2949367373709385e-06,
+      "loss": 0.5468,
+      "step": 4502
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4042996491593032,
+      "learning_rate": 4.292947156960942e-06,
+      "loss": 0.5222,
+      "step": 4503
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.763208800475825,
+      "learning_rate": 4.290957690781561e-06,
+      "loss": 0.5531,
+      "step": 4504
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7865389229207818,
+      "learning_rate": 4.28896833915421e-06,
+      "loss": 0.4983,
+      "step": 4505
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.170165864220015,
+      "learning_rate": 4.286979102400286e-06,
+      "loss": 0.4926,
+      "step": 4506
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4323082470024437,
+      "learning_rate": 4.2849899808411665e-06,
+      "loss": 0.5024,
+      "step": 4507
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4889225319510406,
+      "learning_rate": 4.2830009747982115e-06,
+      "loss": 0.4891,
+      "step": 4508
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3382345132730324,
+      "learning_rate": 4.281012084592766e-06,
+      "loss": 0.5222,
+      "step": 4509
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.675762563925517,
+      "learning_rate": 4.27902331054615e-06,
+      "loss": 0.5036,
+      "step": 4510
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6067386368305003,
+      "learning_rate": 4.277034652979668e-06,
+      "loss": 0.4876,
+      "step": 4511
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.611708514685027,
+      "learning_rate": 4.275046112214604e-06,
+      "loss": 0.5042,
+      "step": 4512
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.494644895380749,
+      "learning_rate": 4.273057688572227e-06,
+      "loss": 0.5389,
+      "step": 4513
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5774346058796291,
+      "learning_rate": 4.271069382373783e-06,
+      "loss": 0.4572,
+      "step": 4514
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4256215511273256,
+      "learning_rate": 4.2690811939405e-06,
+      "loss": 0.523,
+      "step": 4515
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3297159375719316,
+      "learning_rate": 4.267093123593585e-06,
+      "loss": 0.4912,
+      "step": 4516
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4035373991816045,
+      "learning_rate": 4.265105171654233e-06,
+      "loss": 0.4572,
+      "step": 4517
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3611163373689483,
+      "learning_rate": 4.263117338443612e-06,
+      "loss": 0.5006,
+      "step": 4518
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4072504029722148,
+      "learning_rate": 4.261129624282876e-06,
+      "loss": 0.4834,
+      "step": 4519
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5658032422045698,
+      "learning_rate": 4.2591420294931565e-06,
+      "loss": 0.5035,
+      "step": 4520
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.823794325382853,
+      "learning_rate": 4.257154554395566e-06,
+      "loss": 0.516,
+      "step": 4521
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6114850530253642,
+      "learning_rate": 4.2551671993112e-06,
+      "loss": 0.483,
+      "step": 4522
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4416077224034098,
+      "learning_rate": 4.253179964561133e-06,
+      "loss": 0.5327,
+      "step": 4523
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.233556266526752,
+      "learning_rate": 4.25119285046642e-06,
+      "loss": 0.5014,
+      "step": 4524
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4926199002802785,
+      "learning_rate": 4.249205857348097e-06,
+      "loss": 0.509,
+      "step": 4525
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5610452174421434,
+      "learning_rate": 4.247218985527179e-06,
+      "loss": 0.5521,
+      "step": 4526
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7556944477904484,
+      "learning_rate": 4.245232235324666e-06,
+      "loss": 0.4896,
+      "step": 4527
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4150439393082945,
+      "learning_rate": 4.243245607061531e-06,
+      "loss": 0.4713,
+      "step": 4528
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3719288884987528,
+      "learning_rate": 4.241259101058734e-06,
+      "loss": 0.4368,
+      "step": 4529
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4296319411129048,
+      "learning_rate": 4.239272717637212e-06,
+      "loss": 0.532,
+      "step": 4530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6677207912248433,
+      "learning_rate": 4.237286457117882e-06,
+      "loss": 0.4865,
+      "step": 4531
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4600157709673596,
+      "learning_rate": 4.2353003198216405e-06,
+      "loss": 0.5526,
+      "step": 4532
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1718828387548506,
+      "learning_rate": 4.233314306069369e-06,
+      "loss": 0.4848,
+      "step": 4533
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2922949432077968,
+      "learning_rate": 4.231328416181923e-06,
+      "loss": 0.4936,
+      "step": 4534
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6398879189413544,
+      "learning_rate": 4.229342650480143e-06,
+      "loss": 0.5041,
+      "step": 4535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.175714480238288,
+      "learning_rate": 4.227357009284843e-06,
+      "loss": 0.4687,
+      "step": 4536
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7546675858146281,
+      "learning_rate": 4.225371492916824e-06,
+      "loss": 0.4849,
+      "step": 4537
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4694551238357199,
+      "learning_rate": 4.223386101696863e-06,
+      "loss": 0.538,
+      "step": 4538
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5947169512599413,
+      "learning_rate": 4.221400835945716e-06,
+      "loss": 0.5196,
+      "step": 4539
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7358708080151768,
+      "learning_rate": 4.2194156959841215e-06,
+      "loss": 0.4918,
+      "step": 4540
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3877507453717872,
+      "learning_rate": 4.217430682132796e-06,
+      "loss": 0.4993,
+      "step": 4541
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2591813594630503,
+      "learning_rate": 4.215445794712436e-06,
+      "loss": 0.5215,
+      "step": 4542
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4205477304663412,
+      "learning_rate": 4.213461034043719e-06,
+      "loss": 0.5416,
+      "step": 4543
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6272911494152797,
+      "learning_rate": 4.211476400447298e-06,
+      "loss": 0.5314,
+      "step": 4544
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4961401717993632,
+      "learning_rate": 4.20949189424381e-06,
+      "loss": 0.527,
+      "step": 4545
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3452851390092455,
+      "learning_rate": 4.207507515753867e-06,
+      "loss": 0.5442,
+      "step": 4546
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.654315036033042,
+      "learning_rate": 4.205523265298066e-06,
+      "loss": 0.5496,
+      "step": 4547
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.346401650183757,
+      "learning_rate": 4.203539143196978e-06,
+      "loss": 0.5013,
+      "step": 4548
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3989470161936295,
+      "learning_rate": 4.201555149771155e-06,
+      "loss": 0.4853,
+      "step": 4549
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3946104740281104,
+      "learning_rate": 4.199571285341131e-06,
+      "loss": 0.4594,
+      "step": 4550
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6578275793558712,
+      "learning_rate": 4.197587550227416e-06,
+      "loss": 0.5044,
+      "step": 4551
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2876289276965982,
+      "learning_rate": 4.195603944750498e-06,
+      "loss": 0.4791,
+      "step": 4552
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5370400838336578,
+      "learning_rate": 4.193620469230848e-06,
+      "loss": 0.537,
+      "step": 4553
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3196531138427288,
+      "learning_rate": 4.191637123988913e-06,
+      "loss": 0.4833,
+      "step": 4554
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2417498417268902,
+      "learning_rate": 4.18965390934512e-06,
+      "loss": 0.4718,
+      "step": 4555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6394348825494676,
+      "learning_rate": 4.187670825619875e-06,
+      "loss": 0.495,
+      "step": 4556
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3141667309696983,
+      "learning_rate": 4.185687873133561e-06,
+      "loss": 0.5267,
+      "step": 4557
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.690376529517557,
+      "learning_rate": 4.1837050522065434e-06,
+      "loss": 0.5207,
+      "step": 4558
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3277323468444318,
+      "learning_rate": 4.181722363159165e-06,
+      "loss": 0.5052,
+      "step": 4559
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4801879039268384,
+      "learning_rate": 4.179739806311746e-06,
+      "loss": 0.4945,
+      "step": 4560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.042987090195881,
+      "learning_rate": 4.177757381984584e-06,
+      "loss": 0.4719,
+      "step": 4561
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4214907749637817,
+      "learning_rate": 4.17577509049796e-06,
+      "loss": 0.5609,
+      "step": 4562
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3860726873778075,
+      "learning_rate": 4.173792932172128e-06,
+      "loss": 0.4199,
+      "step": 4563
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6571916732434602,
+      "learning_rate": 4.171810907327325e-06,
+      "loss": 0.5922,
+      "step": 4564
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.51516415102186,
+      "learning_rate": 4.169829016283762e-06,
+      "loss": 0.5021,
+      "step": 4565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4737232507571463,
+      "learning_rate": 4.167847259361636e-06,
+      "loss": 0.4847,
+      "step": 4566
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.555560655088034,
+      "learning_rate": 4.165865636881113e-06,
+      "loss": 0.5149,
+      "step": 4567
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6523285718797909,
+      "learning_rate": 4.163884149162342e-06,
+      "loss": 0.4794,
+      "step": 4568
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2591279740887935,
+      "learning_rate": 4.161902796525452e-06,
+      "loss": 0.4776,
+      "step": 4569
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.627216635272726,
+      "learning_rate": 4.159921579290546e-06,
+      "loss": 0.5305,
+      "step": 4570
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4736015643294451,
+      "learning_rate": 4.157940497777708e-06,
+      "loss": 0.5281,
+      "step": 4571
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0493794792054776,
+      "learning_rate": 4.155959552306998e-06,
+      "loss": 0.5319,
+      "step": 4572
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7710477417592,
+      "learning_rate": 4.153978743198454e-06,
+      "loss": 0.5209,
+      "step": 4573
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4132708828952816,
+      "learning_rate": 4.151998070772098e-06,
+      "loss": 0.5137,
+      "step": 4574
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3970788930640707,
+      "learning_rate": 4.150017535347922e-06,
+      "loss": 0.5331,
+      "step": 4575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6491027358822592,
+      "learning_rate": 4.148037137245899e-06,
+      "loss": 0.5045,
+      "step": 4576
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.19176300713811,
+      "learning_rate": 4.1460568767859795e-06,
+      "loss": 0.4623,
+      "step": 4577
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6787985574674565,
+      "learning_rate": 4.144076754288093e-06,
+      "loss": 0.4907,
+      "step": 4578
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.327884011769675,
+      "learning_rate": 4.142096770072144e-06,
+      "loss": 0.525,
+      "step": 4579
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6676423532750408,
+      "learning_rate": 4.140116924458018e-06,
+      "loss": 0.5443,
+      "step": 4580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.441471086599204,
+      "learning_rate": 4.138137217765577e-06,
+      "loss": 0.5027,
+      "step": 4581
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5039303844416383,
+      "learning_rate": 4.1361576503146564e-06,
+      "loss": 0.5751,
+      "step": 4582
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.394640112008406,
+      "learning_rate": 4.134178222425077e-06,
+      "loss": 0.5296,
+      "step": 4583
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3152535316930878,
+      "learning_rate": 4.1321989344166315e-06,
+      "loss": 0.5119,
+      "step": 4584
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2997952010293106,
+      "learning_rate": 4.130219786609092e-06,
+      "loss": 0.5026,
+      "step": 4585
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9590631327681363,
+      "learning_rate": 4.128240779322206e-06,
+      "loss": 0.5002,
+      "step": 4586
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8870920809940253,
+      "learning_rate": 4.1262619128757e-06,
+      "loss": 0.4644,
+      "step": 4587
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5428964670246743,
+      "learning_rate": 4.1242831875892755e-06,
+      "loss": 0.5423,
+      "step": 4588
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3085757589141793,
+      "learning_rate": 4.122304603782616e-06,
+      "loss": 0.4827,
+      "step": 4589
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4333341177657333,
+      "learning_rate": 4.120326161775375e-06,
+      "loss": 0.4813,
+      "step": 4590
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6827500733524555,
+      "learning_rate": 4.118347861887193e-06,
+      "loss": 0.4964,
+      "step": 4591
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6543162660363646,
+      "learning_rate": 4.116369704437678e-06,
+      "loss": 0.5188,
+      "step": 4592
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6541812074781275,
+      "learning_rate": 4.1143916897464204e-06,
+      "loss": 0.5051,
+      "step": 4593
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3791053313562698,
+      "learning_rate": 4.112413818132986e-06,
+      "loss": 0.5294,
+      "step": 4594
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3454682491868881,
+      "learning_rate": 4.110436089916915e-06,
+      "loss": 0.5059,
+      "step": 4595
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3218337421022976,
+      "learning_rate": 4.108458505417728e-06,
+      "loss": 0.4583,
+      "step": 4596
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5307112265445209,
+      "learning_rate": 4.1064810649549216e-06,
+      "loss": 0.5283,
+      "step": 4597
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8299476585591976,
+      "learning_rate": 4.104503768847967e-06,
+      "loss": 0.5274,
+      "step": 4598
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.364327794991199,
+      "learning_rate": 4.102526617416317e-06,
+      "loss": 0.4876,
+      "step": 4599
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6838317119053259,
+      "learning_rate": 4.100549610979396e-06,
+      "loss": 0.4948,
+      "step": 4600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5079001057864923,
+      "learning_rate": 4.098572749856607e-06,
+      "loss": 0.489,
+      "step": 4601
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.440960720058185,
+      "learning_rate": 4.09659603436733e-06,
+      "loss": 0.5237,
+      "step": 4602
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6452758498878746,
+      "learning_rate": 4.09461946483092e-06,
+      "loss": 0.4759,
+      "step": 4603
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4792873873986112,
+      "learning_rate": 4.092643041566709e-06,
+      "loss": 0.4775,
+      "step": 4604
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0103496990842897,
+      "learning_rate": 4.090666764894007e-06,
+      "loss": 0.5322,
+      "step": 4605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3916054186675135,
+      "learning_rate": 4.088690635132094e-06,
+      "loss": 0.5386,
+      "step": 4606
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.225499772634962,
+      "learning_rate": 4.0867146526002384e-06,
+      "loss": 0.5163,
+      "step": 4607
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7005086325318728,
+      "learning_rate": 4.084738817617673e-06,
+      "loss": 0.5397,
+      "step": 4608
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.330519772925582,
+      "learning_rate": 4.082763130503613e-06,
+      "loss": 0.4989,
+      "step": 4609
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6195287767062285,
+      "learning_rate": 4.080787591577247e-06,
+      "loss": 0.5027,
+      "step": 4610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6955980481374432,
+      "learning_rate": 4.078812201157743e-06,
+      "loss": 0.468,
+      "step": 4611
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6819958162274402,
+      "learning_rate": 4.0768369595642396e-06,
+      "loss": 0.5542,
+      "step": 4612
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3882961570525039,
+      "learning_rate": 4.074861867115856e-06,
+      "loss": 0.4712,
+      "step": 4613
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.418499308360564,
+      "learning_rate": 4.072886924131685e-06,
+      "loss": 0.5266,
+      "step": 4614
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3503995058949858,
+      "learning_rate": 4.070912130930798e-06,
+      "loss": 0.4793,
+      "step": 4615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6011790859547999,
+      "learning_rate": 4.068937487832239e-06,
+      "loss": 0.4503,
+      "step": 4616
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7033286082288611,
+      "learning_rate": 4.066962995155028e-06,
+      "loss": 0.5097,
+      "step": 4617
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6431022028485887,
+      "learning_rate": 4.064988653218163e-06,
+      "loss": 0.52,
+      "step": 4618
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4719628028092255,
+      "learning_rate": 4.063014462340616e-06,
+      "loss": 0.5023,
+      "step": 4619
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9883852076554343,
+      "learning_rate": 4.061040422841334e-06,
+      "loss": 0.4708,
+      "step": 4620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7468912437684312,
+      "learning_rate": 4.059066535039242e-06,
+      "loss": 0.5219,
+      "step": 4621
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5013487110806898,
+      "learning_rate": 4.057092799253239e-06,
+      "loss": 0.4259,
+      "step": 4622
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.549834919534906,
+      "learning_rate": 4.055119215802196e-06,
+      "loss": 0.6012,
+      "step": 4623
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.181190433647081,
+      "learning_rate": 4.053145785004968e-06,
+      "loss": 0.5504,
+      "step": 4624
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6775103648742493,
+      "learning_rate": 4.051172507180376e-06,
+      "loss": 0.4888,
+      "step": 4625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1712565831079453,
+      "learning_rate": 4.049199382647224e-06,
+      "loss": 0.4742,
+      "step": 4626
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3516211442014003,
+      "learning_rate": 4.0472264117242845e-06,
+      "loss": 0.5361,
+      "step": 4627
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4519730976420815,
+      "learning_rate": 4.045253594730309e-06,
+      "loss": 0.4768,
+      "step": 4628
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4421565763159003,
+      "learning_rate": 4.043280931984025e-06,
+      "loss": 0.4832,
+      "step": 4629
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5819635692994547,
+      "learning_rate": 4.041308423804132e-06,
+      "loss": 0.5221,
+      "step": 4630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.366048129453399,
+      "learning_rate": 4.039336070509305e-06,
+      "loss": 0.4943,
+      "step": 4631
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4460322616649448,
+      "learning_rate": 4.037363872418199e-06,
+      "loss": 0.4793,
+      "step": 4632
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9828714102023244,
+      "learning_rate": 4.035391829849436e-06,
+      "loss": 0.5577,
+      "step": 4633
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7198579219425816,
+      "learning_rate": 4.033419943121619e-06,
+      "loss": 0.5233,
+      "step": 4634
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.427904785615395,
+      "learning_rate": 4.0314482125533235e-06,
+      "loss": 0.5267,
+      "step": 4635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4270553507828687,
+      "learning_rate": 4.029476638463099e-06,
+      "loss": 0.5129,
+      "step": 4636
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3353797022534557,
+      "learning_rate": 4.027505221169471e-06,
+      "loss": 0.4807,
+      "step": 4637
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9051587551980136,
+      "learning_rate": 4.02553396099094e-06,
+      "loss": 0.4763,
+      "step": 4638
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3909718159931581,
+      "learning_rate": 4.023562858245979e-06,
+      "loss": 0.5409,
+      "step": 4639
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4703475909795112,
+      "learning_rate": 4.021591913253039e-06,
+      "loss": 0.4978,
+      "step": 4640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4012920248011926,
+      "learning_rate": 4.019621126330541e-06,
+      "loss": 0.4711,
+      "step": 4641
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6960206540230156,
+      "learning_rate": 4.017650497796886e-06,
+      "loss": 0.5333,
+      "step": 4642
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.567076240933028,
+      "learning_rate": 4.015680027970445e-06,
+      "loss": 0.4917,
+      "step": 4643
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9646219421988003,
+      "learning_rate": 4.013709717169563e-06,
+      "loss": 0.5112,
+      "step": 4644
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3626883688448634,
+      "learning_rate": 4.011739565712564e-06,
+      "loss": 0.4741,
+      "step": 4645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4192552372878515,
+      "learning_rate": 4.009769573917741e-06,
+      "loss": 0.5132,
+      "step": 4646
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5294985071234206,
+      "learning_rate": 4.007799742103365e-06,
+      "loss": 0.4852,
+      "step": 4647
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.103961404045815,
+      "learning_rate": 4.005830070587679e-06,
+      "loss": 0.4851,
+      "step": 4648
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3305211645597752,
+      "learning_rate": 4.003860559688902e-06,
+      "loss": 0.4964,
+      "step": 4649
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4043067447743847,
+      "learning_rate": 4.001891209725224e-06,
+      "loss": 0.5026,
+      "step": 4650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9045569050008961,
+      "learning_rate": 3.999922021014812e-06,
+      "loss": 0.4974,
+      "step": 4651
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.513471288790594,
+      "learning_rate": 3.997952993875805e-06,
+      "loss": 0.4903,
+      "step": 4652
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3190837653397103,
+      "learning_rate": 3.995984128626317e-06,
+      "loss": 0.5218,
+      "step": 4653
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2803737183980128,
+      "learning_rate": 3.9940154255844355e-06,
+      "loss": 0.4901,
+      "step": 4654
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.324017478162515,
+      "learning_rate": 3.992046885068221e-06,
+      "loss": 0.5455,
+      "step": 4655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.405213306853622,
+      "learning_rate": 3.99007850739571e-06,
+      "loss": 0.5239,
+      "step": 4656
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4332496852812633,
+      "learning_rate": 3.988110292884912e-06,
+      "loss": 0.5049,
+      "step": 4657
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5801465841808873,
+      "learning_rate": 3.986142241853808e-06,
+      "loss": 0.5177,
+      "step": 4658
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5017776976421946,
+      "learning_rate": 3.984174354620353e-06,
+      "loss": 0.5474,
+      "step": 4659
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.264765259932856,
+      "learning_rate": 3.982206631502478e-06,
+      "loss": 0.4865,
+      "step": 4660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8198470905922306,
+      "learning_rate": 3.980239072818086e-06,
+      "loss": 0.4662,
+      "step": 4661
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4756018404030682,
+      "learning_rate": 3.9782716788850525e-06,
+      "loss": 0.5083,
+      "step": 4662
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4214216069678782,
+      "learning_rate": 3.9763044500212285e-06,
+      "loss": 0.5023,
+      "step": 4663
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4574395662824027,
+      "learning_rate": 3.974337386544436e-06,
+      "loss": 0.4644,
+      "step": 4664
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4535700532434277,
+      "learning_rate": 3.972370488772472e-06,
+      "loss": 0.4926,
+      "step": 4665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5898246354626455,
+      "learning_rate": 3.9704037570231055e-06,
+      "loss": 0.5456,
+      "step": 4666
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6957198744230172,
+      "learning_rate": 3.968437191614081e-06,
+      "loss": 0.5038,
+      "step": 4667
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.535459803714931,
+      "learning_rate": 3.966470792863113e-06,
+      "loss": 0.5132,
+      "step": 4668
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6608193469359567,
+      "learning_rate": 3.964504561087891e-06,
+      "loss": 0.4485,
+      "step": 4669
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.349331946559182,
+      "learning_rate": 3.962538496606077e-06,
+      "loss": 0.5023,
+      "step": 4670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.580207427022666,
+      "learning_rate": 3.960572599735306e-06,
+      "loss": 0.4777,
+      "step": 4671
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6782527319642873,
+      "learning_rate": 3.958606870793184e-06,
+      "loss": 0.488,
+      "step": 4672
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3316424558043596,
+      "learning_rate": 3.956641310097296e-06,
+      "loss": 0.5741,
+      "step": 4673
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3524018346110955,
+      "learning_rate": 3.954675917965194e-06,
+      "loss": 0.5005,
+      "step": 4674
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.684657285155044,
+      "learning_rate": 3.952710694714403e-06,
+      "loss": 0.4864,
+      "step": 4675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.510587123243163,
+      "learning_rate": 3.9507456406624235e-06,
+      "loss": 0.5308,
+      "step": 4676
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4383162007434367,
+      "learning_rate": 3.948780756126726e-06,
+      "loss": 0.5088,
+      "step": 4677
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2909856888220486,
+      "learning_rate": 3.946816041424756e-06,
+      "loss": 0.5403,
+      "step": 4678
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.522311388231281,
+      "learning_rate": 3.94485149687393e-06,
+      "loss": 0.4866,
+      "step": 4679
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5130355478873698,
+      "learning_rate": 3.942887122791636e-06,
+      "loss": 0.5284,
+      "step": 4680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3370427571409305,
+      "learning_rate": 3.940922919495239e-06,
+      "loss": 0.481,
+      "step": 4681
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2204372209939194,
+      "learning_rate": 3.938958887302072e-06,
+      "loss": 0.5164,
+      "step": 4682
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8622732708112701,
+      "learning_rate": 3.9369950265294415e-06,
+      "loss": 0.4943,
+      "step": 4683
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6782135724024395,
+      "learning_rate": 3.9350313374946275e-06,
+      "loss": 0.4423,
+      "step": 4684
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0674398268960297,
+      "learning_rate": 3.933067820514879e-06,
+      "loss": 0.5081,
+      "step": 4685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5423215799903884,
+      "learning_rate": 3.931104475907423e-06,
+      "loss": 0.4722,
+      "step": 4686
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.455909664098324,
+      "learning_rate": 3.92914130398945e-06,
+      "loss": 0.5115,
+      "step": 4687
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4171266629705266,
+      "learning_rate": 3.92717830507813e-06,
+      "loss": 0.4569,
+      "step": 4688
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5722967736044233,
+      "learning_rate": 3.925215479490605e-06,
+      "loss": 0.519,
+      "step": 4689
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8243100670739596,
+      "learning_rate": 3.923252827543986e-06,
+      "loss": 0.4902,
+      "step": 4690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4910512694948186,
+      "learning_rate": 3.921290349555355e-06,
+      "loss": 0.5111,
+      "step": 4691
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6551753415818495,
+      "learning_rate": 3.9193280458417685e-06,
+      "loss": 0.527,
+      "step": 4692
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.660451709064471,
+      "learning_rate": 3.917365916720255e-06,
+      "loss": 0.5305,
+      "step": 4693
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.64326306596274,
+      "learning_rate": 3.915403962507812e-06,
+      "loss": 0.5228,
+      "step": 4694
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.601187680203445,
+      "learning_rate": 3.9134421835214105e-06,
+      "loss": 0.5226,
+      "step": 4695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.490661965402863,
+      "learning_rate": 3.911480580077992e-06,
+      "loss": 0.5178,
+      "step": 4696
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.683396309148199,
+      "learning_rate": 3.909519152494475e-06,
+      "loss": 0.5278,
+      "step": 4697
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.812167758657147,
+      "learning_rate": 3.907557901087743e-06,
+      "loss": 0.5433,
+      "step": 4698
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.466698533298413,
+      "learning_rate": 3.905596826174654e-06,
+      "loss": 0.518,
+      "step": 4699
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7310256508734216,
+      "learning_rate": 3.903635928072035e-06,
+      "loss": 0.4678,
+      "step": 4700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4647284525075164,
+      "learning_rate": 3.901675207096689e-06,
+      "loss": 0.5054,
+      "step": 4701
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4310062514790178,
+      "learning_rate": 3.899714663565386e-06,
+      "loss": 0.4812,
+      "step": 4702
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4360374783010557,
+      "learning_rate": 3.897754297794869e-06,
+      "loss": 0.5054,
+      "step": 4703
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4011598640375142,
+      "learning_rate": 3.895794110101854e-06,
+      "loss": 0.4953,
+      "step": 4704
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3910790082629851,
+      "learning_rate": 3.893834100803023e-06,
+      "loss": 0.4697,
+      "step": 4705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.384821909994724,
+      "learning_rate": 3.891874270215037e-06,
+      "loss": 0.5501,
+      "step": 4706
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.358041590073802,
+      "learning_rate": 3.889914618654522e-06,
+      "loss": 0.5326,
+      "step": 4707
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7752222465176617,
+      "learning_rate": 3.887955146438077e-06,
+      "loss": 0.5044,
+      "step": 4708
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5484824637392638,
+      "learning_rate": 3.885995853882273e-06,
+      "loss": 0.4953,
+      "step": 4709
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.1610373215701917,
+      "learning_rate": 3.884036741303649e-06,
+      "loss": 0.45,
+      "step": 4710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.276864536888107,
+      "learning_rate": 3.8820778090187185e-06,
+      "loss": 0.5035,
+      "step": 4711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9254796816869655,
+      "learning_rate": 3.880119057343965e-06,
+      "loss": 0.5279,
+      "step": 4712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6475310297625902,
+      "learning_rate": 3.878160486595837e-06,
+      "loss": 0.5642,
+      "step": 4713
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4280491895781326,
+      "learning_rate": 3.876202097090765e-06,
+      "loss": 0.4728,
+      "step": 4714
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2032458161147614,
+      "learning_rate": 3.874243889145142e-06,
+      "loss": 0.528,
+      "step": 4715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4660037572297808,
+      "learning_rate": 3.872285863075334e-06,
+      "loss": 0.5171,
+      "step": 4716
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.445124971634497,
+      "learning_rate": 3.8703280191976764e-06,
+      "loss": 0.4733,
+      "step": 4717
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6264348500908541,
+      "learning_rate": 3.868370357828476e-06,
+      "loss": 0.513,
+      "step": 4718
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7280898543065133,
+      "learning_rate": 3.866412879284011e-06,
+      "loss": 0.5092,
+      "step": 4719
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7655472224257922,
+      "learning_rate": 3.864455583880529e-06,
+      "loss": 0.5303,
+      "step": 4720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3950462536660542,
+      "learning_rate": 3.862498471934248e-06,
+      "loss": 0.454,
+      "step": 4721
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5446451372213836,
+      "learning_rate": 3.860541543761358e-06,
+      "loss": 0.4389,
+      "step": 4722
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.202339874550893,
+      "learning_rate": 3.858584799678017e-06,
+      "loss": 0.5332,
+      "step": 4723
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7557477548315092,
+      "learning_rate": 3.8566282400003545e-06,
+      "loss": 0.5418,
+      "step": 4724
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2698707672176435,
+      "learning_rate": 3.854671865044469e-06,
+      "loss": 0.4851,
+      "step": 4725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6644076653329196,
+      "learning_rate": 3.852715675126431e-06,
+      "loss": 0.4325,
+      "step": 4726
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2646724952132085,
+      "learning_rate": 3.85075967056228e-06,
+      "loss": 0.5054,
+      "step": 4727
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5026733833240709,
+      "learning_rate": 3.848803851668026e-06,
+      "loss": 0.4743,
+      "step": 4728
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4012847881227701,
+      "learning_rate": 3.8468482187596475e-06,
+      "loss": 0.4744,
+      "step": 4729
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5112355270821844,
+      "learning_rate": 3.844892772153097e-06,
+      "loss": 0.5427,
+      "step": 4730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7444916837957993,
+      "learning_rate": 3.842937512164292e-06,
+      "loss": 0.5219,
+      "step": 4731
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3893452549099528,
+      "learning_rate": 3.840982439109122e-06,
+      "loss": 0.4558,
+      "step": 4732
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7086719823044167,
+      "learning_rate": 3.839027553303447e-06,
+      "loss": 0.5013,
+      "step": 4733
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4469872927187486,
+      "learning_rate": 3.837072855063097e-06,
+      "loss": 0.5551,
+      "step": 4734
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2457288599780056,
+      "learning_rate": 3.835118344703868e-06,
+      "loss": 0.4889,
+      "step": 4735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2896620308751967,
+      "learning_rate": 3.833164022541529e-06,
+      "loss": 0.4871,
+      "step": 4736
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.03442397818743,
+      "learning_rate": 3.83120988889182e-06,
+      "loss": 0.5277,
+      "step": 4737
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3622425574642558,
+      "learning_rate": 3.829255944070445e-06,
+      "loss": 0.4914,
+      "step": 4738
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9296216351758595,
+      "learning_rate": 3.8273021883930865e-06,
+      "loss": 0.4908,
+      "step": 4739
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5253787207780585,
+      "learning_rate": 3.825348622175386e-06,
+      "loss": 0.4846,
+      "step": 4740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.187766035709169,
+      "learning_rate": 3.823395245732961e-06,
+      "loss": 0.4719,
+      "step": 4741
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4975496186557569,
+      "learning_rate": 3.8214420593813975e-06,
+      "loss": 0.5083,
+      "step": 4742
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7297899522931308,
+      "learning_rate": 3.819489063436248e-06,
+      "loss": 0.5002,
+      "step": 4743
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.484072419508185,
+      "learning_rate": 3.817536258213038e-06,
+      "loss": 0.4853,
+      "step": 4744
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5021104486952073,
+      "learning_rate": 3.81558364402726e-06,
+      "loss": 0.4509,
+      "step": 4745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6685918184482609,
+      "learning_rate": 3.8136312211943736e-06,
+      "loss": 0.5322,
+      "step": 4746
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6368358729433512,
+      "learning_rate": 3.8116789900298135e-06,
+      "loss": 0.4764,
+      "step": 4747
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.43761500492689,
+      "learning_rate": 3.809726950848979e-06,
+      "loss": 0.4895,
+      "step": 4748
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5978788025111923,
+      "learning_rate": 3.8077751039672377e-06,
+      "loss": 0.5373,
+      "step": 4749
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2847629859841037,
+      "learning_rate": 3.805823449699929e-06,
+      "loss": 0.4763,
+      "step": 4750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7553604371536287,
+      "learning_rate": 3.80387198836236e-06,
+      "loss": 0.477,
+      "step": 4751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8653801195128799,
+      "learning_rate": 3.801920720269805e-06,
+      "loss": 0.4816,
+      "step": 4752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6215137768031846,
+      "learning_rate": 3.7999696457375094e-06,
+      "loss": 0.4689,
+      "step": 4753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5537199741329855,
+      "learning_rate": 3.7980187650806855e-06,
+      "loss": 0.497,
+      "step": 4754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3683582861339472,
+      "learning_rate": 3.7960680786145177e-06,
+      "loss": 0.4956,
+      "step": 4755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3744261784431637,
+      "learning_rate": 3.7941175866541545e-06,
+      "loss": 0.4658,
+      "step": 4756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.834936259971978,
+      "learning_rate": 3.7921672895147154e-06,
+      "loss": 0.4878,
+      "step": 4757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6618545833319962,
+      "learning_rate": 3.7902171875112893e-06,
+      "loss": 0.4895,
+      "step": 4758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5550754129562883,
+      "learning_rate": 3.7882672809589303e-06,
+      "loss": 0.5166,
+      "step": 4759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6752689455613092,
+      "learning_rate": 3.786317570172665e-06,
+      "loss": 0.4803,
+      "step": 4760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.817067473755128,
+      "learning_rate": 3.784368055467485e-06,
+      "loss": 0.4799,
+      "step": 4761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.796483415824333,
+      "learning_rate": 3.78241873715835e-06,
+      "loss": 0.4372,
+      "step": 4762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4581358695876308,
+      "learning_rate": 3.7804696155601946e-06,
+      "loss": 0.5114,
+      "step": 4763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5162840470971275,
+      "learning_rate": 3.7785206909879125e-06,
+      "loss": 0.5593,
+      "step": 4764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7687354735721073,
+      "learning_rate": 3.7765719637563704e-06,
+      "loss": 0.4865,
+      "step": 4765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6016449273847941,
+      "learning_rate": 3.7746234341804023e-06,
+      "loss": 0.4856,
+      "step": 4766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5925792667624954,
+      "learning_rate": 3.772675102574811e-06,
+      "loss": 0.4904,
+      "step": 4767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5491375986862115,
+      "learning_rate": 3.7707269692543657e-06,
+      "loss": 0.4987,
+      "step": 4768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.024481064899184,
+      "learning_rate": 3.7687790345338054e-06,
+      "loss": 0.5366,
+      "step": 4769
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5008907912251712,
+      "learning_rate": 3.7668312987278333e-06,
+      "loss": 0.4803,
+      "step": 4770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4008267317117975,
+      "learning_rate": 3.764883762151128e-06,
+      "loss": 0.4885,
+      "step": 4771
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3949664117379335,
+      "learning_rate": 3.762936425118329e-06,
+      "loss": 0.5142,
+      "step": 4772
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3049393678095358,
+      "learning_rate": 3.760989287944047e-06,
+      "loss": 0.5097,
+      "step": 4773
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5641320129466818,
+      "learning_rate": 3.7590423509428557e-06,
+      "loss": 0.5167,
+      "step": 4774
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4951590978042526,
+      "learning_rate": 3.7570956144293025e-06,
+      "loss": 0.5374,
+      "step": 4775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.330005347816297,
+      "learning_rate": 3.7551490787178996e-06,
+      "loss": 0.5422,
+      "step": 4776
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3316052030847458,
+      "learning_rate": 3.753202744123126e-06,
+      "loss": 0.453,
+      "step": 4777
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6324208318735371,
+      "learning_rate": 3.7512566109594308e-06,
+      "loss": 0.5317,
+      "step": 4778
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5633976503408085,
+      "learning_rate": 3.7493106795412254e-06,
+      "loss": 0.4704,
+      "step": 4779
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4535976872534775,
+      "learning_rate": 3.747364950182897e-06,
+      "loss": 0.5258,
+      "step": 4780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.231741944089487,
+      "learning_rate": 3.7454194231987927e-06,
+      "loss": 0.4684,
+      "step": 4781
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2879488975320648,
+      "learning_rate": 3.7434740989032316e-06,
+      "loss": 0.5012,
+      "step": 4782
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7339082361721024,
+      "learning_rate": 3.7415289776104953e-06,
+      "loss": 0.4991,
+      "step": 4783
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8207277057420377,
+      "learning_rate": 3.739584059634836e-06,
+      "loss": 0.4733,
+      "step": 4784
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.816680471066737,
+      "learning_rate": 3.7376393452904725e-06,
+      "loss": 0.5114,
+      "step": 4785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4300159454723032,
+      "learning_rate": 3.7356948348915913e-06,
+      "loss": 0.4716,
+      "step": 4786
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.466517194302703,
+      "learning_rate": 3.7337505287523413e-06,
+      "loss": 0.5039,
+      "step": 4787
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6332080984609707,
+      "learning_rate": 3.731806427186848e-06,
+      "loss": 0.4736,
+      "step": 4788
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.521013037464149,
+      "learning_rate": 3.7298625305091963e-06,
+      "loss": 0.4883,
+      "step": 4789
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5662771829569915,
+      "learning_rate": 3.7279188390334385e-06,
+      "loss": 0.4823,
+      "step": 4790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4664761110857254,
+      "learning_rate": 3.7259753530735964e-06,
+      "loss": 0.5037,
+      "step": 4791
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.406015218659607,
+      "learning_rate": 3.7240320729436575e-06,
+      "loss": 0.5087,
+      "step": 4792
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4644074732784458,
+      "learning_rate": 3.7220889989575737e-06,
+      "loss": 0.4667,
+      "step": 4793
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6162804324068345,
+      "learning_rate": 3.7201461314292665e-06,
+      "loss": 0.5539,
+      "step": 4794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9052845018302897,
+      "learning_rate": 3.7182034706726224e-06,
+      "loss": 0.4905,
+      "step": 4795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3987498526053814,
+      "learning_rate": 3.716261017001498e-06,
+      "loss": 0.4913,
+      "step": 4796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.653132447704072,
+      "learning_rate": 3.714318770729713e-06,
+      "loss": 0.5096,
+      "step": 4797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4485827282911432,
+      "learning_rate": 3.7123767321710538e-06,
+      "loss": 0.5258,
+      "step": 4798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3441229709454283,
+      "learning_rate": 3.710434901639274e-06,
+      "loss": 0.5023,
+      "step": 4799
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5096540584449896,
+      "learning_rate": 3.708493279448093e-06,
+      "loss": 0.4974,
+      "step": 4800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3348271010756676,
+      "learning_rate": 3.7065518659111982e-06,
+      "loss": 0.5133,
+      "step": 4801
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3362769274646589,
+      "learning_rate": 3.7046106613422404e-06,
+      "loss": 0.4978,
+      "step": 4802
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6586552427143136,
+      "learning_rate": 3.7026696660548365e-06,
+      "loss": 0.5204,
+      "step": 4803
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2970674817252872,
+      "learning_rate": 3.700728880362577e-06,
+      "loss": 0.5126,
+      "step": 4804
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2692411052824943,
+      "learning_rate": 3.6987883045790093e-06,
+      "loss": 0.4919,
+      "step": 4805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.544857137361854,
+      "learning_rate": 3.6968479390176503e-06,
+      "loss": 0.4943,
+      "step": 4806
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6111067019105714,
+      "learning_rate": 3.6949077839919852e-06,
+      "loss": 0.5665,
+      "step": 4807
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5148276199786361,
+      "learning_rate": 3.692967839815461e-06,
+      "loss": 0.5128,
+      "step": 4808
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9289750603078077,
+      "learning_rate": 3.6910281068014935e-06,
+      "loss": 0.5046,
+      "step": 4809
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3571727913310128,
+      "learning_rate": 3.689088585263464e-06,
+      "loss": 0.544,
+      "step": 4810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.514263873049238,
+      "learning_rate": 3.687149275514718e-06,
+      "loss": 0.4944,
+      "step": 4811
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.320448446471603,
+      "learning_rate": 3.68521017786857e-06,
+      "loss": 0.4634,
+      "step": 4812
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4253421646515063,
+      "learning_rate": 3.6832712926382978e-06,
+      "loss": 0.4323,
+      "step": 4813
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6248156405188252,
+      "learning_rate": 3.6813326201371448e-06,
+      "loss": 0.4576,
+      "step": 4814
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5491421709460802,
+      "learning_rate": 3.679394160678321e-06,
+      "loss": 0.4942,
+      "step": 4815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5293882527430553,
+      "learning_rate": 3.677455914575001e-06,
+      "loss": 0.5424,
+      "step": 4816
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6610870035965135,
+      "learning_rate": 3.675517882140326e-06,
+      "loss": 0.491,
+      "step": 4817
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5876662699587831,
+      "learning_rate": 3.673580063687402e-06,
+      "loss": 0.5051,
+      "step": 4818
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.750905294107381,
+      "learning_rate": 3.671642459529301e-06,
+      "loss": 0.4937,
+      "step": 4819
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7565409567351697,
+      "learning_rate": 3.6697050699790586e-06,
+      "loss": 0.5273,
+      "step": 4820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5630185373429069,
+      "learning_rate": 3.6677678953496797e-06,
+      "loss": 0.536,
+      "step": 4821
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5729595543148676,
+      "learning_rate": 3.66583093595413e-06,
+      "loss": 0.5653,
+      "step": 4822
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3837247868854905,
+      "learning_rate": 3.663894192105344e-06,
+      "loss": 0.5577,
+      "step": 4823
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7157111912006533,
+      "learning_rate": 3.6619576641162176e-06,
+      "loss": 0.5539,
+      "step": 4824
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6563532770825058,
+      "learning_rate": 3.660021352299615e-06,
+      "loss": 0.5398,
+      "step": 4825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4097431835544845,
+      "learning_rate": 3.658085256968365e-06,
+      "loss": 0.5385,
+      "step": 4826
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.360389593574117,
+      "learning_rate": 3.65614937843526e-06,
+      "loss": 0.5052,
+      "step": 4827
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.353844494224524,
+      "learning_rate": 3.6542137170130576e-06,
+      "loss": 0.4953,
+      "step": 4828
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.422660488634164,
+      "learning_rate": 3.652278273014482e-06,
+      "loss": 0.5504,
+      "step": 4829
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5049937847673278,
+      "learning_rate": 3.650343046752222e-06,
+      "loss": 0.4411,
+      "step": 4830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8728735408056143,
+      "learning_rate": 3.6484080385389296e-06,
+      "loss": 0.4558,
+      "step": 4831
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3355117575474822,
+      "learning_rate": 3.6464732486872216e-06,
+      "loss": 0.4604,
+      "step": 4832
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2747248934156539,
+      "learning_rate": 3.6445386775096813e-06,
+      "loss": 0.4732,
+      "step": 4833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3518302410791765,
+      "learning_rate": 3.642604325318856e-06,
+      "loss": 0.4815,
+      "step": 4834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4051502217298926,
+      "learning_rate": 3.640670192427257e-06,
+      "loss": 0.4661,
+      "step": 4835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4809715490321091,
+      "learning_rate": 3.6387362791473583e-06,
+      "loss": 0.5039,
+      "step": 4836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.396861806595747,
+      "learning_rate": 3.6368025857916044e-06,
+      "loss": 0.4719,
+      "step": 4837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4157885612242203,
+      "learning_rate": 3.6348691126723984e-06,
+      "loss": 0.5259,
+      "step": 4838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7313886903818687,
+      "learning_rate": 3.6329358601021103e-06,
+      "loss": 0.5609,
+      "step": 4839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4253943377103053,
+      "learning_rate": 3.6310028283930743e-06,
+      "loss": 0.5139,
+      "step": 4840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.849074190948096,
+      "learning_rate": 3.629070017857588e-06,
+      "loss": 0.4811,
+      "step": 4841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3268550582958893,
+      "learning_rate": 3.6271374288079142e-06,
+      "loss": 0.4985,
+      "step": 4842
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3070111115301872,
+      "learning_rate": 3.6252050615562805e-06,
+      "loss": 0.4931,
+      "step": 4843
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3391127564835532,
+      "learning_rate": 3.623272916414876e-06,
+      "loss": 0.5093,
+      "step": 4844
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3718861741774446,
+      "learning_rate": 3.621340993695858e-06,
+      "loss": 0.5138,
+      "step": 4845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4276519023118137,
+      "learning_rate": 3.6194092937113444e-06,
+      "loss": 0.5519,
+      "step": 4846
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.672717431707128,
+      "learning_rate": 3.617477816773419e-06,
+      "loss": 0.4756,
+      "step": 4847
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.588454212554117,
+      "learning_rate": 3.6155465631941276e-06,
+      "loss": 0.5124,
+      "step": 4848
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7098411838422147,
+      "learning_rate": 3.6136155332854835e-06,
+      "loss": 0.5085,
+      "step": 4849
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4768604915329149,
+      "learning_rate": 3.611684727359459e-06,
+      "loss": 0.513,
+      "step": 4850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6237177436087704,
+      "learning_rate": 3.6097541457279945e-06,
+      "loss": 0.5036,
+      "step": 4851
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.008599890001384,
+      "learning_rate": 3.607823788702991e-06,
+      "loss": 0.5041,
+      "step": 4852
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.33530351725401,
+      "learning_rate": 3.6058936565963158e-06,
+      "loss": 0.5178,
+      "step": 4853
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5829084651240015,
+      "learning_rate": 3.6039637497197995e-06,
+      "loss": 0.4938,
+      "step": 4854
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9056804696592824,
+      "learning_rate": 3.602034068385234e-06,
+      "loss": 0.5461,
+      "step": 4855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9163992244200692,
+      "learning_rate": 3.6001046129043778e-06,
+      "loss": 0.4962,
+      "step": 4856
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7212350091995289,
+      "learning_rate": 3.59817538358895e-06,
+      "loss": 0.5098,
+      "step": 4857
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9415046450085558,
+      "learning_rate": 3.596246380750635e-06,
+      "loss": 0.5487,
+      "step": 4858
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5949102498616452,
+      "learning_rate": 3.5943176047010807e-06,
+      "loss": 0.6143,
+      "step": 4859
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.630007389484059,
+      "learning_rate": 3.592389055751897e-06,
+      "loss": 0.5065,
+      "step": 4860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.476053795816333,
+      "learning_rate": 3.590460734214657e-06,
+      "loss": 0.4923,
+      "step": 4861
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8988176215501187,
+      "learning_rate": 3.5885326404009022e-06,
+      "loss": 0.5095,
+      "step": 4862
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7075130802141478,
+      "learning_rate": 3.5866047746221294e-06,
+      "loss": 0.5574,
+      "step": 4863
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.96164473093287,
+      "learning_rate": 3.5846771371898037e-06,
+      "loss": 0.5252,
+      "step": 4864
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7917743538361852,
+      "learning_rate": 3.582749728415351e-06,
+      "loss": 0.5252,
+      "step": 4865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.645792885373291,
+      "learning_rate": 3.580822548610162e-06,
+      "loss": 0.4615,
+      "step": 4866
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4344210864773772,
+      "learning_rate": 3.5788955980855894e-06,
+      "loss": 0.4666,
+      "step": 4867
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3952139809581494,
+      "learning_rate": 3.5769688771529486e-06,
+      "loss": 0.4653,
+      "step": 4868
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6897052423443377,
+      "learning_rate": 3.575042386123517e-06,
+      "loss": 0.5299,
+      "step": 4869
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5339288039723356,
+      "learning_rate": 3.57311612530854e-06,
+      "loss": 0.4769,
+      "step": 4870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.188281057473945,
+      "learning_rate": 3.5711900950192204e-06,
+      "loss": 0.4984,
+      "step": 4871
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.511980292628914,
+      "learning_rate": 3.5692642955667235e-06,
+      "loss": 0.48,
+      "step": 4872
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.362985578204832,
+      "learning_rate": 3.5673387272621805e-06,
+      "loss": 0.415,
+      "step": 4873
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7036428859732629,
+      "learning_rate": 3.565413390416684e-06,
+      "loss": 0.5269,
+      "step": 4874
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.012695203425002,
+      "learning_rate": 3.56348828534129e-06,
+      "loss": 0.4847,
+      "step": 4875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4546222976760228,
+      "learning_rate": 3.5615634123470143e-06,
+      "loss": 0.4663,
+      "step": 4876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5513891298669515,
+      "learning_rate": 3.5596387717448354e-06,
+      "loss": 0.5032,
+      "step": 4877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0380830917483737,
+      "learning_rate": 3.5577143638457014e-06,
+      "loss": 0.4912,
+      "step": 4878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.600213084538027,
+      "learning_rate": 3.555790188960514e-06,
+      "loss": 0.5027,
+      "step": 4879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7253184447834027,
+      "learning_rate": 3.5538662474001414e-06,
+      "loss": 0.4657,
+      "step": 4880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.390200812178653,
+      "learning_rate": 3.551942539475414e-06,
+      "loss": 0.5211,
+      "step": 4881
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3614362906935606,
+      "learning_rate": 3.550019065497121e-06,
+      "loss": 0.5094,
+      "step": 4882
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4460947953267989,
+      "learning_rate": 3.5480958257760185e-06,
+      "loss": 0.4452,
+      "step": 4883
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7139647753932892,
+      "learning_rate": 3.546172820622823e-06,
+      "loss": 0.5304,
+      "step": 4884
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6558235733830828,
+      "learning_rate": 3.54425005034821e-06,
+      "loss": 0.442,
+      "step": 4885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5570124837515966,
+      "learning_rate": 3.5423275152628245e-06,
+      "loss": 0.5011,
+      "step": 4886
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4332172878692637,
+      "learning_rate": 3.540405215677267e-06,
+      "loss": 0.5,
+      "step": 4887
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6731989318783418,
+      "learning_rate": 3.5384831519021024e-06,
+      "loss": 0.5592,
+      "step": 4888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3510517557984159,
+      "learning_rate": 3.536561324247856e-06,
+      "loss": 0.527,
+      "step": 4889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6372062487205699,
+      "learning_rate": 3.5346397330250176e-06,
+      "loss": 0.505,
+      "step": 4890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4613237336873037,
+      "learning_rate": 3.532718378544035e-06,
+      "loss": 0.5313,
+      "step": 4891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5685237927072886,
+      "learning_rate": 3.530797261115321e-06,
+      "loss": 0.5362,
+      "step": 4892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.340615457625864,
+      "learning_rate": 3.5288763810492486e-06,
+      "loss": 0.5179,
+      "step": 4893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.48299723682518,
+      "learning_rate": 3.5269557386561524e-06,
+      "loss": 0.4949,
+      "step": 4894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1982134842143222,
+      "learning_rate": 3.5250353342463318e-06,
+      "loss": 0.4733,
+      "step": 4895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3323531179160906,
+      "learning_rate": 3.5231151681300426e-06,
+      "loss": 0.4817,
+      "step": 4896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1072391564702007,
+      "learning_rate": 3.5211952406175056e-06,
+      "loss": 0.5162,
+      "step": 4897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9360651071059258,
+      "learning_rate": 3.5192755520189013e-06,
+      "loss": 0.4815,
+      "step": 4898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.316168543249597,
+      "learning_rate": 3.5173561026443737e-06,
+      "loss": 0.4946,
+      "step": 4899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5539248601419484,
+      "learning_rate": 3.5154368928040255e-06,
+      "loss": 0.5328,
+      "step": 4900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.452887193496379,
+      "learning_rate": 3.513517922807922e-06,
+      "loss": 0.4714,
+      "step": 4901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2772497753613659,
+      "learning_rate": 3.511599192966087e-06,
+      "loss": 0.5088,
+      "step": 4902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6841954429457333,
+      "learning_rate": 3.5096807035885134e-06,
+      "loss": 0.5319,
+      "step": 4903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6568211084906097,
+      "learning_rate": 3.5077624549851472e-06,
+      "loss": 0.4766,
+      "step": 4904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.251958650680514,
+      "learning_rate": 3.5058444474659e-06,
+      "loss": 0.4197,
+      "step": 4905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3744636066787468,
+      "learning_rate": 3.503926681340641e-06,
+      "loss": 0.4747,
+      "step": 4906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7970179104924786,
+      "learning_rate": 3.5020091569192025e-06,
+      "loss": 0.468,
+      "step": 4907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1002716500361447,
+      "learning_rate": 3.500091874511379e-06,
+      "loss": 0.4674,
+      "step": 4908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7344236020353803,
+      "learning_rate": 3.4981748344269218e-06,
+      "loss": 0.5158,
+      "step": 4909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.33071499329348,
+      "learning_rate": 3.496258036975547e-06,
+      "loss": 0.5131,
+      "step": 4910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5623268444562706,
+      "learning_rate": 3.494341482466931e-06,
+      "loss": 0.5587,
+      "step": 4911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2844120808059343,
+      "learning_rate": 3.492425171210708e-06,
+      "loss": 0.5187,
+      "step": 4912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0085068307255383,
+      "learning_rate": 3.4905091035164775e-06,
+      "loss": 0.5232,
+      "step": 4913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3519355908870683,
+      "learning_rate": 3.4885932796937948e-06,
+      "loss": 0.5681,
+      "step": 4914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.623680657057162,
+      "learning_rate": 3.48667770005218e-06,
+      "loss": 0.4876,
+      "step": 4915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3509431507059002,
+      "learning_rate": 3.4847623649011104e-06,
+      "loss": 0.5221,
+      "step": 4916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7033636091227293,
+      "learning_rate": 3.4828472745500256e-06,
+      "loss": 0.5763,
+      "step": 4917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4902431059393642,
+      "learning_rate": 3.4809324293083244e-06,
+      "loss": 0.4778,
+      "step": 4918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.663639956345997,
+      "learning_rate": 3.479017829485369e-06,
+      "loss": 0.4622,
+      "step": 4919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2922645060372164,
+      "learning_rate": 3.4771034753904798e-06,
+      "loss": 0.4629,
+      "step": 4920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7922928399840536,
+      "learning_rate": 3.4751893673329363e-06,
+      "loss": 0.4866,
+      "step": 4921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5676006214673102,
+      "learning_rate": 3.473275505621979e-06,
+      "loss": 0.5136,
+      "step": 4922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3092971392638122,
+      "learning_rate": 3.4713618905668103e-06,
+      "loss": 0.5068,
+      "step": 4923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4406852634601057,
+      "learning_rate": 3.469448522476592e-06,
+      "loss": 0.5243,
+      "step": 4924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5603319740314827,
+      "learning_rate": 3.4675354016604433e-06,
+      "loss": 0.5204,
+      "step": 4925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.663970478353552,
+      "learning_rate": 3.465622528427447e-06,
+      "loss": 0.4964,
+      "step": 4926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3333166367010005,
+      "learning_rate": 3.463709903086646e-06,
+      "loss": 0.5025,
+      "step": 4927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.271444599665555,
+      "learning_rate": 3.4617975259470403e-06,
+      "loss": 0.4348,
+      "step": 4928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3694773642972735,
+      "learning_rate": 3.459885397317592e-06,
+      "loss": 0.4865,
+      "step": 4929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4734405859367252,
+      "learning_rate": 3.4579735175072217e-06,
+      "loss": 0.4868,
+      "step": 4930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5059492843366027,
+      "learning_rate": 3.456061886824811e-06,
+      "loss": 0.5086,
+      "step": 4931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5004990367175046,
+      "learning_rate": 3.4541505055791993e-06,
+      "loss": 0.46,
+      "step": 4932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6902157867585337,
+      "learning_rate": 3.4522393740791887e-06,
+      "loss": 0.4743,
+      "step": 4933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3336540061732831,
+      "learning_rate": 3.4503284926335385e-06,
+      "loss": 0.4962,
+      "step": 4934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4185962631748819,
+      "learning_rate": 3.4484178615509676e-06,
+      "loss": 0.4756,
+      "step": 4935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.443001312748826,
+      "learning_rate": 3.446507481140157e-06,
+      "loss": 0.4472,
+      "step": 4936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6777553610363183,
+      "learning_rate": 3.4445973517097453e-06,
+      "loss": 0.5026,
+      "step": 4937
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4188816330877883,
+      "learning_rate": 3.44268747356833e-06,
+      "loss": 0.4978,
+      "step": 4938
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5110459009189268,
+      "learning_rate": 3.440777847024469e-06,
+      "loss": 0.5127,
+      "step": 4939
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6410841022146765,
+      "learning_rate": 3.438868472386679e-06,
+      "loss": 0.4696,
+      "step": 4940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4175618287038916,
+      "learning_rate": 3.436959349963437e-06,
+      "loss": 0.5367,
+      "step": 4941
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6666608715251021,
+      "learning_rate": 3.4350504800631775e-06,
+      "loss": 0.5037,
+      "step": 4942
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.542889725936689,
+      "learning_rate": 3.433141862994295e-06,
+      "loss": 0.5429,
+      "step": 4943
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7142075551377389,
+      "learning_rate": 3.431233499065145e-06,
+      "loss": 0.4819,
+      "step": 4944
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5725999569088291,
+      "learning_rate": 3.4293253885840395e-06,
+      "loss": 0.5112,
+      "step": 4945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3023802101493411,
+      "learning_rate": 3.427417531859251e-06,
+      "loss": 0.4196,
+      "step": 4946
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6768819589739452,
+      "learning_rate": 3.42550992919901e-06,
+      "loss": 0.5163,
+      "step": 4947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7029471252807711,
+      "learning_rate": 3.4236025809115076e-06,
+      "loss": 0.4865,
+      "step": 4948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3768694181253318,
+      "learning_rate": 3.4216954873048912e-06,
+      "loss": 0.5202,
+      "step": 4949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9456364823131251,
+      "learning_rate": 3.41978864868727e-06,
+      "loss": 0.5039,
+      "step": 4950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7185901326118083,
+      "learning_rate": 3.4178820653667085e-06,
+      "loss": 0.4309,
+      "step": 4951
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.518680620525008,
+      "learning_rate": 3.415975737651234e-06,
+      "loss": 0.5213,
+      "step": 4952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.153241301832152,
+      "learning_rate": 3.4140696658488304e-06,
+      "loss": 0.4299,
+      "step": 4953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5080971432334338,
+      "learning_rate": 3.4121638502674397e-06,
+      "loss": 0.552,
+      "step": 4954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3662961579017965,
+      "learning_rate": 3.4102582912149636e-06,
+      "loss": 0.5008,
+      "step": 4955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5083284681008178,
+      "learning_rate": 3.4083529889992617e-06,
+      "loss": 0.4877,
+      "step": 4956
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6526344945720335,
+      "learning_rate": 3.406447943928152e-06,
+      "loss": 0.4756,
+      "step": 4957
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6386548943776154,
+      "learning_rate": 3.4045431563094123e-06,
+      "loss": 0.4902,
+      "step": 4958
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4213765901889195,
+      "learning_rate": 3.402638626450775e-06,
+      "loss": 0.486,
+      "step": 4959
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.848591393086902,
+      "learning_rate": 3.4007343546599384e-06,
+      "loss": 0.4606,
+      "step": 4960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4880124777484838,
+      "learning_rate": 3.3988303412445518e-06,
+      "loss": 0.4948,
+      "step": 4961
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.277388551560406,
+      "learning_rate": 3.3969265865122247e-06,
+      "loss": 0.468,
+      "step": 4962
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2436123249223645,
+      "learning_rate": 3.395023090770526e-06,
+      "loss": 0.4636,
+      "step": 4963
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4514812397300938,
+      "learning_rate": 3.3931198543269835e-06,
+      "loss": 0.5198,
+      "step": 4964
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6008351074207805,
+      "learning_rate": 3.3912168774890797e-06,
+      "loss": 0.513,
+      "step": 4965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6545897409823502,
+      "learning_rate": 3.3893141605642586e-06,
+      "loss": 0.5582,
+      "step": 4966
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4886116131312268,
+      "learning_rate": 3.3874117038599186e-06,
+      "loss": 0.516,
+      "step": 4967
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.60481546255909,
+      "learning_rate": 3.3855095076834216e-06,
+      "loss": 0.4751,
+      "step": 4968
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5747293988329985,
+      "learning_rate": 3.3836075723420836e-06,
+      "loss": 0.544,
+      "step": 4969
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5765305986762645,
+      "learning_rate": 3.3817058981431784e-06,
+      "loss": 0.5053,
+      "step": 4970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5653479677087112,
+      "learning_rate": 3.3798044853939375e-06,
+      "loss": 0.5167,
+      "step": 4971
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6885810336216072,
+      "learning_rate": 3.3779033344015515e-06,
+      "loss": 0.5074,
+      "step": 4972
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.565508621884468,
+      "learning_rate": 3.3760024454731677e-06,
+      "loss": 0.5037,
+      "step": 4973
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.588896209724684,
+      "learning_rate": 3.374101818915892e-06,
+      "loss": 0.5485,
+      "step": 4974
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5535713099568533,
+      "learning_rate": 3.372201455036787e-06,
+      "loss": 0.4844,
+      "step": 4975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6389578924823094,
+      "learning_rate": 3.370301354142871e-06,
+      "loss": 0.5083,
+      "step": 4976
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.416294243738421,
+      "learning_rate": 3.3684015165411264e-06,
+      "loss": 0.4844,
+      "step": 4977
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.1980295333717508,
+      "learning_rate": 3.366501942538487e-06,
+      "loss": 0.5276,
+      "step": 4978
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2092080604491835,
+      "learning_rate": 3.3646026324418456e-06,
+      "loss": 0.4384,
+      "step": 4979
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.073980580745277,
+      "learning_rate": 3.3627035865580525e-06,
+      "loss": 0.5405,
+      "step": 4980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6017686291826134,
+      "learning_rate": 3.3608048051939147e-06,
+      "loss": 0.5,
+      "step": 4981
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3082842776433188,
+      "learning_rate": 3.358906288656197e-06,
+      "loss": 0.4998,
+      "step": 4982
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4383088949447156,
+      "learning_rate": 3.3570080372516213e-06,
+      "loss": 0.5247,
+      "step": 4983
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3887107654466078,
+      "learning_rate": 3.3551100512868663e-06,
+      "loss": 0.5017,
+      "step": 4984
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.38804570646994,
+      "learning_rate": 3.3532123310685706e-06,
+      "loss": 0.5464,
+      "step": 4985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4133582163794047,
+      "learning_rate": 3.3513148769033265e-06,
+      "loss": 0.4914,
+      "step": 4986
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5795226928379709,
+      "learning_rate": 3.3494176890976847e-06,
+      "loss": 0.4805,
+      "step": 4987
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.415716935901781,
+      "learning_rate": 3.3475207679581524e-06,
+      "loss": 0.5058,
+      "step": 4988
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7829808944908172,
+      "learning_rate": 3.3456241137911947e-06,
+      "loss": 0.4917,
+      "step": 4989
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5579286305124613,
+      "learning_rate": 3.343727726903231e-06,
+      "loss": 0.4855,
+      "step": 4990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9304590658531278,
+      "learning_rate": 3.3418316076006394e-06,
+      "loss": 0.4925,
+      "step": 4991
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6227599018995744,
+      "learning_rate": 3.3399357561897545e-06,
+      "loss": 0.5281,
+      "step": 4992
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.677743562278396,
+      "learning_rate": 3.33804017297687e-06,
+      "loss": 0.4952,
+      "step": 4993
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4807846872178518,
+      "learning_rate": 3.3361448582682333e-06,
+      "loss": 0.4668,
+      "step": 4994
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.526223315842072,
+      "learning_rate": 3.3342498123700484e-06,
+      "loss": 0.4895,
+      "step": 4995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0357225736417877,
+      "learning_rate": 3.3323550355884767e-06,
+      "loss": 0.5361,
+      "step": 4996
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5860862144050736,
+      "learning_rate": 3.3304605282296367e-06,
+      "loss": 0.4871,
+      "step": 4997
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6424354224213873,
+      "learning_rate": 3.328566290599602e-06,
+      "loss": 0.5181,
+      "step": 4998
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.648090373847227,
+      "learning_rate": 3.326672323004405e-06,
+      "loss": 0.5208,
+      "step": 4999
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.248674939834039,
+      "learning_rate": 3.324778625750028e-06,
+      "loss": 0.5155,
+      "step": 5000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2753266896871043,
+      "learning_rate": 3.3228851991424203e-06,
+      "loss": 0.4403,
+      "step": 5001
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.464593760439206,
+      "learning_rate": 3.32099204348748e-06,
+      "loss": 0.4549,
+      "step": 5002
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.000264604751063,
+      "learning_rate": 3.319099159091062e-06,
+      "loss": 0.4919,
+      "step": 5003
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8468860351733025,
+      "learning_rate": 3.3172065462589786e-06,
+      "loss": 0.5091,
+      "step": 5004
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5417003575190913,
+      "learning_rate": 3.315314205296999e-06,
+      "loss": 0.5152,
+      "step": 5005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4809068825565626,
+      "learning_rate": 3.3134221365108466e-06,
+      "loss": 0.509,
+      "step": 5006
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3551581168063283,
+      "learning_rate": 3.311530340206202e-06,
+      "loss": 0.5268,
+      "step": 5007
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4109766464130555,
+      "learning_rate": 3.3096388166887007e-06,
+      "loss": 0.5166,
+      "step": 5008
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3582712933874206,
+      "learning_rate": 3.307747566263937e-06,
+      "loss": 0.5598,
+      "step": 5009
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4505331602252791,
+      "learning_rate": 3.3058565892374584e-06,
+      "loss": 0.5152,
+      "step": 5010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.819511271452722,
+      "learning_rate": 3.3039658859147683e-06,
+      "loss": 0.4617,
+      "step": 5011
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.710710478621651,
+      "learning_rate": 3.3020754566013256e-06,
+      "loss": 0.521,
+      "step": 5012
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.821480222707406,
+      "learning_rate": 3.300185301602549e-06,
+      "loss": 0.5599,
+      "step": 5013
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.206968709126485,
+      "learning_rate": 3.2982954212238056e-06,
+      "loss": 0.5515,
+      "step": 5014
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.397308898229944,
+      "learning_rate": 3.2964058157704247e-06,
+      "loss": 0.4342,
+      "step": 5015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.333446236835325,
+      "learning_rate": 3.2945164855476885e-06,
+      "loss": 0.4645,
+      "step": 5016
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8032287899771318,
+      "learning_rate": 3.292627430860833e-06,
+      "loss": 0.4875,
+      "step": 5017
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5778051329605842,
+      "learning_rate": 3.2907386520150553e-06,
+      "loss": 0.4935,
+      "step": 5018
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4134204464925468,
+      "learning_rate": 3.288850149315501e-06,
+      "loss": 0.5004,
+      "step": 5019
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3615187047308002,
+      "learning_rate": 3.2869619230672765e-06,
+      "loss": 0.5062,
+      "step": 5020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4682549324756706,
+      "learning_rate": 3.28507397357544e-06,
+      "loss": 0.4583,
+      "step": 5021
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5559248236760226,
+      "learning_rate": 3.2831863011450073e-06,
+      "loss": 0.4816,
+      "step": 5022
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.475983413954976,
+      "learning_rate": 3.2812989060809474e-06,
+      "loss": 0.5502,
+      "step": 5023
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3817847332999171,
+      "learning_rate": 3.279411788688186e-06,
+      "loss": 0.4537,
+      "step": 5024
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4282415662444998,
+      "learning_rate": 3.2775249492716034e-06,
+      "loss": 0.4906,
+      "step": 5025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6435460317787725,
+      "learning_rate": 3.275638388136036e-06,
+      "loss": 0.5215,
+      "step": 5026
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6332556687334305,
+      "learning_rate": 3.2737521055862744e-06,
+      "loss": 0.501,
+      "step": 5027
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.655276725205797,
+      "learning_rate": 3.2718661019270624e-06,
+      "loss": 0.5005,
+      "step": 5028
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3042666406087373,
+      "learning_rate": 3.269980377463103e-06,
+      "loss": 0.472,
+      "step": 5029
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.918025369784106,
+      "learning_rate": 3.2680949324990497e-06,
+      "loss": 0.5144,
+      "step": 5030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4987614591058473,
+      "learning_rate": 3.2662097673395134e-06,
+      "loss": 0.4655,
+      "step": 5031
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6840112024608185,
+      "learning_rate": 3.264324882289058e-06,
+      "loss": 0.484,
+      "step": 5032
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.361880385408784,
+      "learning_rate": 3.2624402776522046e-06,
+      "loss": 0.5415,
+      "step": 5033
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2987058792137038,
+      "learning_rate": 3.260555953733428e-06,
+      "loss": 0.4534,
+      "step": 5034
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6933632604516184,
+      "learning_rate": 3.2586719108371567e-06,
+      "loss": 0.4754,
+      "step": 5035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.621284804628328,
+      "learning_rate": 3.2567881492677746e-06,
+      "loss": 0.4625,
+      "step": 5036
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4867385124248171,
+      "learning_rate": 3.2549046693296198e-06,
+      "loss": 0.4758,
+      "step": 5037
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4309626921575582,
+      "learning_rate": 3.2530214713269853e-06,
+      "loss": 0.463,
+      "step": 5038
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3982991375504024,
+      "learning_rate": 3.251138555564118e-06,
+      "loss": 0.5258,
+      "step": 5039
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8473583948807533,
+      "learning_rate": 3.2492559223452192e-06,
+      "loss": 0.5246,
+      "step": 5040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6653642314561725,
+      "learning_rate": 3.247373571974445e-06,
+      "loss": 0.4932,
+      "step": 5041
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.430380196755407,
+      "learning_rate": 3.2454915047559064e-06,
+      "loss": 0.5089,
+      "step": 5042
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.314125842643514,
+      "learning_rate": 3.2436097209936678e-06,
+      "loss": 0.4556,
+      "step": 5043
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8527143155929457,
+      "learning_rate": 3.241728220991748e-06,
+      "loss": 0.4847,
+      "step": 5044
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7999529352552075,
+      "learning_rate": 3.2398470050541187e-06,
+      "loss": 0.5358,
+      "step": 5045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7267699225161,
+      "learning_rate": 3.2379660734847085e-06,
+      "loss": 0.5211,
+      "step": 5046
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6039494417327065,
+      "learning_rate": 3.2360854265873975e-06,
+      "loss": 0.4832,
+      "step": 5047
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6272571061404455,
+      "learning_rate": 3.234205064666021e-06,
+      "loss": 0.5201,
+      "step": 5048
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5485993016422799,
+      "learning_rate": 3.232324988024368e-06,
+      "loss": 0.4914,
+      "step": 5049
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6654638703390399,
+      "learning_rate": 3.230445196966181e-06,
+      "loss": 0.5096,
+      "step": 5050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7206764273073127,
+      "learning_rate": 3.228565691795158e-06,
+      "loss": 0.4857,
+      "step": 5051
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6704633349447144,
+      "learning_rate": 3.226686472814948e-06,
+      "loss": 0.5206,
+      "step": 5052
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6874928554487323,
+      "learning_rate": 3.2248075403291573e-06,
+      "loss": 0.4941,
+      "step": 5053
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.448843695999027,
+      "learning_rate": 3.222928894641343e-06,
+      "loss": 0.4932,
+      "step": 5054
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9136320642533393,
+      "learning_rate": 3.2210505360550157e-06,
+      "loss": 0.5233,
+      "step": 5055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4908347517455087,
+      "learning_rate": 3.2191724648736434e-06,
+      "loss": 0.4807,
+      "step": 5056
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2934989746182866,
+      "learning_rate": 3.217294681400643e-06,
+      "loss": 0.4955,
+      "step": 5057
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.580504193882281,
+      "learning_rate": 3.2154171859393847e-06,
+      "loss": 0.4419,
+      "step": 5058
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6964366519154959,
+      "learning_rate": 3.213539978793201e-06,
+      "loss": 0.4858,
+      "step": 5059
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5524515714090368,
+      "learning_rate": 3.2116630602653665e-06,
+      "loss": 0.475,
+      "step": 5060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.167181040290641,
+      "learning_rate": 3.2097864306591143e-06,
+      "loss": 0.4805,
+      "step": 5061
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5429939849232062,
+      "learning_rate": 3.2079100902776318e-06,
+      "loss": 0.5109,
+      "step": 5062
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4068352002330462,
+      "learning_rate": 3.2060340394240567e-06,
+      "loss": 0.4986,
+      "step": 5063
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4641161058301413,
+      "learning_rate": 3.204158278401483e-06,
+      "loss": 0.4867,
+      "step": 5064
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5267975713122197,
+      "learning_rate": 3.2022828075129553e-06,
+      "loss": 0.5088,
+      "step": 5065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4808147883075506,
+      "learning_rate": 3.2004076270614714e-06,
+      "loss": 0.4332,
+      "step": 5066
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6800443743499096,
+      "learning_rate": 3.1985327373499864e-06,
+      "loss": 0.4791,
+      "step": 5067
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.136655646383344,
+      "learning_rate": 3.196658138681404e-06,
+      "loss": 0.5061,
+      "step": 5068
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.450993950615321,
+      "learning_rate": 3.1947838313585823e-06,
+      "loss": 0.5267,
+      "step": 5069
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5025753880323134,
+      "learning_rate": 3.1929098156843307e-06,
+      "loss": 0.5506,
+      "step": 5070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.525847759768212,
+      "learning_rate": 3.1910360919614135e-06,
+      "loss": 0.4512,
+      "step": 5071
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4025262095606097,
+      "learning_rate": 3.189162660492548e-06,
+      "loss": 0.4771,
+      "step": 5072
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5297814691298464,
+      "learning_rate": 3.1872895215804035e-06,
+      "loss": 0.5194,
+      "step": 5073
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.647810789148842,
+      "learning_rate": 3.1854166755275982e-06,
+      "loss": 0.4734,
+      "step": 5074
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0943659345513552,
+      "learning_rate": 3.1835441226367137e-06,
+      "loss": 0.5102,
+      "step": 5075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4940861348841228,
+      "learning_rate": 3.181671863210274e-06,
+      "loss": 0.5145,
+      "step": 5076
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.642438898262156,
+      "learning_rate": 3.1797998975507594e-06,
+      "loss": 0.5052,
+      "step": 5077
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7155625181979648,
+      "learning_rate": 3.1779282259606026e-06,
+      "loss": 0.5192,
+      "step": 5078
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4464801924254893,
+      "learning_rate": 3.1760568487421873e-06,
+      "loss": 0.5187,
+      "step": 5079
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2066580365905597,
+      "learning_rate": 3.1741857661978528e-06,
+      "loss": 0.4903,
+      "step": 5080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.528292843093238,
+      "learning_rate": 3.1723149786298867e-06,
+      "loss": 0.5126,
+      "step": 5081
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5730967438515793,
+      "learning_rate": 3.1704444863405314e-06,
+      "loss": 0.4854,
+      "step": 5082
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4127354692919383,
+      "learning_rate": 3.1685742896319847e-06,
+      "loss": 0.5259,
+      "step": 5083
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5901344549251768,
+      "learning_rate": 3.1667043888063914e-06,
+      "loss": 0.5187,
+      "step": 5084
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.510968311326373,
+      "learning_rate": 3.16483478416585e-06,
+      "loss": 0.5035,
+      "step": 5085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.103291137686788,
+      "learning_rate": 3.1629654760124117e-06,
+      "loss": 0.4973,
+      "step": 5086
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.362060770332344,
+      "learning_rate": 3.1610964646480806e-06,
+      "loss": 0.4657,
+      "step": 5087
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3181780493491968,
+      "learning_rate": 3.159227750374812e-06,
+      "loss": 0.4677,
+      "step": 5088
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6546426254608904,
+      "learning_rate": 3.157359333494511e-06,
+      "loss": 0.5134,
+      "step": 5089
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.43936413406887,
+      "learning_rate": 3.155491214309039e-06,
+      "loss": 0.4698,
+      "step": 5090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2386937886476648,
+      "learning_rate": 3.153623393120203e-06,
+      "loss": 0.462,
+      "step": 5091
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4513888189237607,
+      "learning_rate": 3.1517558702297724e-06,
+      "loss": 0.4814,
+      "step": 5092
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3730832063641922,
+      "learning_rate": 3.1498886459394585e-06,
+      "loss": 0.4829,
+      "step": 5093
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4175167888787246,
+      "learning_rate": 3.1480217205509282e-06,
+      "loss": 0.5328,
+      "step": 5094
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8484807418961515,
+      "learning_rate": 3.1461550943657996e-06,
+      "loss": 0.4671,
+      "step": 5095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7497109505535595,
+      "learning_rate": 3.144288767685643e-06,
+      "loss": 0.5283,
+      "step": 5096
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3495861137244243,
+      "learning_rate": 3.1424227408119803e-06,
+      "loss": 0.5009,
+      "step": 5097
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.373311925624968,
+      "learning_rate": 3.1405570140462833e-06,
+      "loss": 0.4901,
+      "step": 5098
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2854736848654358,
+      "learning_rate": 3.138691587689975e-06,
+      "loss": 0.4316,
+      "step": 5099
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.506737083470091,
+      "learning_rate": 3.1368264620444356e-06,
+      "loss": 0.4619,
+      "step": 5100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.885061084921074,
+      "learning_rate": 3.1349616374109903e-06,
+      "loss": 0.5372,
+      "step": 5101
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4289873958250099,
+      "learning_rate": 3.1330971140909184e-06,
+      "loss": 0.4549,
+      "step": 5102
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3060382248016653,
+      "learning_rate": 3.131232892385451e-06,
+      "loss": 0.5197,
+      "step": 5103
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0688664902976086,
+      "learning_rate": 3.1293689725957667e-06,
+      "loss": 0.5058,
+      "step": 5104
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.829403295529987,
+      "learning_rate": 3.1275053550230005e-06,
+      "loss": 0.4975,
+      "step": 5105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.430231383079677,
+      "learning_rate": 3.1256420399682365e-06,
+      "loss": 0.4455,
+      "step": 5106
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9110115547985471,
+      "learning_rate": 3.1237790277325076e-06,
+      "loss": 0.4466,
+      "step": 5107
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3324536100635709,
+      "learning_rate": 3.121916318616801e-06,
+      "loss": 0.5126,
+      "step": 5108
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3753593719641783,
+      "learning_rate": 3.1200539129220548e-06,
+      "loss": 0.5434,
+      "step": 5109
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3814478368048255,
+      "learning_rate": 3.118191810949156e-06,
+      "loss": 0.5164,
+      "step": 5110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9334061670531075,
+      "learning_rate": 3.1163300129989434e-06,
+      "loss": 0.5299,
+      "step": 5111
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.300266009222235,
+      "learning_rate": 3.114468519372207e-06,
+      "loss": 0.5049,
+      "step": 5112
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3880272247312915,
+      "learning_rate": 3.1126073303696873e-06,
+      "loss": 0.4898,
+      "step": 5113
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3431655578387383,
+      "learning_rate": 3.1107464462920767e-06,
+      "loss": 0.5187,
+      "step": 5114
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3546211679493592,
+      "learning_rate": 3.108885867440015e-06,
+      "loss": 0.531,
+      "step": 5115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.371488891276787,
+      "learning_rate": 3.1070255941140987e-06,
+      "loss": 0.4707,
+      "step": 5116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3519675577207597,
+      "learning_rate": 3.105165626614869e-06,
+      "loss": 0.5278,
+      "step": 5117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4497383499852987,
+      "learning_rate": 3.103305965242821e-06,
+      "loss": 0.461,
+      "step": 5118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6053194736104737,
+      "learning_rate": 3.1014466102983986e-06,
+      "loss": 0.4781,
+      "step": 5119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.649216465541498,
+      "learning_rate": 3.099587562081997e-06,
+      "loss": 0.4484,
+      "step": 5120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7489882471719203,
+      "learning_rate": 3.0977288208939627e-06,
+      "loss": 0.4445,
+      "step": 5121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7986665377692794,
+      "learning_rate": 3.09587038703459e-06,
+      "loss": 0.4882,
+      "step": 5122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3269475764699166,
+      "learning_rate": 3.094012260804127e-06,
+      "loss": 0.4894,
+      "step": 5123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3517778780960439,
+      "learning_rate": 3.0921544425027695e-06,
+      "loss": 0.5173,
+      "step": 5124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.334966564160536,
+      "learning_rate": 3.0902969324306643e-06,
+      "loss": 0.4906,
+      "step": 5125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6757784490718937,
+      "learning_rate": 3.0884397308879098e-06,
+      "loss": 0.5075,
+      "step": 5126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6295839395714828,
+      "learning_rate": 3.0865828381745515e-06,
+      "loss": 0.4894,
+      "step": 5127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5048313436145033,
+      "learning_rate": 3.0847262545905882e-06,
+      "loss": 0.5114,
+      "step": 5128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4335249322231058,
+      "learning_rate": 3.0828699804359663e-06,
+      "loss": 0.4923,
+      "step": 5129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3866191188244976,
+      "learning_rate": 3.081014016010584e-06,
+      "loss": 0.5375,
+      "step": 5130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4240045872604474,
+      "learning_rate": 3.0791583616142883e-06,
+      "loss": 0.4682,
+      "step": 5131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.161237880252194,
+      "learning_rate": 3.0773030175468754e-06,
+      "loss": 0.4879,
+      "step": 5132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7629290611916298,
+      "learning_rate": 3.0754479841080943e-06,
+      "loss": 0.4959,
+      "step": 5133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4955674586673207,
+      "learning_rate": 3.0735932615976416e-06,
+      "loss": 0.5465,
+      "step": 5134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3894155976672427,
+      "learning_rate": 3.071738850315164e-06,
+      "loss": 0.4957,
+      "step": 5135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.442652357430899,
+      "learning_rate": 3.0698847505602576e-06,
+      "loss": 0.5341,
+      "step": 5136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7858544192469936,
+      "learning_rate": 3.0680309626324685e-06,
+      "loss": 0.5084,
+      "step": 5137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2466849264870778,
+      "learning_rate": 3.0661774868312928e-06,
+      "loss": 0.5389,
+      "step": 5138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6570292869455046,
+      "learning_rate": 3.064324323456176e-06,
+      "loss": 0.5112,
+      "step": 5139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3190113984903318,
+      "learning_rate": 3.0624714728065106e-06,
+      "loss": 0.4613,
+      "step": 5140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4251908409537135,
+      "learning_rate": 3.060618935181645e-06,
+      "loss": 0.4501,
+      "step": 5141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4261908907765468,
+      "learning_rate": 3.0587667108808706e-06,
+      "loss": 0.4794,
+      "step": 5142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.536541083783513,
+      "learning_rate": 3.056914800203431e-06,
+      "loss": 0.516,
+      "step": 5143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2948676856612353,
+      "learning_rate": 3.0550632034485186e-06,
+      "loss": 0.495,
+      "step": 5144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.949593745483904,
+      "learning_rate": 3.053211920915275e-06,
+      "loss": 0.4635,
+      "step": 5145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4042463519036232,
+      "learning_rate": 3.0513609529027914e-06,
+      "loss": 0.4489,
+      "step": 5146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5353044882980715,
+      "learning_rate": 3.049510299710108e-06,
+      "loss": 0.5417,
+      "step": 5147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.368567437939403,
+      "learning_rate": 3.0476599616362136e-06,
+      "loss": 0.5106,
+      "step": 5148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.602691605831889,
+      "learning_rate": 3.045809938980047e-06,
+      "loss": 0.4857,
+      "step": 5149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4870178731861237,
+      "learning_rate": 3.0439602320404964e-06,
+      "loss": 0.4919,
+      "step": 5150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5277486373103906,
+      "learning_rate": 3.0421108411163975e-06,
+      "loss": 0.4446,
+      "step": 5151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7414214443891542,
+      "learning_rate": 3.040261766506536e-06,
+      "loss": 0.507,
+      "step": 5152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7320551130735706,
+      "learning_rate": 3.038413008509645e-06,
+      "loss": 0.491,
+      "step": 5153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5798003564239524,
+      "learning_rate": 3.0365645674244094e-06,
+      "loss": 0.4709,
+      "step": 5154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6953870389910313,
+      "learning_rate": 3.03471644354946e-06,
+      "loss": 0.5186,
+      "step": 5155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2829814597360416,
+      "learning_rate": 3.0328686371833765e-06,
+      "loss": 0.4842,
+      "step": 5156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.520157858202863,
+      "learning_rate": 3.031021148624691e-06,
+      "loss": 0.4392,
+      "step": 5157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.584303013751112,
+      "learning_rate": 3.0291739781718808e-06,
+      "loss": 0.5274,
+      "step": 5158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4803637139074282,
+      "learning_rate": 3.0273271261233718e-06,
+      "loss": 0.4807,
+      "step": 5159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5198555339306656,
+      "learning_rate": 3.025480592777539e-06,
+      "loss": 0.5198,
+      "step": 5160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7245927757697032,
+      "learning_rate": 3.023634378432706e-06,
+      "loss": 0.4865,
+      "step": 5161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5536795950544744,
+      "learning_rate": 3.021788483387146e-06,
+      "loss": 0.5101,
+      "step": 5162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.739620434569134,
+      "learning_rate": 3.019942907939079e-06,
+      "loss": 0.5367,
+      "step": 5163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3241117537551723,
+      "learning_rate": 3.0180976523866717e-06,
+      "loss": 0.4565,
+      "step": 5164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3465344224268267,
+      "learning_rate": 3.016252717028046e-06,
+      "loss": 0.4703,
+      "step": 5165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3999767438476154,
+      "learning_rate": 3.0144081021612648e-06,
+      "loss": 0.4976,
+      "step": 5166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3766289113380032,
+      "learning_rate": 3.0125638080843435e-06,
+      "loss": 0.4953,
+      "step": 5167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6140554152088646,
+      "learning_rate": 3.0107198350952415e-06,
+      "loss": 0.5032,
+      "step": 5168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6590837178206382,
+      "learning_rate": 3.0088761834918706e-06,
+      "loss": 0.5041,
+      "step": 5169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.092104546557681,
+      "learning_rate": 3.0070328535720884e-06,
+      "loss": 0.5199,
+      "step": 5170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4407452582905922,
+      "learning_rate": 3.0051898456337013e-06,
+      "loss": 0.5068,
+      "step": 5171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9815869758106484,
+      "learning_rate": 3.003347159974463e-06,
+      "loss": 0.5009,
+      "step": 5172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2769554752225565,
+      "learning_rate": 3.001504796892074e-06,
+      "loss": 0.4757,
+      "step": 5173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.483988926560004,
+      "learning_rate": 2.9996627566841886e-06,
+      "loss": 0.4817,
+      "step": 5174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3562102230857813,
+      "learning_rate": 2.9978210396484013e-06,
+      "loss": 0.4965,
+      "step": 5175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7509090416933895,
+      "learning_rate": 2.99597964608226e-06,
+      "loss": 0.4859,
+      "step": 5176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3193288728949342,
+      "learning_rate": 2.9941385762832563e-06,
+      "loss": 0.4898,
+      "step": 5177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4786685504426724,
+      "learning_rate": 2.9922978305488317e-06,
+      "loss": 0.5607,
+      "step": 5178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4479667494823674,
+      "learning_rate": 2.990457409176375e-06,
+      "loss": 0.5253,
+      "step": 5179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4814821436756966,
+      "learning_rate": 2.9886173124632213e-06,
+      "loss": 0.4418,
+      "step": 5180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.674004839372321,
+      "learning_rate": 2.986777540706654e-06,
+      "loss": 0.4746,
+      "step": 5181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3409942229031568,
+      "learning_rate": 2.984938094203908e-06,
+      "loss": 0.492,
+      "step": 5182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3204528480802598,
+      "learning_rate": 2.98309897325216e-06,
+      "loss": 0.482,
+      "step": 5183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2290584335361103,
+      "learning_rate": 2.9812601781485356e-06,
+      "loss": 0.5398,
+      "step": 5184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4015321794285387,
+      "learning_rate": 2.9794217091901094e-06,
+      "loss": 0.5118,
+      "step": 5185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1052379988251193,
+      "learning_rate": 2.9775835666739028e-06,
+      "loss": 0.4802,
+      "step": 5186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4781251630529064,
+      "learning_rate": 2.975745750896881e-06,
+      "loss": 0.4244,
+      "step": 5187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5160119248805892,
+      "learning_rate": 2.973908262155962e-06,
+      "loss": 0.4722,
+      "step": 5188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0726917074692084,
+      "learning_rate": 2.9720711007480056e-06,
+      "loss": 0.5402,
+      "step": 5189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8211951470508632,
+      "learning_rate": 2.9702342669698247e-06,
+      "loss": 0.4835,
+      "step": 5190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5279731366187965,
+      "learning_rate": 2.968397761118175e-06,
+      "loss": 0.4654,
+      "step": 5191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4769202027950719,
+      "learning_rate": 2.9665615834897597e-06,
+      "loss": 0.4981,
+      "step": 5192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.429049044873586,
+      "learning_rate": 2.9647257343812298e-06,
+      "loss": 0.4863,
+      "step": 5193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.712926248039558,
+      "learning_rate": 2.9628902140891823e-06,
+      "loss": 0.5002,
+      "step": 5194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4039040905005618,
+      "learning_rate": 2.961055022910162e-06,
+      "loss": 0.4909,
+      "step": 5195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5767933463615202,
+      "learning_rate": 2.9592201611406606e-06,
+      "loss": 0.5375,
+      "step": 5196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.450513560065177,
+      "learning_rate": 2.957385629077113e-06,
+      "loss": 0.4729,
+      "step": 5197
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3789791613221432,
+      "learning_rate": 2.955551427015909e-06,
+      "loss": 0.5128,
+      "step": 5198
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0296334637657703,
+      "learning_rate": 2.953717555253378e-06,
+      "loss": 0.5196,
+      "step": 5199
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4936191092661981,
+      "learning_rate": 2.951884014085798e-06,
+      "loss": 0.4448,
+      "step": 5200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4903974201120964,
+      "learning_rate": 2.9500508038093932e-06,
+      "loss": 0.5354,
+      "step": 5201
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.674069417016388,
+      "learning_rate": 2.9482179247203357e-06,
+      "loss": 0.5179,
+      "step": 5202
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3382165010372378,
+      "learning_rate": 2.9463853771147434e-06,
+      "loss": 0.5242,
+      "step": 5203
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4771001395371512,
+      "learning_rate": 2.94455316128868e-06,
+      "loss": 0.4246,
+      "step": 5204
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.437699263205684,
+      "learning_rate": 2.942721277538154e-06,
+      "loss": 0.4561,
+      "step": 5205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5707752606080958,
+      "learning_rate": 2.940889726159127e-06,
+      "loss": 0.5656,
+      "step": 5206
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3425068014198047,
+      "learning_rate": 2.939058507447499e-06,
+      "loss": 0.5571,
+      "step": 5207
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6390851225460445,
+      "learning_rate": 2.9372276216991204e-06,
+      "loss": 0.509,
+      "step": 5208
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4889246574510457,
+      "learning_rate": 2.9353970692097865e-06,
+      "loss": 0.4664,
+      "step": 5209
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.421107639733984,
+      "learning_rate": 2.9335668502752395e-06,
+      "loss": 0.5479,
+      "step": 5210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5940583579838103,
+      "learning_rate": 2.9317369651911677e-06,
+      "loss": 0.465,
+      "step": 5211
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5987081821051372,
+      "learning_rate": 2.9299074142532045e-06,
+      "loss": 0.4642,
+      "step": 5212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2117746474994493,
+      "learning_rate": 2.9280781977569306e-06,
+      "loss": 0.4773,
+      "step": 5213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8658518759658125,
+      "learning_rate": 2.9262493159978703e-06,
+      "loss": 0.468,
+      "step": 5214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3225637881539518,
+      "learning_rate": 2.924420769271499e-06,
+      "loss": 0.469,
+      "step": 5215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3203399947443093,
+      "learning_rate": 2.922592557873231e-06,
+      "loss": 0.4908,
+      "step": 5216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.328462249497007,
+      "learning_rate": 2.9207646820984325e-06,
+      "loss": 0.5081,
+      "step": 5217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6711124618967272,
+      "learning_rate": 2.9189371422424123e-06,
+      "loss": 0.5171,
+      "step": 5218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6230161561337815,
+      "learning_rate": 2.917109938600423e-06,
+      "loss": 0.4466,
+      "step": 5219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3219255341465752,
+      "learning_rate": 2.9152830714676706e-06,
+      "loss": 0.4642,
+      "step": 5220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4107900141016096,
+      "learning_rate": 2.9134565411392958e-06,
+      "loss": 0.5056,
+      "step": 5221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4295451490426954,
+      "learning_rate": 2.9116303479103934e-06,
+      "loss": 0.5116,
+      "step": 5222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.856583146682241,
+      "learning_rate": 2.909804492076001e-06,
+      "loss": 0.493,
+      "step": 5223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.729174337478149,
+      "learning_rate": 2.9079789739311037e-06,
+      "loss": 0.5561,
+      "step": 5224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6528515005726849,
+      "learning_rate": 2.906153793770626e-06,
+      "loss": 0.5052,
+      "step": 5225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.393222031708734,
+      "learning_rate": 2.904328951889447e-06,
+      "loss": 0.4537,
+      "step": 5226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3073884655399801,
+      "learning_rate": 2.9025044485823815e-06,
+      "loss": 0.5334,
+      "step": 5227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.284878231455661,
+      "learning_rate": 2.900680284144194e-06,
+      "loss": 0.5229,
+      "step": 5228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0539141190578754,
+      "learning_rate": 2.898856458869597e-06,
+      "loss": 0.4971,
+      "step": 5229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4554782786293619,
+      "learning_rate": 2.897032973053241e-06,
+      "loss": 0.487,
+      "step": 5230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5009432100606122,
+      "learning_rate": 2.895209826989733e-06,
+      "loss": 0.5017,
+      "step": 5231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3348701171976673,
+      "learning_rate": 2.8933870209736136e-06,
+      "loss": 0.4668,
+      "step": 5232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4133037167718843,
+      "learning_rate": 2.8915645552993756e-06,
+      "loss": 0.484,
+      "step": 5233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2891795152863093,
+      "learning_rate": 2.889742430261452e-06,
+      "loss": 0.4565,
+      "step": 5234
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5537923383669643,
+      "learning_rate": 2.8879206461542253e-06,
+      "loss": 0.4766,
+      "step": 5235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.482406989146018,
+      "learning_rate": 2.8860992032720204e-06,
+      "loss": 0.479,
+      "step": 5236
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3677960555410495,
+      "learning_rate": 2.8842781019091037e-06,
+      "loss": 0.5023,
+      "step": 5237
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4134634784188358,
+      "learning_rate": 2.8824573423596946e-06,
+      "loss": 0.4854,
+      "step": 5238
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7260629179279658,
+      "learning_rate": 2.8806369249179513e-06,
+      "loss": 0.5403,
+      "step": 5239
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5421852186392826,
+      "learning_rate": 2.87881684987798e-06,
+      "loss": 0.5302,
+      "step": 5240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7102444314625498,
+      "learning_rate": 2.876997117533826e-06,
+      "loss": 0.508,
+      "step": 5241
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.350860727657689,
+      "learning_rate": 2.875177728179488e-06,
+      "loss": 0.5409,
+      "step": 5242
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6972621262407982,
+      "learning_rate": 2.8733586821088998e-06,
+      "loss": 0.5073,
+      "step": 5243
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4389568481054311,
+      "learning_rate": 2.871539979615948e-06,
+      "loss": 0.475,
+      "step": 5244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.44297013306258,
+      "learning_rate": 2.8697216209944585e-06,
+      "loss": 0.5036,
+      "step": 5245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6421364177474709,
+      "learning_rate": 2.8679036065382003e-06,
+      "loss": 0.4985,
+      "step": 5246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3111731981965287,
+      "learning_rate": 2.8660859365408934e-06,
+      "loss": 0.5307,
+      "step": 5247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.370112143441277,
+      "learning_rate": 2.8642686112961964e-06,
+      "loss": 0.5473,
+      "step": 5248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6811937781641748,
+      "learning_rate": 2.8624516310977172e-06,
+      "loss": 0.4813,
+      "step": 5249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5384833532133326,
+      "learning_rate": 2.860634996239001e-06,
+      "loss": 0.5112,
+      "step": 5250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3948737627803693,
+      "learning_rate": 2.858818707013545e-06,
+      "loss": 0.4642,
+      "step": 5251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2792735675727973,
+      "learning_rate": 2.8570027637147835e-06,
+      "loss": 0.5108,
+      "step": 5252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.366292951276149,
+      "learning_rate": 2.8551871666361e-06,
+      "loss": 0.5163,
+      "step": 5253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4864720122990032,
+      "learning_rate": 2.8533719160708186e-06,
+      "loss": 0.5041,
+      "step": 5254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.570123238810715,
+      "learning_rate": 2.8515570123122094e-06,
+      "loss": 0.5027,
+      "step": 5255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.351478202009039,
+      "learning_rate": 2.8497424556534893e-06,
+      "loss": 0.4334,
+      "step": 5256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3568254180690646,
+      "learning_rate": 2.84792824638781e-06,
+      "loss": 0.526,
+      "step": 5257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3958890681102574,
+      "learning_rate": 2.8461143848082793e-06,
+      "loss": 0.4824,
+      "step": 5258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.56032635116221,
+      "learning_rate": 2.844300871207937e-06,
+      "loss": 0.5152,
+      "step": 5259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4501231717596466,
+      "learning_rate": 2.842487705879777e-06,
+      "loss": 0.4875,
+      "step": 5260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4516955521205268,
+      "learning_rate": 2.840674889116728e-06,
+      "loss": 0.492,
+      "step": 5261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.1679179628063483,
+      "learning_rate": 2.83886242121167e-06,
+      "loss": 0.4231,
+      "step": 5262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.391145838108834,
+      "learning_rate": 2.8370503024574192e-06,
+      "loss": 0.5274,
+      "step": 5263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.432913453073426,
+      "learning_rate": 2.835238533146741e-06,
+      "loss": 0.5189,
+      "step": 5264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.306301674370133,
+      "learning_rate": 2.8334271135723468e-06,
+      "loss": 0.5035,
+      "step": 5265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6885766126397033,
+      "learning_rate": 2.8316160440268813e-06,
+      "loss": 0.5766,
+      "step": 5266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.407825610945716,
+      "learning_rate": 2.8298053248029434e-06,
+      "loss": 0.4842,
+      "step": 5267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5167359860431884,
+      "learning_rate": 2.8279949561930665e-06,
+      "loss": 0.5149,
+      "step": 5268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.384774347864669,
+      "learning_rate": 2.8261849384897353e-06,
+      "loss": 0.4946,
+      "step": 5269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7555210761499582,
+      "learning_rate": 2.8243752719853714e-06,
+      "loss": 0.4819,
+      "step": 5270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.27797541813699,
+      "learning_rate": 2.822565956972342e-06,
+      "loss": 0.4937,
+      "step": 5271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.917424442556268,
+      "learning_rate": 2.8207569937429626e-06,
+      "loss": 0.4998,
+      "step": 5272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6865994988327913,
+      "learning_rate": 2.8189483825894813e-06,
+      "loss": 0.4762,
+      "step": 5273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.480227158958637,
+      "learning_rate": 2.8171401238041007e-06,
+      "loss": 0.5159,
+      "step": 5274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3659359585215833,
+      "learning_rate": 2.8153322176789556e-06,
+      "loss": 0.4841,
+      "step": 5275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8040659390676708,
+      "learning_rate": 2.813524664506133e-06,
+      "loss": 0.4936,
+      "step": 5276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2706477624521206,
+      "learning_rate": 2.811717464577657e-06,
+      "loss": 0.4966,
+      "step": 5277
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.493513927680801,
+      "learning_rate": 2.8099106181854974e-06,
+      "loss": 0.5047,
+      "step": 5278
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8631817433842472,
+      "learning_rate": 2.8081041256215654e-06,
+      "loss": 0.489,
+      "step": 5279
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1856851549371594,
+      "learning_rate": 2.8062979871777157e-06,
+      "loss": 0.4861,
+      "step": 5280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.471930383855249,
+      "learning_rate": 2.8044922031457487e-06,
+      "loss": 0.5283,
+      "step": 5281
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.025917538098957,
+      "learning_rate": 2.8026867738174013e-06,
+      "loss": 0.5313,
+      "step": 5282
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.358502126189759,
+      "learning_rate": 2.8008816994843592e-06,
+      "loss": 0.5362,
+      "step": 5283
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.526702975365077,
+      "learning_rate": 2.7990769804382446e-06,
+      "loss": 0.5274,
+      "step": 5284
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.7768441835174045,
+      "learning_rate": 2.79727261697063e-06,
+      "loss": 0.4959,
+      "step": 5285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5861308351808623,
+      "learning_rate": 2.7954686093730216e-06,
+      "loss": 0.4983,
+      "step": 5286
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2829002137593568,
+      "learning_rate": 2.7936649579368776e-06,
+      "loss": 0.4543,
+      "step": 5287
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3304196406298983,
+      "learning_rate": 2.791861662953589e-06,
+      "loss": 0.4866,
+      "step": 5288
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5286558611134053,
+      "learning_rate": 2.790058724714496e-06,
+      "loss": 0.4656,
+      "step": 5289
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3558140277966517,
+      "learning_rate": 2.7882561435108823e-06,
+      "loss": 0.4728,
+      "step": 5290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.453425915868669,
+      "learning_rate": 2.7864539196339658e-06,
+      "loss": 0.4677,
+      "step": 5291
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3025199072127507,
+      "learning_rate": 2.784652053374915e-06,
+      "loss": 0.478,
+      "step": 5292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.35504074531167,
+      "learning_rate": 2.7828505450248343e-06,
+      "loss": 0.4876,
+      "step": 5293
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4178035232568393,
+      "learning_rate": 2.7810493948747775e-06,
+      "loss": 0.4972,
+      "step": 5294
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4695984310786603,
+      "learning_rate": 2.779248603215731e-06,
+      "loss": 0.5301,
+      "step": 5295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4901468728111962,
+      "learning_rate": 2.777448170338632e-06,
+      "loss": 0.4498,
+      "step": 5296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3709843878126562,
+      "learning_rate": 2.775648096534357e-06,
+      "loss": 0.4982,
+      "step": 5297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6853526249118413,
+      "learning_rate": 2.7738483820937208e-06,
+      "loss": 0.4892,
+      "step": 5298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5074452665631897,
+      "learning_rate": 2.7720490273074865e-06,
+      "loss": 0.5364,
+      "step": 5299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.555266495159016,
+      "learning_rate": 2.7702500324663518e-06,
+      "loss": 0.54,
+      "step": 5300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.48752313122983,
+      "learning_rate": 2.768451397860964e-06,
+      "loss": 0.5542,
+      "step": 5301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.351707167167519,
+      "learning_rate": 2.766653123781905e-06,
+      "loss": 0.502,
+      "step": 5302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.080333246439908,
+      "learning_rate": 2.7648552105197046e-06,
+      "loss": 0.5655,
+      "step": 5303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.567104012954525,
+      "learning_rate": 2.763057658364827e-06,
+      "loss": 0.4765,
+      "step": 5304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3716735973388143,
+      "learning_rate": 2.7612604676076902e-06,
+      "loss": 0.5053,
+      "step": 5305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.367634803659491,
+      "learning_rate": 2.759463638538642e-06,
+      "loss": 0.4825,
+      "step": 5306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3321793751367244,
+      "learning_rate": 2.757667171447973e-06,
+      "loss": 0.4742,
+      "step": 5307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4559183914666705,
+      "learning_rate": 2.7558710666259235e-06,
+      "loss": 0.5147,
+      "step": 5308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6637424420725844,
+      "learning_rate": 2.754075324362666e-06,
+      "loss": 0.5307,
+      "step": 5309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7309577384722221,
+      "learning_rate": 2.7522799449483224e-06,
+      "loss": 0.5454,
+      "step": 5310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7461041363532903,
+      "learning_rate": 2.7504849286729475e-06,
+      "loss": 0.5696,
+      "step": 5311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.360186930190867,
+      "learning_rate": 2.7486902758265445e-06,
+      "loss": 0.4063,
+      "step": 5312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4725073849409913,
+      "learning_rate": 2.7468959866990554e-06,
+      "loss": 0.5528,
+      "step": 5313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.83191227764115,
+      "learning_rate": 2.745102061580365e-06,
+      "loss": 0.4732,
+      "step": 5314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5576463111550587,
+      "learning_rate": 2.7433085007602955e-06,
+      "loss": 0.4874,
+      "step": 5315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.615574991409325,
+      "learning_rate": 2.7415153045286108e-06,
+      "loss": 0.5217,
+      "step": 5316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7472158798068373,
+      "learning_rate": 2.7397224731750215e-06,
+      "loss": 0.5164,
+      "step": 5317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.564034329369173,
+      "learning_rate": 2.737930006989172e-06,
+      "loss": 0.5332,
+      "step": 5318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3528228639339923,
+      "learning_rate": 2.7361379062606545e-06,
+      "loss": 0.4847,
+      "step": 5319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3151718626427842,
+      "learning_rate": 2.734346171278992e-06,
+      "loss": 0.5129,
+      "step": 5320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6482216627008905,
+      "learning_rate": 2.7325548023336645e-06,
+      "loss": 0.4735,
+      "step": 5321
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3814757439240166,
+      "learning_rate": 2.7307637997140757e-06,
+      "loss": 0.5166,
+      "step": 5322
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4763492860259062,
+      "learning_rate": 2.728973163709583e-06,
+      "loss": 0.4862,
+      "step": 5323
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.74270202227617,
+      "learning_rate": 2.7271828946094753e-06,
+      "loss": 0.4746,
+      "step": 5324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4641364925937064,
+      "learning_rate": 2.7253929927029897e-06,
+      "loss": 0.4816,
+      "step": 5325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5149325381015646,
+      "learning_rate": 2.7236034582793e-06,
+      "loss": 0.448,
+      "step": 5326
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5913256590745417,
+      "learning_rate": 2.7218142916275174e-06,
+      "loss": 0.5024,
+      "step": 5327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0427625447444466,
+      "learning_rate": 2.720025493036703e-06,
+      "loss": 0.5191,
+      "step": 5328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3152223721438832,
+      "learning_rate": 2.718237062795846e-06,
+      "loss": 0.5157,
+      "step": 5329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5793215264008136,
+      "learning_rate": 2.7164490011938915e-06,
+      "loss": 0.4871,
+      "step": 5330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.41395011217096,
+      "learning_rate": 2.714661308519711e-06,
+      "loss": 0.5235,
+      "step": 5331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5978172239163189,
+      "learning_rate": 2.7128739850621255e-06,
+      "loss": 0.4837,
+      "step": 5332
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.564577961403666,
+      "learning_rate": 2.7110870311098884e-06,
+      "loss": 0.527,
+      "step": 5333
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6697258115680298,
+      "learning_rate": 2.7093004469517027e-06,
+      "loss": 0.5105,
+      "step": 5334
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6120972779403429,
+      "learning_rate": 2.707514232876204e-06,
+      "loss": 0.4547,
+      "step": 5335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7313556878552248,
+      "learning_rate": 2.7057283891719703e-06,
+      "loss": 0.5047,
+      "step": 5336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7503742181212245,
+      "learning_rate": 2.703942916127521e-06,
+      "loss": 0.4938,
+      "step": 5337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8591629343416418,
+      "learning_rate": 2.7021578140313155e-06,
+      "loss": 0.4805,
+      "step": 5338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.280015653078731,
+      "learning_rate": 2.7003730831717545e-06,
+      "loss": 0.4685,
+      "step": 5339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3874985784665126,
+      "learning_rate": 2.6985887238371736e-06,
+      "loss": 0.4851,
+      "step": 5340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4078872115531422,
+      "learning_rate": 2.6968047363158556e-06,
+      "loss": 0.4947,
+      "step": 5341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5915183243210775,
+      "learning_rate": 2.6950211208960147e-06,
+      "loss": 0.5,
+      "step": 5342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.597352790740755,
+      "learning_rate": 2.693237877865814e-06,
+      "loss": 0.4887,
+      "step": 5343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9562463930653387,
+      "learning_rate": 2.6914550075133506e-06,
+      "loss": 0.4693,
+      "step": 5344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.407055832699335,
+      "learning_rate": 2.6896725101266584e-06,
+      "loss": 0.5051,
+      "step": 5345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.657836405659061,
+      "learning_rate": 2.6878903859937245e-06,
+      "loss": 0.4861,
+      "step": 5346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4575434472330062,
+      "learning_rate": 2.686108635402459e-06,
+      "loss": 0.5044,
+      "step": 5347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2441405862275843,
+      "learning_rate": 2.684327258640725e-06,
+      "loss": 0.4904,
+      "step": 5348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3436498811887572,
+      "learning_rate": 2.6825462559963144e-06,
+      "loss": 0.4471,
+      "step": 5349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5582742686000106,
+      "learning_rate": 2.6807656277569694e-06,
+      "loss": 0.5168,
+      "step": 5350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6881226866019212,
+      "learning_rate": 2.67898537421036e-06,
+      "loss": 0.501,
+      "step": 5351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2901602306826498,
+      "learning_rate": 2.677205495644108e-06,
+      "loss": 0.4999,
+      "step": 5352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4483734768467285,
+      "learning_rate": 2.675425992345763e-06,
+      "loss": 0.4864,
+      "step": 5353
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.358456513957977,
+      "learning_rate": 2.673646864602822e-06,
+      "loss": 0.4564,
+      "step": 5354
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3999051743052522,
+      "learning_rate": 2.671868112702721e-06,
+      "loss": 0.5713,
+      "step": 5355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7831927687168119,
+      "learning_rate": 2.6700897369328286e-06,
+      "loss": 0.5226,
+      "step": 5356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4520519300332069,
+      "learning_rate": 2.668311737580461e-06,
+      "loss": 0.5093,
+      "step": 5357
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5009878577871714,
+      "learning_rate": 2.6665341149328667e-06,
+      "loss": 0.5286,
+      "step": 5358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7095770926619317,
+      "learning_rate": 2.6647568692772386e-06,
+      "loss": 0.4946,
+      "step": 5359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5466040717092826,
+      "learning_rate": 2.662980000900704e-06,
+      "loss": 0.5143,
+      "step": 5360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3673672944178537,
+      "learning_rate": 2.661203510090332e-06,
+      "loss": 0.5227,
+      "step": 5361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5160648696253634,
+      "learning_rate": 2.659427397133134e-06,
+      "loss": 0.4959,
+      "step": 5362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5195269244707366,
+      "learning_rate": 2.6576516623160515e-06,
+      "loss": 0.4796,
+      "step": 5363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0608837465904233,
+      "learning_rate": 2.6558763059259745e-06,
+      "loss": 0.4623,
+      "step": 5364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4251657068323844,
+      "learning_rate": 2.6541013282497234e-06,
+      "loss": 0.52,
+      "step": 5365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.582165778558871,
+      "learning_rate": 2.6523267295740663e-06,
+      "loss": 0.4503,
+      "step": 5366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7127631256028564,
+      "learning_rate": 2.6505525101857e-06,
+      "loss": 0.5577,
+      "step": 5367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6587222980531775,
+      "learning_rate": 2.6487786703712692e-06,
+      "loss": 0.4739,
+      "step": 5368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4720907931346792,
+      "learning_rate": 2.6470052104173504e-06,
+      "loss": 0.5673,
+      "step": 5369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7165887666323725,
+      "learning_rate": 2.6452321306104634e-06,
+      "loss": 0.4357,
+      "step": 5370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6261364104457807,
+      "learning_rate": 2.6434594312370664e-06,
+      "loss": 0.5145,
+      "step": 5371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8610267248885162,
+      "learning_rate": 2.641687112583551e-06,
+      "loss": 0.4855,
+      "step": 5372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6204122173348687,
+      "learning_rate": 2.639915174936254e-06,
+      "loss": 0.5469,
+      "step": 5373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3873663278041979,
+      "learning_rate": 2.638143618581445e-06,
+      "loss": 0.5024,
+      "step": 5374
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7635110022929765,
+      "learning_rate": 2.6363724438053377e-06,
+      "loss": 0.4929,
+      "step": 5375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5521889396540993,
+      "learning_rate": 2.6346016508940776e-06,
+      "loss": 0.4955,
+      "step": 5376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6675794203169498,
+      "learning_rate": 2.632831240133754e-06,
+      "loss": 0.5452,
+      "step": 5377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7268314997834946,
+      "learning_rate": 2.631061211810391e-06,
+      "loss": 0.4706,
+      "step": 5378
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1173737686340006,
+      "learning_rate": 2.629291566209952e-06,
+      "loss": 0.4931,
+      "step": 5379
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7066487562954822,
+      "learning_rate": 2.627522303618343e-06,
+      "loss": 0.4422,
+      "step": 5380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.496890984897217,
+      "learning_rate": 2.6257534243213977e-06,
+      "loss": 0.4978,
+      "step": 5381
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2373340535763007,
+      "learning_rate": 2.623984928604899e-06,
+      "loss": 0.4543,
+      "step": 5382
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3134546338998108,
+      "learning_rate": 2.6222168167545603e-06,
+      "loss": 0.4877,
+      "step": 5383
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4630151280361863,
+      "learning_rate": 2.6204490890560376e-06,
+      "loss": 0.5555,
+      "step": 5384
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.451411796522502,
+      "learning_rate": 2.6186817457949203e-06,
+      "loss": 0.5066,
+      "step": 5385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3195842991720612,
+      "learning_rate": 2.6169147872567398e-06,
+      "loss": 0.5071,
+      "step": 5386
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3759847548738064,
+      "learning_rate": 2.6151482137269652e-06,
+      "loss": 0.4938,
+      "step": 5387
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3174398231895585,
+      "learning_rate": 2.6133820254909993e-06,
+      "loss": 0.5007,
+      "step": 5388
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9484527254681077,
+      "learning_rate": 2.611616222834188e-06,
+      "loss": 0.4696,
+      "step": 5389
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3472055337152538,
+      "learning_rate": 2.6098508060418094e-06,
+      "loss": 0.5497,
+      "step": 5390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3918179486748368,
+      "learning_rate": 2.6080857753990853e-06,
+      "loss": 0.4979,
+      "step": 5391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3937102693000716,
+      "learning_rate": 2.6063211311911677e-06,
+      "loss": 0.5036,
+      "step": 5392
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.479782076219447,
+      "learning_rate": 2.6045568737031557e-06,
+      "loss": 0.4686,
+      "step": 5393
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3786810743205558,
+      "learning_rate": 2.6027930032200744e-06,
+      "loss": 0.4774,
+      "step": 5394
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6375943044791605,
+      "learning_rate": 2.6010295200268993e-06,
+      "loss": 0.4739,
+      "step": 5395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2752704209922996,
+      "learning_rate": 2.5992664244085337e-06,
+      "loss": 0.4696,
+      "step": 5396
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6687567800175958,
+      "learning_rate": 2.597503716649819e-06,
+      "loss": 0.5181,
+      "step": 5397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0429923657351186,
+      "learning_rate": 2.5957413970355404e-06,
+      "loss": 0.5373,
+      "step": 5398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5646069102263396,
+      "learning_rate": 2.5939794658504113e-06,
+      "loss": 0.5409,
+      "step": 5399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6190220992841864,
+      "learning_rate": 2.592217923379093e-06,
+      "loss": 0.4874,
+      "step": 5400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4753599318825783,
+      "learning_rate": 2.590456769906172e-06,
+      "loss": 0.4838,
+      "step": 5401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4759908583800505,
+      "learning_rate": 2.588696005716184e-06,
+      "loss": 0.5297,
+      "step": 5402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4459072953716825,
+      "learning_rate": 2.586935631093588e-06,
+      "loss": 0.5382,
+      "step": 5403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5083852254491767,
+      "learning_rate": 2.5851756463227985e-06,
+      "loss": 0.5376,
+      "step": 5404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4886043444239714,
+      "learning_rate": 2.5834160516881503e-06,
+      "loss": 0.5065,
+      "step": 5405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4808882691319507,
+      "learning_rate": 2.5816568474739205e-06,
+      "loss": 0.4836,
+      "step": 5406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8579835845587833,
+      "learning_rate": 2.579898033964328e-06,
+      "loss": 0.4827,
+      "step": 5407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3889553938886416,
+      "learning_rate": 2.578139611443521e-06,
+      "loss": 0.474,
+      "step": 5408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3946539677304366,
+      "learning_rate": 2.5763815801955906e-06,
+      "loss": 0.4721,
+      "step": 5409
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3500902016545895,
+      "learning_rate": 2.5746239405045592e-06,
+      "loss": 0.4449,
+      "step": 5410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.1094960897936739,
+      "learning_rate": 2.572866692654392e-06,
+      "loss": 0.4604,
+      "step": 5411
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7659196268114692,
+      "learning_rate": 2.5711098369289867e-06,
+      "loss": 0.4616,
+      "step": 5412
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.494765942335093,
+      "learning_rate": 2.56935337361218e-06,
+      "loss": 0.5224,
+      "step": 5413
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4371359529712553,
+      "learning_rate": 2.5675973029877437e-06,
+      "loss": 0.4767,
+      "step": 5414
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2718173194398794,
+      "learning_rate": 2.565841625339384e-06,
+      "loss": 0.4592,
+      "step": 5415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5232337238041327,
+      "learning_rate": 2.5640863409507497e-06,
+      "loss": 0.4844,
+      "step": 5416
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6899511433971866,
+      "learning_rate": 2.5623314501054187e-06,
+      "loss": 0.5109,
+      "step": 5417
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3912369568606449,
+      "learning_rate": 2.560576953086913e-06,
+      "loss": 0.4751,
+      "step": 5418
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6970311944964789,
+      "learning_rate": 2.5588228501786804e-06,
+      "loss": 0.4954,
+      "step": 5419
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.728631154719974,
+      "learning_rate": 2.5570691416641215e-06,
+      "loss": 0.5504,
+      "step": 5420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.588536882862975,
+      "learning_rate": 2.5553158278265553e-06,
+      "loss": 0.5101,
+      "step": 5421
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2162522648632335,
+      "learning_rate": 2.5535629089492496e-06,
+      "loss": 0.5346,
+      "step": 5422
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3282845832866839,
+      "learning_rate": 2.551810385315403e-06,
+      "loss": 0.4933,
+      "step": 5423
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8115742362225473,
+      "learning_rate": 2.550058257208149e-06,
+      "loss": 0.4784,
+      "step": 5424
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4077049409848308,
+      "learning_rate": 2.5483065249105614e-06,
+      "loss": 0.4772,
+      "step": 5425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3679985824357763,
+      "learning_rate": 2.546555188705646e-06,
+      "loss": 0.5134,
+      "step": 5426
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6740862704181385,
+      "learning_rate": 2.544804248876348e-06,
+      "loss": 0.4975,
+      "step": 5427
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3944543755158425,
+      "learning_rate": 2.5430537057055466e-06,
+      "loss": 0.4992,
+      "step": 5428
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4906121548513718,
+      "learning_rate": 2.54130355947606e-06,
+      "loss": 0.535,
+      "step": 5429
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3062691441609326,
+      "learning_rate": 2.539553810470636e-06,
+      "loss": 0.4755,
+      "step": 5430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2265555064473412,
+      "learning_rate": 2.537804458971965e-06,
+      "loss": 0.4526,
+      "step": 5431
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6179227016248734,
+      "learning_rate": 2.5360555052626666e-06,
+      "loss": 0.4724,
+      "step": 5432
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7545604440470863,
+      "learning_rate": 2.534306949625305e-06,
+      "loss": 0.5003,
+      "step": 5433
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4398909017521693,
+      "learning_rate": 2.53255879234237e-06,
+      "loss": 0.482,
+      "step": 5434
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5461741379825082,
+      "learning_rate": 2.5308110336962904e-06,
+      "loss": 0.4763,
+      "step": 5435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.481190626860878,
+      "learning_rate": 2.5290636739694384e-06,
+      "loss": 0.5068,
+      "step": 5436
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7330072487063952,
+      "learning_rate": 2.5273167134441107e-06,
+      "loss": 0.4936,
+      "step": 5437
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7135444486315234,
+      "learning_rate": 2.5255701524025466e-06,
+      "loss": 0.5632,
+      "step": 5438
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3555355930905109,
+      "learning_rate": 2.523823991126916e-06,
+      "loss": 0.4596,
+      "step": 5439
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8351747571846526,
+      "learning_rate": 2.5220782298993297e-06,
+      "loss": 0.4841,
+      "step": 5440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3244901585305506,
+      "learning_rate": 2.5203328690018266e-06,
+      "loss": 0.5499,
+      "step": 5441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2873125222445203,
+      "learning_rate": 2.5185879087163896e-06,
+      "loss": 0.522,
+      "step": 5442
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4651537968091042,
+      "learning_rate": 2.51684334932493e-06,
+      "loss": 0.5391,
+      "step": 5443
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4053443389505933,
+      "learning_rate": 2.5150991911092935e-06,
+      "loss": 0.4871,
+      "step": 5444
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.492951892417051,
+      "learning_rate": 2.513355434351271e-06,
+      "loss": 0.5062,
+      "step": 5445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.943793634105126,
+      "learning_rate": 2.511612079332577e-06,
+      "loss": 0.4549,
+      "step": 5446
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3936898928307528,
+      "learning_rate": 2.5098691263348697e-06,
+      "loss": 0.4575,
+      "step": 5447
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.26250521984885,
+      "learning_rate": 2.508126575639733e-06,
+      "loss": 0.4571,
+      "step": 5448
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4848240903706404,
+      "learning_rate": 2.5063844275286974e-06,
+      "loss": 0.5088,
+      "step": 5449
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5001599313859406,
+      "learning_rate": 2.5046426822832175e-06,
+      "loss": 0.463,
+      "step": 5450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5160790102059807,
+      "learning_rate": 2.5029013401846913e-06,
+      "loss": 0.4656,
+      "step": 5451
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.123264081371732,
+      "learning_rate": 2.5011604015144435e-06,
+      "loss": 0.4472,
+      "step": 5452
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3789951391327997,
+      "learning_rate": 2.499419866553741e-06,
+      "loss": 0.5015,
+      "step": 5453
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4603123809144725,
+      "learning_rate": 2.4976797355837845e-06,
+      "loss": 0.5077,
+      "step": 5454
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7650029454656442,
+      "learning_rate": 2.495940008885703e-06,
+      "loss": 0.5056,
+      "step": 5455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4849664616453735,
+      "learning_rate": 2.4942006867405685e-06,
+      "loss": 0.531,
+      "step": 5456
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.493518897193527,
+      "learning_rate": 2.492461769429381e-06,
+      "loss": 0.5234,
+      "step": 5457
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3869502079430214,
+      "learning_rate": 2.49072325723308e-06,
+      "loss": 0.5068,
+      "step": 5458
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.681702737575638,
+      "learning_rate": 2.4889851504325348e-06,
+      "loss": 0.5342,
+      "step": 5459
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6611963051545626,
+      "learning_rate": 2.487247449308554e-06,
+      "loss": 0.5225,
+      "step": 5460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4589571409635018,
+      "learning_rate": 2.4855101541418797e-06,
+      "loss": 0.5014,
+      "step": 5461
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2537043292301941,
+      "learning_rate": 2.483773265213184e-06,
+      "loss": 0.4961,
+      "step": 5462
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.32920342648201,
+      "learning_rate": 2.48203678280308e-06,
+      "loss": 0.4458,
+      "step": 5463
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3681830900025551,
+      "learning_rate": 2.4803007071921083e-06,
+      "loss": 0.4675,
+      "step": 5464
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4756488701641188,
+      "learning_rate": 2.478565038660751e-06,
+      "loss": 0.452,
+      "step": 5465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6817738940482172,
+      "learning_rate": 2.4768297774894157e-06,
+      "loss": 0.5072,
+      "step": 5466
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5245969438197655,
+      "learning_rate": 2.4750949239584543e-06,
+      "loss": 0.475,
+      "step": 5467
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.280761509374414,
+      "learning_rate": 2.4733604783481436e-06,
+      "loss": 0.4787,
+      "step": 5468
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4899473963878174,
+      "learning_rate": 2.4716264409387005e-06,
+      "loss": 0.496,
+      "step": 5469
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7960664090989007,
+      "learning_rate": 2.469892812010275e-06,
+      "loss": 0.5228,
+      "step": 5470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2990665854475318,
+      "learning_rate": 2.4681595918429473e-06,
+      "loss": 0.4516,
+      "step": 5471
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.717509107476117,
+      "learning_rate": 2.466426780716738e-06,
+      "loss": 0.5279,
+      "step": 5472
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.820181283099114,
+      "learning_rate": 2.4646943789115947e-06,
+      "loss": 0.5585,
+      "step": 5473
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4540625568618581,
+      "learning_rate": 2.4629623867074043e-06,
+      "loss": 0.5351,
+      "step": 5474
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3742096801051715,
+      "learning_rate": 2.4612308043839835e-06,
+      "loss": 0.5146,
+      "step": 5475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5797660549317376,
+      "learning_rate": 2.459499632221085e-06,
+      "loss": 0.5133,
+      "step": 5476
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.776131095900743,
+      "learning_rate": 2.4577688704983984e-06,
+      "loss": 0.4713,
+      "step": 5477
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.483890797314977,
+      "learning_rate": 2.456038519495538e-06,
+      "loss": 0.4944,
+      "step": 5478
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.958313627579842,
+      "learning_rate": 2.4543085794920616e-06,
+      "loss": 0.5002,
+      "step": 5479
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6271441054448663,
+      "learning_rate": 2.4525790507674543e-06,
+      "loss": 0.5027,
+      "step": 5480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6516835480074524,
+      "learning_rate": 2.4508499336011377e-06,
+      "loss": 0.4594,
+      "step": 5481
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5440243026978204,
+      "learning_rate": 2.4491212282724637e-06,
+      "loss": 0.5465,
+      "step": 5482
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4122249610320332,
+      "learning_rate": 2.447392935060723e-06,
+      "loss": 0.4976,
+      "step": 5483
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4013850774891232,
+      "learning_rate": 2.4456650542451333e-06,
+      "loss": 0.4864,
+      "step": 5484
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.77467430479846,
+      "learning_rate": 2.443937586104851e-06,
+      "loss": 0.5007,
+      "step": 5485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3686948184301855,
+      "learning_rate": 2.4422105309189646e-06,
+      "loss": 0.4656,
+      "step": 5486
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6813398955724435,
+      "learning_rate": 2.4404838889664923e-06,
+      "loss": 0.51,
+      "step": 5487
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3466499318485836,
+      "learning_rate": 2.4387576605263924e-06,
+      "loss": 0.4295,
+      "step": 5488
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.312175726859582,
+      "learning_rate": 2.4370318458775472e-06,
+      "loss": 0.4509,
+      "step": 5489
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4006040680258878,
+      "learning_rate": 2.435306445298782e-06,
+      "loss": 0.4833,
+      "step": 5490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.960510313343304,
+      "learning_rate": 2.4335814590688465e-06,
+      "loss": 0.4776,
+      "step": 5491
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3678018833794279,
+      "learning_rate": 2.4318568874664318e-06,
+      "loss": 0.4486,
+      "step": 5492
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.786463696861167,
+      "learning_rate": 2.4301327307701526e-06,
+      "loss": 0.4443,
+      "step": 5493
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9480246804073973,
+      "learning_rate": 2.4284089892585642e-06,
+      "loss": 0.4685,
+      "step": 5494
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3265526795294944,
+      "learning_rate": 2.426685663210154e-06,
+      "loss": 0.4253,
+      "step": 5495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.341929758596841,
+      "learning_rate": 2.424962752903337e-06,
+      "loss": 0.5203,
+      "step": 5496
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4182009629873498,
+      "learning_rate": 2.4232402586164677e-06,
+      "loss": 0.4407,
+      "step": 5497
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.742253522033818,
+      "learning_rate": 2.421518180627827e-06,
+      "loss": 0.5481,
+      "step": 5498
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6532235894146765,
+      "learning_rate": 2.4197965192156354e-06,
+      "loss": 0.4956,
+      "step": 5499
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3161751128320978,
+      "learning_rate": 2.418075274658039e-06,
+      "loss": 0.4593,
+      "step": 5500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6334983346371813,
+      "learning_rate": 2.4163544472331207e-06,
+      "loss": 0.4933,
+      "step": 5501
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6149656215565829,
+      "learning_rate": 2.4146340372188964e-06,
+      "loss": 0.4689,
+      "step": 5502
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7919360703700247,
+      "learning_rate": 2.412914044893316e-06,
+      "loss": 0.5363,
+      "step": 5503
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.457696564614905,
+      "learning_rate": 2.4111944705342554e-06,
+      "loss": 0.5046,
+      "step": 5504
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8186669373877837,
+      "learning_rate": 2.409475314419527e-06,
+      "loss": 0.4993,
+      "step": 5505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3221148999267518,
+      "learning_rate": 2.407756576826879e-06,
+      "loss": 0.4626,
+      "step": 5506
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6052732752272765,
+      "learning_rate": 2.4060382580339842e-06,
+      "loss": 0.4899,
+      "step": 5507
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5767022145696568,
+      "learning_rate": 2.404320358318456e-06,
+      "loss": 0.5222,
+      "step": 5508
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5095809579755186,
+      "learning_rate": 2.4026028779578315e-06,
+      "loss": 0.4617,
+      "step": 5509
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.97793047083496,
+      "learning_rate": 2.400885817229591e-06,
+      "loss": 0.4826,
+      "step": 5510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3588500491345321,
+      "learning_rate": 2.3991691764111365e-06,
+      "loss": 0.5548,
+      "step": 5511
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6518198586561694,
+      "learning_rate": 2.3974529557798095e-06,
+      "loss": 0.4769,
+      "step": 5512
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.390464564482471,
+      "learning_rate": 2.395737155612878e-06,
+      "loss": 0.5309,
+      "step": 5513
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3918943127026036,
+      "learning_rate": 2.3940217761875433e-06,
+      "loss": 0.51,
+      "step": 5514
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6977411274786531,
+      "learning_rate": 2.392306817780945e-06,
+      "loss": 0.4702,
+      "step": 5515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6922389974526388,
+      "learning_rate": 2.390592280670144e-06,
+      "loss": 0.4979,
+      "step": 5516
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3787793285566037,
+      "learning_rate": 2.388878165132142e-06,
+      "loss": 0.4635,
+      "step": 5517
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.646674192788396,
+      "learning_rate": 2.387164471443869e-06,
+      "loss": 0.4554,
+      "step": 5518
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3207958483783915,
+      "learning_rate": 2.3854511998821894e-06,
+      "loss": 0.5156,
+      "step": 5519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.227695167477801,
+      "learning_rate": 2.3837383507238936e-06,
+      "loss": 0.5333,
+      "step": 5520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4267150504517874,
+      "learning_rate": 2.382025924245711e-06,
+      "loss": 0.5071,
+      "step": 5521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3844833274592145,
+      "learning_rate": 2.3803139207242974e-06,
+      "loss": 0.4785,
+      "step": 5522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4789844668096037,
+      "learning_rate": 2.37860234043624e-06,
+      "loss": 0.5884,
+      "step": 5523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4224685107437975,
+      "learning_rate": 2.3768911836580645e-06,
+      "loss": 0.5148,
+      "step": 5524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9225409466991767,
+      "learning_rate": 2.3751804506662174e-06,
+      "loss": 0.5083,
+      "step": 5525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5887749841143064,
+      "learning_rate": 2.3734701417370866e-06,
+      "loss": 0.4917,
+      "step": 5526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7569083490679074,
+      "learning_rate": 2.3717602571469865e-06,
+      "loss": 0.4632,
+      "step": 5527
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4140716874174504,
+      "learning_rate": 2.3700507971721663e-06,
+      "loss": 0.427,
+      "step": 5528
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.609909140872164,
+      "learning_rate": 2.3683417620888003e-06,
+      "loss": 0.5611,
+      "step": 5529
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5878396327956295,
+      "learning_rate": 2.3666331521730026e-06,
+      "loss": 0.4592,
+      "step": 5530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7151379897062956,
+      "learning_rate": 2.3649249677008097e-06,
+      "loss": 0.5009,
+      "step": 5531
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2996372953750224,
+      "learning_rate": 2.3632172089481973e-06,
+      "loss": 0.546,
+      "step": 5532
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.077678869251373,
+      "learning_rate": 2.361509876191068e-06,
+      "loss": 0.4523,
+      "step": 5533
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.747621552902312,
+      "learning_rate": 2.3598029697052522e-06,
+      "loss": 0.4888,
+      "step": 5534
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3678679837061298,
+      "learning_rate": 2.358096489766523e-06,
+      "loss": 0.5107,
+      "step": 5535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6739888453600988,
+      "learning_rate": 2.3563904366505717e-06,
+      "loss": 0.4862,
+      "step": 5536
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8955093820392193,
+      "learning_rate": 2.3546848106330308e-06,
+      "loss": 0.5101,
+      "step": 5537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6366006897451797,
+      "learning_rate": 2.352979611989454e-06,
+      "loss": 0.5044,
+      "step": 5538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5750665629516203,
+      "learning_rate": 2.3512748409953358e-06,
+      "loss": 0.489,
+      "step": 5539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4189184004954971,
+      "learning_rate": 2.3495704979260924e-06,
+      "loss": 0.4985,
+      "step": 5540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.666993666611936,
+      "learning_rate": 2.3478665830570807e-06,
+      "loss": 0.4446,
+      "step": 5541
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4494929286590599,
+      "learning_rate": 2.3461630966635784e-06,
+      "loss": 0.497,
+      "step": 5542
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.487393842420435,
+      "learning_rate": 2.3444600390208007e-06,
+      "loss": 0.4965,
+      "step": 5543
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5902444904379578,
+      "learning_rate": 2.3427574104038934e-06,
+      "loss": 0.4581,
+      "step": 5544
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.161066446508982,
+      "learning_rate": 2.3410552110879277e-06,
+      "loss": 0.4993,
+      "step": 5545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3929795053155725,
+      "learning_rate": 2.339353441347912e-06,
+      "loss": 0.5134,
+      "step": 5546
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.37935134864808,
+      "learning_rate": 2.3376521014587787e-06,
+      "loss": 0.504,
+      "step": 5547
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0874526415093935,
+      "learning_rate": 2.335951191695399e-06,
+      "loss": 0.53,
+      "step": 5548
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7303232882596309,
+      "learning_rate": 2.334250712332565e-06,
+      "loss": 0.4948,
+      "step": 5549
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8220759392032708,
+      "learning_rate": 2.3325506636450056e-06,
+      "loss": 0.5296,
+      "step": 5550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4777829873197916,
+      "learning_rate": 2.3308510459073817e-06,
+      "loss": 0.4775,
+      "step": 5551
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5027960581625106,
+      "learning_rate": 2.3291518593942774e-06,
+      "loss": 0.5385,
+      "step": 5552
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5217717603564733,
+      "learning_rate": 2.3274531043802148e-06,
+      "loss": 0.4594,
+      "step": 5553
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5906557921951225,
+      "learning_rate": 2.325754781139638e-06,
+      "loss": 0.4724,
+      "step": 5554
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.329096145339709,
+      "learning_rate": 2.3240568899469317e-06,
+      "loss": 0.4716,
+      "step": 5555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4012838734313477,
+      "learning_rate": 2.322359431076401e-06,
+      "loss": 0.4911,
+      "step": 5556
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5384021925882139,
+      "learning_rate": 2.3206624048022872e-06,
+      "loss": 0.506,
+      "step": 5557
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.655449448220534,
+      "learning_rate": 2.318965811398759e-06,
+      "loss": 0.5081,
+      "step": 5558
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5825164260224847,
+      "learning_rate": 2.317269651139916e-06,
+      "loss": 0.5677,
+      "step": 5559
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6821233375146059,
+      "learning_rate": 2.3155739242997893e-06,
+      "loss": 0.5366,
+      "step": 5560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5594929350983364,
+      "learning_rate": 2.3138786311523364e-06,
+      "loss": 0.4455,
+      "step": 5561
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4382628232628942,
+      "learning_rate": 2.3121837719714496e-06,
+      "loss": 0.4402,
+      "step": 5562
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4816789194218292,
+      "learning_rate": 2.310489347030945e-06,
+      "loss": 0.4789,
+      "step": 5563
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5188438068192134,
+      "learning_rate": 2.3087953566045743e-06,
+      "loss": 0.4484,
+      "step": 5564
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5098238352487316,
+      "learning_rate": 2.307101800966015e-06,
+      "loss": 0.4725,
+      "step": 5565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7038045623006983,
+      "learning_rate": 2.305408680388878e-06,
+      "loss": 0.4944,
+      "step": 5566
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2637120328181233,
+      "learning_rate": 2.303715995146699e-06,
+      "loss": 0.5007,
+      "step": 5567
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4973349132088374,
+      "learning_rate": 2.3020237455129473e-06,
+      "loss": 0.5026,
+      "step": 5568
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5771957423178784,
+      "learning_rate": 2.3003319317610232e-06,
+      "loss": 0.441,
+      "step": 5569
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9968666654182445,
+      "learning_rate": 2.298640554164251e-06,
+      "loss": 0.5056,
+      "step": 5570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5088225959683377,
+      "learning_rate": 2.2969496129958896e-06,
+      "loss": 0.4609,
+      "step": 5571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5688355879342355,
+      "learning_rate": 2.295259108529123e-06,
+      "loss": 0.5082,
+      "step": 5572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1218690144924017,
+      "learning_rate": 2.29356904103707e-06,
+      "loss": 0.477,
+      "step": 5573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3317460113932156,
+      "learning_rate": 2.291879410792774e-06,
+      "loss": 0.5049,
+      "step": 5574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6361394777722302,
+      "learning_rate": 2.2901902180692094e-06,
+      "loss": 0.4927,
+      "step": 5575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2600474098118004,
+      "learning_rate": 2.288501463139282e-06,
+      "loss": 0.4862,
+      "step": 5576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8835019069260908,
+      "learning_rate": 2.286813146275823e-06,
+      "loss": 0.465,
+      "step": 5577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8127391882820307,
+      "learning_rate": 2.285125267751598e-06,
+      "loss": 0.5114,
+      "step": 5578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8926521832120369,
+      "learning_rate": 2.2834378278392937e-06,
+      "loss": 0.5105,
+      "step": 5579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4873921582045961,
+      "learning_rate": 2.2817508268115364e-06,
+      "loss": 0.4717,
+      "step": 5580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.819933821860014,
+      "learning_rate": 2.280064264940871e-06,
+      "loss": 0.5442,
+      "step": 5581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.683485669972065,
+      "learning_rate": 2.27837814249978e-06,
+      "loss": 0.5305,
+      "step": 5582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6440117666442702,
+      "learning_rate": 2.2766924597606686e-06,
+      "loss": 0.499,
+      "step": 5583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6879618779299213,
+      "learning_rate": 2.2750072169958754e-06,
+      "loss": 0.5379,
+      "step": 5584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9671847158512492,
+      "learning_rate": 2.273322414477668e-06,
+      "loss": 0.4351,
+      "step": 5585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.466482216219935,
+      "learning_rate": 2.271638052478236e-06,
+      "loss": 0.4478,
+      "step": 5586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9616512482384683,
+      "learning_rate": 2.269954131269708e-06,
+      "loss": 0.4975,
+      "step": 5587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3639134192599855,
+      "learning_rate": 2.268270651124133e-06,
+      "loss": 0.4843,
+      "step": 5588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3553964640402418,
+      "learning_rate": 2.266587612313495e-06,
+      "loss": 0.492,
+      "step": 5589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5577828659517323,
+      "learning_rate": 2.2649050151096994e-06,
+      "loss": 0.5129,
+      "step": 5590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6522603447358237,
+      "learning_rate": 2.2632228597845878e-06,
+      "loss": 0.4852,
+      "step": 5591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6798186942226545,
+      "learning_rate": 2.2615411466099283e-06,
+      "loss": 0.5052,
+      "step": 5592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7162936489412086,
+      "learning_rate": 2.259859875857413e-06,
+      "loss": 0.4911,
+      "step": 5593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4025456726402974,
+      "learning_rate": 2.2581790477986692e-06,
+      "loss": 0.5039,
+      "step": 5594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5845470815346903,
+      "learning_rate": 2.2564986627052463e-06,
+      "loss": 0.5274,
+      "step": 5595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6416412378921217,
+      "learning_rate": 2.2548187208486293e-06,
+      "loss": 0.5115,
+      "step": 5596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3451856385540713,
+      "learning_rate": 2.2531392225002236e-06,
+      "loss": 0.4623,
+      "step": 5597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3698888882150644,
+      "learning_rate": 2.25146016793137e-06,
+      "loss": 0.4535,
+      "step": 5598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6976438584873728,
+      "learning_rate": 2.2497815574133313e-06,
+      "loss": 0.5107,
+      "step": 5599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1391226535547676,
+      "learning_rate": 2.2481033912173044e-06,
+      "loss": 0.5465,
+      "step": 5600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4528519800705884,
+      "learning_rate": 2.2464256696144106e-06,
+      "loss": 0.5174,
+      "step": 5601
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4471062589320642,
+      "learning_rate": 2.2447483928757034e-06,
+      "loss": 0.5343,
+      "step": 5602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7013573683309002,
+      "learning_rate": 2.243071561272159e-06,
+      "loss": 0.51,
+      "step": 5603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6929219375700657,
+      "learning_rate": 2.241395175074683e-06,
+      "loss": 0.5223,
+      "step": 5604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.582600692096603,
+      "learning_rate": 2.2397192345541146e-06,
+      "loss": 0.5067,
+      "step": 5605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3480484291333905,
+      "learning_rate": 2.238043739981212e-06,
+      "loss": 0.4497,
+      "step": 5606
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.346904838587814,
+      "learning_rate": 2.2363686916266696e-06,
+      "loss": 0.5279,
+      "step": 5607
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4764789978750652,
+      "learning_rate": 2.234694089761102e-06,
+      "loss": 0.5216,
+      "step": 5608
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7128029989775837,
+      "learning_rate": 2.2330199346550624e-06,
+      "loss": 0.5023,
+      "step": 5609
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4965838514525496,
+      "learning_rate": 2.2313462265790198e-06,
+      "loss": 0.5115,
+      "step": 5610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6073821385339875,
+      "learning_rate": 2.22967296580338e-06,
+      "loss": 0.4895,
+      "step": 5611
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.479216078712145,
+      "learning_rate": 2.2280001525984718e-06,
+      "loss": 0.4812,
+      "step": 5612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.412021542856434,
+      "learning_rate": 2.2263277872345505e-06,
+      "loss": 0.482,
+      "step": 5613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4877331023293925,
+      "learning_rate": 2.2246558699818056e-06,
+      "loss": 0.5364,
+      "step": 5614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3517184707148822,
+      "learning_rate": 2.222984401110346e-06,
+      "loss": 0.467,
+      "step": 5615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4341668244370513,
+      "learning_rate": 2.2213133808902143e-06,
+      "loss": 0.5515,
+      "step": 5616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2868126565065974,
+      "learning_rate": 2.219642809591378e-06,
+      "loss": 0.5173,
+      "step": 5617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3828616571993415,
+      "learning_rate": 2.2179726874837353e-06,
+      "loss": 0.4831,
+      "step": 5618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6489534862379591,
+      "learning_rate": 2.2163030148371044e-06,
+      "loss": 0.4489,
+      "step": 5619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1143036312350367,
+      "learning_rate": 2.214633791921241e-06,
+      "loss": 0.4652,
+      "step": 5620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6425173069258526,
+      "learning_rate": 2.2129650190058188e-06,
+      "loss": 0.5185,
+      "step": 5621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7896536582636093,
+      "learning_rate": 2.211296696360442e-06,
+      "loss": 0.4961,
+      "step": 5622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.363767251022316,
+      "learning_rate": 2.2096288242546464e-06,
+      "loss": 0.5514,
+      "step": 5623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6085365905066162,
+      "learning_rate": 2.2079614029578865e-06,
+      "loss": 0.5116,
+      "step": 5624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6616446613161735,
+      "learning_rate": 2.206294432739556e-06,
+      "loss": 0.4866,
+      "step": 5625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.405353254193536,
+      "learning_rate": 2.2046279138689617e-06,
+      "loss": 0.5026,
+      "step": 5626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6128351079408425,
+      "learning_rate": 2.20296184661535e-06,
+      "loss": 0.4926,
+      "step": 5627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4587630016659268,
+      "learning_rate": 2.201296231247884e-06,
+      "loss": 0.4461,
+      "step": 5628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3835200685248408,
+      "learning_rate": 2.1996310680356623e-06,
+      "loss": 0.4562,
+      "step": 5629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3805367109816795,
+      "learning_rate": 2.1979663572477057e-06,
+      "loss": 0.4576,
+      "step": 5630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6463950616780922,
+      "learning_rate": 2.19630209915296e-06,
+      "loss": 0.4814,
+      "step": 5631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6488872376356858,
+      "learning_rate": 2.1946382940203024e-06,
+      "loss": 0.4789,
+      "step": 5632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3765869540183546,
+      "learning_rate": 2.1929749421185363e-06,
+      "loss": 0.4682,
+      "step": 5633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7079525292081255,
+      "learning_rate": 2.191312043716392e-06,
+      "loss": 0.5344,
+      "step": 5634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5976843490539296,
+      "learning_rate": 2.1896495990825224e-06,
+      "loss": 0.5024,
+      "step": 5635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9632934935601438,
+      "learning_rate": 2.187987608485513e-06,
+      "loss": 0.461,
+      "step": 5636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2722730035586236,
+      "learning_rate": 2.1863260721938696e-06,
+      "loss": 0.4721,
+      "step": 5637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4855676434755076,
+      "learning_rate": 2.1846649904760315e-06,
+      "loss": 0.4987,
+      "step": 5638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4833522957894028,
+      "learning_rate": 2.1830043636003574e-06,
+      "loss": 0.4885,
+      "step": 5639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3258031890663018,
+      "learning_rate": 2.1813441918351407e-06,
+      "loss": 0.5451,
+      "step": 5640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.644381239071562,
+      "learning_rate": 2.179684475448592e-06,
+      "loss": 0.5008,
+      "step": 5641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4396401967283894,
+      "learning_rate": 2.1780252147088555e-06,
+      "loss": 0.5547,
+      "step": 5642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.235056697104862,
+      "learning_rate": 2.1763664098840013e-06,
+      "loss": 0.4742,
+      "step": 5643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.538756193902995,
+      "learning_rate": 2.1747080612420202e-06,
+      "loss": 0.479,
+      "step": 5644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.489100563897908,
+      "learning_rate": 2.1730501690508363e-06,
+      "loss": 0.4465,
+      "step": 5645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4644523867767538,
+      "learning_rate": 2.1713927335782934e-06,
+      "loss": 0.4719,
+      "step": 5646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.821065940683249,
+      "learning_rate": 2.169735755092168e-06,
+      "loss": 0.5072,
+      "step": 5647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5885001554236768,
+      "learning_rate": 2.168079233860157e-06,
+      "loss": 0.5026,
+      "step": 5648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7164720858256994,
+      "learning_rate": 2.166423170149887e-06,
+      "loss": 0.5201,
+      "step": 5649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7594213840845911,
+      "learning_rate": 2.164767564228911e-06,
+      "loss": 0.5021,
+      "step": 5650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.028797252998077,
+      "learning_rate": 2.1631124163647043e-06,
+      "loss": 0.4976,
+      "step": 5651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5397052115546976,
+      "learning_rate": 2.1614577268246735e-06,
+      "loss": 0.4835,
+      "step": 5652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2110349025128313,
+      "learning_rate": 2.1598034958761448e-06,
+      "loss": 0.4917,
+      "step": 5653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5698386403094702,
+      "learning_rate": 2.1581497237863767e-06,
+      "loss": 0.4759,
+      "step": 5654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4109353451334927,
+      "learning_rate": 2.1564964108225485e-06,
+      "loss": 0.488,
+      "step": 5655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6598736568588375,
+      "learning_rate": 2.15484355725177e-06,
+      "loss": 0.5005,
+      "step": 5656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4919745602635004,
+      "learning_rate": 2.153191163341071e-06,
+      "loss": 0.486,
+      "step": 5657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2253434355089252,
+      "learning_rate": 2.151539229357412e-06,
+      "loss": 0.4571,
+      "step": 5658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5071311879513096,
+      "learning_rate": 2.149887755567679e-06,
+      "loss": 0.5155,
+      "step": 5659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4334540922292345,
+      "learning_rate": 2.148236742238679e-06,
+      "loss": 0.4797,
+      "step": 5660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6044020149090967,
+      "learning_rate": 2.1465861896371514e-06,
+      "loss": 0.5131,
+      "step": 5661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3693514507790747,
+      "learning_rate": 2.1449360980297536e-06,
+      "loss": 0.4986,
+      "step": 5662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.230734930065685,
+      "learning_rate": 2.143286467683076e-06,
+      "loss": 0.4817,
+      "step": 5663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3873301908337472,
+      "learning_rate": 2.1416372988636275e-06,
+      "loss": 0.4873,
+      "step": 5664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6559315099206287,
+      "learning_rate": 2.1399885918378478e-06,
+      "loss": 0.4808,
+      "step": 5665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.19002529984588,
+      "learning_rate": 2.1383403468721013e-06,
+      "loss": 0.5296,
+      "step": 5666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4688445935033496,
+      "learning_rate": 2.1366925642326735e-06,
+      "loss": 0.5354,
+      "step": 5667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.079621728241784,
+      "learning_rate": 2.135045244185781e-06,
+      "loss": 0.5183,
+      "step": 5668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.413368987131197,
+      "learning_rate": 2.13339838699756e-06,
+      "loss": 0.4898,
+      "step": 5669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3210859609376642,
+      "learning_rate": 2.1317519929340787e-06,
+      "loss": 0.4783,
+      "step": 5670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.32159694606616,
+      "learning_rate": 2.130106062261322e-06,
+      "loss": 0.5103,
+      "step": 5671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7155454747055365,
+      "learning_rate": 2.128460595245208e-06,
+      "loss": 0.4686,
+      "step": 5672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4536064375277995,
+      "learning_rate": 2.126815592151574e-06,
+      "loss": 0.4817,
+      "step": 5673
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7002556576948114,
+      "learning_rate": 2.1251710532461854e-06,
+      "loss": 0.4708,
+      "step": 5674
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3823878012680884,
+      "learning_rate": 2.1235269787947345e-06,
+      "loss": 0.505,
+      "step": 5675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3071705033544256,
+      "learning_rate": 2.121883369062832e-06,
+      "loss": 0.5239,
+      "step": 5676
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6916273684756746,
+      "learning_rate": 2.1202402243160215e-06,
+      "loss": 0.4552,
+      "step": 5677
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8044470154991188,
+      "learning_rate": 2.118597544819763e-06,
+      "loss": 0.4711,
+      "step": 5678
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4612633988316264,
+      "learning_rate": 2.11695533083945e-06,
+      "loss": 0.4819,
+      "step": 5679
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4045791920997903,
+      "learning_rate": 2.1153135826403936e-06,
+      "loss": 0.5078,
+      "step": 5680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4976793521690002,
+      "learning_rate": 2.1136723004878356e-06,
+      "loss": 0.4956,
+      "step": 5681
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7046915634044535,
+      "learning_rate": 2.1120314846469364e-06,
+      "loss": 0.5065,
+      "step": 5682
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4735756528319723,
+      "learning_rate": 2.1103911353827855e-06,
+      "loss": 0.5042,
+      "step": 5683
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6003853911474244,
+      "learning_rate": 2.1087512529603984e-06,
+      "loss": 0.546,
+      "step": 5684
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6980308517680083,
+      "learning_rate": 2.1071118376447074e-06,
+      "loss": 0.5027,
+      "step": 5685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6285847762124912,
+      "learning_rate": 2.10547288970058e-06,
+      "loss": 0.504,
+      "step": 5686
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7985510618826166,
+      "learning_rate": 2.1038344093927983e-06,
+      "loss": 0.5049,
+      "step": 5687
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4584743770881174,
+      "learning_rate": 2.102196396986076e-06,
+      "loss": 0.4688,
+      "step": 5688
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.062882331433079,
+      "learning_rate": 2.100558852745046e-06,
+      "loss": 0.5158,
+      "step": 5689
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3176991940677245,
+      "learning_rate": 2.098921776934269e-06,
+      "loss": 0.5022,
+      "step": 5690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6950672852350026,
+      "learning_rate": 2.097285169818232e-06,
+      "loss": 0.4937,
+      "step": 5691
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4397600009557168,
+      "learning_rate": 2.0956490316613375e-06,
+      "loss": 0.4694,
+      "step": 5692
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4476077867770372,
+      "learning_rate": 2.094013362727924e-06,
+      "loss": 0.514,
+      "step": 5693
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4207692375720573,
+      "learning_rate": 2.0923781632822434e-06,
+      "loss": 0.4678,
+      "step": 5694
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2564321877149407,
+      "learning_rate": 2.09074343358848e-06,
+      "loss": 0.5346,
+      "step": 5695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.416804884151017,
+      "learning_rate": 2.0891091739107355e-06,
+      "loss": 0.494,
+      "step": 5696
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6357332631911095,
+      "learning_rate": 2.087475384513043e-06,
+      "loss": 0.515,
+      "step": 5697
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5157776993178462,
+      "learning_rate": 2.08584206565935e-06,
+      "loss": 0.4975,
+      "step": 5698
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6552230500936123,
+      "learning_rate": 2.0842092176135396e-06,
+      "loss": 0.4694,
+      "step": 5699
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0776540845584193,
+      "learning_rate": 2.082576840639411e-06,
+      "loss": 0.4866,
+      "step": 5700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4391172169598452,
+      "learning_rate": 2.080944935000686e-06,
+      "loss": 0.4897,
+      "step": 5701
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.508862587297614,
+      "learning_rate": 2.0793135009610173e-06,
+      "loss": 0.5091,
+      "step": 5702
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4673804733296882,
+      "learning_rate": 2.077682538783974e-06,
+      "loss": 0.4757,
+      "step": 5703
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.345254379836162,
+      "learning_rate": 2.0760520487330554e-06,
+      "loss": 0.5242,
+      "step": 5704
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5063600999543156,
+      "learning_rate": 2.074422031071679e-06,
+      "loss": 0.5018,
+      "step": 5705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4460384916405886,
+      "learning_rate": 2.0727924860631886e-06,
+      "loss": 0.4711,
+      "step": 5706
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.1349399254738972,
+      "learning_rate": 2.071163413970854e-06,
+      "loss": 0.4204,
+      "step": 5707
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.421480116647311,
+      "learning_rate": 2.0695348150578655e-06,
+      "loss": 0.4944,
+      "step": 5708
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.32601532841179,
+      "learning_rate": 2.0679066895873358e-06,
+      "loss": 0.4766,
+      "step": 5709
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5325685134493685,
+      "learning_rate": 2.066279037822305e-06,
+      "loss": 0.5301,
+      "step": 5710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3833534069309887,
+      "learning_rate": 2.0646518600257343e-06,
+      "loss": 0.495,
+      "step": 5711
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4562439694652227,
+      "learning_rate": 2.0630251564605053e-06,
+      "loss": 0.4554,
+      "step": 5712
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8753303528850167,
+      "learning_rate": 2.0613989273894313e-06,
+      "loss": 0.5322,
+      "step": 5713
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4374425915890892,
+      "learning_rate": 2.059773173075239e-06,
+      "loss": 0.4604,
+      "step": 5714
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.435434545196034,
+      "learning_rate": 2.0581478937805864e-06,
+      "loss": 0.5105,
+      "step": 5715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.781414285188784,
+      "learning_rate": 2.056523089768051e-06,
+      "loss": 0.5118,
+      "step": 5716
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4164073183621666,
+      "learning_rate": 2.054898761300136e-06,
+      "loss": 0.4928,
+      "step": 5717
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8871129395490855,
+      "learning_rate": 2.0532749086392625e-06,
+      "loss": 0.5301,
+      "step": 5718
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7594277639683007,
+      "learning_rate": 2.0516515320477813e-06,
+      "loss": 0.4697,
+      "step": 5719
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 26.873164838615025,
+      "learning_rate": 2.0500286317879626e-06,
+      "loss": 0.4826,
+      "step": 5720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5551879147629808,
+      "learning_rate": 2.0484062081219967e-06,
+      "loss": 0.5253,
+      "step": 5721
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5665900419493022,
+      "learning_rate": 2.046784261312006e-06,
+      "loss": 0.5023,
+      "step": 5722
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5347720498202977,
+      "learning_rate": 2.0451627916200236e-06,
+      "loss": 0.5354,
+      "step": 5723
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4266227402970557,
+      "learning_rate": 2.0435417993080194e-06,
+      "loss": 0.5164,
+      "step": 5724
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2459870240656372,
+      "learning_rate": 2.041921284637874e-06,
+      "loss": 0.5072,
+      "step": 5725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4351773132797392,
+      "learning_rate": 2.040301247871399e-06,
+      "loss": 0.4542,
+      "step": 5726
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5153000504442997,
+      "learning_rate": 2.0386816892703225e-06,
+      "loss": 0.4969,
+      "step": 5727
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.994793007552151,
+      "learning_rate": 2.037062609096302e-06,
+      "loss": 0.4812,
+      "step": 5728
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3214414084860961,
+      "learning_rate": 2.035444007610912e-06,
+      "loss": 0.4687,
+      "step": 5729
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6187642823487383,
+      "learning_rate": 2.0338258850756507e-06,
+      "loss": 0.4927,
+      "step": 5730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4220341032570714,
+      "learning_rate": 2.032208241751941e-06,
+      "loss": 0.4744,
+      "step": 5731
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3246981718775324,
+      "learning_rate": 2.0305910779011277e-06,
+      "loss": 0.5096,
+      "step": 5732
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2892581949710045,
+      "learning_rate": 2.02897439378448e-06,
+      "loss": 0.5107,
+      "step": 5733
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5615295073884667,
+      "learning_rate": 2.0273581896631837e-06,
+      "loss": 0.4812,
+      "step": 5734
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4455183398778615,
+      "learning_rate": 2.025742465798355e-06,
+      "loss": 0.502,
+      "step": 5735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.557923520928341,
+      "learning_rate": 2.0241272224510235e-06,
+      "loss": 0.5193,
+      "step": 5736
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3621564814195775,
+      "learning_rate": 2.0225124598821498e-06,
+      "loss": 0.4869,
+      "step": 5737
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.45337552752045,
+      "learning_rate": 2.02089817835261e-06,
+      "loss": 0.5248,
+      "step": 5738
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.233319918986319,
+      "learning_rate": 2.019284378123207e-06,
+      "loss": 0.4867,
+      "step": 5739
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3556952200964876,
+      "learning_rate": 2.017671059454667e-06,
+      "loss": 0.471,
+      "step": 5740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4902769625106627,
+      "learning_rate": 2.0160582226076304e-06,
+      "loss": 0.4932,
+      "step": 5741
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7142701712089856,
+      "learning_rate": 2.0144458678426705e-06,
+      "loss": 0.4717,
+      "step": 5742
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3939041873730047,
+      "learning_rate": 2.0128339954202734e-06,
+      "loss": 0.5065,
+      "step": 5743
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8410248810440564,
+      "learning_rate": 2.0112226056008547e-06,
+      "loss": 0.4957,
+      "step": 5744
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6987436213955667,
+      "learning_rate": 2.009611698644745e-06,
+      "loss": 0.4998,
+      "step": 5745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3997597838397018,
+      "learning_rate": 2.008001274812204e-06,
+      "loss": 0.4884,
+      "step": 5746
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8261750841737288,
+      "learning_rate": 2.006391334363407e-06,
+      "loss": 0.5535,
+      "step": 5747
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6321622352936611,
+      "learning_rate": 2.004781877558455e-06,
+      "loss": 0.4615,
+      "step": 5748
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.914268686757844,
+      "learning_rate": 2.003172904657372e-06,
+      "loss": 0.4293,
+      "step": 5749
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3570114851918123,
+      "learning_rate": 2.0015644159200974e-06,
+      "loss": 0.5056,
+      "step": 5750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0706697522578237,
+      "learning_rate": 1.9999564116065017e-06,
+      "loss": 0.4964,
+      "step": 5751
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0556824675288645,
+      "learning_rate": 1.998348891976368e-06,
+      "loss": 0.4605,
+      "step": 5752
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9574956421434813,
+      "learning_rate": 1.9967418572894087e-06,
+      "loss": 0.4795,
+      "step": 5753
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6449227143739888,
+      "learning_rate": 1.995135307805251e-06,
+      "loss": 0.5001,
+      "step": 5754
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.46664255915646,
+      "learning_rate": 1.9935292437834508e-06,
+      "loss": 0.5121,
+      "step": 5755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3813895466955008,
+      "learning_rate": 1.9919236654834776e-06,
+      "loss": 0.5109,
+      "step": 5756
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0464129272443827,
+      "learning_rate": 1.9903185731647294e-06,
+      "loss": 0.4965,
+      "step": 5757
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.575954407433671,
+      "learning_rate": 1.988713967086524e-06,
+      "loss": 0.5024,
+      "step": 5758
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5791710315234764,
+      "learning_rate": 1.9871098475080968e-06,
+      "loss": 0.5293,
+      "step": 5759
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5824864416558846,
+      "learning_rate": 1.9855062146886104e-06,
+      "loss": 0.526,
+      "step": 5760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7775269855188194,
+      "learning_rate": 1.9839030688871432e-06,
+      "loss": 0.5091,
+      "step": 5761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5349204131009164,
+      "learning_rate": 1.9823004103626996e-06,
+      "loss": 0.478,
+      "step": 5762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3719854482227383,
+      "learning_rate": 1.980698239374201e-06,
+      "loss": 0.4685,
+      "step": 5763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6852175119165915,
+      "learning_rate": 1.979096556180493e-06,
+      "loss": 0.4881,
+      "step": 5764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8625313877636154,
+      "learning_rate": 1.9774953610403443e-06,
+      "loss": 0.5216,
+      "step": 5765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.450107885202208,
+      "learning_rate": 1.975894654212438e-06,
+      "loss": 0.4513,
+      "step": 5766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5230450906221689,
+      "learning_rate": 1.9742944359553855e-06,
+      "loss": 0.5149,
+      "step": 5767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.873593720362626,
+      "learning_rate": 1.972694706527714e-06,
+      "loss": 0.4503,
+      "step": 5768
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.623455636704811,
+      "learning_rate": 1.971095466187876e-06,
+      "loss": 0.538,
+      "step": 5769
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.696139524516132,
+      "learning_rate": 1.9694967151942403e-06,
+      "loss": 0.471,
+      "step": 5770
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9568653647405692,
+      "learning_rate": 1.9678984538051015e-06,
+      "loss": 0.5261,
+      "step": 5771
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5443916493642844,
+      "learning_rate": 1.966300682278671e-06,
+      "loss": 0.4776,
+      "step": 5772
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6470232629154051,
+      "learning_rate": 1.964703400873083e-06,
+      "loss": 0.465,
+      "step": 5773
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4860505755950013,
+      "learning_rate": 1.963106609846395e-06,
+      "loss": 0.4593,
+      "step": 5774
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6271448336615271,
+      "learning_rate": 1.9615103094565798e-06,
+      "loss": 0.5128,
+      "step": 5775
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6066932286908366,
+      "learning_rate": 1.9599144999615355e-06,
+      "loss": 0.513,
+      "step": 5776
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1186709048927224,
+      "learning_rate": 1.9583191816190773e-06,
+      "loss": 0.5115,
+      "step": 5777
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4370821816965451,
+      "learning_rate": 1.9567243546869453e-06,
+      "loss": 0.4406,
+      "step": 5778
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.352541970965906,
+      "learning_rate": 1.955130019422795e-06,
+      "loss": 0.5274,
+      "step": 5779
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3274741708035742,
+      "learning_rate": 1.953536176084207e-06,
+      "loss": 0.4426,
+      "step": 5780
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4631876821754177,
+      "learning_rate": 1.9519428249286825e-06,
+      "loss": 0.5083,
+      "step": 5781
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4116956697135918,
+      "learning_rate": 1.9503499662136378e-06,
+      "loss": 0.5186,
+      "step": 5782
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.684331510368117,
+      "learning_rate": 1.9487576001964166e-06,
+      "loss": 0.4865,
+      "step": 5783
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4742398214878512,
+      "learning_rate": 1.947165727134276e-06,
+      "loss": 0.5243,
+      "step": 5784
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6588770030376725,
+      "learning_rate": 1.945574347284401e-06,
+      "loss": 0.5375,
+      "step": 5785
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4088544461510655,
+      "learning_rate": 1.9439834609038893e-06,
+      "loss": 0.4529,
+      "step": 5786
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4820327420256838,
+      "learning_rate": 1.9423930682497664e-06,
+      "loss": 0.4909,
+      "step": 5787
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4534489856721242,
+      "learning_rate": 1.940803169578969e-06,
+      "loss": 0.5388,
+      "step": 5788
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4831040727804403,
+      "learning_rate": 1.939213765148366e-06,
+      "loss": 0.5078,
+      "step": 5789
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3836192912332803,
+      "learning_rate": 1.937624855214736e-06,
+      "loss": 0.5005,
+      "step": 5790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8974608544636409,
+      "learning_rate": 1.9360364400347803e-06,
+      "loss": 0.5136,
+      "step": 5791
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.59025346888609,
+      "learning_rate": 1.9344485198651243e-06,
+      "loss": 0.5284,
+      "step": 5792
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.767797328488242,
+      "learning_rate": 1.9328610949623068e-06,
+      "loss": 0.5235,
+      "step": 5793
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7296551897243873,
+      "learning_rate": 1.9312741655827945e-06,
+      "loss": 0.5297,
+      "step": 5794
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8958100276090986,
+      "learning_rate": 1.9296877319829656e-06,
+      "loss": 0.4773,
+      "step": 5795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8361969548943964,
+      "learning_rate": 1.928101794419126e-06,
+      "loss": 0.4559,
+      "step": 5796
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.215608924820692,
+      "learning_rate": 1.9265163531474935e-06,
+      "loss": 0.5468,
+      "step": 5797
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.1295612581503611,
+      "learning_rate": 1.924931408424216e-06,
+      "loss": 0.4339,
+      "step": 5798
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4019818248835516,
+      "learning_rate": 1.923346960505353e-06,
+      "loss": 0.5145,
+      "step": 5799
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6831507214857904,
+      "learning_rate": 1.9217630096468824e-06,
+      "loss": 0.5159,
+      "step": 5800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4923805376132564,
+      "learning_rate": 1.920179556104711e-06,
+      "loss": 0.5629,
+      "step": 5801
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2287299430523901,
+      "learning_rate": 1.9185966001346546e-06,
+      "loss": 0.4693,
+      "step": 5802
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6705904758601706,
+      "learning_rate": 1.9170141419924583e-06,
+      "loss": 0.4705,
+      "step": 5803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.447868152017363,
+      "learning_rate": 1.915432181933778e-06,
+      "loss": 0.4798,
+      "step": 5804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7060735277920716,
+      "learning_rate": 1.9138507202141947e-06,
+      "loss": 0.4894,
+      "step": 5805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4861249493666067,
+      "learning_rate": 1.912269757089208e-06,
+      "loss": 0.5246,
+      "step": 5806
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.505205873541527,
+      "learning_rate": 1.9106892928142383e-06,
+      "loss": 0.4855,
+      "step": 5807
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0736622759742143,
+      "learning_rate": 1.9091093276446197e-06,
+      "loss": 0.4664,
+      "step": 5808
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5728866068387426,
+      "learning_rate": 1.9075298618356134e-06,
+      "loss": 0.4832,
+      "step": 5809
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.567346546602158,
+      "learning_rate": 1.905950895642394e-06,
+      "loss": 0.523,
+      "step": 5810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4303099435053412,
+      "learning_rate": 1.9043724293200556e-06,
+      "loss": 0.5129,
+      "step": 5811
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0313109936019274,
+      "learning_rate": 1.9027944631236161e-06,
+      "loss": 0.533,
+      "step": 5812
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6289956972165225,
+      "learning_rate": 1.9012169973080064e-06,
+      "loss": 0.4587,
+      "step": 5813
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4720872598727337,
+      "learning_rate": 1.899640032128085e-06,
+      "loss": 0.4889,
+      "step": 5814
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3338535876848339,
+      "learning_rate": 1.8980635678386206e-06,
+      "loss": 0.4974,
+      "step": 5815
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4441314017374718,
+      "learning_rate": 1.8964876046943081e-06,
+      "loss": 0.484,
+      "step": 5816
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5841626988615176,
+      "learning_rate": 1.8949121429497546e-06,
+      "loss": 0.5346,
+      "step": 5817
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.51894023016548,
+      "learning_rate": 1.8933371828594932e-06,
+      "loss": 0.4579,
+      "step": 5818
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3708186159993732,
+      "learning_rate": 1.8917627246779708e-06,
+      "loss": 0.4752,
+      "step": 5819
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.351325739538545,
+      "learning_rate": 1.890188768659554e-06,
+      "loss": 0.5004,
+      "step": 5820
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4460543738772533,
+      "learning_rate": 1.8886153150585295e-06,
+      "loss": 0.5045,
+      "step": 5821
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4696479913824958,
+      "learning_rate": 1.8870423641291042e-06,
+      "loss": 0.4601,
+      "step": 5822
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5593078416590118,
+      "learning_rate": 1.8854699161254031e-06,
+      "loss": 0.4386,
+      "step": 5823
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6144109272623572,
+      "learning_rate": 1.8838979713014654e-06,
+      "loss": 0.4931,
+      "step": 5824
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.021351446475987,
+      "learning_rate": 1.8823265299112564e-06,
+      "loss": 0.4625,
+      "step": 5825
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5024912284123797,
+      "learning_rate": 1.880755592208653e-06,
+      "loss": 0.4997,
+      "step": 5826
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7213862970566716,
+      "learning_rate": 1.879185158447457e-06,
+      "loss": 0.4717,
+      "step": 5827
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6218774973615702,
+      "learning_rate": 1.8776152288813842e-06,
+      "loss": 0.474,
+      "step": 5828
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9492154048572368,
+      "learning_rate": 1.8760458037640677e-06,
+      "loss": 0.4841,
+      "step": 5829
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.374140362550844,
+      "learning_rate": 1.874476883349068e-06,
+      "loss": 0.4599,
+      "step": 5830
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8033689611937722,
+      "learning_rate": 1.8729084678898534e-06,
+      "loss": 0.4844,
+      "step": 5831
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.441908053408819,
+      "learning_rate": 1.8713405576398187e-06,
+      "loss": 0.4918,
+      "step": 5832
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6312171403721705,
+      "learning_rate": 1.8697731528522694e-06,
+      "loss": 0.5077,
+      "step": 5833
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.715085014776812,
+      "learning_rate": 1.868206253780438e-06,
+      "loss": 0.4836,
+      "step": 5834
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3763809397355025,
+      "learning_rate": 1.8666398606774667e-06,
+      "loss": 0.4653,
+      "step": 5835
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4817881047709291,
+      "learning_rate": 1.865073973796424e-06,
+      "loss": 0.5401,
+      "step": 5836
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8522709107388002,
+      "learning_rate": 1.8635085933902907e-06,
+      "loss": 0.5189,
+      "step": 5837
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6687465006836861,
+      "learning_rate": 1.8619437197119644e-06,
+      "loss": 0.4893,
+      "step": 5838
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8708791015591064,
+      "learning_rate": 1.860379353014271e-06,
+      "loss": 0.5368,
+      "step": 5839
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7487248472707533,
+      "learning_rate": 1.858815493549943e-06,
+      "loss": 0.5026,
+      "step": 5840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 17.919683857302203,
+      "learning_rate": 1.8572521415716387e-06,
+      "loss": 0.4393,
+      "step": 5841
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.462150126869578,
+      "learning_rate": 1.8556892973319284e-06,
+      "loss": 0.4983,
+      "step": 5842
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.334792323329871,
+      "learning_rate": 1.8541269610833061e-06,
+      "loss": 0.5266,
+      "step": 5843
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2295138381557114,
+      "learning_rate": 1.852565133078178e-06,
+      "loss": 0.5002,
+      "step": 5844
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3606810914725282,
+      "learning_rate": 1.851003813568874e-06,
+      "loss": 0.5448,
+      "step": 5845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3410900367471983,
+      "learning_rate": 1.8494430028076372e-06,
+      "loss": 0.4918,
+      "step": 5846
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.353687681132741,
+      "learning_rate": 1.84788270104663e-06,
+      "loss": 0.4591,
+      "step": 5847
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2867728993341863,
+      "learning_rate": 1.846322908537936e-06,
+      "loss": 0.4626,
+      "step": 5848
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5021326759171463,
+      "learning_rate": 1.8447636255335488e-06,
+      "loss": 0.5241,
+      "step": 5849
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5503247553351363,
+      "learning_rate": 1.8432048522853891e-06,
+      "loss": 0.5121,
+      "step": 5850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.851781475388872,
+      "learning_rate": 1.8416465890452862e-06,
+      "loss": 0.4923,
+      "step": 5851
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7091706381910228,
+      "learning_rate": 1.8400888360649949e-06,
+      "loss": 0.477,
+      "step": 5852
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4134026379502893,
+      "learning_rate": 1.8385315935961805e-06,
+      "loss": 0.5358,
+      "step": 5853
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0054461199669316,
+      "learning_rate": 1.836974861890431e-06,
+      "loss": 0.5161,
+      "step": 5854
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3766422435819043,
+      "learning_rate": 1.8354186411992514e-06,
+      "loss": 0.4829,
+      "step": 5855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3201172847650677,
+      "learning_rate": 1.8338629317740598e-06,
+      "loss": 0.4482,
+      "step": 5856
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0163699674700495,
+      "learning_rate": 1.8323077338661981e-06,
+      "loss": 0.5215,
+      "step": 5857
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6190726538042708,
+      "learning_rate": 1.8307530477269192e-06,
+      "loss": 0.4652,
+      "step": 5858
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4391573430666145,
+      "learning_rate": 1.829198873607399e-06,
+      "loss": 0.5407,
+      "step": 5859
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5196588134185403,
+      "learning_rate": 1.8276452117587252e-06,
+      "loss": 0.5364,
+      "step": 5860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4095757168875038,
+      "learning_rate": 1.8260920624319084e-06,
+      "loss": 0.5028,
+      "step": 5861
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.915322471299682,
+      "learning_rate": 1.824539425877871e-06,
+      "loss": 0.4948,
+      "step": 5862
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3294204484252052,
+      "learning_rate": 1.822987302347456e-06,
+      "loss": 0.5076,
+      "step": 5863
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.472479235952562,
+      "learning_rate": 1.8214356920914244e-06,
+      "loss": 0.4979,
+      "step": 5864
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3967686022768886,
+      "learning_rate": 1.8198845953604494e-06,
+      "loss": 0.4981,
+      "step": 5865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.250885405548494,
+      "learning_rate": 1.818334012405128e-06,
+      "loss": 0.4818,
+      "step": 5866
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5312106093910633,
+      "learning_rate": 1.8167839434759665e-06,
+      "loss": 0.4943,
+      "step": 5867
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5033972291446327,
+      "learning_rate": 1.8152343888233965e-06,
+      "loss": 0.5091,
+      "step": 5868
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4241722148533027,
+      "learning_rate": 1.8136853486977575e-06,
+      "loss": 0.4618,
+      "step": 5869
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.837726102143834,
+      "learning_rate": 1.8121368233493154e-06,
+      "loss": 0.4202,
+      "step": 5870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3763583291820833,
+      "learning_rate": 1.8105888130282433e-06,
+      "loss": 0.4719,
+      "step": 5871
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3135692863116784,
+      "learning_rate": 1.8090413179846383e-06,
+      "loss": 0.4808,
+      "step": 5872
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6601278979307443,
+      "learning_rate": 1.8074943384685139e-06,
+      "loss": 0.4719,
+      "step": 5873
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.533500102581177,
+      "learning_rate": 1.8059478747297942e-06,
+      "loss": 0.5315,
+      "step": 5874
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9325656771854214,
+      "learning_rate": 1.8044019270183278e-06,
+      "loss": 0.4732,
+      "step": 5875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6384871820813307,
+      "learning_rate": 1.802856495583873e-06,
+      "loss": 0.4881,
+      "step": 5876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3817610947695655,
+      "learning_rate": 1.8013115806761105e-06,
+      "loss": 0.5039,
+      "step": 5877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9646188909105635,
+      "learning_rate": 1.7997671825446323e-06,
+      "loss": 0.5361,
+      "step": 5878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.658577168296523,
+      "learning_rate": 1.798223301438951e-06,
+      "loss": 0.5399,
+      "step": 5879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6561815290671748,
+      "learning_rate": 1.7966799376084954e-06,
+      "loss": 0.5165,
+      "step": 5880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.698367245602605,
+      "learning_rate": 1.7951370913026067e-06,
+      "loss": 0.5214,
+      "step": 5881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4701773401872964,
+      "learning_rate": 1.7935947627705485e-06,
+      "loss": 0.4431,
+      "step": 5882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4422387143040183,
+      "learning_rate": 1.7920529522614943e-06,
+      "loss": 0.5262,
+      "step": 5883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3581629909265374,
+      "learning_rate": 1.7905116600245404e-06,
+      "loss": 0.5287,
+      "step": 5884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5816641254198855,
+      "learning_rate": 1.788970886308693e-06,
+      "loss": 0.5171,
+      "step": 5885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6127891341786535,
+      "learning_rate": 1.7874306313628802e-06,
+      "loss": 0.4907,
+      "step": 5886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3761269959792422,
+      "learning_rate": 1.78589089543594e-06,
+      "loss": 0.4984,
+      "step": 5887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6734571305492423,
+      "learning_rate": 1.7843516787766357e-06,
+      "loss": 0.5091,
+      "step": 5888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4534246916626015,
+      "learning_rate": 1.7828129816336387e-06,
+      "loss": 0.5099,
+      "step": 5889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4742219251314783,
+      "learning_rate": 1.7812748042555378e-06,
+      "loss": 0.4914,
+      "step": 5890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.357326060264499,
+      "learning_rate": 1.7797371468908414e-06,
+      "loss": 0.461,
+      "step": 5891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6500132546473334,
+      "learning_rate": 1.7782000097879692e-06,
+      "loss": 0.4542,
+      "step": 5892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4501522402373412,
+      "learning_rate": 1.776663393195262e-06,
+      "loss": 0.4509,
+      "step": 5893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3054125980508295,
+      "learning_rate": 1.7751272973609707e-06,
+      "loss": 0.4766,
+      "step": 5894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.982710616249453,
+      "learning_rate": 1.7735917225332666e-06,
+      "loss": 0.5887,
+      "step": 5895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.888839005725094,
+      "learning_rate": 1.7720566689602354e-06,
+      "loss": 0.4827,
+      "step": 5896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4611341133548106,
+      "learning_rate": 1.77052213688988e-06,
+      "loss": 0.4747,
+      "step": 5897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.894171439291106,
+      "learning_rate": 1.768988126570116e-06,
+      "loss": 0.5112,
+      "step": 5898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2186601395328758,
+      "learning_rate": 1.767454638248775e-06,
+      "loss": 0.4769,
+      "step": 5899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2891341736086237,
+      "learning_rate": 1.7659216721736082e-06,
+      "loss": 0.4466,
+      "step": 5900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4263591816159855,
+      "learning_rate": 1.764389228592277e-06,
+      "loss": 0.4862,
+      "step": 5901
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.997366258609366,
+      "learning_rate": 1.7628573077523647e-06,
+      "loss": 0.4851,
+      "step": 5902
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6842837922339248,
+      "learning_rate": 1.7613259099013608e-06,
+      "loss": 0.4807,
+      "step": 5903
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4564574019636156,
+      "learning_rate": 1.7597950352866833e-06,
+      "loss": 0.4421,
+      "step": 5904
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.783210696583975,
+      "learning_rate": 1.7582646841556533e-06,
+      "loss": 0.5436,
+      "step": 5905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3190447863888473,
+      "learning_rate": 1.756734856755516e-06,
+      "loss": 0.4655,
+      "step": 5906
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8206358433489251,
+      "learning_rate": 1.7552055533334268e-06,
+      "loss": 0.4928,
+      "step": 5907
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4403153492787868,
+      "learning_rate": 1.7536767741364558e-06,
+      "loss": 0.4993,
+      "step": 5908
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3428632704917827,
+      "learning_rate": 1.752148519411595e-06,
+      "loss": 0.4975,
+      "step": 5909
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3080469443346712,
+      "learning_rate": 1.7506207894057442e-06,
+      "loss": 0.4692,
+      "step": 5910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2920041190174922,
+      "learning_rate": 1.7490935843657242e-06,
+      "loss": 0.5237,
+      "step": 5911
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4014102888616162,
+      "learning_rate": 1.7475669045382636e-06,
+      "loss": 0.4767,
+      "step": 5912
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4549220607061533,
+      "learning_rate": 1.7460407501700178e-06,
+      "loss": 0.5206,
+      "step": 5913
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.503657221600372,
+      "learning_rate": 1.7445151215075456e-06,
+      "loss": 0.4472,
+      "step": 5914
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6978670048145843,
+      "learning_rate": 1.7429900187973287e-06,
+      "loss": 0.496,
+      "step": 5915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9133555141324627,
+      "learning_rate": 1.741465442285758e-06,
+      "loss": 0.5304,
+      "step": 5916
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3521218724701396,
+      "learning_rate": 1.7399413922191455e-06,
+      "loss": 0.4711,
+      "step": 5917
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2834337970618197,
+      "learning_rate": 1.7384178688437132e-06,
+      "loss": 0.4776,
+      "step": 5918
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4133634953894676,
+      "learning_rate": 1.7368948724055974e-06,
+      "loss": 0.5234,
+      "step": 5919
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7352465408962612,
+      "learning_rate": 1.735372403150854e-06,
+      "loss": 0.499,
+      "step": 5920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5830249891711863,
+      "learning_rate": 1.7338504613254515e-06,
+      "loss": 0.5055,
+      "step": 5921
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6650457763157955,
+      "learning_rate": 1.7323290471752741e-06,
+      "loss": 0.4851,
+      "step": 5922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.181933619004897,
+      "learning_rate": 1.7308081609461163e-06,
+      "loss": 0.4511,
+      "step": 5923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7000840553200878,
+      "learning_rate": 1.7292878028836946e-06,
+      "loss": 0.4752,
+      "step": 5924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3554970293049347,
+      "learning_rate": 1.7277679732336328e-06,
+      "loss": 0.474,
+      "step": 5925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.67754838277534,
+      "learning_rate": 1.7262486722414752e-06,
+      "loss": 0.5077,
+      "step": 5926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3496912569635924,
+      "learning_rate": 1.7247299001526773e-06,
+      "loss": 0.4886,
+      "step": 5927
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.27870766288199,
+      "learning_rate": 1.7232116572126067e-06,
+      "loss": 0.423,
+      "step": 5928
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8641894032739894,
+      "learning_rate": 1.7216939436665558e-06,
+      "loss": 0.5036,
+      "step": 5929
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6300684737499331,
+      "learning_rate": 1.7201767597597197e-06,
+      "loss": 0.4689,
+      "step": 5930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2278131681948699,
+      "learning_rate": 1.7186601057372155e-06,
+      "loss": 0.4723,
+      "step": 5931
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.031259017202795,
+      "learning_rate": 1.7171439818440688e-06,
+      "loss": 0.4803,
+      "step": 5932
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6538675916367489,
+      "learning_rate": 1.7156283883252268e-06,
+      "loss": 0.4797,
+      "step": 5933
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6663573116080352,
+      "learning_rate": 1.7141133254255426e-06,
+      "loss": 0.4742,
+      "step": 5934
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3804574862462933,
+      "learning_rate": 1.712598793389792e-06,
+      "loss": 0.4476,
+      "step": 5935
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7024565514569163,
+      "learning_rate": 1.7110847924626578e-06,
+      "loss": 0.5029,
+      "step": 5936
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4488627553590303,
+      "learning_rate": 1.7095713228887411e-06,
+      "loss": 0.4519,
+      "step": 5937
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5775295336664255,
+      "learning_rate": 1.7080583849125588e-06,
+      "loss": 0.511,
+      "step": 5938
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6119925589353812,
+      "learning_rate": 1.7065459787785355e-06,
+      "loss": 0.4613,
+      "step": 5939
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6115711077895967,
+      "learning_rate": 1.705034104731017e-06,
+      "loss": 0.5089,
+      "step": 5940
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7297349527982928,
+      "learning_rate": 1.703522763014257e-06,
+      "loss": 0.5235,
+      "step": 5941
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.353036605970803,
+      "learning_rate": 1.702011953872429e-06,
+      "loss": 0.462,
+      "step": 5942
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9813637490464453,
+      "learning_rate": 1.7005016775496135e-06,
+      "loss": 0.5045,
+      "step": 5943
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5793417303571762,
+      "learning_rate": 1.6989919342898127e-06,
+      "loss": 0.4784,
+      "step": 5944
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.50614086925251,
+      "learning_rate": 1.697482724336938e-06,
+      "loss": 0.4773,
+      "step": 5945
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7080841760436045,
+      "learning_rate": 1.695974047934814e-06,
+      "loss": 0.5019,
+      "step": 5946
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0029589857718344,
+      "learning_rate": 1.694465905327184e-06,
+      "loss": 0.5011,
+      "step": 5947
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.684124017120539,
+      "learning_rate": 1.6929582967576975e-06,
+      "loss": 0.5296,
+      "step": 5948
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2969706559048375,
+      "learning_rate": 1.6914512224699253e-06,
+      "loss": 0.514,
+      "step": 5949
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5848640384886945,
+      "learning_rate": 1.6899446827073458e-06,
+      "loss": 0.5012,
+      "step": 5950
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9123582479570977,
+      "learning_rate": 1.6884386777133566e-06,
+      "loss": 0.5222,
+      "step": 5951
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6449177798951945,
+      "learning_rate": 1.6869332077312634e-06,
+      "loss": 0.4658,
+      "step": 5952
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4782830397628994,
+      "learning_rate": 1.6854282730042893e-06,
+      "loss": 0.4986,
+      "step": 5953
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.667017827478795,
+      "learning_rate": 1.6839238737755715e-06,
+      "loss": 0.4877,
+      "step": 5954
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.203265992788667,
+      "learning_rate": 1.682420010288155e-06,
+      "loss": 0.4829,
+      "step": 5955
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5407269389663272,
+      "learning_rate": 1.680916682785007e-06,
+      "loss": 0.5296,
+      "step": 5956
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.962889161814955,
+      "learning_rate": 1.6794138915089987e-06,
+      "loss": 0.4829,
+      "step": 5957
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6374891209770737,
+      "learning_rate": 1.677911636702923e-06,
+      "loss": 0.5128,
+      "step": 5958
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3436117385055986,
+      "learning_rate": 1.676409918609479e-06,
+      "loss": 0.4986,
+      "step": 5959
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5132495078968142,
+      "learning_rate": 1.6749087374712858e-06,
+      "loss": 0.5377,
+      "step": 5960
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.465158233052855,
+      "learning_rate": 1.6734080935308694e-06,
+      "loss": 0.512,
+      "step": 5961
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.704589541225872,
+      "learning_rate": 1.6719079870306737e-06,
+      "loss": 0.4977,
+      "step": 5962
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3285426383237866,
+      "learning_rate": 1.6704084182130552e-06,
+      "loss": 0.4882,
+      "step": 5963
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.767399204910095,
+      "learning_rate": 1.66890938732028e-06,
+      "loss": 0.4877,
+      "step": 5964
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.677426192641951,
+      "learning_rate": 1.6674108945945323e-06,
+      "loss": 0.4759,
+      "step": 5965
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3255144212986685,
+      "learning_rate": 1.6659129402779034e-06,
+      "loss": 0.4964,
+      "step": 5966
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4742160373416446,
+      "learning_rate": 1.664415524612405e-06,
+      "loss": 0.4928,
+      "step": 5967
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1391153981085873,
+      "learning_rate": 1.6629186478399538e-06,
+      "loss": 0.508,
+      "step": 5968
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3940825832138908,
+      "learning_rate": 1.6614223102023857e-06,
+      "loss": 0.4659,
+      "step": 5969
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.777478443620191,
+      "learning_rate": 1.6599265119414487e-06,
+      "loss": 0.4779,
+      "step": 5970
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4064111179425418,
+      "learning_rate": 1.658431253298799e-06,
+      "loss": 0.4892,
+      "step": 5971
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3686681256777833,
+      "learning_rate": 1.6569365345160116e-06,
+      "loss": 0.5192,
+      "step": 5972
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6184845033964539,
+      "learning_rate": 1.6554423558345683e-06,
+      "loss": 0.4287,
+      "step": 5973
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9290552096867895,
+      "learning_rate": 1.6539487174958706e-06,
+      "loss": 0.4356,
+      "step": 5974
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3287246157659554,
+      "learning_rate": 1.652455619741225e-06,
+      "loss": 0.4179,
+      "step": 5975
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3576869075341544,
+      "learning_rate": 1.6509630628118584e-06,
+      "loss": 0.4915,
+      "step": 5976
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7120315215609672,
+      "learning_rate": 1.6494710469489033e-06,
+      "loss": 0.5053,
+      "step": 5977
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.845271579003526,
+      "learning_rate": 1.6479795723934088e-06,
+      "loss": 0.5596,
+      "step": 5978
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5981560815448819,
+      "learning_rate": 1.646488639386339e-06,
+      "loss": 0.49,
+      "step": 5979
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3663767878155546,
+      "learning_rate": 1.6449982481685616e-06,
+      "loss": 0.4987,
+      "step": 5980
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3410677975477996,
+      "learning_rate": 1.6435083989808682e-06,
+      "loss": 0.5274,
+      "step": 5981
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4945763766755613,
+      "learning_rate": 1.6420190920639522e-06,
+      "loss": 0.544,
+      "step": 5982
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3177394972577257,
+      "learning_rate": 1.6405303276584277e-06,
+      "loss": 0.482,
+      "step": 5983
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7405380423742085,
+      "learning_rate": 1.6390421060048151e-06,
+      "loss": 0.4612,
+      "step": 5984
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3509356339041592,
+      "learning_rate": 1.6375544273435506e-06,
+      "loss": 0.5186,
+      "step": 5985
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3877589841195326,
+      "learning_rate": 1.636067291914983e-06,
+      "loss": 0.4668,
+      "step": 5986
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9029029071359749,
+      "learning_rate": 1.6345806999593717e-06,
+      "loss": 0.5034,
+      "step": 5987
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6081825938648133,
+      "learning_rate": 1.633094651716889e-06,
+      "loss": 0.4844,
+      "step": 5988
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4427022899923019,
+      "learning_rate": 1.6316091474276163e-06,
+      "loss": 0.4864,
+      "step": 5989
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4622955643128948,
+      "learning_rate": 1.6301241873315544e-06,
+      "loss": 0.5059,
+      "step": 5990
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2968792017603343,
+      "learning_rate": 1.628639771668607e-06,
+      "loss": 0.5585,
+      "step": 5991
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.873706603490066,
+      "learning_rate": 1.627155900678598e-06,
+      "loss": 0.4722,
+      "step": 5992
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.139979943714902,
+      "learning_rate": 1.625672574601258e-06,
+      "loss": 0.4448,
+      "step": 5993
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6664026791604412,
+      "learning_rate": 1.6241897936762313e-06,
+      "loss": 0.5281,
+      "step": 5994
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3131089978810235,
+      "learning_rate": 1.6227075581430751e-06,
+      "loss": 0.5188,
+      "step": 5995
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.287306397039178,
+      "learning_rate": 1.621225868241259e-06,
+      "loss": 0.4818,
+      "step": 5996
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4776254966103886,
+      "learning_rate": 1.6197447242101615e-06,
+      "loss": 0.4644,
+      "step": 5997
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6944930673044044,
+      "learning_rate": 1.618264126289073e-06,
+      "loss": 0.5063,
+      "step": 5998
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.554341842980303,
+      "learning_rate": 1.6167840747171997e-06,
+      "loss": 0.4816,
+      "step": 5999
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4297897583485726,
+      "learning_rate": 1.6153045697336544e-06,
+      "loss": 0.4987,
+      "step": 6000
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.810195614493212,
+      "learning_rate": 1.6138256115774676e-06,
+      "loss": 0.5221,
+      "step": 6001
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.166741697516714,
+      "learning_rate": 1.6123472004875724e-06,
+      "loss": 0.4781,
+      "step": 6002
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3194032375621023,
+      "learning_rate": 1.610869336702826e-06,
+      "loss": 0.459,
+      "step": 6003
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3722221268424892,
+      "learning_rate": 1.6093920204619856e-06,
+      "loss": 0.4974,
+      "step": 6004
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.340858799507641,
+      "learning_rate": 1.6079152520037277e-06,
+      "loss": 0.5013,
+      "step": 6005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1658509794232037,
+      "learning_rate": 1.6064390315666356e-06,
+      "loss": 0.4842,
+      "step": 6006
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.843094029528777,
+      "learning_rate": 1.6049633593892045e-06,
+      "loss": 0.4917,
+      "step": 6007
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6668625189375703,
+      "learning_rate": 1.6034882357098447e-06,
+      "loss": 0.5162,
+      "step": 6008
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6536351282188947,
+      "learning_rate": 1.6020136607668724e-06,
+      "loss": 0.4686,
+      "step": 6009
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7594181925466228,
+      "learning_rate": 1.6005396347985204e-06,
+      "loss": 0.5251,
+      "step": 6010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3645245329180153,
+      "learning_rate": 1.59906615804293e-06,
+      "loss": 0.4973,
+      "step": 6011
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 13.455150978415842,
+      "learning_rate": 1.5975932307381564e-06,
+      "loss": 0.5131,
+      "step": 6012
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7661889663932724,
+      "learning_rate": 1.59612085312216e-06,
+      "loss": 0.5674,
+      "step": 6013
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5823722119837638,
+      "learning_rate": 1.5946490254328207e-06,
+      "loss": 0.517,
+      "step": 6014
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2825435915322208,
+      "learning_rate": 1.5931777479079224e-06,
+      "loss": 0.4547,
+      "step": 6015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4780593108877558,
+      "learning_rate": 1.591707020785162e-06,
+      "loss": 0.4743,
+      "step": 6016
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5317792626925832,
+      "learning_rate": 1.5902368443021515e-06,
+      "loss": 0.4931,
+      "step": 6017
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5205937863915808,
+      "learning_rate": 1.5887672186964066e-06,
+      "loss": 0.5418,
+      "step": 6018
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5859984688186561,
+      "learning_rate": 1.5872981442053643e-06,
+      "loss": 0.4967,
+      "step": 6019
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2044996149466667,
+      "learning_rate": 1.585829621066361e-06,
+      "loss": 0.5056,
+      "step": 6020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.523560594149025,
+      "learning_rate": 1.5843616495166536e-06,
+      "loss": 0.4778,
+      "step": 6021
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3660556794601857,
+      "learning_rate": 1.5828942297934018e-06,
+      "loss": 0.4743,
+      "step": 6022
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5846245967103545,
+      "learning_rate": 1.581427362133685e-06,
+      "loss": 0.4882,
+      "step": 6023
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.292360003145945,
+      "learning_rate": 1.5799610467744836e-06,
+      "loss": 0.4528,
+      "step": 6024
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.007002426929482,
+      "learning_rate": 1.578495283952698e-06,
+      "loss": 0.4652,
+      "step": 6025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7468838114684746,
+      "learning_rate": 1.577030073905133e-06,
+      "loss": 0.4918,
+      "step": 6026
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7271485836999365,
+      "learning_rate": 1.5755654168685037e-06,
+      "loss": 0.4931,
+      "step": 6027
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3111986543695473,
+      "learning_rate": 1.5741013130794435e-06,
+      "loss": 0.4633,
+      "step": 6028
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6461414512441257,
+      "learning_rate": 1.5726377627744877e-06,
+      "loss": 0.4677,
+      "step": 6029
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3608209292052993,
+      "learning_rate": 1.5711747661900884e-06,
+      "loss": 0.5287,
+      "step": 6030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.47618381807499,
+      "learning_rate": 1.5697123235626021e-06,
+      "loss": 0.51,
+      "step": 6031
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.524575288440333,
+      "learning_rate": 1.5682504351283034e-06,
+      "loss": 0.5236,
+      "step": 6032
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9620674437293155,
+      "learning_rate": 1.5667891011233688e-06,
+      "loss": 0.5136,
+      "step": 6033
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.429051606799475,
+      "learning_rate": 1.5653283217838938e-06,
+      "loss": 0.48,
+      "step": 6034
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9913598932383123,
+      "learning_rate": 1.563868097345877e-06,
+      "loss": 0.4949,
+      "step": 6035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.724628914180807,
+      "learning_rate": 1.5624084280452313e-06,
+      "loss": 0.5289,
+      "step": 6036
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3544155289688808,
+      "learning_rate": 1.5609493141177817e-06,
+      "loss": 0.4698,
+      "step": 6037
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4219588068053217,
+      "learning_rate": 1.5594907557992568e-06,
+      "loss": 0.5146,
+      "step": 6038
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.409410630606562,
+      "learning_rate": 1.5580327533253037e-06,
+      "loss": 0.487,
+      "step": 6039
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5799718286887938,
+      "learning_rate": 1.5565753069314716e-06,
+      "loss": 0.4632,
+      "step": 6040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6423879889232786,
+      "learning_rate": 1.555118416853227e-06,
+      "loss": 0.5132,
+      "step": 6041
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6570357501497963,
+      "learning_rate": 1.5536620833259409e-06,
+      "loss": 0.5067,
+      "step": 6042
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6347678528254872,
+      "learning_rate": 1.5522063065848974e-06,
+      "loss": 0.4755,
+      "step": 6043
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9050361222724341,
+      "learning_rate": 1.5507510868652919e-06,
+      "loss": 0.5277,
+      "step": 6044
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4023645235573734,
+      "learning_rate": 1.5492964244022258e-06,
+      "loss": 0.5248,
+      "step": 6045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5803261516406482,
+      "learning_rate": 1.5478423194307147e-06,
+      "loss": 0.5049,
+      "step": 6046
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.440458581359572,
+      "learning_rate": 1.546388772185679e-06,
+      "loss": 0.4499,
+      "step": 6047
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5016619536805895,
+      "learning_rate": 1.5449357829019556e-06,
+      "loss": 0.4498,
+      "step": 6048
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6320705758129939,
+      "learning_rate": 1.5434833518142839e-06,
+      "loss": 0.4782,
+      "step": 6049
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.790727339102416,
+      "learning_rate": 1.542031479157321e-06,
+      "loss": 0.4854,
+      "step": 6050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7130602317036618,
+      "learning_rate": 1.5405801651656266e-06,
+      "loss": 0.5126,
+      "step": 6051
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4593958980420332,
+      "learning_rate": 1.5391294100736736e-06,
+      "loss": 0.4715,
+      "step": 6052
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6284291543273373,
+      "learning_rate": 1.5376792141158469e-06,
+      "loss": 0.4736,
+      "step": 6053
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8665726653903791,
+      "learning_rate": 1.5362295775264353e-06,
+      "loss": 0.5013,
+      "step": 6054
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9314327501677002,
+      "learning_rate": 1.5347805005396427e-06,
+      "loss": 0.5416,
+      "step": 6055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5284492197838069,
+      "learning_rate": 1.5333319833895776e-06,
+      "loss": 0.4701,
+      "step": 6056
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4593419373204979,
+      "learning_rate": 1.5318840263102641e-06,
+      "loss": 0.5075,
+      "step": 6057
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7455863857710499,
+      "learning_rate": 1.5304366295356287e-06,
+      "loss": 0.5364,
+      "step": 6058
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.464678343219715,
+      "learning_rate": 1.5289897932995134e-06,
+      "loss": 0.4746,
+      "step": 6059
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8595107075944286,
+      "learning_rate": 1.5275435178356685e-06,
+      "loss": 0.5084,
+      "step": 6060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9740787321561957,
+      "learning_rate": 1.526097803377749e-06,
+      "loss": 0.46,
+      "step": 6061
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6755881265433279,
+      "learning_rate": 1.5246526501593262e-06,
+      "loss": 0.4791,
+      "step": 6062
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5588452764686183,
+      "learning_rate": 1.5232080584138748e-06,
+      "loss": 0.5,
+      "step": 6063
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4090921609069993,
+      "learning_rate": 1.5217640283747836e-06,
+      "loss": 0.4794,
+      "step": 6064
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4383211702238097,
+      "learning_rate": 1.5203205602753462e-06,
+      "loss": 0.4716,
+      "step": 6065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6150559905537978,
+      "learning_rate": 1.5188776543487699e-06,
+      "loss": 0.5223,
+      "step": 6066
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6791885235010544,
+      "learning_rate": 1.5174353108281653e-06,
+      "loss": 0.4273,
+      "step": 6067
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6686708077913581,
+      "learning_rate": 1.515993529946559e-06,
+      "loss": 0.5173,
+      "step": 6068
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4998571863692616,
+      "learning_rate": 1.5145523119368832e-06,
+      "loss": 0.4399,
+      "step": 6069
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3938713925748085,
+      "learning_rate": 1.513111657031977e-06,
+      "loss": 0.477,
+      "step": 6070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.639811660627296,
+      "learning_rate": 1.5116715654645941e-06,
+      "loss": 0.5163,
+      "step": 6071
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8654325248173858,
+      "learning_rate": 1.5102320374673907e-06,
+      "loss": 0.4395,
+      "step": 6072
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.502720874368011,
+      "learning_rate": 1.5087930732729378e-06,
+      "loss": 0.5218,
+      "step": 6073
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.716755769956245,
+      "learning_rate": 1.5073546731137105e-06,
+      "loss": 0.5174,
+      "step": 6074
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2313240862812058,
+      "learning_rate": 1.5059168372220984e-06,
+      "loss": 0.452,
+      "step": 6075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4806778543728099,
+      "learning_rate": 1.5044795658303924e-06,
+      "loss": 0.5089,
+      "step": 6076
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9318546882145895,
+      "learning_rate": 1.503042859170798e-06,
+      "loss": 0.5295,
+      "step": 6077
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3992136102391561,
+      "learning_rate": 1.5016067174754301e-06,
+      "loss": 0.5428,
+      "step": 6078
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5946679371637276,
+      "learning_rate": 1.5001711409763065e-06,
+      "loss": 0.5025,
+      "step": 6079
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3330513773776378,
+      "learning_rate": 1.4987361299053599e-06,
+      "loss": 0.4952,
+      "step": 6080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.763017399981389,
+      "learning_rate": 1.4973016844944265e-06,
+      "loss": 0.4714,
+      "step": 6081
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.412498599014573,
+      "learning_rate": 1.4958678049752568e-06,
+      "loss": 0.4751,
+      "step": 6082
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.077612354528001,
+      "learning_rate": 1.4944344915795034e-06,
+      "loss": 0.5186,
+      "step": 6083
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.636665581614744,
+      "learning_rate": 1.4930017445387317e-06,
+      "loss": 0.5439,
+      "step": 6084
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3895732747424572,
+      "learning_rate": 1.4915695640844174e-06,
+      "loss": 0.4998,
+      "step": 6085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8684005695710397,
+      "learning_rate": 1.4901379504479374e-06,
+      "loss": 0.5444,
+      "step": 6086
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5776228687137166,
+      "learning_rate": 1.488706903860585e-06,
+      "loss": 0.4587,
+      "step": 6087
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5765276916924604,
+      "learning_rate": 1.487276424553556e-06,
+      "loss": 0.4779,
+      "step": 6088
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5607512892815054,
+      "learning_rate": 1.4858465127579585e-06,
+      "loss": 0.5154,
+      "step": 6089
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5630673033403901,
+      "learning_rate": 1.484417168704806e-06,
+      "loss": 0.459,
+      "step": 6090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.608404412381052,
+      "learning_rate": 1.4829883926250238e-06,
+      "loss": 0.5575,
+      "step": 6091
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.239238780180317,
+      "learning_rate": 1.4815601847494388e-06,
+      "loss": 0.5174,
+      "step": 6092
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.386493163183553,
+      "learning_rate": 1.4801325453087967e-06,
+      "loss": 0.4844,
+      "step": 6093
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2527730800500003,
+      "learning_rate": 1.4787054745337403e-06,
+      "loss": 0.5151,
+      "step": 6094
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4224906006832736,
+      "learning_rate": 1.4772789726548293e-06,
+      "loss": 0.47,
+      "step": 6095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4452147059422642,
+      "learning_rate": 1.475853039902525e-06,
+      "loss": 0.4718,
+      "step": 6096
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3464263529106104,
+      "learning_rate": 1.4744276765071991e-06,
+      "loss": 0.4993,
+      "step": 6097
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4470631571011383,
+      "learning_rate": 1.4730028826991338e-06,
+      "loss": 0.5509,
+      "step": 6098
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2049191188781294,
+      "learning_rate": 1.471578658708514e-06,
+      "loss": 0.4637,
+      "step": 6099
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5432552984631496,
+      "learning_rate": 1.4701550047654378e-06,
+      "loss": 0.5719,
+      "step": 6100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7224810211178514,
+      "learning_rate": 1.4687319210999085e-06,
+      "loss": 0.5144,
+      "step": 6101
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3750536849035844,
+      "learning_rate": 1.46730940794184e-06,
+      "loss": 0.4745,
+      "step": 6102
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2787319080534196,
+      "learning_rate": 1.4658874655210476e-06,
+      "loss": 0.5074,
+      "step": 6103
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8074307407464616,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.5201,
+      "step": 6104
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3816429645259922,
+      "learning_rate": 1.4630452938101187e-06,
+      "loss": 0.4676,
+      "step": 6105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4917974592944019,
+      "learning_rate": 1.4616250649791563e-06,
+      "loss": 0.4782,
+      "step": 6106
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7272759649744858,
+      "learning_rate": 1.4602054078038296e-06,
+      "loss": 0.5241,
+      "step": 6107
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7922423149902295,
+      "learning_rate": 1.4587863225134934e-06,
+      "loss": 0.5275,
+      "step": 6108
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0917670324044777,
+      "learning_rate": 1.457367809337415e-06,
+      "loss": 0.491,
+      "step": 6109
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.545252583770274,
+      "learning_rate": 1.4559498685047679e-06,
+      "loss": 0.5203,
+      "step": 6110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.549050685649776,
+      "learning_rate": 1.4545325002446337e-06,
+      "loss": 0.5277,
+      "step": 6111
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4392223393324777,
+      "learning_rate": 1.453115704785999e-06,
+      "loss": 0.485,
+      "step": 6112
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4131394903061087,
+      "learning_rate": 1.4516994823577613e-06,
+      "loss": 0.479,
+      "step": 6113
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.386963970835647,
+      "learning_rate": 1.4502838331887232e-06,
+      "loss": 0.5144,
+      "step": 6114
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5075390793671843,
+      "learning_rate": 1.4488687575075932e-06,
+      "loss": 0.4816,
+      "step": 6115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2656131827967814,
+      "learning_rate": 1.4474542555429927e-06,
+      "loss": 0.4684,
+      "step": 6116
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6712536801227746,
+      "learning_rate": 1.4460403275234425e-06,
+      "loss": 0.4901,
+      "step": 6117
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6216114085232909,
+      "learning_rate": 1.4446269736773805e-06,
+      "loss": 0.4703,
+      "step": 6118
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4842233786242254,
+      "learning_rate": 1.4432141942331423e-06,
+      "loss": 0.5212,
+      "step": 6119
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.319943915507911,
+      "learning_rate": 1.4418019894189783e-06,
+      "loss": 0.4533,
+      "step": 6120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5602930700038722,
+      "learning_rate": 1.4403903594630387e-06,
+      "loss": 0.495,
+      "step": 6121
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3866372003821805,
+      "learning_rate": 1.4389793045933887e-06,
+      "loss": 0.542,
+      "step": 6122
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6214768264161543,
+      "learning_rate": 1.4375688250379927e-06,
+      "loss": 0.4438,
+      "step": 6123
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5641216385268033,
+      "learning_rate": 1.43615892102473e-06,
+      "loss": 0.5375,
+      "step": 6124
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5477655317021919,
+      "learning_rate": 1.4347495927813797e-06,
+      "loss": 0.5442,
+      "step": 6125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7259272691075274,
+      "learning_rate": 1.4333408405356325e-06,
+      "loss": 0.5016,
+      "step": 6126
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5600227007704104,
+      "learning_rate": 1.4319326645150861e-06,
+      "loss": 0.5079,
+      "step": 6127
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7365409956549633,
+      "learning_rate": 1.4305250649472414e-06,
+      "loss": 0.5232,
+      "step": 6128
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.392625333117977,
+      "learning_rate": 1.4291180420595109e-06,
+      "loss": 0.5264,
+      "step": 6129
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5697261875302033,
+      "learning_rate": 1.4277115960792082e-06,
+      "loss": 0.487,
+      "step": 6130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5277808999870366,
+      "learning_rate": 1.4263057272335612e-06,
+      "loss": 0.4424,
+      "step": 6131
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4738957764427278,
+      "learning_rate": 1.4249004357496965e-06,
+      "loss": 0.5131,
+      "step": 6132
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5456403435516437,
+      "learning_rate": 1.4234957218546531e-06,
+      "loss": 0.4809,
+      "step": 6133
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6002757304383604,
+      "learning_rate": 1.4220915857753765e-06,
+      "loss": 0.4778,
+      "step": 6134
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9512042216256624,
+      "learning_rate": 1.4206880277387143e-06,
+      "loss": 0.5432,
+      "step": 6135
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.780164308223862,
+      "learning_rate": 1.419285047971426e-06,
+      "loss": 0.501,
+      "step": 6136
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4504743857264533,
+      "learning_rate": 1.4178826467001733e-06,
+      "loss": 0.524,
+      "step": 6137
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.146544480364636,
+      "learning_rate": 1.4164808241515287e-06,
+      "loss": 0.5136,
+      "step": 6138
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6272967301645165,
+      "learning_rate": 1.4150795805519663e-06,
+      "loss": 0.4752,
+      "step": 6139
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5723827414808167,
+      "learning_rate": 1.4136789161278724e-06,
+      "loss": 0.4918,
+      "step": 6140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3826988338638708,
+      "learning_rate": 1.4122788311055335e-06,
+      "loss": 0.4696,
+      "step": 6141
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3597633810300407,
+      "learning_rate": 1.410879325711147e-06,
+      "loss": 0.5011,
+      "step": 6142
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6124829679414978,
+      "learning_rate": 1.4094804001708174e-06,
+      "loss": 0.5958,
+      "step": 6143
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.542078020146645,
+      "learning_rate": 1.4080820547105501e-06,
+      "loss": 0.531,
+      "step": 6144
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2953216401574426,
+      "learning_rate": 1.4066842895562631e-06,
+      "loss": 0.4607,
+      "step": 6145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7238163480996251,
+      "learning_rate": 1.4052871049337752e-06,
+      "loss": 0.4902,
+      "step": 6146
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6268366494492497,
+      "learning_rate": 1.4038905010688159e-06,
+      "loss": 0.476,
+      "step": 6147
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4296291021875145,
+      "learning_rate": 1.4024944781870164e-06,
+      "loss": 0.4979,
+      "step": 6148
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3735107930351835,
+      "learning_rate": 1.4010990365139194e-06,
+      "loss": 0.474,
+      "step": 6149
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5927504311548404,
+      "learning_rate": 1.399704176274968e-06,
+      "loss": 0.4923,
+      "step": 6150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4353602620740302,
+      "learning_rate": 1.3983098976955157e-06,
+      "loss": 0.4654,
+      "step": 6151
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3544221085197357,
+      "learning_rate": 1.3969162010008213e-06,
+      "loss": 0.4948,
+      "step": 6152
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7144003386631187,
+      "learning_rate": 1.3955230864160462e-06,
+      "loss": 0.5,
+      "step": 6153
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.350318088129962,
+      "learning_rate": 1.394130554166264e-06,
+      "loss": 0.4938,
+      "step": 6154
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.629170310706614,
+      "learning_rate": 1.3927386044764468e-06,
+      "loss": 0.5517,
+      "step": 6155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7072373672078287,
+      "learning_rate": 1.3913472375714792e-06,
+      "loss": 0.4441,
+      "step": 6156
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9292365455123768,
+      "learning_rate": 1.389956453676146e-06,
+      "loss": 0.4999,
+      "step": 6157
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9975739951652836,
+      "learning_rate": 1.3885662530151422e-06,
+      "loss": 0.5228,
+      "step": 6158
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4946002774292722,
+      "learning_rate": 1.3871766358130683e-06,
+      "loss": 0.5346,
+      "step": 6159
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3701032802565098,
+      "learning_rate": 1.3857876022944266e-06,
+      "loss": 0.5035,
+      "step": 6160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7139531071014826,
+      "learning_rate": 1.3843991526836303e-06,
+      "loss": 0.5413,
+      "step": 6161
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.266923031675677,
+      "learning_rate": 1.3830112872049929e-06,
+      "loss": 0.4469,
+      "step": 6162
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5573888558700415,
+      "learning_rate": 1.3816240060827385e-06,
+      "loss": 0.4741,
+      "step": 6163
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7676773716867706,
+      "learning_rate": 1.3802373095409926e-06,
+      "loss": 0.5104,
+      "step": 6164
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5647130625334107,
+      "learning_rate": 1.378851197803791e-06,
+      "loss": 0.4796,
+      "step": 6165
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2424931016563316,
+      "learning_rate": 1.377465671095069e-06,
+      "loss": 0.4755,
+      "step": 6166
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3848258055425164,
+      "learning_rate": 1.3760807296386725e-06,
+      "loss": 0.4646,
+      "step": 6167
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.527042221518362,
+      "learning_rate": 1.374696373658352e-06,
+      "loss": 0.4795,
+      "step": 6168
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2861442316512977,
+      "learning_rate": 1.3733126033777599e-06,
+      "loss": 0.488,
+      "step": 6169
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.724289675612865,
+      "learning_rate": 1.371929419020459e-06,
+      "loss": 0.5291,
+      "step": 6170
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7075573869726784,
+      "learning_rate": 1.370546820809912e-06,
+      "loss": 0.4787,
+      "step": 6171
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4051071751574435,
+      "learning_rate": 1.3691648089694931e-06,
+      "loss": 0.4678,
+      "step": 6172
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4999698282828469,
+      "learning_rate": 1.3677833837224757e-06,
+      "loss": 0.4626,
+      "step": 6173
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.407283713252255,
+      "learning_rate": 1.3664025452920421e-06,
+      "loss": 0.4922,
+      "step": 6174
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5992010864639077,
+      "learning_rate": 1.3650222939012809e-06,
+      "loss": 0.4771,
+      "step": 6175
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6736440131195913,
+      "learning_rate": 1.3636426297731803e-06,
+      "loss": 0.5068,
+      "step": 6176
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4355258158852493,
+      "learning_rate": 1.3622635531306405e-06,
+      "loss": 0.545,
+      "step": 6177
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8904563872315112,
+      "learning_rate": 1.3608850641964604e-06,
+      "loss": 0.4989,
+      "step": 6178
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3131385472915835,
+      "learning_rate": 1.3595071631933504e-06,
+      "loss": 0.4593,
+      "step": 6179
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6578571419253867,
+      "learning_rate": 1.3581298503439193e-06,
+      "loss": 0.4936,
+      "step": 6180
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.517459438778973,
+      "learning_rate": 1.3567531258706874e-06,
+      "loss": 0.4905,
+      "step": 6181
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.281799319717632,
+      "learning_rate": 1.3553769899960716e-06,
+      "loss": 0.5137,
+      "step": 6182
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5282590270708574,
+      "learning_rate": 1.3540014429424049e-06,
+      "loss": 0.542,
+      "step": 6183
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4806309356958096,
+      "learning_rate": 1.3526264849319166e-06,
+      "loss": 0.413,
+      "step": 6184
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4836012581304867,
+      "learning_rate": 1.351252116186741e-06,
+      "loss": 0.4292,
+      "step": 6185
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3555241248525332,
+      "learning_rate": 1.3498783369289226e-06,
+      "loss": 0.4885,
+      "step": 6186
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2483446342138798,
+      "learning_rate": 1.348505147380405e-06,
+      "loss": 0.4379,
+      "step": 6187
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.493588619382274,
+      "learning_rate": 1.347132547763042e-06,
+      "loss": 0.4947,
+      "step": 6188
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.088968788701897,
+      "learning_rate": 1.3457605382985862e-06,
+      "loss": 0.4351,
+      "step": 6189
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3924279918426934,
+      "learning_rate": 1.3443891192087e-06,
+      "loss": 0.4355,
+      "step": 6190
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3716290693646196,
+      "learning_rate": 1.3430182907149447e-06,
+      "loss": 0.4726,
+      "step": 6191
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3336070112616283,
+      "learning_rate": 1.3416480530387955e-06,
+      "loss": 0.4395,
+      "step": 6192
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6298967280551317,
+      "learning_rate": 1.340278406401621e-06,
+      "loss": 0.5065,
+      "step": 6193
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4516402621866342,
+      "learning_rate": 1.3389093510247043e-06,
+      "loss": 0.4969,
+      "step": 6194
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3431009790179682,
+      "learning_rate": 1.3375408871292256e-06,
+      "loss": 0.4955,
+      "step": 6195
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5731051836007435,
+      "learning_rate": 1.3361730149362706e-06,
+      "loss": 0.4964,
+      "step": 6196
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.331718298876939,
+      "learning_rate": 1.3348057346668353e-06,
+      "loss": 0.5137,
+      "step": 6197
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6625395922594366,
+      "learning_rate": 1.3334390465418122e-06,
+      "loss": 0.4958,
+      "step": 6198
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4321764259365748,
+      "learning_rate": 1.3320729507820029e-06,
+      "loss": 0.5206,
+      "step": 6199
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4436428193547688,
+      "learning_rate": 1.3307074476081127e-06,
+      "loss": 0.4634,
+      "step": 6200
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4385882583497545,
+      "learning_rate": 1.3293425372407526e-06,
+      "loss": 0.4517,
+      "step": 6201
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9622624537904163,
+      "learning_rate": 1.3279782199004321e-06,
+      "loss": 0.5084,
+      "step": 6202
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2241283119219653,
+      "learning_rate": 1.3266144958075717e-06,
+      "loss": 0.4693,
+      "step": 6203
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5993684273325757,
+      "learning_rate": 1.325251365182492e-06,
+      "loss": 0.4913,
+      "step": 6204
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4483264471948116,
+      "learning_rate": 1.323888828245417e-06,
+      "loss": 0.5085,
+      "step": 6205
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.916745646680996,
+      "learning_rate": 1.3225268852164797e-06,
+      "loss": 0.5189,
+      "step": 6206
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4769884347476094,
+      "learning_rate": 1.3211655363157094e-06,
+      "loss": 0.4837,
+      "step": 6207
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5140704961515885,
+      "learning_rate": 1.31980478176305e-06,
+      "loss": 0.522,
+      "step": 6208
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6809489514739075,
+      "learning_rate": 1.3184446217783387e-06,
+      "loss": 0.4498,
+      "step": 6209
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4330904014880177,
+      "learning_rate": 1.3170850565813243e-06,
+      "loss": 0.5133,
+      "step": 6210
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4340701334442918,
+      "learning_rate": 1.3157260863916544e-06,
+      "loss": 0.5312,
+      "step": 6211
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4960968171210913,
+      "learning_rate": 1.3143677114288845e-06,
+      "loss": 0.5202,
+      "step": 6212
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5065093070867135,
+      "learning_rate": 1.3130099319124706e-06,
+      "loss": 0.531,
+      "step": 6213
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6062720817845197,
+      "learning_rate": 1.3116527480617735e-06,
+      "loss": 0.4971,
+      "step": 6214
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4525219409854628,
+      "learning_rate": 1.3102961600960584e-06,
+      "loss": 0.5125,
+      "step": 6215
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.333691087311118,
+      "learning_rate": 1.3089401682344955e-06,
+      "loss": 0.5467,
+      "step": 6216
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.738900694292726,
+      "learning_rate": 1.3075847726961571e-06,
+      "loss": 0.52,
+      "step": 6217
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2815307166243293,
+      "learning_rate": 1.3062299737000173e-06,
+      "loss": 0.5316,
+      "step": 6218
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.543516672086808,
+      "learning_rate": 1.3048757714649585e-06,
+      "loss": 0.5369,
+      "step": 6219
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6412206549282686,
+      "learning_rate": 1.3035221662097614e-06,
+      "loss": 0.4903,
+      "step": 6220
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4833664255463928,
+      "learning_rate": 1.302169158153115e-06,
+      "loss": 0.5045,
+      "step": 6221
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5423653118353517,
+      "learning_rate": 1.300816747513609e-06,
+      "loss": 0.5146,
+      "step": 6222
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5360441497197643,
+      "learning_rate": 1.2994649345097354e-06,
+      "loss": 0.4826,
+      "step": 6223
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9787375634564193,
+      "learning_rate": 1.298113719359893e-06,
+      "loss": 0.5337,
+      "step": 6224
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4540743401928375,
+      "learning_rate": 1.2967631022823824e-06,
+      "loss": 0.4575,
+      "step": 6225
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.262005236895857,
+      "learning_rate": 1.29541308349541e-06,
+      "loss": 0.5949,
+      "step": 6226
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6664454512775195,
+      "learning_rate": 1.294063663217079e-06,
+      "loss": 0.4938,
+      "step": 6227
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6280218423673491,
+      "learning_rate": 1.2927148416654033e-06,
+      "loss": 0.5394,
+      "step": 6228
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.430640150277463,
+      "learning_rate": 1.2913666190582947e-06,
+      "loss": 0.492,
+      "step": 6229
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.607236994045575,
+      "learning_rate": 1.2900189956135728e-06,
+      "loss": 0.4961,
+      "step": 6230
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6616581269513682,
+      "learning_rate": 1.288671971548956e-06,
+      "loss": 0.4807,
+      "step": 6231
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1011305393950583,
+      "learning_rate": 1.2873255470820677e-06,
+      "loss": 0.5271,
+      "step": 6232
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3148633012922666,
+      "learning_rate": 1.2859797224304378e-06,
+      "loss": 0.4674,
+      "step": 6233
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5799807696563415,
+      "learning_rate": 1.2846344978114921e-06,
+      "loss": 0.4335,
+      "step": 6234
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2282627622228512,
+      "learning_rate": 1.283289873442567e-06,
+      "loss": 0.4612,
+      "step": 6235
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5334044812396002,
+      "learning_rate": 1.2819458495408948e-06,
+      "loss": 0.4714,
+      "step": 6236
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3003968312361376,
+      "learning_rate": 1.2806024263236178e-06,
+      "loss": 0.4699,
+      "step": 6237
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.214653471241372,
+      "learning_rate": 1.2792596040077753e-06,
+      "loss": 0.4541,
+      "step": 6238
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4826821480376464,
+      "learning_rate": 1.277917382810314e-06,
+      "loss": 0.5076,
+      "step": 6239
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.823356149131492,
+      "learning_rate": 1.27657576294808e-06,
+      "loss": 0.505,
+      "step": 6240
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6382532688362617,
+      "learning_rate": 1.2752347446378238e-06,
+      "loss": 0.5678,
+      "step": 6241
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4384221892090772,
+      "learning_rate": 1.2738943280962013e-06,
+      "loss": 0.4839,
+      "step": 6242
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.342711899294759,
+      "learning_rate": 1.272554513539765e-06,
+      "loss": 0.481,
+      "step": 6243
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.044459208891881,
+      "learning_rate": 1.271215301184977e-06,
+      "loss": 0.4773,
+      "step": 6244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3545482257587682,
+      "learning_rate": 1.2698766912481958e-06,
+      "loss": 0.491,
+      "step": 6245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3167855784532667,
+      "learning_rate": 1.2685386839456886e-06,
+      "loss": 0.5057,
+      "step": 6246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8376158472318225,
+      "learning_rate": 1.2672012794936195e-06,
+      "loss": 0.4646,
+      "step": 6247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4815672785751532,
+      "learning_rate": 1.2658644781080593e-06,
+      "loss": 0.4576,
+      "step": 6248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8421058516258766,
+      "learning_rate": 1.2645282800049812e-06,
+      "loss": 0.5177,
+      "step": 6249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8411185131379155,
+      "learning_rate": 1.2631926854002574e-06,
+      "loss": 0.5084,
+      "step": 6250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3816045998659126,
+      "learning_rate": 1.2618576945096671e-06,
+      "loss": 0.5492,
+      "step": 6251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3650484318670382,
+      "learning_rate": 1.2605233075488877e-06,
+      "loss": 0.4511,
+      "step": 6252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4297217792305226,
+      "learning_rate": 1.259189524733504e-06,
+      "loss": 0.5025,
+      "step": 6253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7866400137312053,
+      "learning_rate": 1.2578563462789967e-06,
+      "loss": 0.5479,
+      "step": 6254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.288055058437218,
+      "learning_rate": 1.2565237724007563e-06,
+      "loss": 0.5066,
+      "step": 6255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5601250969802751,
+      "learning_rate": 1.2551918033140687e-06,
+      "loss": 0.525,
+      "step": 6256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2428756455250791,
+      "learning_rate": 1.2538604392341264e-06,
+      "loss": 0.4523,
+      "step": 6257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.306311357857756,
+      "learning_rate": 1.2525296803760246e-06,
+      "loss": 0.4819,
+      "step": 6258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5382898567298926,
+      "learning_rate": 1.2511995269547566e-06,
+      "loss": 0.5187,
+      "step": 6259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.528938983892805,
+      "learning_rate": 1.2498699791852225e-06,
+      "loss": 0.4815,
+      "step": 6260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4201500893708132,
+      "learning_rate": 1.2485410372822205e-06,
+      "loss": 0.4682,
+      "step": 6261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4320467050540584,
+      "learning_rate": 1.247212701460455e-06,
+      "loss": 0.5452,
+      "step": 6262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5668199288766043,
+      "learning_rate": 1.2458849719345279e-06,
+      "loss": 0.515,
+      "step": 6263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.504159923138868,
+      "learning_rate": 1.244557848918948e-06,
+      "loss": 0.5079,
+      "step": 6264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3188449069712342,
+      "learning_rate": 1.2432313326281215e-06,
+      "loss": 0.4766,
+      "step": 6265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.44142880046114,
+      "learning_rate": 1.2419054232763595e-06,
+      "loss": 0.4971,
+      "step": 6266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5180567972225074,
+      "learning_rate": 1.2405801210778762e-06,
+      "loss": 0.4639,
+      "step": 6267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3471618262889407,
+      "learning_rate": 1.2392554262467833e-06,
+      "loss": 0.4822,
+      "step": 6268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7091520425341729,
+      "learning_rate": 1.2379313389970992e-06,
+      "loss": 0.5229,
+      "step": 6269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.570282113403896,
+      "learning_rate": 1.2366078595427389e-06,
+      "loss": 0.5271,
+      "step": 6270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3949975771134056,
+      "learning_rate": 1.235284988097526e-06,
+      "loss": 0.4658,
+      "step": 6271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.605511267248311,
+      "learning_rate": 1.2339627248751778e-06,
+      "loss": 0.5136,
+      "step": 6272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8807963700316133,
+      "learning_rate": 1.2326410700893198e-06,
+      "loss": 0.4902,
+      "step": 6273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.465254290911377,
+      "learning_rate": 1.231320023953479e-06,
+      "loss": 0.4896,
+      "step": 6274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.324978655566404,
+      "learning_rate": 1.2299995866810777e-06,
+      "loss": 0.4975,
+      "step": 6275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.428055739224206,
+      "learning_rate": 1.2286797584854482e-06,
+      "loss": 0.5418,
+      "step": 6276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4792979849123264,
+      "learning_rate": 1.2273605395798165e-06,
+      "loss": 0.5015,
+      "step": 6277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3062914972392994,
+      "learning_rate": 1.226041930177318e-06,
+      "loss": 0.5003,
+      "step": 6278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3227903716457101,
+      "learning_rate": 1.2247239304909825e-06,
+      "loss": 0.4546,
+      "step": 6279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5132406211568603,
+      "learning_rate": 1.223406540733747e-06,
+      "loss": 0.512,
+      "step": 6280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5284383420872536,
+      "learning_rate": 1.2220897611184429e-06,
+      "loss": 0.4794,
+      "step": 6281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6839892207048355,
+      "learning_rate": 1.220773591857814e-06,
+      "loss": 0.5089,
+      "step": 6282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.652141145541304,
+      "learning_rate": 1.2194580331644955e-06,
+      "loss": 0.4869,
+      "step": 6283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.323465813497087,
+      "learning_rate": 1.2181430852510268e-06,
+      "loss": 0.5281,
+      "step": 6284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5468279291555211,
+      "learning_rate": 1.2168287483298514e-06,
+      "loss": 0.4227,
+      "step": 6285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5243814666158966,
+      "learning_rate": 1.2155150226133094e-06,
+      "loss": 0.5128,
+      "step": 6286
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6629313928964435,
+      "learning_rate": 1.2142019083136475e-06,
+      "loss": 0.4846,
+      "step": 6287
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.453438332306153,
+      "learning_rate": 1.2128894056430084e-06,
+      "loss": 0.5031,
+      "step": 6288
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.169209631478575,
+      "learning_rate": 1.2115775148134402e-06,
+      "loss": 0.5078,
+      "step": 6289
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7260170535617319,
+      "learning_rate": 1.2102662360368893e-06,
+      "loss": 0.5449,
+      "step": 6290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8277114237596315,
+      "learning_rate": 1.2089555695252064e-06,
+      "loss": 0.513,
+      "step": 6291
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5156474325060718,
+      "learning_rate": 1.20764551549014e-06,
+      "loss": 0.5189,
+      "step": 6292
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.402604227104481,
+      "learning_rate": 1.2063360741433393e-06,
+      "loss": 0.5027,
+      "step": 6293
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.506607537413577,
+      "learning_rate": 1.205027245696359e-06,
+      "loss": 0.4435,
+      "step": 6294
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2969275416545543,
+      "learning_rate": 1.2037190303606489e-06,
+      "loss": 0.4377,
+      "step": 6295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7527021815831167,
+      "learning_rate": 1.2024114283475652e-06,
+      "loss": 0.5807,
+      "step": 6296
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3670555930535397,
+      "learning_rate": 1.2011044398683586e-06,
+      "loss": 0.4939,
+      "step": 6297
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2207355669928845,
+      "learning_rate": 1.1997980651341901e-06,
+      "loss": 0.4313,
+      "step": 6298
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6068755034967181,
+      "learning_rate": 1.1984923043561124e-06,
+      "loss": 0.4666,
+      "step": 6299
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3700754190571853,
+      "learning_rate": 1.197187157745085e-06,
+      "loss": 0.4371,
+      "step": 6300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5182100119914927,
+      "learning_rate": 1.1958826255119626e-06,
+      "loss": 0.4806,
+      "step": 6301
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6868763537558658,
+      "learning_rate": 1.1945787078675075e-06,
+      "loss": 0.4545,
+      "step": 6302
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2726361777698496,
+      "learning_rate": 1.1932754050223772e-06,
+      "loss": 0.437,
+      "step": 6303
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7997900027014602,
+      "learning_rate": 1.1919727171871303e-06,
+      "loss": 0.4812,
+      "step": 6304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7049836031903862,
+      "learning_rate": 1.1906706445722304e-06,
+      "loss": 0.5468,
+      "step": 6305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.482623305676932,
+      "learning_rate": 1.1893691873880348e-06,
+      "loss": 0.4679,
+      "step": 6306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7546459793761726,
+      "learning_rate": 1.1880683458448111e-06,
+      "loss": 0.5173,
+      "step": 6307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0376418787164963,
+      "learning_rate": 1.186768120152717e-06,
+      "loss": 0.4762,
+      "step": 6308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5897112053677294,
+      "learning_rate": 1.1854685105218183e-06,
+      "loss": 0.5106,
+      "step": 6309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.554638495563267,
+      "learning_rate": 1.184169517162076e-06,
+      "loss": 0.4453,
+      "step": 6310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4527954328226296,
+      "learning_rate": 1.1828711402833559e-06,
+      "loss": 0.4581,
+      "step": 6311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.324097258318579,
+      "learning_rate": 1.1815733800954221e-06,
+      "loss": 0.4468,
+      "step": 6312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6615006974970603,
+      "learning_rate": 1.1802762368079368e-06,
+      "loss": 0.488,
+      "step": 6313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.1942935989812935,
+      "learning_rate": 1.1789797106304663e-06,
+      "loss": 0.4155,
+      "step": 6314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6700048498627162,
+      "learning_rate": 1.1776838017724762e-06,
+      "loss": 0.4806,
+      "step": 6315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.426500155071954,
+      "learning_rate": 1.1763885104433331e-06,
+      "loss": 0.5418,
+      "step": 6316
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5091385876369006,
+      "learning_rate": 1.1750938368523e-06,
+      "loss": 0.5207,
+      "step": 6317
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3085262428892623,
+      "learning_rate": 1.1737997812085468e-06,
+      "loss": 0.4497,
+      "step": 6318
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4515944463730204,
+      "learning_rate": 1.1725063437211349e-06,
+      "loss": 0.4652,
+      "step": 6319
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2357108620539061,
+      "learning_rate": 1.1712135245990342e-06,
+      "loss": 0.465,
+      "step": 6320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.774785089411395,
+      "learning_rate": 1.1699213240511092e-06,
+      "loss": 0.4741,
+      "step": 6321
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.101966765503848,
+      "learning_rate": 1.1686297422861241e-06,
+      "loss": 0.4625,
+      "step": 6322
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.892565752711171,
+      "learning_rate": 1.1673387795127505e-06,
+      "loss": 0.467,
+      "step": 6323
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6317216482292651,
+      "learning_rate": 1.1660484359395514e-06,
+      "loss": 0.563,
+      "step": 6324
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.440182272936019,
+      "learning_rate": 1.1647587117749942e-06,
+      "loss": 0.5386,
+      "step": 6325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6785539736597066,
+      "learning_rate": 1.1634696072274444e-06,
+      "loss": 0.5317,
+      "step": 6326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5445312568704768,
+      "learning_rate": 1.1621811225051698e-06,
+      "loss": 0.5147,
+      "step": 6327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7406342594313629,
+      "learning_rate": 1.160893257816334e-06,
+      "loss": 0.4757,
+      "step": 6328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3839958819337508,
+      "learning_rate": 1.1596060133690057e-06,
+      "loss": 0.5226,
+      "step": 6329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3961518978650618,
+      "learning_rate": 1.1583193893711475e-06,
+      "loss": 0.5169,
+      "step": 6330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6338408284297716,
+      "learning_rate": 1.1570333860306265e-06,
+      "loss": 0.5181,
+      "step": 6331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9004636131185335,
+      "learning_rate": 1.1557480035552093e-06,
+      "loss": 0.5224,
+      "step": 6332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8167801921999878,
+      "learning_rate": 1.1544632421525576e-06,
+      "loss": 0.5135,
+      "step": 6333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5028563153474552,
+      "learning_rate": 1.1531791020302391e-06,
+      "loss": 0.4808,
+      "step": 6334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4995582428022929,
+      "learning_rate": 1.1518955833957153e-06,
+      "loss": 0.5058,
+      "step": 6335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.564743351446164,
+      "learning_rate": 1.1506126864563522e-06,
+      "loss": 0.5488,
+      "step": 6336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5702469015325229,
+      "learning_rate": 1.149330411419411e-06,
+      "loss": 0.5287,
+      "step": 6337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4508956069268264,
+      "learning_rate": 1.1480487584920553e-06,
+      "loss": 0.4943,
+      "step": 6338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4344574161100998,
+      "learning_rate": 1.146767727881349e-06,
+      "loss": 0.4971,
+      "step": 6339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7641317544981567,
+      "learning_rate": 1.1454873197942507e-06,
+      "loss": 0.5236,
+      "step": 6340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.693249372026301,
+      "learning_rate": 1.1442075344376253e-06,
+      "loss": 0.535,
+      "step": 6341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5097494619684362,
+      "learning_rate": 1.14292837201823e-06,
+      "loss": 0.5031,
+      "step": 6342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6040740006328453,
+      "learning_rate": 1.1416498327427278e-06,
+      "loss": 0.5009,
+      "step": 6343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5067558032158819,
+      "learning_rate": 1.1403719168176747e-06,
+      "loss": 0.4709,
+      "step": 6344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.2037937435097756,
+      "learning_rate": 1.1390946244495327e-06,
+      "loss": 0.5293,
+      "step": 6345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4940316941859748,
+      "learning_rate": 1.1378179558446561e-06,
+      "loss": 0.481,
+      "step": 6346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4547452020022014,
+      "learning_rate": 1.136541911209304e-06,
+      "loss": 0.5031,
+      "step": 6347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5577078497772556,
+      "learning_rate": 1.135266490749634e-06,
+      "loss": 0.5071,
+      "step": 6348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4071597547018102,
+      "learning_rate": 1.1339916946716984e-06,
+      "loss": 0.5052,
+      "step": 6349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.237185372105183,
+      "learning_rate": 1.1327175231814547e-06,
+      "loss": 0.474,
+      "step": 6350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5312058336458136,
+      "learning_rate": 1.131443976484754e-06,
+      "loss": 0.4986,
+      "step": 6351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.278808877803244,
+      "learning_rate": 1.1301710547873512e-06,
+      "loss": 0.4917,
+      "step": 6352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.471052130513398,
+      "learning_rate": 1.1288987582948956e-06,
+      "loss": 0.4279,
+      "step": 6353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9214340906088334,
+      "learning_rate": 1.1276270872129408e-06,
+      "loss": 0.4989,
+      "step": 6354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6966020038555492,
+      "learning_rate": 1.1263560417469332e-06,
+      "loss": 0.4959,
+      "step": 6355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9936299392743482,
+      "learning_rate": 1.1250856221022233e-06,
+      "loss": 0.4889,
+      "step": 6356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2150697051618198,
+      "learning_rate": 1.1238158284840594e-06,
+      "loss": 0.4787,
+      "step": 6357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.376086418479404,
+      "learning_rate": 1.1225466610975854e-06,
+      "loss": 0.4775,
+      "step": 6358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6770030416826525,
+      "learning_rate": 1.1212781201478496e-06,
+      "loss": 0.4983,
+      "step": 6359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3716829551919227,
+      "learning_rate": 1.1200102058397927e-06,
+      "loss": 0.5119,
+      "step": 6360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7293039440124984,
+      "learning_rate": 1.11874291837826e-06,
+      "loss": 0.4778,
+      "step": 6361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2486910175729624,
+      "learning_rate": 1.117476257967991e-06,
+      "loss": 0.4769,
+      "step": 6362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9053404412571056,
+      "learning_rate": 1.1162102248136264e-06,
+      "loss": 0.5044,
+      "step": 6363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6287425902785133,
+      "learning_rate": 1.114944819119707e-06,
+      "loss": 0.4668,
+      "step": 6364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7416565194171552,
+      "learning_rate": 1.1136800410906672e-06,
+      "loss": 0.4914,
+      "step": 6365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.085037618074328,
+      "learning_rate": 1.1124158909308458e-06,
+      "loss": 0.4845,
+      "step": 6366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7980273651122092,
+      "learning_rate": 1.1111523688444741e-06,
+      "loss": 0.4781,
+      "step": 6367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.883979465274335,
+      "learning_rate": 1.1098894750356893e-06,
+      "loss": 0.5242,
+      "step": 6368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.313537169348473,
+      "learning_rate": 1.1086272097085187e-06,
+      "loss": 0.4718,
+      "step": 6369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9156618835053345,
+      "learning_rate": 1.1073655730668965e-06,
+      "loss": 0.5015,
+      "step": 6370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.284402050371876,
+      "learning_rate": 1.1061045653146463e-06,
+      "loss": 0.4395,
+      "step": 6371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4106454134612665,
+      "learning_rate": 1.1048441866555004e-06,
+      "loss": 0.4857,
+      "step": 6372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.622936999846198,
+      "learning_rate": 1.103584437293082e-06,
+      "loss": 0.4752,
+      "step": 6373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5432383878844,
+      "learning_rate": 1.1023253174309128e-06,
+      "loss": 0.4941,
+      "step": 6374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4826265580141185,
+      "learning_rate": 1.1010668272724178e-06,
+      "loss": 0.5047,
+      "step": 6375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4461541616248892,
+      "learning_rate": 1.099808967020914e-06,
+      "loss": 0.5171,
+      "step": 6376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6063058861931088,
+      "learning_rate": 1.0985517368796227e-06,
+      "loss": 0.5259,
+      "step": 6377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6309292102124595,
+      "learning_rate": 1.0972951370516577e-06,
+      "loss": 0.565,
+      "step": 6378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.349710798461493,
+      "learning_rate": 1.0960391677400373e-06,
+      "loss": 0.4981,
+      "step": 6379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4939126100808078,
+      "learning_rate": 1.0947838291476687e-06,
+      "loss": 0.5147,
+      "step": 6380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.443005036475542,
+      "learning_rate": 1.0935291214773703e-06,
+      "loss": 0.4891,
+      "step": 6381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5200846600311144,
+      "learning_rate": 1.0922750449318464e-06,
+      "loss": 0.4843,
+      "step": 6382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9827824644081335,
+      "learning_rate": 1.0910215997137048e-06,
+      "loss": 0.4668,
+      "step": 6383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.380065844859165,
+      "learning_rate": 1.0897687860254514e-06,
+      "loss": 0.4843,
+      "step": 6384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6967284179344584,
+      "learning_rate": 1.0885166040694884e-06,
+      "loss": 0.5127,
+      "step": 6385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6695641296913672,
+      "learning_rate": 1.0872650540481178e-06,
+      "loss": 0.4706,
+      "step": 6386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5456696537393462,
+      "learning_rate": 1.0860141361635374e-06,
+      "loss": 0.4878,
+      "step": 6387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2956173483980842,
+      "learning_rate": 1.0847638506178444e-06,
+      "loss": 0.5433,
+      "step": 6388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3262687673658258,
+      "learning_rate": 1.0835141976130331e-06,
+      "loss": 0.4642,
+      "step": 6389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.64537801718758,
+      "learning_rate": 1.0822651773509984e-06,
+      "loss": 0.5271,
+      "step": 6390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6997776708532146,
+      "learning_rate": 1.0810167900335283e-06,
+      "loss": 0.4831,
+      "step": 6391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6522170913115182,
+      "learning_rate": 1.0797690358623103e-06,
+      "loss": 0.4778,
+      "step": 6392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3692356724278743,
+      "learning_rate": 1.078521915038932e-06,
+      "loss": 0.4815,
+      "step": 6393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3408947630589219,
+      "learning_rate": 1.0772754277648739e-06,
+      "loss": 0.4349,
+      "step": 6394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8953971727722827,
+      "learning_rate": 1.0760295742415205e-06,
+      "loss": 0.4876,
+      "step": 6395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.53510002475653,
+      "learning_rate": 1.0747843546701452e-06,
+      "loss": 0.4715,
+      "step": 6396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6980999378389511,
+      "learning_rate": 1.0735397692519312e-06,
+      "loss": 0.5138,
+      "step": 6397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.558202819615058,
+      "learning_rate": 1.0722958181879466e-06,
+      "loss": 0.4924,
+      "step": 6398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7068775428592562,
+      "learning_rate": 1.0710525016791667e-06,
+      "loss": 0.5051,
+      "step": 6399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6023047898555305,
+      "learning_rate": 1.0698098199264578e-06,
+      "loss": 0.5152,
+      "step": 6400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4369425786969896,
+      "learning_rate": 1.0685677731305855e-06,
+      "loss": 0.4318,
+      "step": 6401
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5998174512047107,
+      "learning_rate": 1.0673263614922152e-06,
+      "loss": 0.558,
+      "step": 6402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7877368017433597,
+      "learning_rate": 1.0660855852119062e-06,
+      "loss": 0.4984,
+      "step": 6403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9796036482449146,
+      "learning_rate": 1.0648454444901179e-06,
+      "loss": 0.4578,
+      "step": 6404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.882827935795179,
+      "learning_rate": 1.0636059395272053e-06,
+      "loss": 0.4866,
+      "step": 6405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.42440315165274,
+      "learning_rate": 1.0623670705234235e-06,
+      "loss": 0.515,
+      "step": 6406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3774620416132493,
+      "learning_rate": 1.0611288376789198e-06,
+      "loss": 0.4862,
+      "step": 6407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3592757543233827,
+      "learning_rate": 1.059891241193744e-06,
+      "loss": 0.5142,
+      "step": 6408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4592048124895594,
+      "learning_rate": 1.058654281267838e-06,
+      "loss": 0.5171,
+      "step": 6409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4238846191505734,
+      "learning_rate": 1.057417958101047e-06,
+      "loss": 0.5033,
+      "step": 6410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4464240773430705,
+      "learning_rate": 1.0561822718931074e-06,
+      "loss": 0.5016,
+      "step": 6411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6974068949415977,
+      "learning_rate": 1.0549472228436535e-06,
+      "loss": 0.4979,
+      "step": 6412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3602079975816623,
+      "learning_rate": 1.0537128111522232e-06,
+      "loss": 0.4518,
+      "step": 6413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6865326422197805,
+      "learning_rate": 1.0524790370182431e-06,
+      "loss": 0.4919,
+      "step": 6414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3161318499696908,
+      "learning_rate": 1.0512459006410424e-06,
+      "loss": 0.4716,
+      "step": 6415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3809273299426368,
+      "learning_rate": 1.0500134022198421e-06,
+      "loss": 0.4707,
+      "step": 6416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.358364784189713,
+      "learning_rate": 1.0487815419537672e-06,
+      "loss": 0.4555,
+      "step": 6417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2188671693165967,
+      "learning_rate": 1.047550320041832e-06,
+      "loss": 0.4906,
+      "step": 6418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8062966823296596,
+      "learning_rate": 1.0463197366829536e-06,
+      "loss": 0.4798,
+      "step": 6419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5575137348150088,
+      "learning_rate": 1.0450897920759422e-06,
+      "loss": 0.515,
+      "step": 6420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4850541577111969,
+      "learning_rate": 1.043860486419505e-06,
+      "loss": 0.4679,
+      "step": 6421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4781014051994306,
+      "learning_rate": 1.042631819912251e-06,
+      "loss": 0.4556,
+      "step": 6422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3726640581035268,
+      "learning_rate": 1.0414037927526782e-06,
+      "loss": 0.4647,
+      "step": 6423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.714250634087497,
+      "learning_rate": 1.0401764051391888e-06,
+      "loss": 0.4789,
+      "step": 6424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4200836700258241,
+      "learning_rate": 1.0389496572700747e-06,
+      "loss": 0.4973,
+      "step": 6425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3985361590853174,
+      "learning_rate": 1.0377235493435301e-06,
+      "loss": 0.4872,
+      "step": 6426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.414545373279879,
+      "learning_rate": 1.0364980815576419e-06,
+      "loss": 0.4704,
+      "step": 6427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3847624550964073,
+      "learning_rate": 1.0352732541103972e-06,
+      "loss": 0.5241,
+      "step": 6428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5204777902527213,
+      "learning_rate": 1.034049067199675e-06,
+      "loss": 0.4473,
+      "step": 6429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3824734739922342,
+      "learning_rate": 1.032825521023255e-06,
+      "loss": 0.4806,
+      "step": 6430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.35143148368396,
+      "learning_rate": 1.031602615778814e-06,
+      "loss": 0.4947,
+      "step": 6431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3017480279577303,
+      "learning_rate": 1.0303803516639187e-06,
+      "loss": 0.5221,
+      "step": 6432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.314222059951463,
+      "learning_rate": 1.0291587288760403e-06,
+      "loss": 0.5007,
+      "step": 6433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7591906723145145,
+      "learning_rate": 1.02793774761254e-06,
+      "loss": 0.4792,
+      "step": 6434
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.495983252455482,
+      "learning_rate": 1.026717408070681e-06,
+      "loss": 0.4376,
+      "step": 6435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.862959961522283,
+      "learning_rate": 1.0254977104476166e-06,
+      "loss": 0.4925,
+      "step": 6436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4445140615487064,
+      "learning_rate": 1.0242786549404015e-06,
+      "loss": 0.515,
+      "step": 6437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5709457175445063,
+      "learning_rate": 1.0230602417459862e-06,
+      "loss": 0.5253,
+      "step": 6438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4846376046602925,
+      "learning_rate": 1.0218424710612135e-06,
+      "loss": 0.539,
+      "step": 6439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.805011782517603,
+      "learning_rate": 1.0206253430828266e-06,
+      "loss": 0.4773,
+      "step": 6440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4557146952088784,
+      "learning_rate": 1.0194088580074618e-06,
+      "loss": 0.465,
+      "step": 6441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.380442482912851,
+      "learning_rate": 1.0181930160316554e-06,
+      "loss": 0.5045,
+      "step": 6442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.321381721792749,
+      "learning_rate": 1.0169778173518347e-06,
+      "loss": 0.4964,
+      "step": 6443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.081757579088301,
+      "learning_rate": 1.0157632621643282e-06,
+      "loss": 0.514,
+      "step": 6444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6201478985872608,
+      "learning_rate": 1.0145493506653548e-06,
+      "loss": 0.5069,
+      "step": 6445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6145747097101975,
+      "learning_rate": 1.0133360830510352e-06,
+      "loss": 0.4738,
+      "step": 6446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4517395862733522,
+      "learning_rate": 1.0121234595173835e-06,
+      "loss": 0.5372,
+      "step": 6447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3009646288536616,
+      "learning_rate": 1.0109114802603081e-06,
+      "loss": 0.4628,
+      "step": 6448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5314688288647906,
+      "learning_rate": 1.0097001454756173e-06,
+      "loss": 0.5247,
+      "step": 6449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3562688096522537,
+      "learning_rate": 1.0084894553590098e-06,
+      "loss": 0.4466,
+      "step": 6450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3705066231074705,
+      "learning_rate": 1.0072794101060861e-06,
+      "loss": 0.4579,
+      "step": 6451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5832753927850254,
+      "learning_rate": 1.0060700099123372e-06,
+      "loss": 0.4806,
+      "step": 6452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8103783315737578,
+      "learning_rate": 1.004861254973154e-06,
+      "loss": 0.4862,
+      "step": 6453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.247061170385736,
+      "learning_rate": 1.0036531454838216e-06,
+      "loss": 0.4913,
+      "step": 6454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.405826263478046,
+      "learning_rate": 1.0024456816395195e-06,
+      "loss": 0.511,
+      "step": 6455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.800506386193242,
+      "learning_rate": 1.001238863635326e-06,
+      "loss": 0.4625,
+      "step": 6456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4412549678615536,
+      "learning_rate": 1.0000326916662112e-06,
+      "loss": 0.519,
+      "step": 6457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3940917749864985,
+      "learning_rate": 9.988271659270444e-07,
+      "loss": 0.4812,
+      "step": 6458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.331626678614534,
+      "learning_rate": 9.976222866125867e-07,
+      "loss": 0.4947,
+      "step": 6459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7731658384307778,
+      "learning_rate": 9.964180539175e-07,
+      "loss": 0.5341,
+      "step": 6460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6299854711262632,
+      "learning_rate": 9.952144680363358e-07,
+      "loss": 0.5018,
+      "step": 6461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7760926328349766,
+      "learning_rate": 9.940115291635448e-07,
+      "loss": 0.4784,
+      "step": 6462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8962216803491019,
+      "learning_rate": 9.92809237493475e-07,
+      "loss": 0.47,
+      "step": 6463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4401955974492697,
+      "learning_rate": 9.916075932203633e-07,
+      "loss": 0.4428,
+      "step": 6464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4001781733936711,
+      "learning_rate": 9.904065965383492e-07,
+      "loss": 0.5315,
+      "step": 6465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7370364220018046,
+      "learning_rate": 9.89206247641461e-07,
+      "loss": 0.566,
+      "step": 6466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3805628292137204,
+      "learning_rate": 9.880065467236289e-07,
+      "loss": 0.4446,
+      "step": 6467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4475562347146755,
+      "learning_rate": 9.868074939786727e-07,
+      "loss": 0.5286,
+      "step": 6468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.884552940235352,
+      "learning_rate": 9.856090896003117e-07,
+      "loss": 0.5077,
+      "step": 6469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2709276223524975,
+      "learning_rate": 9.84411333782157e-07,
+      "loss": 0.4425,
+      "step": 6470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7191717752920916,
+      "learning_rate": 9.83214226717717e-07,
+      "loss": 0.5699,
+      "step": 6471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.426700264224924,
+      "learning_rate": 9.820177686003972e-07,
+      "loss": 0.505,
+      "step": 6472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6161588921373857,
+      "learning_rate": 9.808219596234924e-07,
+      "loss": 0.5044,
+      "step": 6473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.565981060733389,
+      "learning_rate": 9.796267999801995e-07,
+      "loss": 0.5106,
+      "step": 6474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5211022442715696,
+      "learning_rate": 9.784322898636034e-07,
+      "loss": 0.5166,
+      "step": 6475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4123861555634079,
+      "learning_rate": 9.772384294666915e-07,
+      "loss": 0.5146,
+      "step": 6476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.796078323450227,
+      "learning_rate": 9.760452189823394e-07,
+      "loss": 0.5041,
+      "step": 6477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.655518718867122,
+      "learning_rate": 9.748526586033219e-07,
+      "loss": 0.465,
+      "step": 6478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.321255408001864,
+      "learning_rate": 9.736607485223076e-07,
+      "loss": 0.4534,
+      "step": 6479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5973847957315321,
+      "learning_rate": 9.724694889318615e-07,
+      "loss": 0.469,
+      "step": 6480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7206749394132033,
+      "learning_rate": 9.71278880024441e-07,
+      "loss": 0.4653,
+      "step": 6481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.456188547690438,
+      "learning_rate": 9.700889219923975e-07,
+      "loss": 0.4847,
+      "step": 6482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.361231917322826,
+      "learning_rate": 9.688996150279827e-07,
+      "loss": 0.4994,
+      "step": 6483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0411586611106918,
+      "learning_rate": 9.677109593233363e-07,
+      "loss": 0.4838,
+      "step": 6484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6873790949681748,
+      "learning_rate": 9.665229550704991e-07,
+      "loss": 0.5037,
+      "step": 6485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6506240442314934,
+      "learning_rate": 9.653356024613997e-07,
+      "loss": 0.4929,
+      "step": 6486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6794439760619745,
+      "learning_rate": 9.641489016878702e-07,
+      "loss": 0.5285,
+      "step": 6487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4150040353312494,
+      "learning_rate": 9.62962852941629e-07,
+      "loss": 0.5077,
+      "step": 6488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.390843234714611,
+      "learning_rate": 9.61777456414295e-07,
+      "loss": 0.4565,
+      "step": 6489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2235657869784267,
+      "learning_rate": 9.60592712297379e-07,
+      "loss": 0.4837,
+      "step": 6490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4018516966954073,
+      "learning_rate": 9.594086207822845e-07,
+      "loss": 0.4996,
+      "step": 6491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8108744655621944,
+      "learning_rate": 9.58225182060315e-07,
+      "loss": 0.4742,
+      "step": 6492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6366771711601844,
+      "learning_rate": 9.570423963226632e-07,
+      "loss": 0.4805,
+      "step": 6493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6650749303310582,
+      "learning_rate": 9.558602637604192e-07,
+      "loss": 0.5174,
+      "step": 6494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.208185061384778,
+      "learning_rate": 9.546787845645672e-07,
+      "loss": 0.554,
+      "step": 6495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4893339674896775,
+      "learning_rate": 9.534979589259869e-07,
+      "loss": 0.4519,
+      "step": 6496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6165958466542163,
+      "learning_rate": 9.523177870354483e-07,
+      "loss": 0.4916,
+      "step": 6497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7710197037064592,
+      "learning_rate": 9.511382690836213e-07,
+      "loss": 0.5109,
+      "step": 6498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3793051055716132,
+      "learning_rate": 9.499594052610661e-07,
+      "loss": 0.4939,
+      "step": 6499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0578189131865265,
+      "learning_rate": 9.487811957582376e-07,
+      "loss": 0.4535,
+      "step": 6500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5081857196219226,
+      "learning_rate": 9.476036407654881e-07,
+      "loss": 0.4491,
+      "step": 6501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.456257458370949,
+      "learning_rate": 9.464267404730593e-07,
+      "loss": 0.4809,
+      "step": 6502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.474834198528999,
+      "learning_rate": 9.452504950710911e-07,
+      "loss": 0.4986,
+      "step": 6503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7138505129091549,
+      "learning_rate": 9.440749047496168e-07,
+      "loss": 0.5048,
+      "step": 6504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4093392810821976,
+      "learning_rate": 9.428999696985642e-07,
+      "loss": 0.4755,
+      "step": 6505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6833102214769327,
+      "learning_rate": 9.417256901077515e-07,
+      "loss": 0.487,
+      "step": 6506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3875066026251863,
+      "learning_rate": 9.405520661668977e-07,
+      "loss": 0.5175,
+      "step": 6507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4386284006670491,
+      "learning_rate": 9.393790980656081e-07,
+      "loss": 0.4967,
+      "step": 6508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.477341431126602,
+      "learning_rate": 9.382067859933897e-07,
+      "loss": 0.4585,
+      "step": 6509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3587078333246,
+      "learning_rate": 9.37035130139638e-07,
+      "loss": 0.4865,
+      "step": 6510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4243793950309338,
+      "learning_rate": 9.358641306936416e-07,
+      "loss": 0.513,
+      "step": 6511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4195418236296353,
+      "learning_rate": 9.346937878445916e-07,
+      "loss": 0.4805,
+      "step": 6512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6071698001178198,
+      "learning_rate": 9.335241017815627e-07,
+      "loss": 0.5024,
+      "step": 6513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5025394272531454,
+      "learning_rate": 9.323550726935304e-07,
+      "loss": 0.562,
+      "step": 6514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.531377625714681,
+      "learning_rate": 9.311867007693598e-07,
+      "loss": 0.4915,
+      "step": 6515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7240203212532423,
+      "learning_rate": 9.300189861978143e-07,
+      "loss": 0.4966,
+      "step": 6516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6216464797284622,
+      "learning_rate": 9.288519291675451e-07,
+      "loss": 0.4925,
+      "step": 6517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5411812159660658,
+      "learning_rate": 9.276855298671033e-07,
+      "loss": 0.5281,
+      "step": 6518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.580795284714528,
+      "learning_rate": 9.265197884849286e-07,
+      "loss": 0.4922,
+      "step": 6519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5446521341749373,
+      "learning_rate": 9.253547052093587e-07,
+      "loss": 0.4868,
+      "step": 6520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3983084243758115,
+      "learning_rate": 9.241902802286229e-07,
+      "loss": 0.4888,
+      "step": 6521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.308543603312496,
+      "learning_rate": 9.23026513730843e-07,
+      "loss": 0.4892,
+      "step": 6522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7423299968073849,
+      "learning_rate": 9.218634059040377e-07,
+      "loss": 0.493,
+      "step": 6523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3112906377656586,
+      "learning_rate": 9.207009569361153e-07,
+      "loss": 0.5326,
+      "step": 6524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4610564692928665,
+      "learning_rate": 9.195391670148812e-07,
+      "loss": 0.4936,
+      "step": 6525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.890123244074039,
+      "learning_rate": 9.183780363280314e-07,
+      "loss": 0.4905,
+      "step": 6526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3160123399401755,
+      "learning_rate": 9.172175650631576e-07,
+      "loss": 0.4919,
+      "step": 6527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6255882144586874,
+      "learning_rate": 9.160577534077453e-07,
+      "loss": 0.445,
+      "step": 6528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5051745110594048,
+      "learning_rate": 9.148986015491695e-07,
+      "loss": 0.4883,
+      "step": 6529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6865982859562735,
+      "learning_rate": 9.137401096747045e-07,
+      "loss": 0.501,
+      "step": 6530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9954453456068366,
+      "learning_rate": 9.125822779715121e-07,
+      "loss": 0.4812,
+      "step": 6531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.683593194230567,
+      "learning_rate": 9.114251066266527e-07,
+      "loss": 0.5226,
+      "step": 6532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.165224842620645,
+      "learning_rate": 9.102685958270746e-07,
+      "loss": 0.5188,
+      "step": 6533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8715140158538006,
+      "learning_rate": 9.09112745759626e-07,
+      "loss": 0.4523,
+      "step": 6534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.578535773480896,
+      "learning_rate": 9.07957556611041e-07,
+      "loss": 0.4604,
+      "step": 6535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1796850042322746,
+      "learning_rate": 9.068030285679519e-07,
+      "loss": 0.4398,
+      "step": 6536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7277912894351544,
+      "learning_rate": 9.05649161816885e-07,
+      "loss": 0.4712,
+      "step": 6537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6811486351535663,
+      "learning_rate": 9.044959565442541e-07,
+      "loss": 0.5187,
+      "step": 6538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3938299493600599,
+      "learning_rate": 9.033434129363727e-07,
+      "loss": 0.5106,
+      "step": 6539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.865073353995794,
+      "learning_rate": 9.021915311794421e-07,
+      "loss": 0.4647,
+      "step": 6540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3106511091520945,
+      "learning_rate": 9.010403114595612e-07,
+      "loss": 0.4894,
+      "step": 6541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6391450455268382,
+      "learning_rate": 8.998897539627172e-07,
+      "loss": 0.4969,
+      "step": 6542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4456021413174263,
+      "learning_rate": 8.987398588747948e-07,
+      "loss": 0.4843,
+      "step": 6543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5590898039625034,
+      "learning_rate": 8.975906263815681e-07,
+      "loss": 0.5073,
+      "step": 6544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.665956667537492,
+      "learning_rate": 8.96442056668706e-07,
+      "loss": 0.5101,
+      "step": 6545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8921706826375544,
+      "learning_rate": 8.952941499217722e-07,
+      "loss": 0.4772,
+      "step": 6546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2156733101385413,
+      "learning_rate": 8.941469063262182e-07,
+      "loss": 0.4468,
+      "step": 6547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5766155722354083,
+      "learning_rate": 8.930003260673936e-07,
+      "loss": 0.5071,
+      "step": 6548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.545352202694005,
+      "learning_rate": 8.91854409330537e-07,
+      "loss": 0.4885,
+      "step": 6549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.31182920256894,
+      "learning_rate": 8.907091563007824e-07,
+      "loss": 0.4635,
+      "step": 6550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6185349718289779,
+      "learning_rate": 8.895645671631542e-07,
+      "loss": 0.5057,
+      "step": 6551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.540854118339132,
+      "learning_rate": 8.884206421025715e-07,
+      "loss": 0.4671,
+      "step": 6552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4774845635395661,
+      "learning_rate": 8.872773813038466e-07,
+      "loss": 0.5321,
+      "step": 6553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5020064700053972,
+      "learning_rate": 8.861347849516816e-07,
+      "loss": 0.5313,
+      "step": 6554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7142941992414023,
+      "learning_rate": 8.849928532306745e-07,
+      "loss": 0.5107,
+      "step": 6555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.91307148790412,
+      "learning_rate": 8.838515863253128e-07,
+      "loss": 0.5381,
+      "step": 6556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6351845953656223,
+      "learning_rate": 8.827109844199805e-07,
+      "loss": 0.4757,
+      "step": 6557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6147448240438045,
+      "learning_rate": 8.815710476989487e-07,
+      "loss": 0.5138,
+      "step": 6558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1570591241507078,
+      "learning_rate": 8.804317763463877e-07,
+      "loss": 0.4723,
+      "step": 6559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0196638013532677,
+      "learning_rate": 8.792931705463542e-07,
+      "loss": 0.5151,
+      "step": 6560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.358548606885228,
+      "learning_rate": 8.781552304828006e-07,
+      "loss": 0.4373,
+      "step": 6561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5555876320839892,
+      "learning_rate": 8.770179563395725e-07,
+      "loss": 0.5272,
+      "step": 6562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.873411723591231,
+      "learning_rate": 8.758813483004047e-07,
+      "loss": 0.558,
+      "step": 6563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5335119418462217,
+      "learning_rate": 8.74745406548928e-07,
+      "loss": 0.4761,
+      "step": 6564
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4249307697946578,
+      "learning_rate": 8.736101312686618e-07,
+      "loss": 0.525,
+      "step": 6565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7083767116323012,
+      "learning_rate": 8.72475522643022e-07,
+      "loss": 0.4557,
+      "step": 6566
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5754117600788526,
+      "learning_rate": 8.713415808553121e-07,
+      "loss": 0.4816,
+      "step": 6567
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6529075246510267,
+      "learning_rate": 8.702083060887323e-07,
+      "loss": 0.4781,
+      "step": 6568
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5804308238440943,
+      "learning_rate": 8.690756985263737e-07,
+      "loss": 0.4972,
+      "step": 6569
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7295768902466224,
+      "learning_rate": 8.679437583512168e-07,
+      "loss": 0.5174,
+      "step": 6570
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6384249650212698,
+      "learning_rate": 8.668124857461385e-07,
+      "loss": 0.5158,
+      "step": 6571
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3816051165201575,
+      "learning_rate": 8.656818808939038e-07,
+      "loss": 0.4673,
+      "step": 6572
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4216938442872074,
+      "learning_rate": 8.645519439771744e-07,
+      "loss": 0.4542,
+      "step": 6573
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4833293871723272,
+      "learning_rate": 8.634226751784991e-07,
+      "loss": 0.5507,
+      "step": 6574
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.606769771979235,
+      "learning_rate": 8.622940746803238e-07,
+      "loss": 0.5216,
+      "step": 6575
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.348401256388603,
+      "learning_rate": 8.611661426649809e-07,
+      "loss": 0.4688,
+      "step": 6576
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6077809246600447,
+      "learning_rate": 8.600388793146991e-07,
+      "loss": 0.5163,
+      "step": 6577
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4424136275117476,
+      "learning_rate": 8.589122848115977e-07,
+      "loss": 0.5012,
+      "step": 6578
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6050305815590833,
+      "learning_rate": 8.5778635933769e-07,
+      "loss": 0.4884,
+      "step": 6579
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4093501853324268,
+      "learning_rate": 8.566611030748767e-07,
+      "loss": 0.4972,
+      "step": 6580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4112177810139739,
+      "learning_rate": 8.555365162049529e-07,
+      "loss": 0.5028,
+      "step": 6581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0270135079817404,
+      "learning_rate": 8.544125989096063e-07,
+      "loss": 0.5033,
+      "step": 6582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2998776012858089,
+      "learning_rate": 8.532893513704149e-07,
+      "loss": 0.4833,
+      "step": 6583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.483841216926302,
+      "learning_rate": 8.52166773768851e-07,
+      "loss": 0.472,
+      "step": 6584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3508207620418844,
+      "learning_rate": 8.51044866286273e-07,
+      "loss": 0.46,
+      "step": 6585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.653697562371862,
+      "learning_rate": 8.499236291039403e-07,
+      "loss": 0.5096,
+      "step": 6586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5744504468038782,
+      "learning_rate": 8.488030624029947e-07,
+      "loss": 0.5106,
+      "step": 6587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.463803355129119,
+      "learning_rate": 8.476831663644758e-07,
+      "loss": 0.4511,
+      "step": 6588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6581237351175009,
+      "learning_rate": 8.465639411693116e-07,
+      "loss": 0.5617,
+      "step": 6589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5768383400051669,
+      "learning_rate": 8.45445386998322e-07,
+      "loss": 0.5398,
+      "step": 6590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6111736629103783,
+      "learning_rate": 8.443275040322213e-07,
+      "loss": 0.4301,
+      "step": 6591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5076339706999673,
+      "learning_rate": 8.432102924516112e-07,
+      "loss": 0.4785,
+      "step": 6592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8147989208223385,
+      "learning_rate": 8.420937524369876e-07,
+      "loss": 0.5381,
+      "step": 6593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5861599435098752,
+      "learning_rate": 8.409778841687383e-07,
+      "loss": 0.4792,
+      "step": 6594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7391419145258895,
+      "learning_rate": 8.398626878271421e-07,
+      "loss": 0.5066,
+      "step": 6595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3562454526494723,
+      "learning_rate": 8.387481635923667e-07,
+      "loss": 0.4684,
+      "step": 6596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.680088091115047,
+      "learning_rate": 8.376343116444763e-07,
+      "loss": 0.5065,
+      "step": 6597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6406248070930465,
+      "learning_rate": 8.365211321634209e-07,
+      "loss": 0.5093,
+      "step": 6598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4648655817523175,
+      "learning_rate": 8.354086253290445e-07,
+      "loss": 0.4623,
+      "step": 6599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7348006256424257,
+      "learning_rate": 8.342967913210843e-07,
+      "loss": 0.5108,
+      "step": 6600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6226349665887883,
+      "learning_rate": 8.331856303191632e-07,
+      "loss": 0.4747,
+      "step": 6601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3900268223214782,
+      "learning_rate": 8.320751425028039e-07,
+      "loss": 0.5242,
+      "step": 6602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5022996561742294,
+      "learning_rate": 8.309653280514118e-07,
+      "loss": 0.4604,
+      "step": 6603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6514066294972094,
+      "learning_rate": 8.2985618714429e-07,
+      "loss": 0.4741,
+      "step": 6604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3691633878283713,
+      "learning_rate": 8.287477199606276e-07,
+      "loss": 0.4871,
+      "step": 6605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5466999097901013,
+      "learning_rate": 8.276399266795088e-07,
+      "loss": 0.5516,
+      "step": 6606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4721951482256834,
+      "learning_rate": 8.265328074799067e-07,
+      "loss": 0.5031,
+      "step": 6607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6451556978401359,
+      "learning_rate": 8.254263625406856e-07,
+      "loss": 0.5001,
+      "step": 6608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4413842515562194,
+      "learning_rate": 8.243205920406016e-07,
+      "loss": 0.5116,
+      "step": 6609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.287070423719794,
+      "learning_rate": 8.232154961583017e-07,
+      "loss": 0.4357,
+      "step": 6610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.578682068297509,
+      "learning_rate": 8.22111075072326e-07,
+      "loss": 0.4841,
+      "step": 6611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4041042897849592,
+      "learning_rate": 8.210073289611004e-07,
+      "loss": 0.486,
+      "step": 6612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4408155137435312,
+      "learning_rate": 8.19904258002947e-07,
+      "loss": 0.4741,
+      "step": 6613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8826684892795102,
+      "learning_rate": 8.188018623760747e-07,
+      "loss": 0.4876,
+      "step": 6614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4705849420242674,
+      "learning_rate": 8.177001422585873e-07,
+      "loss": 0.4982,
+      "step": 6615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6166224577611759,
+      "learning_rate": 8.165990978284749e-07,
+      "loss": 0.483,
+      "step": 6616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.659097390392933,
+      "learning_rate": 8.154987292636241e-07,
+      "loss": 0.471,
+      "step": 6617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5422971791203492,
+      "learning_rate": 8.143990367418059e-07,
+      "loss": 0.5401,
+      "step": 6618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.252984211458063,
+      "learning_rate": 8.133000204406871e-07,
+      "loss": 0.4784,
+      "step": 6619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4384644065425785,
+      "learning_rate": 8.122016805378241e-07,
+      "loss": 0.5327,
+      "step": 6620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8644675432583973,
+      "learning_rate": 8.11104017210661e-07,
+      "loss": 0.5353,
+      "step": 6621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6273030259451302,
+      "learning_rate": 8.100070306365382e-07,
+      "loss": 0.4845,
+      "step": 6622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5151125444202667,
+      "learning_rate": 8.0891072099268e-07,
+      "loss": 0.4817,
+      "step": 6623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6395495200122552,
+      "learning_rate": 8.07815088456208e-07,
+      "loss": 0.4903,
+      "step": 6624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8833985121582997,
+      "learning_rate": 8.067201332041286e-07,
+      "loss": 0.5114,
+      "step": 6625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2573482842800296,
+      "learning_rate": 8.056258554133433e-07,
+      "loss": 0.4473,
+      "step": 6626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.841027142125862,
+      "learning_rate": 8.045322552606427e-07,
+      "loss": 0.491,
+      "step": 6627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4684161021406357,
+      "learning_rate": 8.034393329227052e-07,
+      "loss": 0.5239,
+      "step": 6628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4732512569164948,
+      "learning_rate": 8.023470885761053e-07,
+      "loss": 0.4898,
+      "step": 6629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2874992148845803,
+      "learning_rate": 8.012555223973012e-07,
+      "loss": 0.4937,
+      "step": 6630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3111270367689432,
+      "learning_rate": 8.001646345626479e-07,
+      "loss": 0.4869,
+      "step": 6631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5229838431069287,
+      "learning_rate": 7.990744252483856e-07,
+      "loss": 0.512,
+      "step": 6632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2912528249724036,
+      "learning_rate": 7.979848946306496e-07,
+      "loss": 0.4723,
+      "step": 6633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5275938148321992,
+      "learning_rate": 7.968960428854605e-07,
+      "loss": 0.5204,
+      "step": 6634
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5890097981606934,
+      "learning_rate": 7.958078701887334e-07,
+      "loss": 0.4763,
+      "step": 6635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2712850160856939,
+      "learning_rate": 7.947203767162737e-07,
+      "loss": 0.4592,
+      "step": 6636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6645130002721502,
+      "learning_rate": 7.936335626437725e-07,
+      "loss": 0.5086,
+      "step": 6637
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5569392743648256,
+      "learning_rate": 7.925474281468165e-07,
+      "loss": 0.479,
+      "step": 6638
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.492305078322501,
+      "learning_rate": 7.914619734008782e-07,
+      "loss": 0.512,
+      "step": 6639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4290671588367234,
+      "learning_rate": 7.903771985813252e-07,
+      "loss": 0.4731,
+      "step": 6640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.48286399941432,
+      "learning_rate": 7.892931038634094e-07,
+      "loss": 0.4881,
+      "step": 6641
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.04536664575143,
+      "learning_rate": 7.882096894222774e-07,
+      "loss": 0.5289,
+      "step": 6642
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.758295482436992,
+      "learning_rate": 7.871269554329652e-07,
+      "loss": 0.5026,
+      "step": 6643
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3460918611676007,
+      "learning_rate": 7.860449020703964e-07,
+      "loss": 0.4616,
+      "step": 6644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.689594433035852,
+      "learning_rate": 7.849635295093877e-07,
+      "loss": 0.4857,
+      "step": 6645
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6720809223765922,
+      "learning_rate": 7.838828379246422e-07,
+      "loss": 0.4977,
+      "step": 6646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.47095461092173,
+      "learning_rate": 7.828028274907579e-07,
+      "loss": 0.5237,
+      "step": 6647
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5393842391301888,
+      "learning_rate": 7.817234983822169e-07,
+      "loss": 0.4689,
+      "step": 6648
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4527967916406928,
+      "learning_rate": 7.806448507733977e-07,
+      "loss": 0.4854,
+      "step": 6649
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9815323734000887,
+      "learning_rate": 7.795668848385624e-07,
+      "loss": 0.5219,
+      "step": 6650
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7985118940600708,
+      "learning_rate": 7.784896007518667e-07,
+      "loss": 0.4939,
+      "step": 6651
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2185676476628795,
+      "learning_rate": 7.774129986873574e-07,
+      "loss": 0.5087,
+      "step": 6652
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.635521286764626,
+      "learning_rate": 7.763370788189656e-07,
+      "loss": 0.5354,
+      "step": 6653
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3406571460882213,
+      "learning_rate": 7.752618413205193e-07,
+      "loss": 0.4728,
+      "step": 6654
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8504903138524444,
+      "learning_rate": 7.74187286365729e-07,
+      "loss": 0.5104,
+      "step": 6655
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.389096639498019,
+      "learning_rate": 7.731134141282015e-07,
+      "loss": 0.5099,
+      "step": 6656
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.5366339284376185,
+      "learning_rate": 7.720402247814279e-07,
+      "loss": 0.4948,
+      "step": 6657
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.842372319506879,
+      "learning_rate": 7.70967718498794e-07,
+      "loss": 0.4862,
+      "step": 6658
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4324724171615184,
+      "learning_rate": 7.698958954535696e-07,
+      "loss": 0.5086,
+      "step": 6659
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.522130642732948,
+      "learning_rate": 7.688247558189194e-07,
+      "loss": 0.5072,
+      "step": 6660
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4951917839977171,
+      "learning_rate": 7.67754299767895e-07,
+      "loss": 0.4972,
+      "step": 6661
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5651674502519535,
+      "learning_rate": 7.666845274734375e-07,
+      "loss": 0.4754,
+      "step": 6662
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5627366549655106,
+      "learning_rate": 7.656154391083786e-07,
+      "loss": 0.4705,
+      "step": 6663
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6720391307867858,
+      "learning_rate": 7.645470348454381e-07,
+      "loss": 0.5589,
+      "step": 6664
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.376154003760944,
+      "learning_rate": 7.634793148572273e-07,
+      "loss": 0.4723,
+      "step": 6665
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8004953958144452,
+      "learning_rate": 7.624122793162442e-07,
+      "loss": 0.5408,
+      "step": 6666
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5105450022050773,
+      "learning_rate": 7.613459283948788e-07,
+      "loss": 0.5245,
+      "step": 6667
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5470366339253507,
+      "learning_rate": 7.602802622654109e-07,
+      "loss": 0.479,
+      "step": 6668
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4887463762505064,
+      "learning_rate": 7.592152811000053e-07,
+      "loss": 0.4652,
+      "step": 6669
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6424167995495772,
+      "learning_rate": 7.581509850707214e-07,
+      "loss": 0.4651,
+      "step": 6670
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.493513154419008,
+      "learning_rate": 7.570873743495039e-07,
+      "loss": 0.5321,
+      "step": 6671
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4814326981200936,
+      "learning_rate": 7.560244491081903e-07,
+      "loss": 0.4731,
+      "step": 6672
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4213082522322267,
+      "learning_rate": 7.549622095185033e-07,
+      "loss": 0.5278,
+      "step": 6673
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.290850849983848,
+      "learning_rate": 7.539006557520595e-07,
+      "loss": 0.4822,
+      "step": 6674
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4676546201481493,
+      "learning_rate": 7.528397879803589e-07,
+      "loss": 0.494,
+      "step": 6675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9356018002672504,
+      "learning_rate": 7.517796063747984e-07,
+      "loss": 0.4418,
+      "step": 6676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3003381438959716,
+      "learning_rate": 7.507201111066581e-07,
+      "loss": 0.5061,
+      "step": 6677
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9853811495656869,
+      "learning_rate": 7.496613023471061e-07,
+      "loss": 0.468,
+      "step": 6678
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9867056108884016,
+      "learning_rate": 7.486031802672061e-07,
+      "loss": 0.4867,
+      "step": 6679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2765450132478833,
+      "learning_rate": 7.475457450379036e-07,
+      "loss": 0.4914,
+      "step": 6680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.230696818398601,
+      "learning_rate": 7.464889968300398e-07,
+      "loss": 0.4248,
+      "step": 6681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4779685084729597,
+      "learning_rate": 7.454329358143392e-07,
+      "loss": 0.5187,
+      "step": 6682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6978034202504741,
+      "learning_rate": 7.44377562161418e-07,
+      "loss": 0.525,
+      "step": 6683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.902718193616752,
+      "learning_rate": 7.433228760417827e-07,
+      "loss": 0.4744,
+      "step": 6684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5683391925887717,
+      "learning_rate": 7.422688776258268e-07,
+      "loss": 0.4921,
+      "step": 6685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7998659135311224,
+      "learning_rate": 7.412155670838311e-07,
+      "loss": 0.5272,
+      "step": 6686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.400954649684802,
+      "learning_rate": 7.401629445859704e-07,
+      "loss": 0.4615,
+      "step": 6687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7090806767697804,
+      "learning_rate": 7.391110103023031e-07,
+      "loss": 0.5015,
+      "step": 6688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3007966513175657,
+      "learning_rate": 7.38059764402777e-07,
+      "loss": 0.4713,
+      "step": 6689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4708094230997422,
+      "learning_rate": 7.370092070572332e-07,
+      "loss": 0.4907,
+      "step": 6690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5038268737583214,
+      "learning_rate": 7.359593384353958e-07,
+      "loss": 0.5051,
+      "step": 6691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.524279014891403,
+      "learning_rate": 7.349101587068819e-07,
+      "loss": 0.5052,
+      "step": 6692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7172922101698749,
+      "learning_rate": 7.338616680411953e-07,
+      "loss": 0.4597,
+      "step": 6693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5930470211813232,
+      "learning_rate": 7.3281386660773e-07,
+      "loss": 0.515,
+      "step": 6694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8029058761861165,
+      "learning_rate": 7.317667545757656e-07,
+      "loss": 0.4888,
+      "step": 6695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4923651663363493,
+      "learning_rate": 7.307203321144746e-07,
+      "loss": 0.4664,
+      "step": 6696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6185986604616538,
+      "learning_rate": 7.296745993929144e-07,
+      "loss": 0.4886,
+      "step": 6697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2183265656631697,
+      "learning_rate": 7.286295565800305e-07,
+      "loss": 0.5012,
+      "step": 6698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0722141533527365,
+      "learning_rate": 7.275852038446623e-07,
+      "loss": 0.4462,
+      "step": 6699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5310571180094792,
+      "learning_rate": 7.265415413555304e-07,
+      "loss": 0.5104,
+      "step": 6700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5599434286674574,
+      "learning_rate": 7.254985692812517e-07,
+      "loss": 0.5282,
+      "step": 6701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.125125729176223,
+      "learning_rate": 7.244562877903244e-07,
+      "loss": 0.5671,
+      "step": 6702
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6534221527354256,
+      "learning_rate": 7.234146970511408e-07,
+      "loss": 0.4942,
+      "step": 6703
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4397086788418985,
+      "learning_rate": 7.22373797231976e-07,
+      "loss": 0.488,
+      "step": 6704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4557116794569993,
+      "learning_rate": 7.213335885010004e-07,
+      "loss": 0.5201,
+      "step": 6705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5797733791302986,
+      "learning_rate": 7.202940710262657e-07,
+      "loss": 0.5304,
+      "step": 6706
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6977736195425923,
+      "learning_rate": 7.192552449757157e-07,
+      "loss": 0.4691,
+      "step": 6707
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5101471426826862,
+      "learning_rate": 7.18217110517182e-07,
+      "loss": 0.5608,
+      "step": 6708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5421036528060617,
+      "learning_rate": 7.171796678183845e-07,
+      "loss": 0.4554,
+      "step": 6709
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8703537498179146,
+      "learning_rate": 7.161429170469325e-07,
+      "loss": 0.4687,
+      "step": 6710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.48883273820731,
+      "learning_rate": 7.151068583703202e-07,
+      "loss": 0.4986,
+      "step": 6711
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6717400136691807,
+      "learning_rate": 7.140714919559339e-07,
+      "loss": 0.5116,
+      "step": 6712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4106732687778272,
+      "learning_rate": 7.13036817971044e-07,
+      "loss": 0.5113,
+      "step": 6713
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5991966251618026,
+      "learning_rate": 7.120028365828136e-07,
+      "loss": 0.5148,
+      "step": 6714
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.17589346803186,
+      "learning_rate": 7.109695479582889e-07,
+      "loss": 0.4669,
+      "step": 6715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.206969652600979,
+      "learning_rate": 7.099369522644084e-07,
+      "loss": 0.4686,
+      "step": 6716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8786885049448072,
+      "learning_rate": 7.089050496679983e-07,
+      "loss": 0.4683,
+      "step": 6717
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7313156690433527,
+      "learning_rate": 7.078738403357682e-07,
+      "loss": 0.4429,
+      "step": 6718
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9177512900258369,
+      "learning_rate": 7.068433244343221e-07,
+      "loss": 0.4881,
+      "step": 6719
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.477028490334779,
+      "learning_rate": 7.058135021301465e-07,
+      "loss": 0.4805,
+      "step": 6720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5125368951161189,
+      "learning_rate": 7.04784373589621e-07,
+      "loss": 0.4786,
+      "step": 6721
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4642545983208444,
+      "learning_rate": 7.037559389790078e-07,
+      "loss": 0.4947,
+      "step": 6722
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3479241995411038,
+      "learning_rate": 7.027281984644613e-07,
+      "loss": 0.4791,
+      "step": 6723
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4424745723584182,
+      "learning_rate": 7.017011522120198e-07,
+      "loss": 0.5482,
+      "step": 6724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6080376385555766,
+      "learning_rate": 7.00674800387614e-07,
+      "loss": 0.4837,
+      "step": 6725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.548812317353501,
+      "learning_rate": 6.996491431570601e-07,
+      "loss": 0.478,
+      "step": 6726
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0592868543349034,
+      "learning_rate": 6.9862418068606e-07,
+      "loss": 0.4935,
+      "step": 6727
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4010202204985316,
+      "learning_rate": 6.975999131402078e-07,
+      "loss": 0.4488,
+      "step": 6728
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3449621662643738,
+      "learning_rate": 6.965763406849807e-07,
+      "loss": 0.5375,
+      "step": 6729
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4492464573692114,
+      "learning_rate": 6.95553463485748e-07,
+      "loss": 0.5089,
+      "step": 6730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.11499981877359,
+      "learning_rate": 6.945312817077627e-07,
+      "loss": 0.5208,
+      "step": 6731
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3102363888371937,
+      "learning_rate": 6.935097955161685e-07,
+      "loss": 0.4738,
+      "step": 6732
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6401365279072655,
+      "learning_rate": 6.924890050759942e-07,
+      "loss": 0.4658,
+      "step": 6733
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7261704136767818,
+      "learning_rate": 6.914689105521583e-07,
+      "loss": 0.5132,
+      "step": 6734
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5893186137882933,
+      "learning_rate": 6.904495121094673e-07,
+      "loss": 0.5327,
+      "step": 6735
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.970506769573906,
+      "learning_rate": 6.894308099126118e-07,
+      "loss": 0.5151,
+      "step": 6736
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5212801434657697,
+      "learning_rate": 6.884128041261739e-07,
+      "loss": 0.4688,
+      "step": 6737
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6441088709954745,
+      "learning_rate": 6.873954949146194e-07,
+      "loss": 0.4809,
+      "step": 6738
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5754933441117371,
+      "learning_rate": 6.863788824423056e-07,
+      "loss": 0.5092,
+      "step": 6739
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5357523625887821,
+      "learning_rate": 6.853629668734735e-07,
+      "loss": 0.4985,
+      "step": 6740
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7619076256572428,
+      "learning_rate": 6.843477483722544e-07,
+      "loss": 0.4898,
+      "step": 6741
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5544198655012471,
+      "learning_rate": 6.83333227102666e-07,
+      "loss": 0.5098,
+      "step": 6742
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5978962889844126,
+      "learning_rate": 6.823194032286117e-07,
+      "loss": 0.5124,
+      "step": 6743
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.71567831075207,
+      "learning_rate": 6.813062769138856e-07,
+      "loss": 0.4856,
+      "step": 6744
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7832381321843704,
+      "learning_rate": 6.80293848322165e-07,
+      "loss": 0.5148,
+      "step": 6745
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4160018873091047,
+      "learning_rate": 6.792821176170189e-07,
+      "loss": 0.4683,
+      "step": 6746
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6406893819086323,
+      "learning_rate": 6.782710849618989e-07,
+      "loss": 0.4772,
+      "step": 6747
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4935104545738345,
+      "learning_rate": 6.772607505201489e-07,
+      "loss": 0.5658,
+      "step": 6748
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.196383469215414,
+      "learning_rate": 6.762511144549949e-07,
+      "loss": 0.4915,
+      "step": 6749
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.548129351775618,
+      "learning_rate": 6.75242176929553e-07,
+      "loss": 0.5091,
+      "step": 6750
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.626900511568277,
+      "learning_rate": 6.74233938106828e-07,
+      "loss": 0.4861,
+      "step": 6751
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3609349319146933,
+      "learning_rate": 6.732263981497073e-07,
+      "loss": 0.5313,
+      "step": 6752
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.650097378062769,
+      "learning_rate": 6.722195572209694e-07,
+      "loss": 0.4674,
+      "step": 6753
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 19.547528360647437,
+      "learning_rate": 6.712134154832773e-07,
+      "loss": 0.4418,
+      "step": 6754
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4892496792692191,
+      "learning_rate": 6.702079730991829e-07,
+      "loss": 0.5574,
+      "step": 6755
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3719467735091257,
+      "learning_rate": 6.692032302311229e-07,
+      "loss": 0.5561,
+      "step": 6756
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.027162632150969,
+      "learning_rate": 6.68199187041424e-07,
+      "loss": 0.4945,
+      "step": 6757
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.517854761362138,
+      "learning_rate": 6.671958436922976e-07,
+      "loss": 0.4923,
+      "step": 6758
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.982350398814244,
+      "learning_rate": 6.661932003458421e-07,
+      "loss": 0.4545,
+      "step": 6759
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7313162513888793,
+      "learning_rate": 6.651912571640451e-07,
+      "loss": 0.5244,
+      "step": 6760
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9088298322723904,
+      "learning_rate": 6.641900143087765e-07,
+      "loss": 0.4495,
+      "step": 6761
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2008124010151733,
+      "learning_rate": 6.631894719417987e-07,
+      "loss": 0.4727,
+      "step": 6762
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4947199189202311,
+      "learning_rate": 6.62189630224756e-07,
+      "loss": 0.484,
+      "step": 6763
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3880490751486438,
+      "learning_rate": 6.611904893191834e-07,
+      "loss": 0.4469,
+      "step": 6764
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5934160169118272,
+      "learning_rate": 6.601920493864983e-07,
+      "loss": 0.5014,
+      "step": 6765
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5999476697876904,
+      "learning_rate": 6.591943105880111e-07,
+      "loss": 0.4419,
+      "step": 6766
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.5642275673683885,
+      "learning_rate": 6.581972730849134e-07,
+      "loss": 0.4993,
+      "step": 6767
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8449583643704757,
+      "learning_rate": 6.572009370382842e-07,
+      "loss": 0.4553,
+      "step": 6768
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3906538181873989,
+      "learning_rate": 6.562053026090931e-07,
+      "loss": 0.4385,
+      "step": 6769
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.352069179403521,
+      "learning_rate": 6.552103699581908e-07,
+      "loss": 0.46,
+      "step": 6770
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3583931022423161,
+      "learning_rate": 6.542161392463197e-07,
+      "loss": 0.5101,
+      "step": 6771
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.297035235675263,
+      "learning_rate": 6.53222610634105e-07,
+      "loss": 0.4222,
+      "step": 6772
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5077869483962805,
+      "learning_rate": 6.522297842820613e-07,
+      "loss": 0.5018,
+      "step": 6773
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6769521844531408,
+      "learning_rate": 6.512376603505855e-07,
+      "loss": 0.5211,
+      "step": 6774
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.700422140158777,
+      "learning_rate": 6.502462389999681e-07,
+      "loss": 0.502,
+      "step": 6775
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.69152040435353,
+      "learning_rate": 6.4925552039038e-07,
+      "loss": 0.4762,
+      "step": 6776
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3741472861278572,
+      "learning_rate": 6.482655046818798e-07,
+      "loss": 0.4722,
+      "step": 6777
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3143366070549236,
+      "learning_rate": 6.472761920344145e-07,
+      "loss": 0.4673,
+      "step": 6778
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7907775755887394,
+      "learning_rate": 6.462875826078152e-07,
+      "loss": 0.4486,
+      "step": 6779
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1548052117194203,
+      "learning_rate": 6.452996765618019e-07,
+      "loss": 0.4873,
+      "step": 6780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4313653704255374,
+      "learning_rate": 6.443124740559775e-07,
+      "loss": 0.5491,
+      "step": 6781
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4906237832229412,
+      "learning_rate": 6.433259752498344e-07,
+      "loss": 0.5144,
+      "step": 6782
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.533754994591377,
+      "learning_rate": 6.423401803027502e-07,
+      "loss": 0.4494,
+      "step": 6783
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7584702113360715,
+      "learning_rate": 6.413550893739901e-07,
+      "loss": 0.4887,
+      "step": 6784
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.887154015113772,
+      "learning_rate": 6.403707026227013e-07,
+      "loss": 0.4916,
+      "step": 6785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3250076607832437,
+      "learning_rate": 6.393870202079228e-07,
+      "loss": 0.5126,
+      "step": 6786
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7756467996112335,
+      "learning_rate": 6.384040422885762e-07,
+      "loss": 0.5473,
+      "step": 6787
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3992838988740335,
+      "learning_rate": 6.374217690234685e-07,
+      "loss": 0.487,
+      "step": 6788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5766772797693478,
+      "learning_rate": 6.364402005712978e-07,
+      "loss": 0.4455,
+      "step": 6789
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7894202623754922,
+      "learning_rate": 6.354593370906414e-07,
+      "loss": 0.452,
+      "step": 6790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3385213206944417,
+      "learning_rate": 6.344791787399707e-07,
+      "loss": 0.449,
+      "step": 6791
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5542387817707777,
+      "learning_rate": 6.334997256776354e-07,
+      "loss": 0.5214,
+      "step": 6792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3368617749906255,
+      "learning_rate": 6.325209780618769e-07,
+      "loss": 0.4635,
+      "step": 6793
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5667271151416986,
+      "learning_rate": 6.31542936050819e-07,
+      "loss": 0.4699,
+      "step": 6794
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.590647491310967,
+      "learning_rate": 6.305655998024751e-07,
+      "loss": 0.4396,
+      "step": 6795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4654616436837038,
+      "learning_rate": 6.295889694747409e-07,
+      "loss": 0.5049,
+      "step": 6796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2355365348432805,
+      "learning_rate": 6.286130452253985e-07,
+      "loss": 0.4376,
+      "step": 6797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.384137486464471,
+      "learning_rate": 6.276378272121187e-07,
+      "loss": 0.4944,
+      "step": 6798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4240883171336052,
+      "learning_rate": 6.266633155924556e-07,
+      "loss": 0.5043,
+      "step": 6799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4914504853613988,
+      "learning_rate": 6.256895105238526e-07,
+      "loss": 0.4874,
+      "step": 6800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6128088096293538,
+      "learning_rate": 6.247164121636335e-07,
+      "loss": 0.4839,
+      "step": 6801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.716479416737869,
+      "learning_rate": 6.237440206690126e-07,
+      "loss": 0.4588,
+      "step": 6802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5119129085354808,
+      "learning_rate": 6.227723361970872e-07,
+      "loss": 0.5006,
+      "step": 6803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.338104109990617,
+      "learning_rate": 6.218013589048427e-07,
+      "loss": 0.4623,
+      "step": 6804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.765034388598217,
+      "learning_rate": 6.208310889491492e-07,
+      "loss": 0.4845,
+      "step": 6805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.43352133288454,
+      "learning_rate": 6.198615264867586e-07,
+      "loss": 0.5209,
+      "step": 6806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9436874397350934,
+      "learning_rate": 6.188926716743176e-07,
+      "loss": 0.5162,
+      "step": 6807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.781493429515346,
+      "learning_rate": 6.179245246683501e-07,
+      "loss": 0.4996,
+      "step": 6808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4249171534997158,
+      "learning_rate": 6.169570856252705e-07,
+      "loss": 0.4656,
+      "step": 6809
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6613674725633242,
+      "learning_rate": 6.159903547013746e-07,
+      "loss": 0.5147,
+      "step": 6810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0691800130247353,
+      "learning_rate": 6.150243320528492e-07,
+      "loss": 0.563,
+      "step": 6811
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8332421182034946,
+      "learning_rate": 6.140590178357613e-07,
+      "loss": 0.4808,
+      "step": 6812
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9932584487443314,
+      "learning_rate": 6.130944122060678e-07,
+      "loss": 0.4274,
+      "step": 6813
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.389824362923636,
+      "learning_rate": 6.121305153196089e-07,
+      "loss": 0.4536,
+      "step": 6814
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6346590063909652,
+      "learning_rate": 6.111673273321078e-07,
+      "loss": 0.4885,
+      "step": 6815
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9257813859222959,
+      "learning_rate": 6.102048483991802e-07,
+      "loss": 0.4537,
+      "step": 6816
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2158152973201872,
+      "learning_rate": 6.092430786763204e-07,
+      "loss": 0.4991,
+      "step": 6817
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7135611233357684,
+      "learning_rate": 6.082820183189125e-07,
+      "loss": 0.495,
+      "step": 6818
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.550687378435678,
+      "learning_rate": 6.073216674822224e-07,
+      "loss": 0.4899,
+      "step": 6819
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6664152226717598,
+      "learning_rate": 6.063620263214054e-07,
+      "loss": 0.4895,
+      "step": 6820
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6455128240796204,
+      "learning_rate": 6.054030949914974e-07,
+      "loss": 0.4846,
+      "step": 6821
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.371850460382941,
+      "learning_rate": 6.044448736474246e-07,
+      "loss": 0.4772,
+      "step": 6822
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4973161988390176,
+      "learning_rate": 6.03487362443994e-07,
+      "loss": 0.5066,
+      "step": 6823
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3825259689837013,
+      "learning_rate": 6.025305615359012e-07,
+      "loss": 0.4931,
+      "step": 6824
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5546522556411053,
+      "learning_rate": 6.015744710777272e-07,
+      "loss": 0.4854,
+      "step": 6825
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3147495445183237,
+      "learning_rate": 6.006190912239335e-07,
+      "loss": 0.502,
+      "step": 6826
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4508606786369576,
+      "learning_rate": 5.996644221288727e-07,
+      "loss": 0.5091,
+      "step": 6827
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4484167775380636,
+      "learning_rate": 5.987104639467789e-07,
+      "loss": 0.5087,
+      "step": 6828
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4993726181242262,
+      "learning_rate": 5.977572168317736e-07,
+      "loss": 0.47,
+      "step": 6829
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4413689595288772,
+      "learning_rate": 5.968046809378603e-07,
+      "loss": 0.4919,
+      "step": 6830
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5899499033265325,
+      "learning_rate": 5.958528564189303e-07,
+      "loss": 0.4734,
+      "step": 6831
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.389840082506576,
+      "learning_rate": 5.949017434287613e-07,
+      "loss": 0.5051,
+      "step": 6832
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3909594420089038,
+      "learning_rate": 5.939513421210108e-07,
+      "loss": 0.4938,
+      "step": 6833
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5849512907541565,
+      "learning_rate": 5.930016526492272e-07,
+      "loss": 0.5064,
+      "step": 6834
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.492599074910955,
+      "learning_rate": 5.920526751668388e-07,
+      "loss": 0.4408,
+      "step": 6835
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5042860804074434,
+      "learning_rate": 5.911044098271634e-07,
+      "loss": 0.5245,
+      "step": 6836
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6754168298874146,
+      "learning_rate": 5.901568567833999e-07,
+      "loss": 0.4731,
+      "step": 6837
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.409445270913483,
+      "learning_rate": 5.892100161886355e-07,
+      "loss": 0.5193,
+      "step": 6838
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6385356392034915,
+      "learning_rate": 5.882638881958391e-07,
+      "loss": 0.4756,
+      "step": 6839
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9955726024647689,
+      "learning_rate": 5.873184729578662e-07,
+      "loss": 0.4707,
+      "step": 6840
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4777711350788643,
+      "learning_rate": 5.863737706274585e-07,
+      "loss": 0.5064,
+      "step": 6841
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.323637595905769,
+      "learning_rate": 5.854297813572391e-07,
+      "loss": 0.463,
+      "step": 6842
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2142497034160975,
+      "learning_rate": 5.844865052997195e-07,
+      "loss": 0.5102,
+      "step": 6843
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6423488316310543,
+      "learning_rate": 5.835439426072925e-07,
+      "loss": 0.4905,
+      "step": 6844
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0563604548077445,
+      "learning_rate": 5.82602093432239e-07,
+      "loss": 0.4863,
+      "step": 6845
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.095485821395087,
+      "learning_rate": 5.816609579267213e-07,
+      "loss": 0.522,
+      "step": 6846
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4831526662475276,
+      "learning_rate": 5.807205362427903e-07,
+      "loss": 0.4434,
+      "step": 6847
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3373394692004903,
+      "learning_rate": 5.797808285323769e-07,
+      "loss": 0.4971,
+      "step": 6848
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.644148456235197,
+      "learning_rate": 5.788418349473007e-07,
+      "loss": 0.4669,
+      "step": 6849
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5073342737537523,
+      "learning_rate": 5.779035556392649e-07,
+      "loss": 0.5087,
+      "step": 6850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.379628713431779,
+      "learning_rate": 5.769659907598552e-07,
+      "loss": 0.4415,
+      "step": 6851
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.564518382684953,
+      "learning_rate": 5.760291404605456e-07,
+      "loss": 0.5007,
+      "step": 6852
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4022145113727098,
+      "learning_rate": 5.750930048926894e-07,
+      "loss": 0.4262,
+      "step": 6853
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.304688007201386,
+      "learning_rate": 5.741575842075315e-07,
+      "loss": 0.4666,
+      "step": 6854
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.545692701820653,
+      "learning_rate": 5.732228785561933e-07,
+      "loss": 0.5042,
+      "step": 6855
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8449247267045616,
+      "learning_rate": 5.722888880896871e-07,
+      "loss": 0.4493,
+      "step": 6856
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8883732598933922,
+      "learning_rate": 5.713556129589076e-07,
+      "loss": 0.477,
+      "step": 6857
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.386474045155772,
+      "learning_rate": 5.704230533146321e-07,
+      "loss": 0.4596,
+      "step": 6858
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6987912717293612,
+      "learning_rate": 5.69491209307525e-07,
+      "loss": 0.4987,
+      "step": 6859
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.060497679451871,
+      "learning_rate": 5.685600810881331e-07,
+      "loss": 0.505,
+      "step": 6860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3193315097978286,
+      "learning_rate": 5.676296688068894e-07,
+      "loss": 0.5181,
+      "step": 6861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.485638345755509,
+      "learning_rate": 5.666999726141086e-07,
+      "loss": 0.5405,
+      "step": 6862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3331619343085594,
+      "learning_rate": 5.65770992659993e-07,
+      "loss": 0.4544,
+      "step": 6863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6910977225802527,
+      "learning_rate": 5.648427290946251e-07,
+      "loss": 0.4943,
+      "step": 6864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5988001915255896,
+      "learning_rate": 5.639151820679778e-07,
+      "loss": 0.5161,
+      "step": 6865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5984809540074646,
+      "learning_rate": 5.629883517299023e-07,
+      "loss": 0.4703,
+      "step": 6866
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0140295213578163,
+      "learning_rate": 5.620622382301349e-07,
+      "loss": 0.4887,
+      "step": 6867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6634956971310924,
+      "learning_rate": 5.611368417183e-07,
+      "loss": 0.4703,
+      "step": 6868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2212580229409449,
+      "learning_rate": 5.602121623439016e-07,
+      "loss": 0.5075,
+      "step": 6869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4546267165288047,
+      "learning_rate": 5.592882002563316e-07,
+      "loss": 0.5057,
+      "step": 6870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7950585081558756,
+      "learning_rate": 5.583649556048615e-07,
+      "loss": 0.4638,
+      "step": 6871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5665975330120265,
+      "learning_rate": 5.574424285386515e-07,
+      "loss": 0.5191,
+      "step": 6872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4222467915785417,
+      "learning_rate": 5.56520619206744e-07,
+      "loss": 0.543,
+      "step": 6873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5842470350852345,
+      "learning_rate": 5.555995277580661e-07,
+      "loss": 0.4898,
+      "step": 6874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5901026069408062,
+      "learning_rate": 5.546791543414271e-07,
+      "loss": 0.496,
+      "step": 6875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5356187517929056,
+      "learning_rate": 5.537594991055207e-07,
+      "loss": 0.4643,
+      "step": 6876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5491125842300477,
+      "learning_rate": 5.52840562198927e-07,
+      "loss": 0.4627,
+      "step": 6877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6161860408966715,
+      "learning_rate": 5.519223437701066e-07,
+      "loss": 0.5372,
+      "step": 6878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.671789743114391,
+      "learning_rate": 5.510048439674082e-07,
+      "loss": 0.4862,
+      "step": 6879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.320335868944061,
+      "learning_rate": 5.500880629390581e-07,
+      "loss": 0.4466,
+      "step": 6880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9753531052116389,
+      "learning_rate": 5.491720008331747e-07,
+      "loss": 0.479,
+      "step": 6881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4582939912346808,
+      "learning_rate": 5.482566577977533e-07,
+      "loss": 0.4891,
+      "step": 6882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5970603233940475,
+      "learning_rate": 5.473420339806768e-07,
+      "loss": 0.527,
+      "step": 6883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4712123078288173,
+      "learning_rate": 5.464281295297108e-07,
+      "loss": 0.5004,
+      "step": 6884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4977850469823213,
+      "learning_rate": 5.455149445925023e-07,
+      "loss": 0.51,
+      "step": 6885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5388104560848637,
+      "learning_rate": 5.446024793165877e-07,
+      "loss": 0.5139,
+      "step": 6886
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7299515295271184,
+      "learning_rate": 5.436907338493813e-07,
+      "loss": 0.5657,
+      "step": 6887
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5580953548348517,
+      "learning_rate": 5.427797083381853e-07,
+      "loss": 0.4934,
+      "step": 6888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3881940205970094,
+      "learning_rate": 5.418694029301808e-07,
+      "loss": 0.4988,
+      "step": 6889
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5982128731435383,
+      "learning_rate": 5.409598177724401e-07,
+      "loss": 0.505,
+      "step": 6890
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.482857028697596,
+      "learning_rate": 5.400509530119119e-07,
+      "loss": 0.4682,
+      "step": 6891
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.502884208589032,
+      "learning_rate": 5.391428087954326e-07,
+      "loss": 0.4585,
+      "step": 6892
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.390095861718777,
+      "learning_rate": 5.382353852697198e-07,
+      "loss": 0.5184,
+      "step": 6893
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.380414240486957,
+      "learning_rate": 5.373286825813767e-07,
+      "loss": 0.4599,
+      "step": 6894
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3751438932439002,
+      "learning_rate": 5.364227008768886e-07,
+      "loss": 0.4775,
+      "step": 6895
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8372806556281678,
+      "learning_rate": 5.355174403026242e-07,
+      "loss": 0.477,
+      "step": 6896
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6275339354203016,
+      "learning_rate": 5.346129010048368e-07,
+      "loss": 0.4783,
+      "step": 6897
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6915073517649066,
+      "learning_rate": 5.337090831296626e-07,
+      "loss": 0.4772,
+      "step": 6898
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.722069587274079,
+      "learning_rate": 5.328059868231228e-07,
+      "loss": 0.4988,
+      "step": 6899
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7647107186011264,
+      "learning_rate": 5.319036122311184e-07,
+      "loss": 0.501,
+      "step": 6900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5874068943608002,
+      "learning_rate": 5.310019594994381e-07,
+      "loss": 0.446,
+      "step": 6901
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7929932108513464,
+      "learning_rate": 5.301010287737496e-07,
+      "loss": 0.5228,
+      "step": 6902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5965039152926832,
+      "learning_rate": 5.292008201996079e-07,
+      "loss": 0.5783,
+      "step": 6903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.427082342840929,
+      "learning_rate": 5.283013339224491e-07,
+      "loss": 0.4827,
+      "step": 6904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6876122832076539,
+      "learning_rate": 5.274025700875912e-07,
+      "loss": 0.4663,
+      "step": 6905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.398426594294747,
+      "learning_rate": 5.265045288402415e-07,
+      "loss": 0.449,
+      "step": 6906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3728899110215704,
+      "learning_rate": 5.256072103254828e-07,
+      "loss": 0.4561,
+      "step": 6907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2854946046937044,
+      "learning_rate": 5.247106146882874e-07,
+      "loss": 0.478,
+      "step": 6908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.279134504546732,
+      "learning_rate": 5.238147420735057e-07,
+      "loss": 0.4776,
+      "step": 6909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8530388685628911,
+      "learning_rate": 5.229195926258762e-07,
+      "loss": 0.5059,
+      "step": 6910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.436969838582139,
+      "learning_rate": 5.220251664900161e-07,
+      "loss": 0.4998,
+      "step": 6911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5032036608395283,
+      "learning_rate": 5.211314638104303e-07,
+      "loss": 0.5714,
+      "step": 6912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0512575600594394,
+      "learning_rate": 5.202384847315017e-07,
+      "loss": 0.4843,
+      "step": 6913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4050897740726445,
+      "learning_rate": 5.193462293974999e-07,
+      "loss": 0.4815,
+      "step": 6914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2195798154838986,
+      "learning_rate": 5.184546979525779e-07,
+      "loss": 0.4516,
+      "step": 6915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4815168906852805,
+      "learning_rate": 5.175638905407682e-07,
+      "loss": 0.4266,
+      "step": 6916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.602255012128706,
+      "learning_rate": 5.166738073059907e-07,
+      "loss": 0.5068,
+      "step": 6917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0514319780060712,
+      "learning_rate": 5.157844483920449e-07,
+      "loss": 0.4871,
+      "step": 6918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3562324046688408,
+      "learning_rate": 5.148958139426152e-07,
+      "loss": 0.5221,
+      "step": 6919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3925440441575618,
+      "learning_rate": 5.140079041012674e-07,
+      "loss": 0.5112,
+      "step": 6920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.463072794109092,
+      "learning_rate": 5.131207190114518e-07,
+      "loss": 0.436,
+      "step": 6921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7851821317438705,
+      "learning_rate": 5.122342588165013e-07,
+      "loss": 0.4625,
+      "step": 6922
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4061491256502077,
+      "learning_rate": 5.113485236596305e-07,
+      "loss": 0.4975,
+      "step": 6923
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7953328392218353,
+      "learning_rate": 5.104635136839392e-07,
+      "loss": 0.5293,
+      "step": 6924
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.883648584810455,
+      "learning_rate": 5.09579229032407e-07,
+      "loss": 0.4316,
+      "step": 6925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4734215631821337,
+      "learning_rate": 5.08695669847899e-07,
+      "loss": 0.5,
+      "step": 6926
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.390752070784834,
+      "learning_rate": 5.078128362731605e-07,
+      "loss": 0.4571,
+      "step": 6927
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3760124274026402,
+      "learning_rate": 5.06930728450823e-07,
+      "loss": 0.4068,
+      "step": 6928
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4446249003683183,
+      "learning_rate": 5.060493465233962e-07,
+      "loss": 0.5094,
+      "step": 6929
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.636254259184282,
+      "learning_rate": 5.051686906332776e-07,
+      "loss": 0.5108,
+      "step": 6930
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.267435946951365,
+      "learning_rate": 5.042887609227442e-07,
+      "loss": 0.4685,
+      "step": 6931
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.754610151389348,
+      "learning_rate": 5.034095575339553e-07,
+      "loss": 0.5163,
+      "step": 6932
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.463126296444401,
+      "learning_rate": 5.025310806089562e-07,
+      "loss": 0.4913,
+      "step": 6933
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.013428114686188,
+      "learning_rate": 5.016533302896698e-07,
+      "loss": 0.5233,
+      "step": 6934
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3545099717129814,
+      "learning_rate": 5.007763067179066e-07,
+      "loss": 0.4909,
+      "step": 6935
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.733885022215113,
+      "learning_rate": 4.999000100353562e-07,
+      "loss": 0.5019,
+      "step": 6936
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8120183817577111,
+      "learning_rate": 4.990244403835936e-07,
+      "loss": 0.4918,
+      "step": 6937
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6096501007633839,
+      "learning_rate": 4.981495979040723e-07,
+      "loss": 0.4859,
+      "step": 6938
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3699844039744509,
+      "learning_rate": 4.972754827381327e-07,
+      "loss": 0.4719,
+      "step": 6939
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6133878006430424,
+      "learning_rate": 4.964020950269955e-07,
+      "loss": 0.4776,
+      "step": 6940
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5580353495575774,
+      "learning_rate": 4.955294349117634e-07,
+      "loss": 0.4682,
+      "step": 6941
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9898604663077117,
+      "learning_rate": 4.946575025334244e-07,
+      "loss": 0.4954,
+      "step": 6942
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6918769235942603,
+      "learning_rate": 4.937862980328439e-07,
+      "loss": 0.4625,
+      "step": 6943
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5010276257602275,
+      "learning_rate": 4.929158215507751e-07,
+      "loss": 0.4708,
+      "step": 6944
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4238385712397128,
+      "learning_rate": 4.920460732278493e-07,
+      "loss": 0.4965,
+      "step": 6945
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8676252330232028,
+      "learning_rate": 4.911770532045828e-07,
+      "loss": 0.425,
+      "step": 6946
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8991603311100667,
+      "learning_rate": 4.903087616213748e-07,
+      "loss": 0.4908,
+      "step": 6947
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8943841158665158,
+      "learning_rate": 4.894411986185027e-07,
+      "loss": 0.5346,
+      "step": 6948
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7537682284072407,
+      "learning_rate": 4.885743643361318e-07,
+      "loss": 0.4971,
+      "step": 6949
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5885510591798264,
+      "learning_rate": 4.877082589143045e-07,
+      "loss": 0.4881,
+      "step": 6950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.647460623372295,
+      "learning_rate": 4.868428824929494e-07,
+      "loss": 0.5323,
+      "step": 6951
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3393433538659767,
+      "learning_rate": 4.859782352118742e-07,
+      "loss": 0.4605,
+      "step": 6952
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.583819537463703,
+      "learning_rate": 4.851143172107719e-07,
+      "loss": 0.5271,
+      "step": 6953
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.594390898281684,
+      "learning_rate": 4.84251128629214e-07,
+      "loss": 0.4707,
+      "step": 6954
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8417235223966422,
+      "learning_rate": 4.833886696066581e-07,
+      "loss": 0.4842,
+      "step": 6955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5608893660067085,
+      "learning_rate": 4.825269402824417e-07,
+      "loss": 0.5323,
+      "step": 6956
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6182598019281986,
+      "learning_rate": 4.816659407957835e-07,
+      "loss": 0.4844,
+      "step": 6957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7888155570652113,
+      "learning_rate": 4.808056712857878e-07,
+      "loss": 0.4965,
+      "step": 6958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.470841132601057,
+      "learning_rate": 4.799461318914366e-07,
+      "loss": 0.4736,
+      "step": 6959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5958096796885273,
+      "learning_rate": 4.790873227515974e-07,
+      "loss": 0.4258,
+      "step": 6960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4668180431159055,
+      "learning_rate": 4.782292440050173e-07,
+      "loss": 0.4788,
+      "step": 6961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5940426778726535,
+      "learning_rate": 4.773718957903267e-07,
+      "loss": 0.484,
+      "step": 6962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2904653920667863,
+      "learning_rate": 4.7651527824603804e-07,
+      "loss": 0.4734,
+      "step": 6963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.552833442468006,
+      "learning_rate": 4.7565939151054675e-07,
+      "loss": 0.5049,
+      "step": 6964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7028529981961882,
+      "learning_rate": 4.748042357221278e-07,
+      "loss": 0.482,
+      "step": 6965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.118028428478353,
+      "learning_rate": 4.739498110189372e-07,
+      "loss": 0.5504,
+      "step": 6966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5657740026232285,
+      "learning_rate": 4.7309611753901806e-07,
+      "loss": 0.4814,
+      "step": 6967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3267340843161755,
+      "learning_rate": 4.7224315542028945e-07,
+      "loss": 0.4931,
+      "step": 6968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3367205730794152,
+      "learning_rate": 4.7139092480055657e-07,
+      "loss": 0.4812,
+      "step": 6969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4363411485262998,
+      "learning_rate": 4.705394258175039e-07,
+      "loss": 0.5002,
+      "step": 6970
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.3872480517546,
+      "learning_rate": 4.696886586086985e-07,
+      "loss": 0.5091,
+      "step": 6971
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2398113701487647,
+      "learning_rate": 4.688386233115899e-07,
+      "loss": 0.476,
+      "step": 6972
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4385066929159718,
+      "learning_rate": 4.679893200635094e-07,
+      "loss": 0.4617,
+      "step": 6973
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5019290438655046,
+      "learning_rate": 4.671407490016688e-07,
+      "loss": 0.4902,
+      "step": 6974
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4173222510567882,
+      "learning_rate": 4.662929102631614e-07,
+      "loss": 0.4585,
+      "step": 6975
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.913621336040645,
+      "learning_rate": 4.654458039849641e-07,
+      "loss": 0.4994,
+      "step": 6976
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5367385729264298,
+      "learning_rate": 4.6459943030393307e-07,
+      "loss": 0.5567,
+      "step": 6977
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1268769033496473,
+      "learning_rate": 4.637537893568095e-07,
+      "loss": 0.4387,
+      "step": 6978
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5296982464759212,
+      "learning_rate": 4.629088812802113e-07,
+      "loss": 0.4426,
+      "step": 6979
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.507568436514036,
+      "learning_rate": 4.620647062106437e-07,
+      "loss": 0.4643,
+      "step": 6980
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5170270026171093,
+      "learning_rate": 4.6122126428448875e-07,
+      "loss": 0.4917,
+      "step": 6981
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3505376511882239,
+      "learning_rate": 4.603785556380136e-07,
+      "loss": 0.4725,
+      "step": 6982
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6328794550198478,
+      "learning_rate": 4.595365804073637e-07,
+      "loss": 0.5333,
+      "step": 6983
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3863656552023294,
+      "learning_rate": 4.5869533872856753e-07,
+      "loss": 0.4819,
+      "step": 6984
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4091310542928852,
+      "learning_rate": 4.578548307375369e-07,
+      "loss": 0.4823,
+      "step": 6985
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3905944890206892,
+      "learning_rate": 4.570150565700604e-07,
+      "loss": 0.4707,
+      "step": 6986
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3268258977962248,
+      "learning_rate": 4.561760163618129e-07,
+      "loss": 0.4803,
+      "step": 6987
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6313590022212238,
+      "learning_rate": 4.553377102483486e-07,
+      "loss": 0.482,
+      "step": 6988
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3878481573742947,
+      "learning_rate": 4.545001383651032e-07,
+      "loss": 0.4823,
+      "step": 6989
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5216919725001328,
+      "learning_rate": 4.536633008473934e-07,
+      "loss": 0.486,
+      "step": 6990
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6149779706162697,
+      "learning_rate": 4.5282719783041883e-07,
+      "loss": 0.518,
+      "step": 6991
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3652599863085746,
+      "learning_rate": 4.519918294492581e-07,
+      "loss": 0.4898,
+      "step": 6992
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4485925323053548,
+      "learning_rate": 4.5115719583887164e-07,
+      "loss": 0.4877,
+      "step": 6993
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5871411096296153,
+      "learning_rate": 4.503232971341037e-07,
+      "loss": 0.4946,
+      "step": 6994
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1794323859003204,
+      "learning_rate": 4.494901334696755e-07,
+      "loss": 0.4527,
+      "step": 6995
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6742379324367876,
+      "learning_rate": 4.4865770498019545e-07,
+      "loss": 0.4836,
+      "step": 6996
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4569596184809448,
+      "learning_rate": 4.478260118001465e-07,
+      "loss": 0.4605,
+      "step": 6997
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4222433310926212,
+      "learning_rate": 4.4699505406389844e-07,
+      "loss": 0.5305,
+      "step": 6998
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9163666236933992,
+      "learning_rate": 4.4616483190569773e-07,
+      "loss": 0.5253,
+      "step": 6999
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4860749254374586,
+      "learning_rate": 4.4533534545967593e-07,
+      "loss": 0.4816,
+      "step": 7000
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.977852488541996,
+      "learning_rate": 4.445065948598426e-07,
+      "loss": 0.5104,
+      "step": 7001
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4291979735578966,
+      "learning_rate": 4.436785802400906e-07,
+      "loss": 0.4516,
+      "step": 7002
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6554197655711902,
+      "learning_rate": 4.428513017341923e-07,
+      "loss": 0.496,
+      "step": 7003
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4514424167573279,
+      "learning_rate": 4.4202475947579984e-07,
+      "loss": 0.495,
+      "step": 7004
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3707185593181825,
+      "learning_rate": 4.411989535984529e-07,
+      "loss": 0.4912,
+      "step": 7005
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5072176077192196,
+      "learning_rate": 4.403738842355643e-07,
+      "loss": 0.4895,
+      "step": 7006
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4085864240181591,
+      "learning_rate": 4.3954955152043346e-07,
+      "loss": 0.4813,
+      "step": 7007
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8764392388366993,
+      "learning_rate": 4.387259555862361e-07,
+      "loss": 0.5289,
+      "step": 7008
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3581652441850534,
+      "learning_rate": 4.3790309656603356e-07,
+      "loss": 0.4587,
+      "step": 7009
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.314257733383638,
+      "learning_rate": 4.3708097459276454e-07,
+      "loss": 0.4924,
+      "step": 7010
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4795930320292494,
+      "learning_rate": 4.3625958979925167e-07,
+      "loss": 0.5266,
+      "step": 7011
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6737719701590443,
+      "learning_rate": 4.354389423181948e-07,
+      "loss": 0.5088,
+      "step": 7012
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0460952816878635,
+      "learning_rate": 4.3461903228217806e-07,
+      "loss": 0.4388,
+      "step": 7013
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4735752787958016,
+      "learning_rate": 4.3379985982366537e-07,
+      "loss": 0.4596,
+      "step": 7014
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.811600709993341,
+      "learning_rate": 4.329814250750003e-07,
+      "loss": 0.4417,
+      "step": 7015
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.59423739643667,
+      "learning_rate": 4.321637281684099e-07,
+      "loss": 0.5272,
+      "step": 7016
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5071981707243922,
+      "learning_rate": 4.313467692359974e-07,
+      "loss": 0.5227,
+      "step": 7017
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3654563630416694,
+      "learning_rate": 4.3053054840975325e-07,
+      "loss": 0.4817,
+      "step": 7018
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.694570195541426,
+      "learning_rate": 4.2971506582154155e-07,
+      "loss": 0.4642,
+      "step": 7019
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3821178205066278,
+      "learning_rate": 4.289003216031129e-07,
+      "loss": 0.5009,
+      "step": 7020
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6802113063898786,
+      "learning_rate": 4.2808631588609704e-07,
+      "loss": 0.4776,
+      "step": 7021
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5209079592364636,
+      "learning_rate": 4.272730488020016e-07,
+      "loss": 0.4953,
+      "step": 7022
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4538938283349732,
+      "learning_rate": 4.2646052048221866e-07,
+      "loss": 0.5018,
+      "step": 7023
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4972582845594824,
+      "learning_rate": 4.2564873105801816e-07,
+      "loss": 0.5248,
+      "step": 7024
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6705282520765294,
+      "learning_rate": 4.2483768066055365e-07,
+      "loss": 0.5378,
+      "step": 7025
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5286983823682547,
+      "learning_rate": 4.240273694208552e-07,
+      "loss": 0.476,
+      "step": 7026
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5329349887772938,
+      "learning_rate": 4.2321779746983706e-07,
+      "loss": 0.5087,
+      "step": 7027
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.142670079847223,
+      "learning_rate": 4.224089649382923e-07,
+      "loss": 0.5523,
+      "step": 7028
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.099696836371022,
+      "learning_rate": 4.216008719568948e-07,
+      "loss": 0.4539,
+      "step": 7029
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3233495269215116,
+      "learning_rate": 4.2079351865620013e-07,
+      "loss": 0.4979,
+      "step": 7030
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.369596831259944,
+      "learning_rate": 4.1998690516664165e-07,
+      "loss": 0.4557,
+      "step": 7031
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6306367826899542,
+      "learning_rate": 4.191810316185374e-07,
+      "loss": 0.5057,
+      "step": 7032
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.498312849387136,
+      "learning_rate": 4.183758981420799e-07,
+      "loss": 0.5031,
+      "step": 7033
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.390904829889811,
+      "learning_rate": 4.17571504867349e-07,
+      "loss": 0.5169,
+      "step": 7034
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6042617395300853,
+      "learning_rate": 4.167678519242985e-07,
+      "loss": 0.4576,
+      "step": 7035
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.941251307886222,
+      "learning_rate": 4.159649394427673e-07,
+      "loss": 0.5456,
+      "step": 7036
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5270596546982536,
+      "learning_rate": 4.151627675524733e-07,
+      "loss": 0.4076,
+      "step": 7037
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4227414585009857,
+      "learning_rate": 4.1436133638301346e-07,
+      "loss": 0.4517,
+      "step": 7038
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6963363973200976,
+      "learning_rate": 4.1356064606386694e-07,
+      "loss": 0.5013,
+      "step": 7039
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3894778049158558,
+      "learning_rate": 4.1276069672439135e-07,
+      "loss": 0.4925,
+      "step": 7040
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6235082248499837,
+      "learning_rate": 4.119614884938261e-07,
+      "loss": 0.4537,
+      "step": 7041
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6274598684100545,
+      "learning_rate": 4.111630215012902e-07,
+      "loss": 0.4917,
+      "step": 7042
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6183380914641294,
+      "learning_rate": 4.103652958757831e-07,
+      "loss": 0.505,
+      "step": 7043
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5196493987413147,
+      "learning_rate": 4.0956831174618403e-07,
+      "loss": 0.4673,
+      "step": 7044
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.383596502398442,
+      "learning_rate": 4.0877206924125337e-07,
+      "loss": 0.4677,
+      "step": 7045
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8790719779796432,
+      "learning_rate": 4.0797656848963086e-07,
+      "loss": 0.5108,
+      "step": 7046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4487511887778446,
+      "learning_rate": 4.0718180961983655e-07,
+      "loss": 0.472,
+      "step": 7047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6373521545515435,
+      "learning_rate": 4.0638779276027106e-07,
+      "loss": 0.5152,
+      "step": 7048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6613826241580701,
+      "learning_rate": 4.05594518039214e-07,
+      "loss": 0.4638,
+      "step": 7049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.698074527625873,
+      "learning_rate": 4.048019855848273e-07,
+      "loss": 0.4555,
+      "step": 7050
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6727169009641928,
+      "learning_rate": 4.040101955251502e-07,
+      "loss": 0.4903,
+      "step": 7051
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5725392983596518,
+      "learning_rate": 4.0321914798810423e-07,
+      "loss": 0.4994,
+      "step": 7052
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4342289383946316,
+      "learning_rate": 4.0242884310148945e-07,
+      "loss": 0.4319,
+      "step": 7053
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5702375613899564,
+      "learning_rate": 4.0163928099298644e-07,
+      "loss": 0.4696,
+      "step": 7054
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3036488520632732,
+      "learning_rate": 4.008504617901571e-07,
+      "loss": 0.4305,
+      "step": 7055
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6565508220144567,
+      "learning_rate": 4.0006238562044054e-07,
+      "loss": 0.476,
+      "step": 7056
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8813873177790947,
+      "learning_rate": 3.992750526111594e-07,
+      "loss": 0.5009,
+      "step": 7057
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.313124075124981,
+      "learning_rate": 3.984884628895119e-07,
+      "loss": 0.4919,
+      "step": 7058
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3541457299851034,
+      "learning_rate": 3.9770261658258093e-07,
+      "loss": 0.5033,
+      "step": 7059
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4366559915321013,
+      "learning_rate": 3.9691751381732423e-07,
+      "loss": 0.4939,
+      "step": 7060
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5378230470666459,
+      "learning_rate": 3.961331547205838e-07,
+      "loss": 0.4741,
+      "step": 7061
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6909535505851894,
+      "learning_rate": 3.953495394190804e-07,
+      "loss": 0.5063,
+      "step": 7062
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3864880882834227,
+      "learning_rate": 3.9456666803941223e-07,
+      "loss": 0.4779,
+      "step": 7063
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.451029212441224,
+      "learning_rate": 3.937845407080604e-07,
+      "loss": 0.4789,
+      "step": 7064
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5687085727703403,
+      "learning_rate": 3.930031575513837e-07,
+      "loss": 0.5399,
+      "step": 7065
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2924570237505366,
+      "learning_rate": 3.9222251869562187e-07,
+      "loss": 0.5184,
+      "step": 7066
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.500961660790129,
+      "learning_rate": 3.9144262426689336e-07,
+      "loss": 0.4618,
+      "step": 7067
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7030292271341143,
+      "learning_rate": 3.90663474391198e-07,
+      "loss": 0.5014,
+      "step": 7068
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2696125199636845,
+      "learning_rate": 3.898850691944123e-07,
+      "loss": 0.4243,
+      "step": 7069
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5740029906704345,
+      "learning_rate": 3.891074088022978e-07,
+      "loss": 0.5329,
+      "step": 7070
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6799717738407427,
+      "learning_rate": 3.8833049334048964e-07,
+      "loss": 0.4688,
+      "step": 7071
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0888448681941014,
+      "learning_rate": 3.875543229345069e-07,
+      "loss": 0.473,
+      "step": 7072
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.220257275918847,
+      "learning_rate": 3.8677889770974584e-07,
+      "loss": 0.4829,
+      "step": 7073
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9022525076310015,
+      "learning_rate": 3.8600421779148303e-07,
+      "loss": 0.5442,
+      "step": 7074
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4515938616789295,
+      "learning_rate": 3.85230283304876e-07,
+      "loss": 0.5364,
+      "step": 7075
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5089744744651061,
+      "learning_rate": 3.844570943749593e-07,
+      "loss": 0.4802,
+      "step": 7076
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5771912659647935,
+      "learning_rate": 3.83684651126649e-07,
+      "loss": 0.506,
+      "step": 7077
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5971997865078935,
+      "learning_rate": 3.829129536847398e-07,
+      "loss": 0.5213,
+      "step": 7078
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8669592100775223,
+      "learning_rate": 3.8214200217390806e-07,
+      "loss": 0.5021,
+      "step": 7079
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.520345410466755,
+      "learning_rate": 3.8137179671870527e-07,
+      "loss": 0.5044,
+      "step": 7080
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3843402750451081,
+      "learning_rate": 3.8060233744356634e-07,
+      "loss": 0.484,
+      "step": 7081
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3083414725989773,
+      "learning_rate": 3.798336244728046e-07,
+      "loss": 0.4468,
+      "step": 7082
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2366682560363815,
+      "learning_rate": 3.790656579306101e-07,
+      "loss": 0.512,
+      "step": 7083
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3774229109773617,
+      "learning_rate": 3.7829843794105703e-07,
+      "loss": 0.4919,
+      "step": 7084
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2780907545876765,
+      "learning_rate": 3.7753196462809504e-07,
+      "loss": 0.4593,
+      "step": 7085
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6647475571312081,
+      "learning_rate": 3.767662381155551e-07,
+      "loss": 0.4928,
+      "step": 7086
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4554371973142786,
+      "learning_rate": 3.760012585271472e-07,
+      "loss": 0.5152,
+      "step": 7087
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2374825063762878,
+      "learning_rate": 3.7523702598646185e-07,
+      "loss": 0.4613,
+      "step": 7088
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.246405451093617,
+      "learning_rate": 3.7447354061696474e-07,
+      "loss": 0.4897,
+      "step": 7089
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.497491611425484,
+      "learning_rate": 3.737108025420061e-07,
+      "loss": 0.5095,
+      "step": 7090
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.448953188662499,
+      "learning_rate": 3.7294881188481235e-07,
+      "loss": 0.5111,
+      "step": 7091
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3265262982691868,
+      "learning_rate": 3.721875687684884e-07,
+      "loss": 0.4747,
+      "step": 7092
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4006392357014956,
+      "learning_rate": 3.714270733160219e-07,
+      "loss": 0.4648,
+      "step": 7093
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4159650763549114,
+      "learning_rate": 3.706673256502746e-07,
+      "loss": 0.5373,
+      "step": 7094
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.71334454708964,
+      "learning_rate": 3.6990832589399396e-07,
+      "loss": 0.4858,
+      "step": 7095
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4613986540202215,
+      "learning_rate": 3.6915007416980067e-07,
+      "loss": 0.4787,
+      "step": 7096
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.439666219997754,
+      "learning_rate": 3.6839257060019895e-07,
+      "loss": 0.5414,
+      "step": 7097
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7148078587442241,
+      "learning_rate": 3.6763581530756766e-07,
+      "loss": 0.4941,
+      "step": 7098
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7614900950085237,
+      "learning_rate": 3.6687980841417005e-07,
+      "loss": 0.5396,
+      "step": 7099
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4900408278461268,
+      "learning_rate": 3.661245500421429e-07,
+      "loss": 0.5185,
+      "step": 7100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.410188054615284,
+      "learning_rate": 3.653700403135074e-07,
+      "loss": 0.478,
+      "step": 7101
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3467691276511204,
+      "learning_rate": 3.6461627935015897e-07,
+      "loss": 0.4917,
+      "step": 7102
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.476580030239518,
+      "learning_rate": 3.638632672738757e-07,
+      "loss": 0.459,
+      "step": 7103
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6438546283641577,
+      "learning_rate": 3.6311100420631405e-07,
+      "loss": 0.4631,
+      "step": 7104
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9040075114903101,
+      "learning_rate": 3.623594902690064e-07,
+      "loss": 0.4624,
+      "step": 7105
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.77764161446545,
+      "learning_rate": 3.616087255833689e-07,
+      "loss": 0.4732,
+      "step": 7106
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5382156221576366,
+      "learning_rate": 3.608587102706923e-07,
+      "loss": 0.4758,
+      "step": 7107
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.880523836838473,
+      "learning_rate": 3.601094444521497e-07,
+      "loss": 0.5175,
+      "step": 7108
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.319183728535753,
+      "learning_rate": 3.5936092824878976e-07,
+      "loss": 0.4914,
+      "step": 7109
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4221183759049574,
+      "learning_rate": 3.586131617815436e-07,
+      "loss": 0.527,
+      "step": 7110
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7200757512072127,
+      "learning_rate": 3.57866145171219e-07,
+      "loss": 0.4504,
+      "step": 7111
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5298123462961473,
+      "learning_rate": 3.571198785385027e-07,
+      "loss": 0.4757,
+      "step": 7112
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4777977381633165,
+      "learning_rate": 3.5637436200396115e-07,
+      "loss": 0.5434,
+      "step": 7113
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6301079688826285,
+      "learning_rate": 3.5562959568803843e-07,
+      "loss": 0.4849,
+      "step": 7114
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3511275125402105,
+      "learning_rate": 3.54885579711059e-07,
+      "loss": 0.5205,
+      "step": 7115
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4338245476889193,
+      "learning_rate": 3.541423141932238e-07,
+      "loss": 0.4767,
+      "step": 7116
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5251989790197307,
+      "learning_rate": 3.5339979925461566e-07,
+      "loss": 0.4685,
+      "step": 7117
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.368497415552152,
+      "learning_rate": 3.526580350151931e-07,
+      "loss": 0.5087,
+      "step": 7118
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4030800124288458,
+      "learning_rate": 3.519170215947948e-07,
+      "loss": 0.5299,
+      "step": 7119
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8438357371279437,
+      "learning_rate": 3.511767591131393e-07,
+      "loss": 0.5338,
+      "step": 7120
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.089390945220485,
+      "learning_rate": 3.504372476898205e-07,
+      "loss": 0.5054,
+      "step": 7121
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6357190621813416,
+      "learning_rate": 3.4969848744431487e-07,
+      "loss": 0.5339,
+      "step": 7122
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5665757084370637,
+      "learning_rate": 3.489604784959744e-07,
+      "loss": 0.4957,
+      "step": 7123
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4908392013334355,
+      "learning_rate": 3.482232209640318e-07,
+      "loss": 0.469,
+      "step": 7124
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5897939270204158,
+      "learning_rate": 3.474867149675959e-07,
+      "loss": 0.5276,
+      "step": 7125
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0001121398745645,
+      "learning_rate": 3.467509606256581e-07,
+      "loss": 0.519,
+      "step": 7126
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7415972576660335,
+      "learning_rate": 3.4601595805708375e-07,
+      "loss": 0.4467,
+      "step": 7127
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.807146073626419,
+      "learning_rate": 3.4528170738062027e-07,
+      "loss": 0.5191,
+      "step": 7128
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6450389504343121,
+      "learning_rate": 3.445482087148927e-07,
+      "loss": 0.4805,
+      "step": 7129
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4979170319680737,
+      "learning_rate": 3.438154621784029e-07,
+      "loss": 0.4916,
+      "step": 7130
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6597403892980278,
+      "learning_rate": 3.4308346788953394e-07,
+      "loss": 0.4517,
+      "step": 7131
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4360278235152508,
+      "learning_rate": 3.423522259665446e-07,
+      "loss": 0.4837,
+      "step": 7132
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.78841914191125,
+      "learning_rate": 3.4162173652757427e-07,
+      "loss": 0.5079,
+      "step": 7133
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7577408413013773,
+      "learning_rate": 3.4089199969063957e-07,
+      "loss": 0.5455,
+      "step": 7134
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8621486529156506,
+      "learning_rate": 3.4016301557363573e-07,
+      "loss": 0.5004,
+      "step": 7135
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3443240630108315,
+      "learning_rate": 3.3943478429433797e-07,
+      "loss": 0.4592,
+      "step": 7136
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5636160522143439,
+      "learning_rate": 3.38707305970396e-07,
+      "loss": 0.4867,
+      "step": 7137
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4797636371493548,
+      "learning_rate": 3.3798058071934315e-07,
+      "loss": 0.4948,
+      "step": 7138
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7344713320689402,
+      "learning_rate": 3.372546086585859e-07,
+      "loss": 0.4734,
+      "step": 7139
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8399894492094186,
+      "learning_rate": 3.365293899054134e-07,
+      "loss": 0.487,
+      "step": 7140
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6768159138284706,
+      "learning_rate": 3.35804924576989e-07,
+      "loss": 0.5342,
+      "step": 7141
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.492573985054435,
+      "learning_rate": 3.3508121279035856e-07,
+      "loss": 0.4673,
+      "step": 7142
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8692692734840493,
+      "learning_rate": 3.3435825466244245e-07,
+      "loss": 0.5184,
+      "step": 7143
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.846394080663176,
+      "learning_rate": 3.3363605031004167e-07,
+      "loss": 0.5093,
+      "step": 7144
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8655230755218826,
+      "learning_rate": 3.329145998498351e-07,
+      "loss": 0.529,
+      "step": 7145
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7549149164049279,
+      "learning_rate": 3.321939033983784e-07,
+      "loss": 0.5134,
+      "step": 7146
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7333913256290412,
+      "learning_rate": 3.314739610721074e-07,
+      "loss": 0.4907,
+      "step": 7147
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9064714456136993,
+      "learning_rate": 3.307547729873339e-07,
+      "loss": 0.4643,
+      "step": 7148
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3797398845123694,
+      "learning_rate": 3.3003633926025126e-07,
+      "loss": 0.4881,
+      "step": 7149
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.465083071290942,
+      "learning_rate": 3.293186600069259e-07,
+      "loss": 0.4659,
+      "step": 7150
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6122896595208838,
+      "learning_rate": 3.2860173534330744e-07,
+      "loss": 0.5195,
+      "step": 7151
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5037231912700562,
+      "learning_rate": 3.2788556538522086e-07,
+      "loss": 0.4601,
+      "step": 7152
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7022555579379326,
+      "learning_rate": 3.2717015024836875e-07,
+      "loss": 0.453,
+      "step": 7153
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.341508810355981,
+      "learning_rate": 3.2645549004833464e-07,
+      "loss": 0.5044,
+      "step": 7154
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.378762619010031,
+      "learning_rate": 3.2574158490057573e-07,
+      "loss": 0.4611,
+      "step": 7155
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.694921735627754,
+      "learning_rate": 3.2502843492043246e-07,
+      "loss": 0.5117,
+      "step": 7156
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6336333360011273,
+      "learning_rate": 3.243160402231177e-07,
+      "loss": 0.4878,
+      "step": 7157
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8195915630189345,
+      "learning_rate": 3.2360440092372716e-07,
+      "loss": 0.491,
+      "step": 7158
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.472846216891738,
+      "learning_rate": 3.228935171372299e-07,
+      "loss": 0.5043,
+      "step": 7159
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5073396580443326,
+      "learning_rate": 3.221833889784792e-07,
+      "loss": 0.4583,
+      "step": 7160
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.628352258905142,
+      "learning_rate": 3.214740165622005e-07,
+      "loss": 0.5192,
+      "step": 7161
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4070354758414292,
+      "learning_rate": 3.2076540000299804e-07,
+      "loss": 0.4868,
+      "step": 7162
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8743589664505378,
+      "learning_rate": 3.200575394153577e-07,
+      "loss": 0.4941,
+      "step": 7163
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6957401814278172,
+      "learning_rate": 3.1935043491363784e-07,
+      "loss": 0.458,
+      "step": 7164
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5618551792774675,
+      "learning_rate": 3.1864408661207935e-07,
+      "loss": 0.4657,
+      "step": 7165
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3676928671793762,
+      "learning_rate": 3.1793849462479817e-07,
+      "loss": 0.4656,
+      "step": 7166
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3381120850497739,
+      "learning_rate": 3.172336590657893e-07,
+      "loss": 0.5007,
+      "step": 7167
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7082831621344305,
+      "learning_rate": 3.165295800489238e-07,
+      "loss": 0.4443,
+      "step": 7168
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.805985383532289,
+      "learning_rate": 3.158262576879545e-07,
+      "loss": 0.4938,
+      "step": 7169
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4085538252156595,
+      "learning_rate": 3.1512369209650673e-07,
+      "loss": 0.5054,
+      "step": 7170
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3710216311892396,
+      "learning_rate": 3.14421883388088e-07,
+      "loss": 0.423,
+      "step": 7171
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4943971761522419,
+      "learning_rate": 3.137208316760809e-07,
+      "loss": 0.4773,
+      "step": 7172
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3146847075274068,
+      "learning_rate": 3.1302053707374605e-07,
+      "loss": 0.5013,
+      "step": 7173
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5673236063433522,
+      "learning_rate": 3.123209996942228e-07,
+      "loss": 0.4954,
+      "step": 7174
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5592195204413026,
+      "learning_rate": 3.1162221965052753e-07,
+      "loss": 0.4778,
+      "step": 7175
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6967523625595746,
+      "learning_rate": 3.109241970555538e-07,
+      "loss": 0.4947,
+      "step": 7176
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4297009420714675,
+      "learning_rate": 3.1022693202207355e-07,
+      "loss": 0.4709,
+      "step": 7177
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7542963468229498,
+      "learning_rate": 3.095304246627373e-07,
+      "loss": 0.4753,
+      "step": 7178
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5299932504585612,
+      "learning_rate": 3.0883467509007005e-07,
+      "loss": 0.4826,
+      "step": 7179
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.199621225915617,
+      "learning_rate": 3.081396834164774e-07,
+      "loss": 0.4789,
+      "step": 7180
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.539053220866323,
+      "learning_rate": 3.0744544975424173e-07,
+      "loss": 0.467,
+      "step": 7181
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5034629040386387,
+      "learning_rate": 3.0675197421552116e-07,
+      "loss": 0.4767,
+      "step": 7182
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7777352958898378,
+      "learning_rate": 3.060592569123544e-07,
+      "loss": 0.4573,
+      "step": 7183
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5106190466274196,
+      "learning_rate": 3.0536729795665355e-07,
+      "loss": 0.5401,
+      "step": 7184
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4379789088763033,
+      "learning_rate": 3.046760974602136e-07,
+      "loss": 0.446,
+      "step": 7185
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2860703238892677,
+      "learning_rate": 3.039856555347026e-07,
+      "loss": 0.4643,
+      "step": 7186
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3835890592762499,
+      "learning_rate": 3.032959722916684e-07,
+      "loss": 0.4692,
+      "step": 7187
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6878989264050003,
+      "learning_rate": 3.026070478425336e-07,
+      "loss": 0.4983,
+      "step": 7188
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6771846553556036,
+      "learning_rate": 3.019188822986024e-07,
+      "loss": 0.4646,
+      "step": 7189
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6440426570746323,
+      "learning_rate": 3.012314757710527e-07,
+      "loss": 0.5425,
+      "step": 7190
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.074279077892209,
+      "learning_rate": 3.0054482837094003e-07,
+      "loss": 0.4386,
+      "step": 7191
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5173362425262247,
+      "learning_rate": 2.9985894020919946e-07,
+      "loss": 0.504,
+      "step": 7192
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4264982703753875,
+      "learning_rate": 2.9917381139664245e-07,
+      "loss": 0.5233,
+      "step": 7193
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 26.549343377170732,
+      "learning_rate": 2.9848944204395827e-07,
+      "loss": 0.5357,
+      "step": 7194
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4586810658158686,
+      "learning_rate": 2.978058322617111e-07,
+      "loss": 0.4955,
+      "step": 7195
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3883893237445497,
+      "learning_rate": 2.9712298216034553e-07,
+      "loss": 0.4755,
+      "step": 7196
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4159920892476956,
+      "learning_rate": 2.96440891850181e-07,
+      "loss": 0.5324,
+      "step": 7197
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.415269551961386,
+      "learning_rate": 2.95759561441416e-07,
+      "loss": 0.5041,
+      "step": 7198
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5364140965549926,
+      "learning_rate": 2.950789910441254e-07,
+      "loss": 0.5059,
+      "step": 7199
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4547641326655592,
+      "learning_rate": 2.9439918076826003e-07,
+      "loss": 0.5275,
+      "step": 7200
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6340603974161589,
+      "learning_rate": 2.937201307236504e-07,
+      "loss": 0.4743,
+      "step": 7201
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6838431804403837,
+      "learning_rate": 2.930418410200031e-07,
+      "loss": 0.4303,
+      "step": 7202
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.59902466849343,
+      "learning_rate": 2.923643117669023e-07,
+      "loss": 0.4981,
+      "step": 7203
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.588337220386599,
+      "learning_rate": 2.91687543073807e-07,
+      "loss": 0.459,
+      "step": 7204
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8065695379788074,
+      "learning_rate": 2.9101153505005697e-07,
+      "loss": 0.4769,
+      "step": 7205
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2622077395268985,
+      "learning_rate": 2.9033628780486603e-07,
+      "loss": 0.4758,
+      "step": 7206
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2822251089559757,
+      "learning_rate": 2.8966180144732735e-07,
+      "loss": 0.4679,
+      "step": 7207
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7942440218571605,
+      "learning_rate": 2.889880760864089e-07,
+      "loss": 0.504,
+      "step": 7208
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5395160273565343,
+      "learning_rate": 2.883151118309574e-07,
+      "loss": 0.4623,
+      "step": 7209
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5010348696910691,
+      "learning_rate": 2.8764290878969757e-07,
+      "loss": 0.4999,
+      "step": 7210
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6289662605286688,
+      "learning_rate": 2.869714670712276e-07,
+      "loss": 0.4778,
+      "step": 7211
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6921986203870252,
+      "learning_rate": 2.8630078678402673e-07,
+      "loss": 0.4852,
+      "step": 7212
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.584016313678421,
+      "learning_rate": 2.856308680364472e-07,
+      "loss": 0.4692,
+      "step": 7213
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4918658623049326,
+      "learning_rate": 2.8496171093672255e-07,
+      "loss": 0.5201,
+      "step": 7214
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5396443968278153,
+      "learning_rate": 2.842933155929589e-07,
+      "loss": 0.4347,
+      "step": 7215
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7675236400198393,
+      "learning_rate": 2.8362568211314334e-07,
+      "loss": 0.4514,
+      "step": 7216
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.568592089505673,
+      "learning_rate": 2.8295881060513565e-07,
+      "loss": 0.5031,
+      "step": 7217
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4069262109055722,
+      "learning_rate": 2.8229270117667686e-07,
+      "loss": 0.5155,
+      "step": 7218
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.39470701848657,
+      "learning_rate": 2.8162735393538197e-07,
+      "loss": 0.5019,
+      "step": 7219
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4321517994755029,
+      "learning_rate": 2.8096276898874344e-07,
+      "loss": 0.4615,
+      "step": 7220
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6475657395982446,
+      "learning_rate": 2.802989464441319e-07,
+      "loss": 0.4951,
+      "step": 7221
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3483897836284766,
+      "learning_rate": 2.796358864087922e-07,
+      "loss": 0.4781,
+      "step": 7222
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2149006601586705,
+      "learning_rate": 2.789735889898493e-07,
+      "loss": 0.501,
+      "step": 7223
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5858220803089245,
+      "learning_rate": 2.7831205429430133e-07,
+      "loss": 0.5271,
+      "step": 7224
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7325956961464044,
+      "learning_rate": 2.776512824290256e-07,
+      "loss": 0.5268,
+      "step": 7225
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7756493090951462,
+      "learning_rate": 2.769912735007768e-07,
+      "loss": 0.5266,
+      "step": 7226
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9930583301822953,
+      "learning_rate": 2.763320276161835e-07,
+      "loss": 0.495,
+      "step": 7227
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8473930690073925,
+      "learning_rate": 2.7567354488175433e-07,
+      "loss": 0.5277,
+      "step": 7228
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.692660262075829,
+      "learning_rate": 2.750158254038715e-07,
+      "loss": 0.525,
+      "step": 7229
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3647834288816196,
+      "learning_rate": 2.7435886928879663e-07,
+      "loss": 0.4567,
+      "step": 7230
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1593877264595376,
+      "learning_rate": 2.737026766426654e-07,
+      "loss": 0.5127,
+      "step": 7231
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3555113213082137,
+      "learning_rate": 2.730472475714929e-07,
+      "loss": 0.4757,
+      "step": 7232
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4278182115249216,
+      "learning_rate": 2.723925821811685e-07,
+      "loss": 0.489,
+      "step": 7233
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8923024576982832,
+      "learning_rate": 2.717386805774591e-07,
+      "loss": 0.5259,
+      "step": 7234
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3922482113704302,
+      "learning_rate": 2.710855428660097e-07,
+      "loss": 0.4448,
+      "step": 7235
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.50956630158752,
+      "learning_rate": 2.7043316915233874e-07,
+      "loss": 0.497,
+      "step": 7236
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.295771368551603,
+      "learning_rate": 2.6978155954184403e-07,
+      "loss": 0.4751,
+      "step": 7237
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3927473634808414,
+      "learning_rate": 2.691307141397975e-07,
+      "loss": 0.4655,
+      "step": 7238
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4048674764987086,
+      "learning_rate": 2.684806330513512e-07,
+      "loss": 0.4892,
+      "step": 7239
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1409357943365657,
+      "learning_rate": 2.6783131638152883e-07,
+      "loss": 0.5432,
+      "step": 7240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.547539150362807,
+      "learning_rate": 2.6718276423523593e-07,
+      "loss": 0.489,
+      "step": 7241
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3301631048062446,
+      "learning_rate": 2.665349767172493e-07,
+      "loss": 0.5291,
+      "step": 7242
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.391096724620809,
+      "learning_rate": 2.6588795393222566e-07,
+      "loss": 0.4868,
+      "step": 7243
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7189858035690682,
+      "learning_rate": 2.6524169598469816e-07,
+      "loss": 0.5221,
+      "step": 7244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4065853979337697,
+      "learning_rate": 2.6459620297907386e-07,
+      "loss": 0.5053,
+      "step": 7245
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.589793540733917,
+      "learning_rate": 2.639514750196398e-07,
+      "loss": 0.5439,
+      "step": 7246
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.694368040776559,
+      "learning_rate": 2.633075122105555e-07,
+      "loss": 0.503,
+      "step": 7247
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3267658279966987,
+      "learning_rate": 2.626643146558605e-07,
+      "loss": 0.5074,
+      "step": 7248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.586372075700058,
+      "learning_rate": 2.6202188245946726e-07,
+      "loss": 0.4962,
+      "step": 7249
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3828261644604938,
+      "learning_rate": 2.6138021572516715e-07,
+      "loss": 0.4942,
+      "step": 7250
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4093827303195634,
+      "learning_rate": 2.607393145566284e-07,
+      "loss": 0.4875,
+      "step": 7251
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9936764361454529,
+      "learning_rate": 2.6009917905739203e-07,
+      "loss": 0.5106,
+      "step": 7252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.422134201103214,
+      "learning_rate": 2.594598093308792e-07,
+      "loss": 0.4947,
+      "step": 7253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2892958431167894,
+      "learning_rate": 2.588212054803846e-07,
+      "loss": 0.4645,
+      "step": 7254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.471846303409366,
+      "learning_rate": 2.5818336760908115e-07,
+      "loss": 0.5207,
+      "step": 7255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3972002061205235,
+      "learning_rate": 2.5754629582001644e-07,
+      "loss": 0.5183,
+      "step": 7256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4677880478954894,
+      "learning_rate": 2.569099902161165e-07,
+      "loss": 0.4941,
+      "step": 7257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.985567571670727,
+      "learning_rate": 2.5627445090017913e-07,
+      "loss": 0.4583,
+      "step": 7258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5235852187263925,
+      "learning_rate": 2.5563967797488487e-07,
+      "loss": 0.4655,
+      "step": 7259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.677128208654452,
+      "learning_rate": 2.5500567154278567e-07,
+      "loss": 0.5045,
+      "step": 7260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3204359303408124,
+      "learning_rate": 2.54372431706309e-07,
+      "loss": 0.5068,
+      "step": 7261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5599365732510986,
+      "learning_rate": 2.537399585677625e-07,
+      "loss": 0.5113,
+      "step": 7262
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4635723228362618,
+      "learning_rate": 2.531082522293266e-07,
+      "loss": 0.5076,
+      "step": 7263
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4179714784792024,
+      "learning_rate": 2.524773127930602e-07,
+      "loss": 0.4573,
+      "step": 7264
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6028209650618992,
+      "learning_rate": 2.518471403608952e-07,
+      "loss": 0.5354,
+      "step": 7265
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6732471951768726,
+      "learning_rate": 2.512177350346434e-07,
+      "loss": 0.5216,
+      "step": 7266
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.705772586239429,
+      "learning_rate": 2.5058909691598966e-07,
+      "loss": 0.5521,
+      "step": 7267
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3672127403570247,
+      "learning_rate": 2.499612261064971e-07,
+      "loss": 0.4545,
+      "step": 7268
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7575170328414065,
+      "learning_rate": 2.4933412270760295e-07,
+      "loss": 0.5077,
+      "step": 7269
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9858809443827865,
+      "learning_rate": 2.487077868206206e-07,
+      "loss": 0.5322,
+      "step": 7270
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4890330148958661,
+      "learning_rate": 2.4808221854674185e-07,
+      "loss": 0.493,
+      "step": 7271
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3164375874859087,
+      "learning_rate": 2.4745741798703094e-07,
+      "loss": 0.4836,
+      "step": 7272
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4383171376568136,
+      "learning_rate": 2.4683338524243204e-07,
+      "loss": 0.4913,
+      "step": 7273
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.578046977346187,
+      "learning_rate": 2.4621012041376013e-07,
+      "loss": 0.4647,
+      "step": 7274
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6640790823621678,
+      "learning_rate": 2.455876236017124e-07,
+      "loss": 0.4695,
+      "step": 7275
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5343616831889717,
+      "learning_rate": 2.4496589490685617e-07,
+      "loss": 0.4824,
+      "step": 7276
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.712511496470845,
+      "learning_rate": 2.4434493442963944e-07,
+      "loss": 0.5031,
+      "step": 7277
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2402210784873,
+      "learning_rate": 2.4372474227038145e-07,
+      "loss": 0.4437,
+      "step": 7278
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4835344448500538,
+      "learning_rate": 2.43105318529282e-07,
+      "loss": 0.4807,
+      "step": 7279
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6450460043573967,
+      "learning_rate": 2.424866633064127e-07,
+      "loss": 0.4643,
+      "step": 7280
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4435211933583696,
+      "learning_rate": 2.418687767017236e-07,
+      "loss": 0.5171,
+      "step": 7281
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4803283735920534,
+      "learning_rate": 2.412516588150393e-07,
+      "loss": 0.4445,
+      "step": 7282
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6756255272826075,
+      "learning_rate": 2.406353097460601e-07,
+      "loss": 0.5409,
+      "step": 7283
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4945053637205326,
+      "learning_rate": 2.400197295943646e-07,
+      "loss": 0.4947,
+      "step": 7284
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5668521984285715,
+      "learning_rate": 2.394049184594027e-07,
+      "loss": 0.4485,
+      "step": 7285
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6935285110259156,
+      "learning_rate": 2.3879087644050503e-07,
+      "loss": 0.5001,
+      "step": 7286
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.478287517350354,
+      "learning_rate": 2.3817760363687382e-07,
+      "loss": 0.5318,
+      "step": 7287
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5620851531817914,
+      "learning_rate": 2.3756510014758926e-07,
+      "loss": 0.5353,
+      "step": 7288
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5637832675029486,
+      "learning_rate": 2.3695336607160668e-07,
+      "loss": 0.4819,
+      "step": 7289
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6615999181013263,
+      "learning_rate": 2.3634240150775645e-07,
+      "loss": 0.5141,
+      "step": 7290
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0650508896497906,
+      "learning_rate": 2.3573220655474572e-07,
+      "loss": 0.5006,
+      "step": 7291
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5867828142828393,
+      "learning_rate": 2.351227813111573e-07,
+      "loss": 0.4795,
+      "step": 7292
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.677807090950925,
+      "learning_rate": 2.3451412587544908e-07,
+      "loss": 0.4779,
+      "step": 7293
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3721886541775619,
+      "learning_rate": 2.3390624034595356e-07,
+      "loss": 0.5185,
+      "step": 7294
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4681451485285608,
+      "learning_rate": 2.3329912482088213e-07,
+      "loss": 0.5293,
+      "step": 7295
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4934042079383747,
+      "learning_rate": 2.3269277939831748e-07,
+      "loss": 0.4729,
+      "step": 7296
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.146107040808323,
+      "learning_rate": 2.3208720417622177e-07,
+      "loss": 0.4677,
+      "step": 7297
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.306195733810811,
+      "learning_rate": 2.3148239925243067e-07,
+      "loss": 0.4468,
+      "step": 7298
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4923955412993348,
+      "learning_rate": 2.3087836472465319e-07,
+      "loss": 0.4738,
+      "step": 7299
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7773578519524955,
+      "learning_rate": 2.302751006904802e-07,
+      "loss": 0.5329,
+      "step": 7300
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4199512023896965,
+      "learning_rate": 2.296726072473726e-07,
+      "loss": 0.4481,
+      "step": 7301
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4258423801499063,
+      "learning_rate": 2.2907088449266867e-07,
+      "loss": 0.5177,
+      "step": 7302
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5231668326365035,
+      "learning_rate": 2.2846993252358119e-07,
+      "loss": 0.5138,
+      "step": 7303
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4672551656917383,
+      "learning_rate": 2.278697514372008e-07,
+      "loss": 0.4154,
+      "step": 7304
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6189360069251224,
+      "learning_rate": 2.2727034133049054e-07,
+      "loss": 0.5066,
+      "step": 7305
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1915601216281146,
+      "learning_rate": 2.2667170230029177e-07,
+      "loss": 0.532,
+      "step": 7306
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.593834724413796,
+      "learning_rate": 2.2607383444331888e-07,
+      "loss": 0.5509,
+      "step": 7307
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7054554094812038,
+      "learning_rate": 2.2547673785616285e-07,
+      "loss": 0.4572,
+      "step": 7308
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8211515336626856,
+      "learning_rate": 2.2488041263529048e-07,
+      "loss": 0.4482,
+      "step": 7309
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.648506058182298,
+      "learning_rate": 2.2428485887704244e-07,
+      "loss": 0.4783,
+      "step": 7310
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7043440299302586,
+      "learning_rate": 2.2369007667763676e-07,
+      "loss": 0.5503,
+      "step": 7311
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.493921056568484,
+      "learning_rate": 2.2309606613316438e-07,
+      "loss": 0.4892,
+      "step": 7312
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5789653486264432,
+      "learning_rate": 2.2250282733959462e-07,
+      "loss": 0.4903,
+      "step": 7313
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2578843844561818,
+      "learning_rate": 2.2191036039276914e-07,
+      "loss": 0.4967,
+      "step": 7314
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4787303637879343,
+      "learning_rate": 2.213186653884064e-07,
+      "loss": 0.554,
+      "step": 7315
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4311610016698448,
+      "learning_rate": 2.2072774242210048e-07,
+      "loss": 0.4968,
+      "step": 7316
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8183962667671558,
+      "learning_rate": 2.2013759158931947e-07,
+      "loss": 0.4779,
+      "step": 7317
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.241204665605121,
+      "learning_rate": 2.1954821298540873e-07,
+      "loss": 0.4723,
+      "step": 7318
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.584852717960534,
+      "learning_rate": 2.1895960670558548e-07,
+      "loss": 0.5204,
+      "step": 7319
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.254694152059909,
+      "learning_rate": 2.1837177284494637e-07,
+      "loss": 0.5354,
+      "step": 7320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5623354567276635,
+      "learning_rate": 2.1778471149845938e-07,
+      "loss": 0.5041,
+      "step": 7321
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.440466493552607,
+      "learning_rate": 2.171984227609708e-07,
+      "loss": 0.4777,
+      "step": 7322
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4591778128934314,
+      "learning_rate": 2.1661290672719993e-07,
+      "loss": 0.4227,
+      "step": 7323
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.335574212901139,
+      "learning_rate": 2.160281634917416e-07,
+      "loss": 0.452,
+      "step": 7324
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5346708718374296,
+      "learning_rate": 2.1544419314906805e-07,
+      "loss": 0.4654,
+      "step": 7325
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.117331882237975,
+      "learning_rate": 2.1486099579352327e-07,
+      "loss": 0.5138,
+      "step": 7326
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.476418156466487,
+      "learning_rate": 2.1427857151932851e-07,
+      "loss": 0.4655,
+      "step": 7327
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0906591898100957,
+      "learning_rate": 2.136969204205791e-07,
+      "loss": 0.5024,
+      "step": 7328
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.491630283610432,
+      "learning_rate": 2.1311604259124653e-07,
+      "loss": 0.4872,
+      "step": 7329
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5464114719601534,
+      "learning_rate": 2.125359381251757e-07,
+      "loss": 0.5307,
+      "step": 7330
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.408479255570862,
+      "learning_rate": 2.1195660711608944e-07,
+      "loss": 0.512,
+      "step": 7331
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.005446125069675,
+      "learning_rate": 2.113780496575818e-07,
+      "loss": 0.4561,
+      "step": 7332
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7105871509101782,
+      "learning_rate": 2.1080026584312407e-07,
+      "loss": 0.4906,
+      "step": 7333
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6551695201751928,
+      "learning_rate": 2.102232557660644e-07,
+      "loss": 0.4628,
+      "step": 7334
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.515246530641419,
+      "learning_rate": 2.0964701951962095e-07,
+      "loss": 0.4869,
+      "step": 7335
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3945580750786308,
+      "learning_rate": 2.0907155719689208e-07,
+      "loss": 0.4854,
+      "step": 7336
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4034244421481015,
+      "learning_rate": 2.0849686889084786e-07,
+      "loss": 0.459,
+      "step": 7337
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5868486583384376,
+      "learning_rate": 2.07922954694334e-07,
+      "loss": 0.4551,
+      "step": 7338
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6710749272297594,
+      "learning_rate": 2.0734981470007133e-07,
+      "loss": 0.5157,
+      "step": 7339
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.874129184801048,
+      "learning_rate": 2.0677744900065633e-07,
+      "loss": 0.5156,
+      "step": 7340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8234472645431714,
+      "learning_rate": 2.0620585768855949e-07,
+      "loss": 0.4908,
+      "step": 7341
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3063947819347423,
+      "learning_rate": 2.0563504085612583e-07,
+      "loss": 0.4893,
+      "step": 7342
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.418876188696213,
+      "learning_rate": 2.0506499859557717e-07,
+      "loss": 0.449,
+      "step": 7343
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4762691073800227,
+      "learning_rate": 2.04495730999007e-07,
+      "loss": 0.493,
+      "step": 7344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.554007050981706,
+      "learning_rate": 2.0392723815838734e-07,
+      "loss": 0.499,
+      "step": 7345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3702325490919527,
+      "learning_rate": 2.0335952016556193e-07,
+      "loss": 0.4782,
+      "step": 7346
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6724464755354924,
+      "learning_rate": 2.0279257711225186e-07,
+      "loss": 0.4829,
+      "step": 7347
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5667356213060892,
+      "learning_rate": 2.0222640909004942e-07,
+      "loss": 0.4719,
+      "step": 7348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3072445321244563,
+      "learning_rate": 2.0166101619042643e-07,
+      "loss": 0.4947,
+      "step": 7349
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2520941586893453,
+      "learning_rate": 2.0109639850472706e-07,
+      "loss": 0.4779,
+      "step": 7350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4133855463711558,
+      "learning_rate": 2.0053255612416832e-07,
+      "loss": 0.4619,
+      "step": 7351
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4728280235748008,
+      "learning_rate": 1.9996948913984626e-07,
+      "loss": 0.4631,
+      "step": 7352
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6765788312000822,
+      "learning_rate": 1.99407197642727e-07,
+      "loss": 0.4636,
+      "step": 7353
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9494597592574943,
+      "learning_rate": 1.988456817236556e-07,
+      "loss": 0.4529,
+      "step": 7354
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.443993759050577,
+      "learning_rate": 1.9828494147334843e-07,
+      "loss": 0.4866,
+      "step": 7355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5748048125426386,
+      "learning_rate": 1.977249769823991e-07,
+      "loss": 0.4593,
+      "step": 7356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3756291053600451,
+      "learning_rate": 1.9716578834127366e-07,
+      "loss": 0.5024,
+      "step": 7357
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.639060653594591,
+      "learning_rate": 1.966073756403153e-07,
+      "loss": 0.4482,
+      "step": 7358
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3952555902533088,
+      "learning_rate": 1.9604973896974022e-07,
+      "loss": 0.4986,
+      "step": 7359
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8896722139396491,
+      "learning_rate": 1.9549287841963915e-07,
+      "loss": 0.502,
+      "step": 7360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5602669134106195,
+      "learning_rate": 1.949367940799779e-07,
+      "loss": 0.5039,
+      "step": 7361
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7246978242341773,
+      "learning_rate": 1.9438148604059627e-07,
+      "loss": 0.4685,
+      "step": 7362
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2521584103521244,
+      "learning_rate": 1.9382695439121024e-07,
+      "loss": 0.4646,
+      "step": 7363
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9572783096995119,
+      "learning_rate": 1.9327319922140819e-07,
+      "loss": 0.5122,
+      "step": 7364
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3215317375129358,
+      "learning_rate": 1.9272022062065466e-07,
+      "loss": 0.4642,
+      "step": 7365
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3334449214336,
+      "learning_rate": 1.9216801867828817e-07,
+      "loss": 0.4742,
+      "step": 7366
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0968260435500023,
+      "learning_rate": 1.916165934835229e-07,
+      "loss": 0.456,
+      "step": 7367
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.390221874520958,
+      "learning_rate": 1.9106594512544485e-07,
+      "loss": 0.472,
+      "step": 7368
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3826685396985134,
+      "learning_rate": 1.9051607369301616e-07,
+      "loss": 0.495,
+      "step": 7369
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6508771302483614,
+      "learning_rate": 1.899669792750747e-07,
+      "loss": 0.5044,
+      "step": 7370
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5330385791261132,
+      "learning_rate": 1.8941866196032998e-07,
+      "loss": 0.4936,
+      "step": 7371
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7707144138711233,
+      "learning_rate": 1.88871121837369e-07,
+      "loss": 0.486,
+      "step": 7372
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.262247496461452,
+      "learning_rate": 1.8832435899464984e-07,
+      "loss": 0.4518,
+      "step": 7373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3873141734746803,
+      "learning_rate": 1.8777837352050908e-07,
+      "loss": 0.47,
+      "step": 7374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3693940975255388,
+      "learning_rate": 1.872331655031534e-07,
+      "loss": 0.5038,
+      "step": 7375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0349987280215167,
+      "learning_rate": 1.8668873503066786e-07,
+      "loss": 0.4869,
+      "step": 7376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6231415884536113,
+      "learning_rate": 1.861450821910088e-07,
+      "loss": 0.4838,
+      "step": 7377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6188759308500793,
+      "learning_rate": 1.856022070720076e-07,
+      "loss": 0.51,
+      "step": 7378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2550870322710346,
+      "learning_rate": 1.8506010976137244e-07,
+      "loss": 0.4649,
+      "step": 7379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3361841719564518,
+      "learning_rate": 1.8451879034668163e-07,
+      "loss": 0.4599,
+      "step": 7380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7603972626571474,
+      "learning_rate": 1.839782489153913e-07,
+      "loss": 0.5608,
+      "step": 7381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.786862186824463,
+      "learning_rate": 1.83438485554831e-07,
+      "loss": 0.5361,
+      "step": 7382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3992880337173184,
+      "learning_rate": 1.828995003522044e-07,
+      "loss": 0.5047,
+      "step": 7383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3833373236184248,
+      "learning_rate": 1.8236129339458787e-07,
+      "loss": 0.4584,
+      "step": 7384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6915427558700735,
+      "learning_rate": 1.8182386476893467e-07,
+      "loss": 0.4778,
+      "step": 7385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5895090237630853,
+      "learning_rate": 1.8128721456207088e-07,
+      "loss": 0.4951,
+      "step": 7386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7293079354326406,
+      "learning_rate": 1.8075134286069718e-07,
+      "loss": 0.4858,
+      "step": 7387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.465841554108314,
+      "learning_rate": 1.802162497513882e-07,
+      "loss": 0.4851,
+      "step": 7388
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6913645552862504,
+      "learning_rate": 1.7968193532059197e-07,
+      "loss": 0.4761,
+      "step": 7389
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3774108786510824,
+      "learning_rate": 1.7914839965463339e-07,
+      "loss": 0.4674,
+      "step": 7390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4307532756136354,
+      "learning_rate": 1.78615642839709e-07,
+      "loss": 0.4621,
+      "step": 7391
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4096128289356364,
+      "learning_rate": 1.780836649618911e-07,
+      "loss": 0.5172,
+      "step": 7392
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7973737870861728,
+      "learning_rate": 1.7755246610712372e-07,
+      "loss": 0.4674,
+      "step": 7393
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.388887248954646,
+      "learning_rate": 1.7702204636122878e-07,
+      "loss": 0.4694,
+      "step": 7394
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4194576297864852,
+      "learning_rate": 1.764924058098988e-07,
+      "loss": 0.5377,
+      "step": 7395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.466166965935639,
+      "learning_rate": 1.7596354453870257e-07,
+      "loss": 0.5266,
+      "step": 7396
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4755667902189378,
+      "learning_rate": 1.7543546263308176e-07,
+      "loss": 0.4729,
+      "step": 7397
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.2663679978132127,
+      "learning_rate": 1.74908160178352e-07,
+      "loss": 0.5259,
+      "step": 7398
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3924679571475973,
+      "learning_rate": 1.7438163725970624e-07,
+      "loss": 0.5001,
+      "step": 7399
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8073122795850416,
+      "learning_rate": 1.7385589396220592e-07,
+      "loss": 0.4704,
+      "step": 7400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3579610533590791,
+      "learning_rate": 1.7333093037079197e-07,
+      "loss": 0.4827,
+      "step": 7401
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4829654561982808,
+      "learning_rate": 1.7280674657027486e-07,
+      "loss": 0.4699,
+      "step": 7402
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4002345695803489,
+      "learning_rate": 1.7228334264534242e-07,
+      "loss": 0.4516,
+      "step": 7403
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.440620235561422,
+      "learning_rate": 1.7176071868055421e-07,
+      "loss": 0.4559,
+      "step": 7404
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1864763417305415,
+      "learning_rate": 1.7123887476034607e-07,
+      "loss": 0.5007,
+      "step": 7405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5826582958870667,
+      "learning_rate": 1.7071781096902497e-07,
+      "loss": 0.4817,
+      "step": 7406
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6371850257202158,
+      "learning_rate": 1.701975273907741e-07,
+      "loss": 0.5102,
+      "step": 7407
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6444450230778677,
+      "learning_rate": 1.6967802410965016e-07,
+      "loss": 0.4651,
+      "step": 7408
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6004728101518253,
+      "learning_rate": 1.6915930120958268e-07,
+      "loss": 0.4958,
+      "step": 7409
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7822773798269116,
+      "learning_rate": 1.6864135877437683e-07,
+      "loss": 0.4633,
+      "step": 7410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.435767134768869,
+      "learning_rate": 1.6812419688770953e-07,
+      "loss": 0.467,
+      "step": 7411
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.372733854208186,
+      "learning_rate": 1.6760781563313399e-07,
+      "loss": 0.4703,
+      "step": 7412
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.526168994526638,
+      "learning_rate": 1.670922150940757e-07,
+      "loss": 0.4873,
+      "step": 7413
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.426658070398611,
+      "learning_rate": 1.665773953538341e-07,
+      "loss": 0.5046,
+      "step": 7414
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7451714554121183,
+      "learning_rate": 1.6606335649558436e-07,
+      "loss": 0.4596,
+      "step": 7415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4002667303381664,
+      "learning_rate": 1.655500986023717e-07,
+      "loss": 0.4861,
+      "step": 7416
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.103839952829796,
+      "learning_rate": 1.650376217571198e-07,
+      "loss": 0.5538,
+      "step": 7417
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8123136742602968,
+      "learning_rate": 1.6452592604262185e-07,
+      "loss": 0.4742,
+      "step": 7418
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.532392411596531,
+      "learning_rate": 1.6401501154154786e-07,
+      "loss": 0.4765,
+      "step": 7419
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.232248589830049,
+      "learning_rate": 1.6350487833644012e-07,
+      "loss": 0.5166,
+      "step": 7420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6485974896546554,
+      "learning_rate": 1.629955265097155e-07,
+      "loss": 0.4424,
+      "step": 7421
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4977480346655385,
+      "learning_rate": 1.6248695614366427e-07,
+      "loss": 0.5384,
+      "step": 7422
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5735385964767112,
+      "learning_rate": 1.619791673204496e-07,
+      "loss": 0.5221,
+      "step": 7423
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5112925459666093,
+      "learning_rate": 1.6147216012211087e-07,
+      "loss": 0.5002,
+      "step": 7424
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9905644059573038,
+      "learning_rate": 1.6096593463055754e-07,
+      "loss": 0.4811,
+      "step": 7425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.881686217710671,
+      "learning_rate": 1.60460490927577e-07,
+      "loss": 0.5035,
+      "step": 7426
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.314687441900597,
+      "learning_rate": 1.599558290948261e-07,
+      "loss": 0.4807,
+      "step": 7427
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6786577910900888,
+      "learning_rate": 1.594519492138391e-07,
+      "loss": 0.459,
+      "step": 7428
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.398277622031906,
+      "learning_rate": 1.589488513660209e-07,
+      "loss": 0.5031,
+      "step": 7429
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.373663322700556,
+      "learning_rate": 1.58446535632652e-07,
+      "loss": 0.4752,
+      "step": 7430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5268937260525122,
+      "learning_rate": 1.579450020948864e-07,
+      "loss": 0.5015,
+      "step": 7431
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7816541926644356,
+      "learning_rate": 1.574442508337498e-07,
+      "loss": 0.4621,
+      "step": 7432
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3089321631636746,
+      "learning_rate": 1.5694428193014477e-07,
+      "loss": 0.4985,
+      "step": 7433
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4269437872264588,
+      "learning_rate": 1.5644509546484387e-07,
+      "loss": 0.5111,
+      "step": 7434
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6211926111031364,
+      "learning_rate": 1.5594669151849706e-07,
+      "loss": 0.52,
+      "step": 7435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9188201433192622,
+      "learning_rate": 1.5544907017162435e-07,
+      "loss": 0.5104,
+      "step": 7436
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4296650370385822,
+      "learning_rate": 1.5495223150462145e-07,
+      "loss": 0.5223,
+      "step": 7437
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.379650631045068,
+      "learning_rate": 1.5445617559775694e-07,
+      "loss": 0.5103,
+      "step": 7438
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5477816134993425,
+      "learning_rate": 1.5396090253117224e-07,
+      "loss": 0.4758,
+      "step": 7439
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3501455563275457,
+      "learning_rate": 1.53466412384885e-07,
+      "loss": 0.4398,
+      "step": 7440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4149457953921925,
+      "learning_rate": 1.529727052387825e-07,
+      "loss": 0.5128,
+      "step": 7441
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3173573220357797,
+      "learning_rate": 1.5247978117262918e-07,
+      "loss": 0.4253,
+      "step": 7442
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.574707130020846,
+      "learning_rate": 1.5198764026605918e-07,
+      "loss": 0.5004,
+      "step": 7443
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4846028133844213,
+      "learning_rate": 1.514962825985844e-07,
+      "loss": 0.5118,
+      "step": 7444
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.147310520687052,
+      "learning_rate": 1.5100570824958638e-07,
+      "loss": 0.498,
+      "step": 7445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3950785800094783,
+      "learning_rate": 1.5051591729832282e-07,
+      "loss": 0.4847,
+      "step": 7446
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.367432471953318,
+      "learning_rate": 1.5002690982392264e-07,
+      "loss": 0.5167,
+      "step": 7447
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1658022233578893,
+      "learning_rate": 1.4953868590538989e-07,
+      "loss": 0.4746,
+      "step": 7448
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7341710818649676,
+      "learning_rate": 1.49051245621602e-07,
+      "loss": 0.5009,
+      "step": 7449
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9017135541191659,
+      "learning_rate": 1.4856458905130823e-07,
+      "loss": 0.5082,
+      "step": 7450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4883425706331486,
+      "learning_rate": 1.480787162731334e-07,
+      "loss": 0.5104,
+      "step": 7451
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7310640824314047,
+      "learning_rate": 1.4759362736557313e-07,
+      "loss": 0.4518,
+      "step": 7452
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6820489174655378,
+      "learning_rate": 1.4710932240699915e-07,
+      "loss": 0.4706,
+      "step": 7453
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.835461807837606,
+      "learning_rate": 1.4662580147565386e-07,
+      "loss": 0.5171,
+      "step": 7454
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5063706588096515,
+      "learning_rate": 1.4614306464965533e-07,
+      "loss": 0.4654,
+      "step": 7455
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6698220323564079,
+      "learning_rate": 1.4566111200699396e-07,
+      "loss": 0.5236,
+      "step": 7456
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3900997845556018,
+      "learning_rate": 1.45179943625533e-07,
+      "loss": 0.5373,
+      "step": 7457
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5143660815460784,
+      "learning_rate": 1.4469955958300974e-07,
+      "loss": 0.4765,
+      "step": 7458
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.626916081300669,
+      "learning_rate": 1.442199599570343e-07,
+      "loss": 0.5145,
+      "step": 7459
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4632897847436472,
+      "learning_rate": 1.4374114482509028e-07,
+      "loss": 0.5348,
+      "step": 7460
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4279017685403526,
+      "learning_rate": 1.432631142645341e-07,
+      "loss": 0.4946,
+      "step": 7461
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2083577425437353,
+      "learning_rate": 1.4278586835259622e-07,
+      "loss": 0.458,
+      "step": 7462
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4784130586185524,
+      "learning_rate": 1.4230940716637943e-07,
+      "loss": 0.5431,
+      "step": 7463
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6445800312409866,
+      "learning_rate": 1.41833730782861e-07,
+      "loss": 0.539,
+      "step": 7464
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2848076635613273,
+      "learning_rate": 1.4135883927889006e-07,
+      "loss": 0.5047,
+      "step": 7465
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4212789763956692,
+      "learning_rate": 1.408847327311902e-07,
+      "loss": 0.5072,
+      "step": 7466
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5426409834820622,
+      "learning_rate": 1.4041141121635737e-07,
+      "loss": 0.5697,
+      "step": 7467
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.507637441455359,
+      "learning_rate": 1.399388748108599e-07,
+      "loss": 0.4793,
+      "step": 7468
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4319365540845996,
+      "learning_rate": 1.394671235910411e-07,
+      "loss": 0.5337,
+      "step": 7469
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5313856313039929,
+      "learning_rate": 1.389961576331156e-07,
+      "loss": 0.492,
+      "step": 7470
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.505239142027003,
+      "learning_rate": 1.3852597701317306e-07,
+      "loss": 0.4844,
+      "step": 7471
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2618369771812814,
+      "learning_rate": 1.380565818071744e-07,
+      "loss": 0.4765,
+      "step": 7472
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.497229469316913,
+      "learning_rate": 1.375879720909562e-07,
+      "loss": 0.4455,
+      "step": 7473
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5833775904859986,
+      "learning_rate": 1.3712014794022455e-07,
+      "loss": 0.4677,
+      "step": 7474
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.567828763742299,
+      "learning_rate": 1.366531094305623e-07,
+      "loss": 0.4709,
+      "step": 7475
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.514977473626188,
+      "learning_rate": 1.3618685663742248e-07,
+      "loss": 0.5325,
+      "step": 7476
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5067250409224675,
+      "learning_rate": 1.3572138963613258e-07,
+      "loss": 0.5067,
+      "step": 7477
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.087781436589823,
+      "learning_rate": 1.35256708501893e-07,
+      "loss": 0.4266,
+      "step": 7478
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.345971248358072,
+      "learning_rate": 1.3479281330977646e-07,
+      "loss": 0.4553,
+      "step": 7479
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.553935766243598,
+      "learning_rate": 1.3432970413472967e-07,
+      "loss": 0.5117,
+      "step": 7480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.414846608514627,
+      "learning_rate": 1.3386738105157281e-07,
+      "loss": 0.5022,
+      "step": 7481
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5417676396175475,
+      "learning_rate": 1.334058441349978e-07,
+      "loss": 0.4737,
+      "step": 7482
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5825303379028746,
+      "learning_rate": 1.3294509345956884e-07,
+      "loss": 0.4209,
+      "step": 7483
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4591388912703152,
+      "learning_rate": 1.3248512909972643e-07,
+      "loss": 0.4908,
+      "step": 7484
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8270115930764523,
+      "learning_rate": 1.3202595112977945e-07,
+      "loss": 0.4857,
+      "step": 7485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.430368141396785,
+      "learning_rate": 1.3156755962391464e-07,
+      "loss": 0.5005,
+      "step": 7486
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3706232110255132,
+      "learning_rate": 1.3110995465618725e-07,
+      "loss": 0.4999,
+      "step": 7487
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.2975987019146755,
+      "learning_rate": 1.3065313630052757e-07,
+      "loss": 0.5456,
+      "step": 7488
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.532224608280657,
+      "learning_rate": 1.3019710463073987e-07,
+      "loss": 0.4886,
+      "step": 7489
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5089578742075092,
+      "learning_rate": 1.2974185972049858e-07,
+      "loss": 0.4434,
+      "step": 7490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7338452161486364,
+      "learning_rate": 1.2928740164335375e-07,
+      "loss": 0.5407,
+      "step": 7491
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8783834907040973,
+      "learning_rate": 1.2883373047272663e-07,
+      "loss": 0.49,
+      "step": 7492
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6643538611499085,
+      "learning_rate": 1.2838084628191195e-07,
+      "loss": 0.4932,
+      "step": 7493
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.998442612275223,
+      "learning_rate": 1.2792874914407617e-07,
+      "loss": 0.466,
+      "step": 7494
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3993279711964997,
+      "learning_rate": 1.2747743913226086e-07,
+      "loss": 0.481,
+      "step": 7495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4127255593346921,
+      "learning_rate": 1.270269163193788e-07,
+      "loss": 0.4884,
+      "step": 7496
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1344092010920592,
+      "learning_rate": 1.2657718077821512e-07,
+      "loss": 0.5471,
+      "step": 7497
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6925433386356652,
+      "learning_rate": 1.2612823258142949e-07,
+      "loss": 0.4819,
+      "step": 7498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.444345067251979,
+      "learning_rate": 1.2568007180155328e-07,
+      "loss": 0.4577,
+      "step": 7499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8671287144526687,
+      "learning_rate": 1.2523269851099085e-07,
+      "loss": 0.51,
+      "step": 7500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5835192422472604,
+      "learning_rate": 1.2478611278201824e-07,
+      "loss": 0.4785,
+      "step": 7501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6868950034259951,
+      "learning_rate": 1.243403146867872e-07,
+      "loss": 0.4806,
+      "step": 7502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5403639562764544,
+      "learning_rate": 1.2389530429731844e-07,
+      "loss": 0.515,
+      "step": 7503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.434244700079374,
+      "learning_rate": 1.2345108168550836e-07,
+      "loss": 0.4656,
+      "step": 7504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.662107659471043,
+      "learning_rate": 1.2300764692312507e-07,
+      "loss": 0.5934,
+      "step": 7505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3295395274177795,
+      "learning_rate": 1.225650000818085e-07,
+      "loss": 0.4394,
+      "step": 7506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6382934586730635,
+      "learning_rate": 1.2212314123307368e-07,
+      "loss": 0.5056,
+      "step": 7507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2936263880044172,
+      "learning_rate": 1.2168207044830572e-07,
+      "loss": 0.4899,
+      "step": 7508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0376476751674737,
+      "learning_rate": 1.2124178779876373e-07,
+      "loss": 0.501,
+      "step": 7509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6932854553594543,
+      "learning_rate": 1.2080229335557858e-07,
+      "loss": 0.5038,
+      "step": 7510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3352372815688656,
+      "learning_rate": 1.2036358718975572e-07,
+      "loss": 0.4864,
+      "step": 7511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5051657873763646,
+      "learning_rate": 1.1992566937217066e-07,
+      "loss": 0.4696,
+      "step": 7512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3730238116729956,
+      "learning_rate": 1.1948853997357402e-07,
+      "loss": 0.5046,
+      "step": 7513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8058338907155638,
+      "learning_rate": 1.1905219906458765e-07,
+      "loss": 0.4973,
+      "step": 7514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5547767592050927,
+      "learning_rate": 1.1861664671570517e-07,
+      "loss": 0.4965,
+      "step": 7515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6235911597108474,
+      "learning_rate": 1.1818188299729583e-07,
+      "loss": 0.4996,
+      "step": 7516
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.452063347559451,
+      "learning_rate": 1.177479079795979e-07,
+      "loss": 0.4445,
+      "step": 7517
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7283130334264376,
+      "learning_rate": 1.1731472173272529e-07,
+      "loss": 0.5076,
+      "step": 7518
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.700430020011993,
+      "learning_rate": 1.1688232432666147e-07,
+      "loss": 0.5086,
+      "step": 7519
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1730840757222,
+      "learning_rate": 1.1645071583126499e-07,
+      "loss": 0.4975,
+      "step": 7520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.403025967928819,
+      "learning_rate": 1.1601989631626565e-07,
+      "loss": 0.4639,
+      "step": 7521
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6772365589825002,
+      "learning_rate": 1.1558986585126608e-07,
+      "loss": 0.4967,
+      "step": 7522
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4835936516919237,
+      "learning_rate": 1.1516062450574239e-07,
+      "loss": 0.4588,
+      "step": 7523
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6819121439101569,
+      "learning_rate": 1.1473217234904133e-07,
+      "loss": 0.4629,
+      "step": 7524
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4333015642586124,
+      "learning_rate": 1.1430450945038363e-07,
+      "loss": 0.4882,
+      "step": 7525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8930763894354676,
+      "learning_rate": 1.1387763587886181e-07,
+      "loss": 0.493,
+      "step": 7526
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8009081752597993,
+      "learning_rate": 1.1345155170344124e-07,
+      "loss": 0.5328,
+      "step": 7527
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3986294432791426,
+      "learning_rate": 1.1302625699295855e-07,
+      "loss": 0.4201,
+      "step": 7528
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4873467528400124,
+      "learning_rate": 1.1260175181612488e-07,
+      "loss": 0.5264,
+      "step": 7529
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5485362346370584,
+      "learning_rate": 1.1217803624152312e-07,
+      "loss": 0.4878,
+      "step": 7530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.542461414453805,
+      "learning_rate": 1.1175511033760688e-07,
+      "loss": 0.4484,
+      "step": 7531
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4454517583014335,
+      "learning_rate": 1.1133297417270539e-07,
+      "loss": 0.5089,
+      "step": 7532
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6930035810814124,
+      "learning_rate": 1.1091162781501685e-07,
+      "loss": 0.5337,
+      "step": 7533
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.248049538526929,
+      "learning_rate": 1.104910713326146e-07,
+      "loss": 0.4951,
+      "step": 7534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8676919633357776,
+      "learning_rate": 1.1007130479344208e-07,
+      "loss": 0.4885,
+      "step": 7535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5895304874985916,
+      "learning_rate": 1.0965232826531725e-07,
+      "loss": 0.5174,
+      "step": 7536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7402560764830721,
+      "learning_rate": 1.0923414181592873e-07,
+      "loss": 0.498,
+      "step": 7537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4436156231144772,
+      "learning_rate": 1.088167455128386e-07,
+      "loss": 0.4194,
+      "step": 7538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5020695344067279,
+      "learning_rate": 1.0840013942348182e-07,
+      "loss": 0.4603,
+      "step": 7539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.905688093155993,
+      "learning_rate": 1.0798432361516287e-07,
+      "loss": 0.4891,
+      "step": 7540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4016492508769964,
+      "learning_rate": 1.075692981550619e-07,
+      "loss": 0.4704,
+      "step": 7541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3829597401522613,
+      "learning_rate": 1.0715506311022972e-07,
+      "loss": 0.486,
+      "step": 7542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5601454546027733,
+      "learning_rate": 1.0674161854758947e-07,
+      "loss": 0.5033,
+      "step": 7543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.342026448883406,
+      "learning_rate": 1.0632896453393605e-07,
+      "loss": 0.481,
+      "step": 7544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4630019635397487,
+      "learning_rate": 1.0591710113593834e-07,
+      "loss": 0.4689,
+      "step": 7545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6749586920383936,
+      "learning_rate": 1.0550602842013647e-07,
+      "loss": 0.4868,
+      "step": 7546
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4416026970217426,
+      "learning_rate": 1.0509574645294173e-07,
+      "loss": 0.5184,
+      "step": 7547
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9095201673801285,
+      "learning_rate": 1.0468625530064058e-07,
+      "loss": 0.5482,
+      "step": 7548
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4417897196789546,
+      "learning_rate": 1.042775550293884e-07,
+      "loss": 0.5059,
+      "step": 7549
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.012208951779018,
+      "learning_rate": 1.0386964570521574e-07,
+      "loss": 0.5351,
+      "step": 7550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.37239360342099,
+      "learning_rate": 1.0346252739402207e-07,
+      "loss": 0.4915,
+      "step": 7551
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1832857772124155,
+      "learning_rate": 1.0305620016158258e-07,
+      "loss": 0.469,
+      "step": 7552
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4342619421717173,
+      "learning_rate": 1.0265066407354196e-07,
+      "loss": 0.4912,
+      "step": 7553
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9992131575406364,
+      "learning_rate": 1.0224591919541837e-07,
+      "loss": 0.5002,
+      "step": 7554
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3660580834026919,
+      "learning_rate": 1.0184196559260229e-07,
+      "loss": 0.4791,
+      "step": 7555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.567805960522422,
+      "learning_rate": 1.0143880333035594e-07,
+      "loss": 0.4781,
+      "step": 7556
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2868250415977684,
+      "learning_rate": 1.010364324738139e-07,
+      "loss": 0.456,
+      "step": 7557
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4103965262983116,
+      "learning_rate": 1.0063485308798193e-07,
+      "loss": 0.4559,
+      "step": 7558
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.629549103806848,
+      "learning_rate": 1.0023406523773981e-07,
+      "loss": 0.4718,
+      "step": 7559
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5189480072759118,
+      "learning_rate": 9.983406898783688e-08,
+      "loss": 0.4718,
+      "step": 7560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.394972304534396,
+      "learning_rate": 9.943486440289751e-08,
+      "loss": 0.4669,
+      "step": 7561
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5152927943746088,
+      "learning_rate": 9.90364515474157e-08,
+      "loss": 0.4931,
+      "step": 7562
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4183916894876742,
+      "learning_rate": 9.86388304857594e-08,
+      "loss": 0.4588,
+      "step": 7563
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6619140841376372,
+      "learning_rate": 9.824200128216665e-08,
+      "loss": 0.4911,
+      "step": 7564
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4373377755286527,
+      "learning_rate": 9.784596400075063e-08,
+      "loss": 0.4961,
+      "step": 7565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3645511041619751,
+      "learning_rate": 9.745071870549295e-08,
+      "loss": 0.489,
+      "step": 7566
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3053146652099252,
+      "learning_rate": 9.705626546024916e-08,
+      "loss": 0.4641,
+      "step": 7567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6799087105795374,
+      "learning_rate": 9.666260432874719e-08,
+      "loss": 0.4837,
+      "step": 7568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3840190159719987,
+      "learning_rate": 9.626973537458561e-08,
+      "loss": 0.4886,
+      "step": 7569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7803774263684327,
+      "learning_rate": 9.5877658661237e-08,
+      "loss": 0.5038,
+      "step": 7570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5466790610753376,
+      "learning_rate": 9.548637425204399e-08,
+      "loss": 0.4751,
+      "step": 7571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.207045566540985,
+      "learning_rate": 9.50958822102227e-08,
+      "loss": 0.5139,
+      "step": 7572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0388938387697997,
+      "learning_rate": 9.470618259885989e-08,
+      "loss": 0.5066,
+      "step": 7573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6693846347721168,
+      "learning_rate": 9.431727548091574e-08,
+      "loss": 0.494,
+      "step": 7574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5269826421316284,
+      "learning_rate": 9.392916091922111e-08,
+      "loss": 0.5288,
+      "step": 7575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.011778871310886,
+      "learning_rate": 9.354183897647917e-08,
+      "loss": 0.4874,
+      "step": 7576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5042801955481406,
+      "learning_rate": 9.315530971526543e-08,
+      "loss": 0.5289,
+      "step": 7577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6520762442159772,
+      "learning_rate": 9.276957319802604e-08,
+      "loss": 0.4724,
+      "step": 7578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7630841557824317,
+      "learning_rate": 9.238462948708227e-08,
+      "loss": 0.5152,
+      "step": 7579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3418160264322265,
+      "learning_rate": 9.20004786446238e-08,
+      "loss": 0.4603,
+      "step": 7580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6097567059166966,
+      "learning_rate": 9.161712073271379e-08,
+      "loss": 0.4397,
+      "step": 7581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4953389541248439,
+      "learning_rate": 9.123455581328711e-08,
+      "loss": 0.5223,
+      "step": 7582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2689736383401649,
+      "learning_rate": 9.085278394815045e-08,
+      "loss": 0.441,
+      "step": 7583
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.752001103494819,
+      "learning_rate": 9.04718051989828e-08,
+      "loss": 0.4691,
+      "step": 7584
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 17.642591161965804,
+      "learning_rate": 9.009161962733327e-08,
+      "loss": 0.4844,
+      "step": 7585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.482388634847006,
+      "learning_rate": 8.971222729462603e-08,
+      "loss": 0.4791,
+      "step": 7586
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3991761059629115,
+      "learning_rate": 8.933362826215374e-08,
+      "loss": 0.4641,
+      "step": 7587
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.344337928641865,
+      "learning_rate": 8.895582259108415e-08,
+      "loss": 0.4699,
+      "step": 7588
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6337283293116154,
+      "learning_rate": 8.857881034245285e-08,
+      "loss": 0.5037,
+      "step": 7589
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5404316087887595,
+      "learning_rate": 8.82025915771717e-08,
+      "loss": 0.5291,
+      "step": 7590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4641869888308263,
+      "learning_rate": 8.78271663560204e-08,
+      "loss": 0.4885,
+      "step": 7591
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9766064882640204,
+      "learning_rate": 8.745253473965376e-08,
+      "loss": 0.5273,
+      "step": 7592
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3515860790010006,
+      "learning_rate": 8.707869678859504e-08,
+      "loss": 0.4739,
+      "step": 7593
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7362237175457764,
+      "learning_rate": 8.67056525632426e-08,
+      "loss": 0.5172,
+      "step": 7594
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3652520626139257,
+      "learning_rate": 8.633340212386431e-08,
+      "loss": 0.4888,
+      "step": 7595
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.562635416592457,
+      "learning_rate": 8.596194553060044e-08,
+      "loss": 0.4362,
+      "step": 7596
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6489051034073675,
+      "learning_rate": 8.559128284346352e-08,
+      "loss": 0.4773,
+      "step": 7597
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3462427241427373,
+      "learning_rate": 8.522141412233676e-08,
+      "loss": 0.4857,
+      "step": 7598
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4627023788707003,
+      "learning_rate": 8.485233942697624e-08,
+      "loss": 0.4767,
+      "step": 7599
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6271714374623194,
+      "learning_rate": 8.448405881700927e-08,
+      "loss": 0.4635,
+      "step": 7600
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2335539334688235,
+      "learning_rate": 8.411657235193438e-08,
+      "loss": 0.4921,
+      "step": 7601
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.406101675022231,
+      "learning_rate": 8.37498800911224e-08,
+      "loss": 0.4775,
+      "step": 7602
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4206446049445998,
+      "learning_rate": 8.338398209381537e-08,
+      "loss": 0.5124,
+      "step": 7603
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4878336030354353,
+      "learning_rate": 8.301887841912881e-08,
+      "loss": 0.4836,
+      "step": 7604
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.389099926530956,
+      "learning_rate": 8.265456912604608e-08,
+      "loss": 0.535,
+      "step": 7605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6773820458682531,
+      "learning_rate": 8.229105427342676e-08,
+      "loss": 0.4777,
+      "step": 7606
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.663011926443581,
+      "learning_rate": 8.192833391999833e-08,
+      "loss": 0.497,
+      "step": 7607
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4798884001272063,
+      "learning_rate": 8.156640812436278e-08,
+      "loss": 0.5235,
+      "step": 7608
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.610926579403922,
+      "learning_rate": 8.120527694499114e-08,
+      "loss": 0.4755,
+      "step": 7609
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4878269166594444,
+      "learning_rate": 8.08449404402284e-08,
+      "loss": 0.4835,
+      "step": 7610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5222386394029344,
+      "learning_rate": 8.048539866828909e-08,
+      "loss": 0.5179,
+      "step": 7611
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6431193155061984,
+      "learning_rate": 8.012665168726063e-08,
+      "loss": 0.4471,
+      "step": 7612
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4935441485253522,
+      "learning_rate": 7.976869955510225e-08,
+      "loss": 0.5175,
+      "step": 7613
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3801497191756107,
+      "learning_rate": 7.941154232964376e-08,
+      "loss": 0.4915,
+      "step": 7614
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6613682876564198,
+      "learning_rate": 7.905518006858792e-08,
+      "loss": 0.5026,
+      "step": 7615
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.811683475249906,
+      "learning_rate": 7.8699612829507e-08,
+      "loss": 0.4692,
+      "step": 7616
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3027553272100894,
+      "learning_rate": 7.834484066984727e-08,
+      "loss": 0.5004,
+      "step": 7617
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9372547346821705,
+      "learning_rate": 7.799086364692343e-08,
+      "loss": 0.4583,
+      "step": 7618
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4098518044530997,
+      "learning_rate": 7.76376818179253e-08,
+      "loss": 0.4473,
+      "step": 7619
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3656186635555496,
+      "learning_rate": 7.728529523991224e-08,
+      "loss": 0.4934,
+      "step": 7620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2652488484325881,
+      "learning_rate": 7.693370396981481e-08,
+      "loss": 0.4483,
+      "step": 7621
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5197178165179743,
+      "learning_rate": 7.658290806443647e-08,
+      "loss": 0.4688,
+      "step": 7622
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7503132976742892,
+      "learning_rate": 7.623290758045021e-08,
+      "loss": 0.5441,
+      "step": 7623
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4192647679144745,
+      "learning_rate": 7.588370257440303e-08,
+      "loss": 0.4673,
+      "step": 7624
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.931190050772453,
+      "learning_rate": 7.553529310271147e-08,
+      "loss": 0.4878,
+      "step": 7625
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.440480842373573,
+      "learning_rate": 7.518767922166381e-08,
+      "loss": 0.5096,
+      "step": 7626
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4898492224534166,
+      "learning_rate": 7.484086098742016e-08,
+      "loss": 0.4761,
+      "step": 7627
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6981654852220975,
+      "learning_rate": 7.449483845601291e-08,
+      "loss": 0.5055,
+      "step": 7628
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2724001278107187,
+      "learning_rate": 7.4149611683344e-08,
+      "loss": 0.4587,
+      "step": 7629
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5613833744236691,
+      "learning_rate": 7.380518072518883e-08,
+      "loss": 0.4533,
+      "step": 7630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6580808902507445,
+      "learning_rate": 7.346154563719232e-08,
+      "loss": 0.4535,
+      "step": 7631
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6613063750238234,
+      "learning_rate": 7.311870647487229e-08,
+      "loss": 0.4727,
+      "step": 7632
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.014698388336979,
+      "learning_rate": 7.277666329361776e-08,
+      "loss": 0.5082,
+      "step": 7633
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2342182090689713,
+      "learning_rate": 7.243541614868787e-08,
+      "loss": 0.4175,
+      "step": 7634
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8184782840077198,
+      "learning_rate": 7.209496509521519e-08,
+      "loss": 0.5313,
+      "step": 7635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4133824416213492,
+      "learning_rate": 7.175531018820125e-08,
+      "loss": 0.5306,
+      "step": 7636
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6655108699828332,
+      "learning_rate": 7.141645148252107e-08,
+      "loss": 0.4792,
+      "step": 7637
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8164807101058542,
+      "learning_rate": 7.107838903292085e-08,
+      "loss": 0.4482,
+      "step": 7638
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1941787148976293,
+      "learning_rate": 7.074112289401636e-08,
+      "loss": 0.4537,
+      "step": 7639
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.378549116427674,
+      "learning_rate": 7.040465312029731e-08,
+      "loss": 0.4766,
+      "step": 7640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1018978404214064,
+      "learning_rate": 7.006897976612192e-08,
+      "loss": 0.4628,
+      "step": 7641
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.44923619105166,
+      "learning_rate": 6.973410288572179e-08,
+      "loss": 0.4854,
+      "step": 7642
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9691991774920328,
+      "learning_rate": 6.940002253319978e-08,
+      "loss": 0.5281,
+      "step": 7643
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.537616321063936,
+      "learning_rate": 6.906673876252822e-08,
+      "loss": 0.5314,
+      "step": 7644
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7123009207082152,
+      "learning_rate": 6.873425162755354e-08,
+      "loss": 0.5079,
+      "step": 7645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.118658183151667,
+      "learning_rate": 6.840256118199051e-08,
+      "loss": 0.4906,
+      "step": 7646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6863783866001274,
+      "learning_rate": 6.807166747942795e-08,
+      "loss": 0.4469,
+      "step": 7647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.524207423173094,
+      "learning_rate": 6.774157057332365e-08,
+      "loss": 0.488,
+      "step": 7648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3968831983784278,
+      "learning_rate": 6.741227051700827e-08,
+      "loss": 0.4683,
+      "step": 7649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.632796678065021,
+      "learning_rate": 6.70837673636826e-08,
+      "loss": 0.4381,
+      "step": 7650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2330369250788322,
+      "learning_rate": 6.675606116642031e-08,
+      "loss": 0.4289,
+      "step": 7651
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5967975832790167,
+      "learning_rate": 6.642915197816347e-08,
+      "loss": 0.4778,
+      "step": 7652
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3934013360644446,
+      "learning_rate": 6.610303985172873e-08,
+      "loss": 0.4686,
+      "step": 7653
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1985070265059699,
+      "learning_rate": 6.577772483980228e-08,
+      "loss": 0.4303,
+      "step": 7654
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4211373045289915,
+      "learning_rate": 6.54532069949404e-08,
+      "loss": 0.4834,
+      "step": 7655
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6044525557085323,
+      "learning_rate": 6.512948636957284e-08,
+      "loss": 0.5273,
+      "step": 7656
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2829303434544845,
+      "learning_rate": 6.480656301599886e-08,
+      "loss": 0.4323,
+      "step": 7657
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6705878337304934,
+      "learning_rate": 6.448443698639062e-08,
+      "loss": 0.4911,
+      "step": 7658
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.310649686550002,
+      "learning_rate": 6.416310833278872e-08,
+      "loss": 0.4296,
+      "step": 7659
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4713096666729257,
+      "learning_rate": 6.384257710710828e-08,
+      "loss": 0.4754,
+      "step": 7660
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0267556929527535,
+      "learning_rate": 6.35228433611329e-08,
+      "loss": 0.5011,
+      "step": 7661
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7116531090517827,
+      "learning_rate": 6.320390714651958e-08,
+      "loss": 0.4999,
+      "step": 7662
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0886268269934094,
+      "learning_rate": 6.288576851479378e-08,
+      "loss": 0.506,
+      "step": 7663
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2864415516373995,
+      "learning_rate": 6.256842751735492e-08,
+      "loss": 0.4921,
+      "step": 7664
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6822313096347454,
+      "learning_rate": 6.22518842054709e-08,
+      "loss": 0.471,
+      "step": 7665
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5822354547473338,
+      "learning_rate": 6.193613863028303e-08,
+      "loss": 0.5187,
+      "step": 7666
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2247180484116207,
+      "learning_rate": 6.162119084280271e-08,
+      "loss": 0.4228,
+      "step": 7667
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.932817199646369,
+      "learning_rate": 6.130704089391259e-08,
+      "loss": 0.535,
+      "step": 7668
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4551478562969469,
+      "learning_rate": 6.09936888343654e-08,
+      "loss": 0.5047,
+      "step": 7669
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5915916781050325,
+      "learning_rate": 6.068113471478676e-08,
+      "loss": 0.5171,
+      "step": 7670
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.898586512014985,
+      "learning_rate": 6.036937858567294e-08,
+      "loss": 0.4807,
+      "step": 7671
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.796050658097426,
+      "learning_rate": 6.00584204973903e-08,
+      "loss": 0.513,
+      "step": 7672
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5586630964181434,
+      "learning_rate": 5.974826050017701e-08,
+      "loss": 0.4912,
+      "step": 7673
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6303241842222034,
+      "learning_rate": 5.943889864414243e-08,
+      "loss": 0.508,
+      "step": 7674
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3696136170171038,
+      "learning_rate": 5.913033497926546e-08,
+      "loss": 0.4956,
+      "step": 7675
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1958120852784795,
+      "learning_rate": 5.8822569555399e-08,
+      "loss": 0.4919,
+      "step": 7676
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3777355475873272,
+      "learning_rate": 5.8515602422263287e-08,
+      "loss": 0.495,
+      "step": 7677
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.391424721471707,
+      "learning_rate": 5.820943362945364e-08,
+      "loss": 0.4778,
+      "step": 7678
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3233746300787803,
+      "learning_rate": 5.790406322643327e-08,
+      "loss": 0.4705,
+      "step": 7679
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7454553079953834,
+      "learning_rate": 5.759949126253772e-08,
+      "loss": 0.559,
+      "step": 7680
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8206013531681495,
+      "learning_rate": 5.7295717786972625e-08,
+      "loss": 0.4776,
+      "step": 7681
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6651423869229104,
+      "learning_rate": 5.6992742848815955e-08,
+      "loss": 0.5239,
+      "step": 7682
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5163882063428316,
+      "learning_rate": 5.669056649701632e-08,
+      "loss": 0.4811,
+      "step": 7683
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6129136935060087,
+      "learning_rate": 5.63891887803919e-08,
+      "loss": 0.5292,
+      "step": 7684
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4545287345230198,
+      "learning_rate": 5.608860974763319e-08,
+      "loss": 0.46,
+      "step": 7685
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.765077600660373,
+      "learning_rate": 5.5788829447301906e-08,
+      "loss": 0.5191,
+      "step": 7686
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4372217535237481,
+      "learning_rate": 5.548984792783041e-08,
+      "loss": 0.4949,
+      "step": 7687
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3253862390810809,
+      "learning_rate": 5.519166523752117e-08,
+      "loss": 0.5215,
+      "step": 7688
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7964409105865975,
+      "learning_rate": 5.4894281424548446e-08,
+      "loss": 0.4757,
+      "step": 7689
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3087364335845009,
+      "learning_rate": 5.459769653695657e-08,
+      "loss": 0.4752,
+      "step": 7690
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.627844249158136,
+      "learning_rate": 5.430191062266277e-08,
+      "loss": 0.4828,
+      "step": 7691
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9735770293227999,
+      "learning_rate": 5.400692372945271e-08,
+      "loss": 0.5109,
+      "step": 7692
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4459533226148706,
+      "learning_rate": 5.371273590498438e-08,
+      "loss": 0.4588,
+      "step": 7693
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.704100185893509,
+      "learning_rate": 5.341934719678699e-08,
+      "loss": 0.5431,
+      "step": 7694
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7613191834669115,
+      "learning_rate": 5.312675765225928e-08,
+      "loss": 0.4894,
+      "step": 7695
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.259121277853792,
+      "learning_rate": 5.283496731867288e-08,
+      "loss": 0.5277,
+      "step": 7696
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.344515438831487,
+      "learning_rate": 5.254397624316731e-08,
+      "loss": 0.4692,
+      "step": 7697
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.458748155261817,
+      "learning_rate": 5.2253784472756084e-08,
+      "loss": 0.5377,
+      "step": 7698
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.480855482106529,
+      "learning_rate": 5.1964392054321686e-08,
+      "loss": 0.4587,
+      "step": 7699
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.33662859192495,
+      "learning_rate": 5.167579903461839e-08,
+      "loss": 0.5017,
+      "step": 7700
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5791003334365201,
+      "learning_rate": 5.1388005460270565e-08,
+      "loss": 0.5071,
+      "step": 7701
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4395511057076358,
+      "learning_rate": 5.110101137777379e-08,
+      "loss": 0.5164,
+      "step": 7702
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2375084850602827,
+      "learning_rate": 5.0814816833494876e-08,
+      "loss": 0.4812,
+      "step": 7703
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3507590154257363,
+      "learning_rate": 5.052942187367016e-08,
+      "loss": 0.4899,
+      "step": 7704
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.315754043115813,
+      "learning_rate": 5.024482654440943e-08,
+      "loss": 0.4567,
+      "step": 7705
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.495124890904158,
+      "learning_rate": 4.99610308916898e-08,
+      "loss": 0.4682,
+      "step": 7706
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3499923101809124,
+      "learning_rate": 4.967803496136181e-08,
+      "loss": 0.4835,
+      "step": 7707
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4715506582456879,
+      "learning_rate": 4.9395838799146114e-08,
+      "loss": 0.4866,
+      "step": 7708
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5842712418136582,
+      "learning_rate": 4.911444245063346e-08,
+      "loss": 0.5489,
+      "step": 7709
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8163171475763948,
+      "learning_rate": 4.8833845961286375e-08,
+      "loss": 0.4716,
+      "step": 7710
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3644745215747869,
+      "learning_rate": 4.855404937643693e-08,
+      "loss": 0.4827,
+      "step": 7711
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4313206438200663,
+      "learning_rate": 4.827505274128952e-08,
+      "loss": 0.4813,
+      "step": 7712
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.315013732174951,
+      "learning_rate": 4.799685610091809e-08,
+      "loss": 0.5051,
+      "step": 7713
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4328290676809463,
+      "learning_rate": 4.771945950026835e-08,
+      "loss": 0.4569,
+      "step": 7714
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6288798079160522,
+      "learning_rate": 4.744286298415557e-08,
+      "loss": 0.4907,
+      "step": 7715
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7069136478926474,
+      "learning_rate": 4.7167066597266776e-08,
+      "loss": 0.4865,
+      "step": 7716
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.308945493190629,
+      "learning_rate": 4.689207038415799e-08,
+      "loss": 0.4076,
+      "step": 7717
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4611682125753729,
+      "learning_rate": 4.6617874389259246e-08,
+      "loss": 0.5038,
+      "step": 7718
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.724751515685758,
+      "learning_rate": 4.634447865686842e-08,
+      "loss": 0.4707,
+      "step": 7719
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3998749280496665,
+      "learning_rate": 4.6071883231154077e-08,
+      "loss": 0.4931,
+      "step": 7720
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3858879631903585,
+      "learning_rate": 4.5800088156158215e-08,
+      "loss": 0.4985,
+      "step": 7721
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.467848420192936,
+      "learning_rate": 4.5529093475790156e-08,
+      "loss": 0.5039,
+      "step": 7722
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5766696103475495,
+      "learning_rate": 4.525889923383264e-08,
+      "loss": 0.4582,
+      "step": 7723
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.330634533354273,
+      "learning_rate": 4.498950547393743e-08,
+      "loss": 0.4371,
+      "step": 7724
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.841870302597123,
+      "learning_rate": 4.472091223962749e-08,
+      "loss": 0.469,
+      "step": 7725
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0793156583564767,
+      "learning_rate": 4.445311957429588e-08,
+      "loss": 0.4845,
+      "step": 7726
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3752340510642325,
+      "learning_rate": 4.418612752120743e-08,
+      "loss": 0.5401,
+      "step": 7727
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6006130720471534,
+      "learning_rate": 4.3919936123497654e-08,
+      "loss": 0.4821,
+      "step": 7728
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3840103206152756,
+      "learning_rate": 4.365454542417047e-08,
+      "loss": 0.5033,
+      "step": 7729
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4145112943608362,
+      "learning_rate": 4.3389955466103804e-08,
+      "loss": 0.4687,
+      "step": 7730
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4189671569908613,
+      "learning_rate": 4.312616629204347e-08,
+      "loss": 0.478,
+      "step": 7731
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.503259140881166,
+      "learning_rate": 4.2863177944607594e-08,
+      "loss": 0.4842,
+      "step": 7732
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2896560391253253,
+      "learning_rate": 4.26009904662833e-08,
+      "loss": 0.4583,
+      "step": 7733
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4580249324439911,
+      "learning_rate": 4.233960389943004e-08,
+      "loss": 0.4627,
+      "step": 7734
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3594047511001994,
+      "learning_rate": 4.2079018286277365e-08,
+      "loss": 0.4613,
+      "step": 7735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0801452303061256,
+      "learning_rate": 4.1819233668924375e-08,
+      "loss": 0.4604,
+      "step": 7736
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2787454071154456,
+      "learning_rate": 4.156025008934195e-08,
+      "loss": 0.4782,
+      "step": 7737
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6044864170581877,
+      "learning_rate": 4.130206758937105e-08,
+      "loss": 0.4766,
+      "step": 7738
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5826720176373226,
+      "learning_rate": 4.1044686210723884e-08,
+      "loss": 0.4529,
+      "step": 7739
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5454458656320416,
+      "learning_rate": 4.078810599498162e-08,
+      "loss": 0.4545,
+      "step": 7740
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8520856327740807,
+      "learning_rate": 4.053232698359832e-08,
+      "loss": 0.5048,
+      "step": 7741
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8044521412502397,
+      "learning_rate": 4.027734921789594e-08,
+      "loss": 0.5048,
+      "step": 7742
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3926948369663805,
+      "learning_rate": 4.002317273906986e-08,
+      "loss": 0.4797,
+      "step": 7743
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5154880391926469,
+      "learning_rate": 3.976979758818389e-08,
+      "loss": 0.4929,
+      "step": 7744
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.227339551797787,
+      "learning_rate": 3.9517223806171956e-08,
+      "loss": 0.4407,
+      "step": 7745
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5725278008193808,
+      "learning_rate": 3.926545143384142e-08,
+      "loss": 0.4647,
+      "step": 7746
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8648068300491099,
+      "learning_rate": 3.901448051186696e-08,
+      "loss": 0.5158,
+      "step": 7747
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.203877062516585,
+      "learning_rate": 3.876431108079615e-08,
+      "loss": 0.4707,
+      "step": 7748
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6137307611030258,
+      "learning_rate": 3.8514943181044984e-08,
+      "loss": 0.5176,
+      "step": 7749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3832513856819502,
+      "learning_rate": 3.826637685290236e-08,
+      "loss": 0.48,
+      "step": 7750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7953339027244035,
+      "learning_rate": 3.8018612136524466e-08,
+      "loss": 0.5226,
+      "step": 7751
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4591121678175754,
+      "learning_rate": 3.777164907194209e-08,
+      "loss": 0.4381,
+      "step": 7752
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6930226256798032,
+      "learning_rate": 3.752548769905273e-08,
+      "loss": 0.4754,
+      "step": 7753
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.077763466139759,
+      "learning_rate": 3.728012805762627e-08,
+      "loss": 0.4937,
+      "step": 7754
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2946149909988884,
+      "learning_rate": 3.7035570187303195e-08,
+      "loss": 0.4801,
+      "step": 7755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9916719865590282,
+      "learning_rate": 3.6791814127593585e-08,
+      "loss": 0.5081,
+      "step": 7756
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.458255870646246,
+      "learning_rate": 3.654885991787816e-08,
+      "loss": 0.4952,
+      "step": 7757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5908084020469206,
+      "learning_rate": 3.630670759740884e-08,
+      "loss": 0.4759,
+      "step": 7758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8546868327967143,
+      "learning_rate": 3.6065357205307125e-08,
+      "loss": 0.4787,
+      "step": 7759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5645776854853,
+      "learning_rate": 3.582480878056516e-08,
+      "loss": 0.5225,
+      "step": 7760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.602317185925536,
+      "learning_rate": 3.5585062362046284e-08,
+      "loss": 0.5119,
+      "step": 7761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.407875740303682,
+      "learning_rate": 3.534611798848286e-08,
+      "loss": 0.4688,
+      "step": 7762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5088178343160268,
+      "learning_rate": 3.510797569847957e-08,
+      "loss": 0.5119,
+      "step": 7763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6905489646928964,
+      "learning_rate": 3.487063553050896e-08,
+      "loss": 0.4833,
+      "step": 7764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3799093539466991,
+      "learning_rate": 3.4634097522916464e-08,
+      "loss": 0.5413,
+      "step": 7765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3819950617851147,
+      "learning_rate": 3.4398361713916526e-08,
+      "loss": 0.4909,
+      "step": 7766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3018371723847642,
+      "learning_rate": 3.416342814159423e-08,
+      "loss": 0.4656,
+      "step": 7767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.29448541490877,
+      "learning_rate": 3.392929684390533e-08,
+      "loss": 0.4505,
+      "step": 7768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5527951190707798,
+      "learning_rate": 3.3695967858675685e-08,
+      "loss": 0.4814,
+      "step": 7769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9497720926334228,
+      "learning_rate": 3.346344122360179e-08,
+      "loss": 0.5077,
+      "step": 7770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3985591816774912,
+      "learning_rate": 3.323171697625027e-08,
+      "loss": 0.4473,
+      "step": 7771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.413311741709775,
+      "learning_rate": 3.300079515405841e-08,
+      "loss": 0.5118,
+      "step": 7772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8534187425790147,
+      "learning_rate": 3.2770675794334126e-08,
+      "loss": 0.4712,
+      "step": 7773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6653584338759195,
+      "learning_rate": 3.254135893425381e-08,
+      "loss": 0.4437,
+      "step": 7774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7163977797998067,
+      "learning_rate": 3.231284461086615e-08,
+      "loss": 0.5157,
+      "step": 7775
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6106701632470143,
+      "learning_rate": 3.208513286109049e-08,
+      "loss": 0.4645,
+      "step": 7776
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7946249031013422,
+      "learning_rate": 3.185822372171521e-08,
+      "loss": 0.5111,
+      "step": 7777
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5144254795001317,
+      "learning_rate": 3.163211722939874e-08,
+      "loss": 0.4857,
+      "step": 7778
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4056587916463201,
+      "learning_rate": 3.140681342067187e-08,
+      "loss": 0.5091,
+      "step": 7779
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6749938480844588,
+      "learning_rate": 3.1182312331933254e-08,
+      "loss": 0.4518,
+      "step": 7780
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.024631804677709,
+      "learning_rate": 3.095861399945388e-08,
+      "loss": 0.4645,
+      "step": 7781
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5604921011767916,
+      "learning_rate": 3.073571845937373e-08,
+      "loss": 0.4301,
+      "step": 7782
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3798606656257772,
+      "learning_rate": 3.0513625747703446e-08,
+      "loss": 0.4492,
+      "step": 7783
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8501868863665396,
+      "learning_rate": 3.029233590032432e-08,
+      "loss": 0.4584,
+      "step": 7784
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6542302064268632,
+      "learning_rate": 3.007184895298776e-08,
+      "loss": 0.4523,
+      "step": 7785
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.515767367030657,
+      "learning_rate": 2.9852164941315266e-08,
+      "loss": 0.5063,
+      "step": 7786
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.492474367639044,
+      "learning_rate": 2.963328390079845e-08,
+      "loss": 0.517,
+      "step": 7787
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7311554006788261,
+      "learning_rate": 2.9415205866800133e-08,
+      "loss": 0.5292,
+      "step": 7788
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6439229672503788,
+      "learning_rate": 2.9197930874551584e-08,
+      "loss": 0.4827,
+      "step": 7789
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3395622287881843,
+      "learning_rate": 2.898145895915694e-08,
+      "loss": 0.4924,
+      "step": 7790
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9080209785459703,
+      "learning_rate": 2.8765790155588223e-08,
+      "loss": 0.5357,
+      "step": 7791
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2865527623095567,
+      "learning_rate": 2.8550924498688127e-08,
+      "loss": 0.4766,
+      "step": 7792
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6958372990417212,
+      "learning_rate": 2.833686202317165e-08,
+      "loss": 0.5184,
+      "step": 7793
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5048457616257143,
+      "learning_rate": 2.812360276362114e-08,
+      "loss": 0.5324,
+      "step": 7794
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3447727190681364,
+      "learning_rate": 2.7911146754491247e-08,
+      "loss": 0.4206,
+      "step": 7795
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.355496696654266,
+      "learning_rate": 2.7699494030105633e-08,
+      "loss": 0.4686,
+      "step": 7796
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7543877273172854,
+      "learning_rate": 2.7488644624659165e-08,
+      "loss": 0.5243,
+      "step": 7797
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3222072531274944,
+      "learning_rate": 2.7278598572216264e-08,
+      "loss": 0.4562,
+      "step": 7798
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3822661254883901,
+      "learning_rate": 2.7069355906711447e-08,
+      "loss": 0.4683,
+      "step": 7799
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2333982581816603,
+      "learning_rate": 2.6860916661949344e-08,
+      "loss": 0.425,
+      "step": 7800
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4989416221823086,
+      "learning_rate": 2.665328087160579e-08,
+      "loss": 0.5049,
+      "step": 7801
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1683052174068702,
+      "learning_rate": 2.6446448569226735e-08,
+      "loss": 0.4297,
+      "step": 7802
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 14.415304359454165,
+      "learning_rate": 2.6240419788226557e-08,
+      "loss": 0.5134,
+      "step": 7803
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4887720785871432,
+      "learning_rate": 2.603519456189141e-08,
+      "loss": 0.5064,
+      "step": 7804
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.269943717078514,
+      "learning_rate": 2.5830772923377544e-08,
+      "loss": 0.5161,
+      "step": 7805
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5539420147292864,
+      "learning_rate": 2.5627154905710772e-08,
+      "loss": 0.5005,
+      "step": 7806
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1212492048854474,
+      "learning_rate": 2.542434054178755e-08,
+      "loss": 0.5048,
+      "step": 7807
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3866045054200524,
+      "learning_rate": 2.5222329864373897e-08,
+      "loss": 0.4657,
+      "step": 7808
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7650098436343118,
+      "learning_rate": 2.5021122906107588e-08,
+      "loss": 0.4732,
+      "step": 7809
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4154524445769976,
+      "learning_rate": 2.4820719699493735e-08,
+      "loss": 0.4984,
+      "step": 7810
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3844456717672102,
+      "learning_rate": 2.462112027691088e-08,
+      "loss": 0.5104,
+      "step": 7811
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.729652869759931,
+      "learning_rate": 2.44223246706049e-08,
+      "loss": 0.4732,
+      "step": 7812
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4065463489571486,
+      "learning_rate": 2.4224332912693438e-08,
+      "loss": 0.497,
+      "step": 7813
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.647805148838936,
+      "learning_rate": 2.4027145035163124e-08,
+      "loss": 0.492,
+      "step": 7814
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.541283830966554,
+      "learning_rate": 2.3830761069872922e-08,
+      "loss": 0.5116,
+      "step": 7815
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5878840426569956,
+      "learning_rate": 2.363518104854856e-08,
+      "loss": 0.5096,
+      "step": 7816
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.367993865190107,
+      "learning_rate": 2.3440405002789214e-08,
+      "loss": 0.4724,
+      "step": 7817
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4697283027569419,
+      "learning_rate": 2.3246432964061928e-08,
+      "loss": 0.5135,
+      "step": 7818
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5397683114755623,
+      "learning_rate": 2.3053264963704413e-08,
+      "loss": 0.5403,
+      "step": 7819
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1536563276716563,
+      "learning_rate": 2.2860901032925597e-08,
+      "loss": 0.4726,
+      "step": 7820
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4454911065646863,
+      "learning_rate": 2.2669341202802285e-08,
+      "loss": 0.503,
+      "step": 7821
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3626319884830418,
+      "learning_rate": 2.2478585504284168e-08,
+      "loss": 0.481,
+      "step": 7822
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.322301916185995,
+      "learning_rate": 2.22886339681877e-08,
+      "loss": 0.4919,
+      "step": 7823
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2743290032865122,
+      "learning_rate": 2.209948662520278e-08,
+      "loss": 0.4753,
+      "step": 7824
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7346529047342756,
+      "learning_rate": 2.1911143505887188e-08,
+      "loss": 0.4928,
+      "step": 7825
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4341911057238383,
+      "learning_rate": 2.1723604640668804e-08,
+      "loss": 0.4799,
+      "step": 7826
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6480767999457715,
+      "learning_rate": 2.1536870059847835e-08,
+      "loss": 0.4848,
+      "step": 7827
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5460470829974238,
+      "learning_rate": 2.1350939793591264e-08,
+      "loss": 0.5232,
+      "step": 7828
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5786795687722468,
+      "learning_rate": 2.116581387193839e-08,
+      "loss": 0.5959,
+      "step": 7829
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3439945497910952,
+      "learning_rate": 2.0981492324798068e-08,
+      "loss": 0.512,
+      "step": 7830
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 22.31225001066693,
+      "learning_rate": 2.079797518194926e-08,
+      "loss": 0.4968,
+      "step": 7831
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2889659992198874,
+      "learning_rate": 2.061526247303991e-08,
+      "loss": 0.4042,
+      "step": 7832
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.571354792462103,
+      "learning_rate": 2.043335422758974e-08,
+      "loss": 0.5359,
+      "step": 7833
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4696440018354946,
+      "learning_rate": 2.0252250474987468e-08,
+      "loss": 0.5617,
+      "step": 7834
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7384848762491998,
+      "learning_rate": 2.007195124449135e-08,
+      "loss": 0.4767,
+      "step": 7835
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5219894565268821,
+      "learning_rate": 1.9892456565230866e-08,
+      "loss": 0.4939,
+      "step": 7836
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.516993884796863,
+      "learning_rate": 1.9713766466205043e-08,
+      "loss": 0.4976,
+      "step": 7837
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6486390321432192,
+      "learning_rate": 1.9535880976283007e-08,
+      "loss": 0.5048,
+      "step": 7838
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6886966252760525,
+      "learning_rate": 1.935880012420288e-08,
+      "loss": 0.5321,
+      "step": 7839
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.881517861414415,
+      "learning_rate": 1.918252393857456e-08,
+      "loss": 0.496,
+      "step": 7840
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.227334661334372,
+      "learning_rate": 1.9007052447876373e-08,
+      "loss": 0.4894,
+      "step": 7841
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5709242214792047,
+      "learning_rate": 1.8832385680457866e-08,
+      "loss": 0.5064,
+      "step": 7842
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6058301693578824,
+      "learning_rate": 1.865852366453702e-08,
+      "loss": 0.4816,
+      "step": 7843
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4782888331040613,
+      "learning_rate": 1.8485466428204145e-08,
+      "loss": 0.4699,
+      "step": 7844
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2843569767793035,
+      "learning_rate": 1.831321399941688e-08,
+      "loss": 0.4522,
+      "step": 7845
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.820698828941528,
+      "learning_rate": 1.814176640600518e-08,
+      "loss": 0.4923,
+      "step": 7846
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3360093042205863,
+      "learning_rate": 1.797112367566689e-08,
+      "loss": 0.4754,
+      "step": 7847
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3404701278536049,
+      "learning_rate": 1.7801285835971627e-08,
+      "loss": 0.4751,
+      "step": 7848
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.645967978646397,
+      "learning_rate": 1.7632252914357994e-08,
+      "loss": 0.4745,
+      "step": 7849
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.816582910025013,
+      "learning_rate": 1.746402493813415e-08,
+      "loss": 0.4759,
+      "step": 7850
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4984141287718775,
+      "learning_rate": 1.7296601934480018e-08,
+      "loss": 0.5054,
+      "step": 7851
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2956987510462552,
+      "learning_rate": 1.7129983930443405e-08,
+      "loss": 0.5119,
+      "step": 7852
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2386501615892918,
+      "learning_rate": 1.696417095294278e-08,
+      "loss": 0.4774,
+      "step": 7853
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.624533253640688,
+      "learning_rate": 1.6799163028767273e-08,
+      "loss": 0.4785,
+      "step": 7854
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3569411519411598,
+      "learning_rate": 1.6634960184575e-08,
+      "loss": 0.4949,
+      "step": 7855
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2807250837591946,
+      "learning_rate": 1.647156244689474e-08,
+      "loss": 0.4969,
+      "step": 7856
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7328035968027572,
+      "learning_rate": 1.6308969842124266e-08,
+      "loss": 0.4832,
+      "step": 7857
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.702462953664373,
+      "learning_rate": 1.6147182396532013e-08,
+      "loss": 0.5399,
+      "step": 7858
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5917224380993937,
+      "learning_rate": 1.5986200136256512e-08,
+      "loss": 0.4511,
+      "step": 7859
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7284270235794241,
+      "learning_rate": 1.5826023087305853e-08,
+      "loss": 0.4766,
+      "step": 7860
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5717015680984,
+      "learning_rate": 1.5666651275557664e-08,
+      "loss": 0.5071,
+      "step": 7861
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5548574313851318,
+      "learning_rate": 1.550808472676024e-08,
+      "loss": 0.5269,
+      "step": 7862
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.540841191632159,
+      "learning_rate": 1.535032346653087e-08,
+      "loss": 0.4903,
+      "step": 7863
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7457093299726929,
+      "learning_rate": 1.5193367520357493e-08,
+      "loss": 0.4557,
+      "step": 7864
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7486686969238991,
+      "learning_rate": 1.5037216913598163e-08,
+      "loss": 0.4794,
+      "step": 7865
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3896525495364775,
+      "learning_rate": 1.488187167147992e-08,
+      "loss": 0.4907,
+      "step": 7866
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3235501419649687,
+      "learning_rate": 1.4727331819100466e-08,
+      "loss": 0.4744,
+      "step": 7867
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.567961452961845,
+      "learning_rate": 1.4573597381427052e-08,
+      "loss": 0.4887,
+      "step": 7868
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3819936970432891,
+      "learning_rate": 1.4420668383296476e-08,
+      "loss": 0.5158,
+      "step": 7869
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4597986889667054,
+      "learning_rate": 1.4268544849416199e-08,
+      "loss": 0.5161,
+      "step": 7870
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7559433201397077,
+      "learning_rate": 1.411722680436267e-08,
+      "loss": 0.4797,
+      "step": 7871
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7793292300008385,
+      "learning_rate": 1.3966714272583004e-08,
+      "loss": 0.5119,
+      "step": 7872
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6959078402252198,
+      "learning_rate": 1.3817007278393302e-08,
+      "loss": 0.4507,
+      "step": 7873
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2366883966054931,
+      "learning_rate": 1.3668105845980883e-08,
+      "loss": 0.4014,
+      "step": 7874
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3055286853521844,
+      "learning_rate": 1.3520009999401506e-08,
+      "loss": 0.5215,
+      "step": 7875
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5049883730512816,
+      "learning_rate": 1.3372719762581587e-08,
+      "loss": 0.4989,
+      "step": 7876
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2385683699552819,
+      "learning_rate": 1.3226235159317091e-08,
+      "loss": 0.497,
+      "step": 7877
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6753540862673787,
+      "learning_rate": 1.3080556213274087e-08,
+      "loss": 0.4658,
+      "step": 7878
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3777811871545333,
+      "learning_rate": 1.2935682947987638e-08,
+      "loss": 0.4743,
+      "step": 7879
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7929465749642053,
+      "learning_rate": 1.279161538686402e-08,
+      "loss": 0.4985,
+      "step": 7880
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5867881570055768,
+      "learning_rate": 1.2648353553178506e-08,
+      "loss": 0.5077,
+      "step": 7881
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.496988232167877,
+      "learning_rate": 1.2505897470075357e-08,
+      "loss": 0.4619,
+      "step": 7882
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.388582596738789,
+      "learning_rate": 1.2364247160571164e-08,
+      "loss": 0.4528,
+      "step": 7883
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3925588795777673,
+      "learning_rate": 1.2223402647549843e-08,
+      "loss": 0.4751,
+      "step": 7884
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.315233321300705,
+      "learning_rate": 1.2083363953766525e-08,
+      "loss": 0.5084,
+      "step": 7885
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4819177522796911,
+      "learning_rate": 1.194413110184478e-08,
+      "loss": 0.5214,
+      "step": 7886
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2908476184298494,
+      "learning_rate": 1.1805704114279948e-08,
+      "loss": 0.4593,
+      "step": 7887
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3019806390033144,
+      "learning_rate": 1.1668083013435804e-08,
+      "loss": 0.5032,
+      "step": 7888
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7250265062600212,
+      "learning_rate": 1.1531267821545678e-08,
+      "loss": 0.4961,
+      "step": 7889
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.506622513165299,
+      "learning_rate": 1.1395258560714106e-08,
+      "loss": 0.5137,
+      "step": 7890
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.388515573850204,
+      "learning_rate": 1.1260055252914071e-08,
+      "loss": 0.5433,
+      "step": 7891
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.648138461522593,
+      "learning_rate": 1.1125657919989208e-08,
+      "loss": 0.5046,
+      "step": 7892
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8268093577316844,
+      "learning_rate": 1.099206658365215e-08,
+      "loss": 0.474,
+      "step": 7893
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.462649421981926,
+      "learning_rate": 1.0859281265486189e-08,
+      "loss": 0.4886,
+      "step": 7894
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8737725933823501,
+      "learning_rate": 1.0727301986943605e-08,
+      "loss": 0.4958,
+      "step": 7895
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4911950734488788,
+      "learning_rate": 1.0596128769347347e-08,
+      "loss": 0.4844,
+      "step": 7896
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.638525328685385,
+      "learning_rate": 1.0465761633888793e-08,
+      "loss": 0.5159,
+      "step": 7897
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3760722132055998,
+      "learning_rate": 1.0336200601630542e-08,
+      "loss": 0.504,
+      "step": 7898
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.525924868180361,
+      "learning_rate": 1.020744569350418e-08,
+      "loss": 0.531,
+      "step": 7899
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6230687050084631,
+      "learning_rate": 1.007949693031085e-08,
+      "loss": 0.4862,
+      "step": 7900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.759421855350358,
+      "learning_rate": 9.952354332722348e-09,
+      "loss": 0.4731,
+      "step": 7901
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7801379899361045,
+      "learning_rate": 9.826017921279463e-09,
+      "loss": 0.4566,
+      "step": 7902
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.52962165080147,
+      "learning_rate": 9.700487716392537e-09,
+      "loss": 0.5379,
+      "step": 7903
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.963892250825473,
+      "learning_rate": 9.57576373834257e-09,
+      "loss": 0.4842,
+      "step": 7904
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.373974698728333,
+      "learning_rate": 9.451846007280107e-09,
+      "loss": 0.5251,
+      "step": 7905
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.716667262902133,
+      "learning_rate": 9.328734543224138e-09,
+      "loss": 0.4694,
+      "step": 7906
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6714175262182387,
+      "learning_rate": 9.206429366065418e-09,
+      "loss": 0.4735,
+      "step": 7907
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.833029537151372,
+      "learning_rate": 9.084930495563138e-09,
+      "loss": 0.5255,
+      "step": 7908
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5869767827391525,
+      "learning_rate": 8.964237951346044e-09,
+      "loss": 0.4631,
+      "step": 7909
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4241008230767385,
+      "learning_rate": 8.844351752913539e-09,
+      "loss": 0.5049,
+      "step": 7910
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5467926959639,
+      "learning_rate": 8.72527191963457e-09,
+      "loss": 0.5108,
+      "step": 7911
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5800485283358539,
+      "learning_rate": 8.606998470746531e-09,
+      "loss": 0.5012,
+      "step": 7912
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2881792953126658,
+      "learning_rate": 8.489531425358576e-09,
+      "loss": 0.4238,
+      "step": 7913
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5260835129310093,
+      "learning_rate": 8.372870802447752e-09,
+      "loss": 0.4783,
+      "step": 7914
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3749626334965497,
+      "learning_rate": 8.257016620862313e-09,
+      "loss": 0.5112,
+      "step": 7915
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3489776684036383,
+      "learning_rate": 8.141968899318953e-09,
+      "loss": 0.5131,
+      "step": 7916
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5186970614708848,
+      "learning_rate": 8.027727656405027e-09,
+      "loss": 0.5219,
+      "step": 7917
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5606982472763145,
+      "learning_rate": 7.914292910576882e-09,
+      "loss": 0.5179,
+      "step": 7918
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.1687304109309202,
+      "learning_rate": 7.801664680161526e-09,
+      "loss": 0.4706,
+      "step": 7919
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9162988368769898,
+      "learning_rate": 7.6898429833544e-09,
+      "loss": 0.4693,
+      "step": 7920
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9704069261789112,
+      "learning_rate": 7.578827838221059e-09,
+      "loss": 0.5034,
+      "step": 7921
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9394228000122264,
+      "learning_rate": 7.468619262697708e-09,
+      "loss": 0.5163,
+      "step": 7922
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7271604647231455,
+      "learning_rate": 7.359217274589547e-09,
+      "loss": 0.4904,
+      "step": 7923
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4181225934413368,
+      "learning_rate": 7.250621891570775e-09,
+      "loss": 0.5419,
+      "step": 7924
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3000477016700693,
+      "learning_rate": 7.142833131186244e-09,
+      "loss": 0.473,
+      "step": 7925
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2807710850926293,
+      "learning_rate": 7.035851010850359e-09,
+      "loss": 0.4878,
+      "step": 7926
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4349642139062453,
+      "learning_rate": 6.929675547846515e-09,
+      "loss": 0.4818,
+      "step": 7927
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4312178958680224,
+      "learning_rate": 6.824306759328769e-09,
+      "loss": 0.5231,
+      "step": 7928
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5932973980498493,
+      "learning_rate": 6.719744662320172e-09,
+      "loss": 0.5194,
+      "step": 7929
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7986015396528459,
+      "learning_rate": 6.6159892737138746e-09,
+      "loss": 0.4934,
+      "step": 7930
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5386905898599568,
+      "learning_rate": 6.5130406102720255e-09,
+      "loss": 0.4808,
+      "step": 7931
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.596510821692502,
+      "learning_rate": 6.4108986886274315e-09,
+      "loss": 0.4887,
+      "step": 7932
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.44232300560664,
+      "learning_rate": 6.309563525281337e-09,
+      "loss": 0.4932,
+      "step": 7933
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8056468590688406,
+      "learning_rate": 6.2090351366062005e-09,
+      "loss": 0.5026,
+      "step": 7934
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4217766938218368,
+      "learning_rate": 6.10931353884292e-09,
+      "loss": 0.4779,
+      "step": 7935
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.289948288498926,
+      "learning_rate": 6.0103987481019425e-09,
+      "loss": 0.4627,
+      "step": 7936
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5237895784464568,
+      "learning_rate": 5.912290780364371e-09,
+      "loss": 0.4737,
+      "step": 7937
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2976296732878239,
+      "learning_rate": 5.814989651480307e-09,
+      "loss": 0.4797,
+      "step": 7938
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4923362635589985,
+      "learning_rate": 5.718495377169398e-09,
+      "loss": 0.5031,
+      "step": 7939
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6018338958446563,
+      "learning_rate": 5.6228079730208386e-09,
+      "loss": 0.5023,
+      "step": 7940
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.453318636362527,
+      "learning_rate": 5.527927454495041e-09,
+      "loss": 0.5132,
+      "step": 7941
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6526477392008091,
+      "learning_rate": 5.43385383691919e-09,
+      "loss": 0.5141,
+      "step": 7942
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.527193581772258,
+      "learning_rate": 5.340587135492792e-09,
+      "loss": 0.4674,
+      "step": 7943
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5351659484527673,
+      "learning_rate": 5.248127365283795e-09,
+      "loss": 0.5062,
+      "step": 7944
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5048077567795504,
+      "learning_rate": 5.156474541229694e-09,
+      "loss": 0.4739,
+      "step": 7945
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5923104159215158,
+      "learning_rate": 5.065628678137535e-09,
+      "loss": 0.442,
+      "step": 7946
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5367669913165014,
+      "learning_rate": 4.975589790684465e-09,
+      "loss": 0.538,
+      "step": 7947
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.710977541997585,
+      "learning_rate": 4.8863578934171815e-09,
+      "loss": 0.5158,
+      "step": 7948
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.408633133018979,
+      "learning_rate": 4.797933000751376e-09,
+      "loss": 0.5119,
+      "step": 7949
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5436028172282805,
+      "learning_rate": 4.710315126973952e-09,
+      "loss": 0.4909,
+      "step": 7950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0541233013668996,
+      "learning_rate": 4.6235042862397e-09,
+      "loss": 0.4652,
+      "step": 7951
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.019705736632635,
+      "learning_rate": 4.53750049257351e-09,
+      "loss": 0.5356,
+      "step": 7952
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4453039044862797,
+      "learning_rate": 4.452303759869825e-09,
+      "loss": 0.4711,
+      "step": 7953
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3822195378285782,
+      "learning_rate": 4.3679141018937445e-09,
+      "loss": 0.4825,
+      "step": 7954
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4037807224753245,
+      "learning_rate": 4.284331532278807e-09,
+      "loss": 0.4862,
+      "step": 7955
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5031811880438142,
+      "learning_rate": 4.201556064528101e-09,
+      "loss": 0.5291,
+      "step": 7956
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.307167431767514,
+      "learning_rate": 4.1195877120153716e-09,
+      "loss": 0.4643,
+      "step": 7957
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5174307919187913,
+      "learning_rate": 4.038426487982805e-09,
+      "loss": 0.4864,
+      "step": 7958
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5627982372472973,
+      "learning_rate": 3.9580724055432455e-09,
+      "loss": 0.5064,
+      "step": 7959
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.398204333643108,
+      "learning_rate": 3.878525477677975e-09,
+      "loss": 0.4993,
+      "step": 7960
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2810352158757143,
+      "learning_rate": 3.799785717238935e-09,
+      "loss": 0.4969,
+      "step": 7961
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.709476098049419,
+      "learning_rate": 3.7218531369476174e-09,
+      "loss": 0.4781,
+      "step": 7962
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5636347778347486,
+      "learning_rate": 3.644727749393395e-09,
+      "loss": 0.4709,
+      "step": 7963
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.63189574021935,
+      "learning_rate": 3.568409567037967e-09,
+      "loss": 0.521,
+      "step": 7964
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6745069970963564,
+      "learning_rate": 3.4928986022103595e-09,
+      "loss": 0.4898,
+      "step": 7965
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2734109830619809,
+      "learning_rate": 3.4181948671102583e-09,
+      "loss": 0.5043,
+      "step": 7966
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7824674763979274,
+      "learning_rate": 3.344298373806898e-09,
+      "loss": 0.4987,
+      "step": 7967
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4767364501336402,
+      "learning_rate": 3.2712091342385065e-09,
+      "loss": 0.5063,
+      "step": 7968
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4764672841746604,
+      "learning_rate": 3.1989271602134163e-09,
+      "loss": 0.4688,
+      "step": 7969
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4718649596668767,
+      "learning_rate": 3.1274524634100634e-09,
+      "loss": 0.4937,
+      "step": 7970
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9146013021572743,
+      "learning_rate": 3.0567850553747667e-09,
+      "loss": 0.4767,
+      "step": 7971
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5376106330790187,
+      "learning_rate": 2.9869249475256155e-09,
+      "loss": 0.4888,
+      "step": 7972
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6629488262473051,
+      "learning_rate": 2.917872151148027e-09,
+      "loss": 0.4718,
+      "step": 7973
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5242519408275426,
+      "learning_rate": 2.8496266773986314e-09,
+      "loss": 0.5081,
+      "step": 7974
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3755774278666677,
+      "learning_rate": 2.7821885373030545e-09,
+      "loss": 0.5311,
+      "step": 7975
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.385990845021792,
+      "learning_rate": 2.715557741756469e-09,
+      "loss": 0.478,
+      "step": 7976
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2967126233409623,
+      "learning_rate": 2.6497343015235987e-09,
+      "loss": 0.4459,
+      "step": 7977
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4684511791320787,
+      "learning_rate": 2.584718227239269e-09,
+      "loss": 0.4752,
+      "step": 7978
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.519683551519481,
+      "learning_rate": 2.520509529406745e-09,
+      "loss": 0.4846,
+      "step": 7979
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5136123778915815,
+      "learning_rate": 2.4571082183993955e-09,
+      "loss": 0.4975,
+      "step": 7980
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.440141895316327,
+      "learning_rate": 2.3945143044612485e-09,
+      "loss": 0.4474,
+      "step": 7981
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7085897376426566,
+      "learning_rate": 2.33272779770366e-09,
+      "loss": 0.4575,
+      "step": 7982
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4363383949547228,
+      "learning_rate": 2.271748708110311e-09,
+      "loss": 0.4879,
+      "step": 7983
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5093556934051071,
+      "learning_rate": 2.211577045531099e-09,
+      "loss": 0.5451,
+      "step": 7984
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4565013995684928,
+      "learning_rate": 2.1522128196888036e-09,
+      "loss": 0.5185,
+      "step": 7985
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4807416392206982,
+      "learning_rate": 2.0936560401735306e-09,
+      "loss": 0.4876,
+      "step": 7986
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7294427571785986,
+      "learning_rate": 2.0359067164454905e-09,
+      "loss": 0.4929,
+      "step": 7987
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.646187430172883,
+      "learning_rate": 1.9789648578349975e-09,
+      "loss": 0.492,
+      "step": 7988
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5770653317502463,
+      "learning_rate": 1.9228304735413594e-09,
+      "loss": 0.5405,
+      "step": 7989
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5488248308485255,
+      "learning_rate": 1.8675035726339883e-09,
+      "loss": 0.4976,
+      "step": 7990
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5937102313405058,
+      "learning_rate": 1.812984164050735e-09,
+      "loss": 0.544,
+      "step": 7991
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5565998451857792,
+      "learning_rate": 1.7592722565995536e-09,
+      "loss": 0.4508,
+      "step": 7992
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.208208978134417,
+      "learning_rate": 1.706367858959057e-09,
+      "loss": 0.5158,
+      "step": 7993
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6943100273181849,
+      "learning_rate": 1.6542709796757429e-09,
+      "loss": 0.4905,
+      "step": 7994
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4754337734216012,
+      "learning_rate": 1.6029816271667664e-09,
+      "loss": 0.4433,
+      "step": 7995
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7784248749244436,
+      "learning_rate": 1.5524998097177225e-09,
+      "loss": 0.5054,
+      "step": 7996
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6584613863369149,
+      "learning_rate": 1.502825535484309e-09,
+      "loss": 0.4693,
+      "step": 7997
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3627989829776788,
+      "learning_rate": 1.4539588124928838e-09,
+      "loss": 0.4536,
+      "step": 7998
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6144518696085406,
+      "learning_rate": 1.4058996486376875e-09,
+      "loss": 0.5055,
+      "step": 7999
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.069795091114158,
+      "learning_rate": 1.3586480516825096e-09,
+      "loss": 0.4743,
+      "step": 8000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 8058,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "total_flos": 4101554005508096.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}