diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,45761 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 6533,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 383.1503991853064,
+      "learning_rate": 1.0204081632653061e-07,
+      "loss": 3.16,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1115.042586838023,
+      "learning_rate": 2.0408163265306121e-07,
+      "loss": 3.4515,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 641.1581033180005,
+      "learning_rate": 3.0612244897959183e-07,
+      "loss": 3.4183,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 543.9001930338954,
+      "learning_rate": 4.0816326530612243e-07,
+      "loss": 3.1405,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 415.6700527460039,
+      "learning_rate": 5.102040816326531e-07,
+      "loss": 2.9573,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 560.8177502149401,
+      "learning_rate": 6.122448979591837e-07,
+      "loss": 3.0618,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 549.9535790740002,
+      "learning_rate": 7.142857142857143e-07,
+      "loss": 2.787,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 135.6880950548416,
+      "learning_rate": 8.163265306122449e-07,
+      "loss": 2.1078,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 304.45317074621744,
+      "learning_rate": 9.183673469387756e-07,
+      "loss": 2.0253,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 129.6075967509114,
+      "learning_rate": 1.0204081632653063e-06,
+      "loss": 1.987,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 116.64503619601689,
+      "learning_rate": 1.122448979591837e-06,
+      "loss": 1.7659,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 82.57909679030715,
+      "learning_rate": 1.2244897959183673e-06,
+      "loss": 1.7313,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 51.7873071107413,
+      "learning_rate": 1.3265306122448982e-06,
+      "loss": 1.6111,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.55146603646816,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 0.6476,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 52.146128170413405,
+      "learning_rate": 1.5306122448979593e-06,
+      "loss": 1.4587,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.018684442954445,
+      "learning_rate": 1.6326530612244897e-06,
+      "loss": 1.5048,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.633909350728741,
+      "learning_rate": 1.7346938775510206e-06,
+      "loss": 0.58,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 31.3525327981405,
+      "learning_rate": 1.8367346938775512e-06,
+      "loss": 1.257,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.10854106994972,
+      "learning_rate": 1.938775510204082e-06,
+      "loss": 1.2033,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.144610314079856,
+      "learning_rate": 2.0408163265306125e-06,
+      "loss": 1.1375,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.783699466031027,
+      "learning_rate": 2.1428571428571427e-06,
+      "loss": 1.1003,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.60475253722536,
+      "learning_rate": 2.244897959183674e-06,
+      "loss": 1.2054,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.340451063129958,
+      "learning_rate": 2.3469387755102044e-06,
+      "loss": 1.055,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.793621550631057,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 1.1943,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 22.91654748783573,
+      "learning_rate": 2.5510204081632657e-06,
+      "loss": 1.1807,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.665266982661313,
+      "learning_rate": 2.6530612244897964e-06,
+      "loss": 1.1996,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.279207516291493,
+      "learning_rate": 2.7551020408163266e-06,
+      "loss": 1.1656,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.967120212081813,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.1532,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.383928985415782,
+      "learning_rate": 2.959183673469388e-06,
+      "loss": 0.9899,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.203716579957373,
+      "learning_rate": 3.0612244897959185e-06,
+      "loss": 1.0944,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.071417491040497,
+      "learning_rate": 3.1632653061224496e-06,
+      "loss": 1.1039,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.075535174989573,
+      "learning_rate": 3.2653061224489794e-06,
+      "loss": 1.0426,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.406595669779419,
+      "learning_rate": 3.3673469387755105e-06,
+      "loss": 1.0862,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.618904672536877,
+      "learning_rate": 3.469387755102041e-06,
+      "loss": 0.4901,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 22.979117055215557,
+      "learning_rate": 3.5714285714285718e-06,
+      "loss": 1.1253,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 98.03089037716222,
+      "learning_rate": 3.6734693877551024e-06,
+      "loss": 1.1556,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.287036563173132,
+      "learning_rate": 3.7755102040816327e-06,
+      "loss": 1.0596,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.815097934554741,
+      "learning_rate": 3.877551020408164e-06,
+      "loss": 1.0359,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 13.600743070021583,
+      "learning_rate": 3.979591836734694e-06,
+      "loss": 1.0887,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.078085686869478,
+      "learning_rate": 4.081632653061225e-06,
+      "loss": 1.0267,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.085579534504056,
+      "learning_rate": 4.183673469387755e-06,
+      "loss": 1.0953,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.251690714050342,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 1.0291,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.157487900257816,
+      "learning_rate": 4.3877551020408165e-06,
+      "loss": 1.1289,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 15.394745194540516,
+      "learning_rate": 4.489795918367348e-06,
+      "loss": 1.1724,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.206250385388731,
+      "learning_rate": 4.591836734693878e-06,
+      "loss": 1.0897,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.385889193455325,
+      "learning_rate": 4.693877551020409e-06,
+      "loss": 1.0748,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.013036055651966,
+      "learning_rate": 4.795918367346939e-06,
+      "loss": 1.0592,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.540522265960379,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 0.9947,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.128507543319763,
+      "learning_rate": 5e-06,
+      "loss": 0.4851,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.456741645011198,
+      "learning_rate": 5.1020408163265315e-06,
+      "loss": 1.1242,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.073951090190663,
+      "learning_rate": 5.204081632653062e-06,
+      "loss": 1.1371,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.848944210654246,
+      "learning_rate": 5.306122448979593e-06,
+      "loss": 0.9847,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.814113508474519,
+      "learning_rate": 5.408163265306123e-06,
+      "loss": 0.965,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.075740772611151,
+      "learning_rate": 5.510204081632653e-06,
+      "loss": 0.4762,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.253826117056938,
+      "learning_rate": 5.6122448979591834e-06,
+      "loss": 1.0542,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.609978272010776,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.0704,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.397718068371473,
+      "learning_rate": 5.816326530612246e-06,
+      "loss": 1.1626,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.260133946689537,
+      "learning_rate": 5.918367346938776e-06,
+      "loss": 1.0113,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.852944108730934,
+      "learning_rate": 6.020408163265307e-06,
+      "loss": 0.4874,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.178769332745159,
+      "learning_rate": 6.122448979591837e-06,
+      "loss": 1.0644,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.103892360369785,
+      "learning_rate": 6.224489795918368e-06,
+      "loss": 1.0928,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.494779873667326,
+      "learning_rate": 6.326530612244899e-06,
+      "loss": 1.0097,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.629387087808804,
+      "learning_rate": 6.4285714285714295e-06,
+      "loss": 1.0211,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.464978091176936,
+      "learning_rate": 6.530612244897959e-06,
+      "loss": 1.0977,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.660183593852527,
+      "learning_rate": 6.63265306122449e-06,
+      "loss": 1.0661,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.831455627108557,
+      "learning_rate": 6.734693877551021e-06,
+      "loss": 1.0992,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.722994865015469,
+      "learning_rate": 6.836734693877551e-06,
+      "loss": 1.1489,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.8110902501706425,
+      "learning_rate": 6.938775510204082e-06,
+      "loss": 1.0966,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.14628522025491,
+      "learning_rate": 7.0408163265306125e-06,
+      "loss": 0.9751,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.698256800772039,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 0.9841,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.727266458576524,
+      "learning_rate": 7.244897959183675e-06,
+      "loss": 1.078,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.712483103175598,
+      "learning_rate": 7.346938775510205e-06,
+      "loss": 0.4774,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.267739903860731,
+      "learning_rate": 7.448979591836736e-06,
+      "loss": 1.0613,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.970987660453837,
+      "learning_rate": 7.551020408163265e-06,
+      "loss": 1.0343,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.088617564760351,
+      "learning_rate": 7.653061224489796e-06,
+      "loss": 0.9876,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.702281289345013,
+      "learning_rate": 7.755102040816327e-06,
+      "loss": 1.055,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.672033119575265,
+      "learning_rate": 7.857142857142858e-06,
+      "loss": 0.9966,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.069957763879359,
+      "learning_rate": 7.959183673469388e-06,
+      "loss": 0.5214,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.247445520922536,
+      "learning_rate": 8.06122448979592e-06,
+      "loss": 1.0514,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.516511852647083,
+      "learning_rate": 8.16326530612245e-06,
+      "loss": 1.0861,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.052749314346663,
+      "learning_rate": 8.26530612244898e-06,
+      "loss": 1.016,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.750995746740932,
+      "learning_rate": 8.36734693877551e-06,
+      "loss": 1.1089,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.713321176065724,
+      "learning_rate": 8.469387755102042e-06,
+      "loss": 1.0066,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.794509522449217,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.012,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.664116358377716,
+      "learning_rate": 8.673469387755103e-06,
+      "loss": 1.0805,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.320697850800005,
+      "learning_rate": 8.775510204081633e-06,
+      "loss": 1.0494,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.006680080855596,
+      "learning_rate": 8.877551020408163e-06,
+      "loss": 1.0981,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.213977754985643,
+      "learning_rate": 8.979591836734695e-06,
+      "loss": 1.0488,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.977104454473886,
+      "learning_rate": 9.081632653061225e-06,
+      "loss": 0.9923,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 26.93170641637084,
+      "learning_rate": 9.183673469387756e-06,
+      "loss": 1.1516,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.731873248521301,
+      "learning_rate": 9.285714285714288e-06,
+      "loss": 1.0429,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.4993308961508784,
+      "learning_rate": 9.387755102040818e-06,
+      "loss": 1.0329,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.509109125779879,
+      "learning_rate": 9.489795918367348e-06,
+      "loss": 1.028,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.2065413570334895,
+      "learning_rate": 9.591836734693878e-06,
+      "loss": 1.0646,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.809366591371024,
+      "learning_rate": 9.693877551020408e-06,
+      "loss": 1.1289,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.476657106326629,
+      "learning_rate": 9.795918367346939e-06,
+      "loss": 1.0345,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.927969816535037,
+      "learning_rate": 9.89795918367347e-06,
+      "loss": 0.996,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 59.64526210339528,
+      "learning_rate": 1e-05,
+      "loss": 1.0362,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.401085627447702,
+      "learning_rate": 1.0102040816326531e-05,
+      "loss": 1.0387,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.258010610505343,
+      "learning_rate": 1.0204081632653063e-05,
+      "loss": 1.145,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.661261424729024,
+      "learning_rate": 1.0306122448979591e-05,
+      "loss": 1.1126,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.257632267247965,
+      "learning_rate": 1.0408163265306123e-05,
+      "loss": 0.9783,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.410703256358511,
+      "learning_rate": 1.0510204081632654e-05,
+      "loss": 1.0303,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.848851190925594,
+      "learning_rate": 1.0612244897959186e-05,
+      "loss": 1.1671,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.5957945532873214,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 1.1229,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.606270565711428,
+      "learning_rate": 1.0816326530612246e-05,
+      "loss": 1.0775,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.7030893221224215,
+      "learning_rate": 1.0918367346938776e-05,
+      "loss": 1.1389,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.739502719203268,
+      "learning_rate": 1.1020408163265306e-05,
+      "loss": 1.0237,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.301768842909257,
+      "learning_rate": 1.1122448979591838e-05,
+      "loss": 0.9889,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.18146117465339,
+      "learning_rate": 1.1224489795918367e-05,
+      "loss": 1.0543,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.047654607041304,
+      "learning_rate": 1.1326530612244899e-05,
+      "loss": 1.09,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.637823055554012,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.0977,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.281759210863109,
+      "learning_rate": 1.1530612244897961e-05,
+      "loss": 1.1304,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.552251083233848,
+      "learning_rate": 1.1632653061224491e-05,
+      "loss": 1.0337,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.73550542774755,
+      "learning_rate": 1.1734693877551021e-05,
+      "loss": 1.0381,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.767902308250887,
+      "learning_rate": 1.1836734693877552e-05,
+      "loss": 1.0879,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.813147914121843,
+      "learning_rate": 1.1938775510204084e-05,
+      "loss": 1.0883,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.756174166724238,
+      "learning_rate": 1.2040816326530614e-05,
+      "loss": 0.9809,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.362500089300721,
+      "learning_rate": 1.2142857142857142e-05,
+      "loss": 1.0745,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.699334275929911,
+      "learning_rate": 1.2244897959183674e-05,
+      "loss": 1.0208,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.953344832033511,
+      "learning_rate": 1.2346938775510204e-05,
+      "loss": 0.9662,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.090119536112503,
+      "learning_rate": 1.2448979591836736e-05,
+      "loss": 1.1166,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.724578786244247,
+      "learning_rate": 1.2551020408163267e-05,
+      "loss": 1.0081,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.82491736656003,
+      "learning_rate": 1.2653061224489798e-05,
+      "loss": 1.1457,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.253776774271056,
+      "learning_rate": 1.2755102040816327e-05,
+      "loss": 1.1222,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.271747960540945,
+      "learning_rate": 1.2857142857142859e-05,
+      "loss": 1.1013,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.69741315980313,
+      "learning_rate": 1.2959183673469389e-05,
+      "loss": 1.0754,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.4875349314231565,
+      "learning_rate": 1.3061224489795918e-05,
+      "loss": 1.102,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.038528782577277,
+      "learning_rate": 1.316326530612245e-05,
+      "loss": 1.1194,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.606252307014983,
+      "learning_rate": 1.326530612244898e-05,
+      "loss": 1.1034,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.327163657525908,
+      "learning_rate": 1.3367346938775512e-05,
+      "loss": 1.2087,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.280116871322719,
+      "learning_rate": 1.3469387755102042e-05,
+      "loss": 1.0586,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.6456091927763365,
+      "learning_rate": 1.3571428571428574e-05,
+      "loss": 1.0209,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.558331129509311,
+      "learning_rate": 1.3673469387755102e-05,
+      "loss": 1.1321,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.498934702458065,
+      "learning_rate": 1.3775510204081634e-05,
+      "loss": 1.0218,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.339930806582817,
+      "learning_rate": 1.3877551020408165e-05,
+      "loss": 1.0357,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.454047705726223,
+      "learning_rate": 1.3979591836734696e-05,
+      "loss": 0.9997,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.979588983206454,
+      "learning_rate": 1.4081632653061225e-05,
+      "loss": 1.1083,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.50575669433855,
+      "learning_rate": 1.4183673469387755e-05,
+      "loss": 1.0733,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.352196459546808,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 1.0983,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.932749572950468,
+      "learning_rate": 1.4387755102040817e-05,
+      "loss": 1.1568,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.83481653533095,
+      "learning_rate": 1.448979591836735e-05,
+      "loss": 1.1989,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.2113642259074915,
+      "learning_rate": 1.4591836734693878e-05,
+      "loss": 1.0765,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.559171013530669,
+      "learning_rate": 1.469387755102041e-05,
+      "loss": 1.2268,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.313659971812374,
+      "learning_rate": 1.479591836734694e-05,
+      "loss": 1.0467,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.643371072712547,
+      "learning_rate": 1.4897959183673472e-05,
+      "loss": 1.1269,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.021005714047092,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.1134,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.373077262357572,
+      "learning_rate": 1.510204081632653e-05,
+      "loss": 1.0595,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.5019027656338215,
+      "learning_rate": 1.5204081632653063e-05,
+      "loss": 1.0591,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.834636688292023,
+      "learning_rate": 1.530612244897959e-05,
+      "loss": 1.0933,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.277045405401534,
+      "learning_rate": 1.5408163265306123e-05,
+      "loss": 1.0802,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.249863285521951,
+      "learning_rate": 1.5510204081632655e-05,
+      "loss": 1.0813,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.709586843024497,
+      "learning_rate": 1.5612244897959187e-05,
+      "loss": 1.0674,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.445331032824768,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 0.9501,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.36900789371636,
+      "learning_rate": 1.5816326530612247e-05,
+      "loss": 1.1602,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.647764431722952,
+      "learning_rate": 1.5918367346938776e-05,
+      "loss": 1.1474,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.312425415130779,
+      "learning_rate": 1.6020408163265308e-05,
+      "loss": 1.0233,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.383622524455289,
+      "learning_rate": 1.612244897959184e-05,
+      "loss": 1.1282,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.665115899124785,
+      "learning_rate": 1.6224489795918368e-05,
+      "loss": 1.0337,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.249438793574112,
+      "learning_rate": 1.63265306122449e-05,
+      "loss": 1.131,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.182023365734747,
+      "learning_rate": 1.642857142857143e-05,
+      "loss": 1.1596,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.233978778860636,
+      "learning_rate": 1.653061224489796e-05,
+      "loss": 1.0368,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.211250025821531,
+      "learning_rate": 1.6632653061224492e-05,
+      "loss": 1.1777,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.193660311427835,
+      "learning_rate": 1.673469387755102e-05,
+      "loss": 1.0773,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.667021464427709,
+      "learning_rate": 1.6836734693877553e-05,
+      "loss": 1.1726,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.144632567111517,
+      "learning_rate": 1.6938775510204085e-05,
+      "loss": 1.028,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.419125839731695,
+      "learning_rate": 1.7040816326530613e-05,
+      "loss": 1.0666,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.5162703152902735,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.9476,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.534476185280114,
+      "learning_rate": 1.7244897959183674e-05,
+      "loss": 1.0293,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.2943079469894,
+      "learning_rate": 1.7346938775510206e-05,
+      "loss": 1.0758,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.076390118325341,
+      "learning_rate": 1.7448979591836738e-05,
+      "loss": 1.204,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.087148129774895,
+      "learning_rate": 1.7551020408163266e-05,
+      "loss": 1.1198,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.071546224881873,
+      "learning_rate": 1.7653061224489798e-05,
+      "loss": 1.0729,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.934479929849503,
+      "learning_rate": 1.7755102040816327e-05,
+      "loss": 1.1263,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.5097148641155105,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 1.0998,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.933561500487126,
+      "learning_rate": 1.795918367346939e-05,
+      "loss": 1.0783,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.1969610822139884,
+      "learning_rate": 1.806122448979592e-05,
+      "loss": 1.0884,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.219171188346219,
+      "learning_rate": 1.816326530612245e-05,
+      "loss": 1.139,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.238974571818712,
+      "learning_rate": 1.826530612244898e-05,
+      "loss": 1.096,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.321624515588689,
+      "learning_rate": 1.836734693877551e-05,
+      "loss": 1.1538,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.287719028369076,
+      "learning_rate": 1.8469387755102043e-05,
+      "loss": 1.1504,
+      "step": 181
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.816101595539008,
+      "learning_rate": 1.8571428571428575e-05,
+      "loss": 1.0856,
+      "step": 182
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.421835257556857,
+      "learning_rate": 1.8673469387755104e-05,
+      "loss": 1.115,
+      "step": 183
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.226520103966431,
+      "learning_rate": 1.8775510204081636e-05,
+      "loss": 1.1927,
+      "step": 184
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.957358233989192,
+      "learning_rate": 1.8877551020408164e-05,
+      "loss": 1.067,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.884140241314185,
+      "learning_rate": 1.8979591836734696e-05,
+      "loss": 1.0874,
+      "step": 186
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.744041077689365,
+      "learning_rate": 1.9081632653061225e-05,
+      "loss": 1.0818,
+      "step": 187
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.5741498454057234,
+      "learning_rate": 1.9183673469387756e-05,
+      "loss": 1.0791,
+      "step": 188
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.252343995041189,
+      "learning_rate": 1.928571428571429e-05,
+      "loss": 1.1573,
+      "step": 189
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.436600844797034,
+      "learning_rate": 1.9387755102040817e-05,
+      "loss": 1.0886,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.686800742042656,
+      "learning_rate": 1.948979591836735e-05,
+      "loss": 1.0898,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.3941981721298164,
+      "learning_rate": 1.9591836734693877e-05,
+      "loss": 1.1444,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.043454999129311,
+      "learning_rate": 1.969387755102041e-05,
+      "loss": 1.0928,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.371089443369653,
+      "learning_rate": 1.979591836734694e-05,
+      "loss": 1.1281,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.203990082279683,
+      "learning_rate": 1.9897959183673473e-05,
+      "loss": 1.1808,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.997545641343556,
+      "learning_rate": 2e-05,
+      "loss": 1.144,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.456843142113312,
+      "learning_rate": 1.999999877114023e-05,
+      "loss": 1.0797,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.453312121763317,
+      "learning_rate": 1.9999995084561225e-05,
+      "loss": 1.0321,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.309253158294316,
+      "learning_rate": 1.9999988940263887e-05,
+      "loss": 0.7624,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.513880161766579,
+      "learning_rate": 1.9999980338249726e-05,
+      "loss": 1.1025,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.668581887524097,
+      "learning_rate": 1.999996927852086e-05,
+      "loss": 1.1191,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.402701531746498,
+      "learning_rate": 1.9999955761080003e-05,
+      "loss": 1.0652,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.9881038307362,
+      "learning_rate": 1.999993978593048e-05,
+      "loss": 1.1601,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.586499941037915,
+      "learning_rate": 1.9999921353076216e-05,
+      "loss": 1.1691,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.46705629801981,
+      "learning_rate": 1.9999900462521743e-05,
+      "loss": 1.1723,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 59.17034445169865,
+      "learning_rate": 1.9999877114272194e-05,
+      "loss": 0.9822,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.176551551254107,
+      "learning_rate": 1.9999851308333305e-05,
+      "loss": 1.1339,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.182693806023538,
+      "learning_rate": 1.9999823044711424e-05,
+      "loss": 0.9911,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.95969238154806,
+      "learning_rate": 1.9999792323413492e-05,
+      "loss": 1.1375,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.304652344445716,
+      "learning_rate": 1.9999759144447062e-05,
+      "loss": 1.0383,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.736874041587823,
+      "learning_rate": 1.9999723507820288e-05,
+      "loss": 1.0764,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.647646408322613,
+      "learning_rate": 1.999968541354193e-05,
+      "loss": 1.2264,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.084198319034713,
+      "learning_rate": 1.9999644861621346e-05,
+      "loss": 1.1324,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.938224249926442,
+      "learning_rate": 1.9999601852068507e-05,
+      "loss": 1.102,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.063465444477739,
+      "learning_rate": 1.9999556384893984e-05,
+      "loss": 1.0721,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.564512063889825,
+      "learning_rate": 1.999950846010895e-05,
+      "loss": 1.1503,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.414783073934225,
+      "learning_rate": 1.999945807772518e-05,
+      "loss": 1.061,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.011468038105515,
+      "learning_rate": 1.999940523775506e-05,
+      "loss": 1.1042,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.833874683418772,
+      "learning_rate": 1.999934994021158e-05,
+      "loss": 0.9945,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.799836412952987,
+      "learning_rate": 1.9999292185108322e-05,
+      "loss": 1.0803,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.554987452781492,
+      "learning_rate": 1.999923197245949e-05,
+      "loss": 1.0829,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.372674266383977,
+      "learning_rate": 1.9999169302279874e-05,
+      "loss": 1.1556,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.657197518806744,
+      "learning_rate": 1.9999104174584885e-05,
+      "loss": 1.0696,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.728835803379367,
+      "learning_rate": 1.999903658939052e-05,
+      "loss": 1.0844,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.828075849629188,
+      "learning_rate": 1.99989665467134e-05,
+      "loss": 1.0699,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.769915650459503,
+      "learning_rate": 1.9998894046570735e-05,
+      "loss": 1.1587,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.069781347033537,
+      "learning_rate": 1.9998819088980338e-05,
+      "loss": 1.211,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.27532701759076,
+      "learning_rate": 1.999874167396064e-05,
+      "loss": 1.1225,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.972816909113727,
+      "learning_rate": 1.999866180153066e-05,
+      "loss": 1.2299,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.328508278998827,
+      "learning_rate": 1.9998579471710037e-05,
+      "loss": 1.1446,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.761237265785631,
+      "learning_rate": 1.9998494684519e-05,
+      "loss": 1.0745,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.0094751510628495,
+      "learning_rate": 1.9998407439978383e-05,
+      "loss": 1.1848,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.687867135519146,
+      "learning_rate": 1.9998317738109638e-05,
+      "loss": 1.1307,
+      "step": 233
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.373549487869426,
+      "learning_rate": 1.99982255789348e-05,
+      "loss": 1.1336,
+      "step": 234
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.491520583566882,
+      "learning_rate": 1.999813096247653e-05,
+      "loss": 1.1269,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5109900214204846,
+      "learning_rate": 1.999803388875808e-05,
+      "loss": 1.1576,
+      "step": 236
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.897322807826128,
+      "learning_rate": 1.9997934357803302e-05,
+      "loss": 1.1426,
+      "step": 237
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.737081073192212,
+      "learning_rate": 1.999783236963666e-05,
+      "loss": 0.9822,
+      "step": 238
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 507.51763352048624,
+      "learning_rate": 1.9997727924283227e-05,
+      "loss": 2.0016,
+      "step": 239
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.720695674654614,
+      "learning_rate": 1.9997621021768663e-05,
+      "loss": 1.0005,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.22021102440788,
+      "learning_rate": 1.9997511662119248e-05,
+      "loss": 1.2467,
+      "step": 241
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.925981261258009,
+      "learning_rate": 1.9997399845361853e-05,
+      "loss": 1.1494,
+      "step": 242
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.326969030434556,
+      "learning_rate": 1.9997285571523966e-05,
+      "loss": 1.2098,
+      "step": 243
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.852488413814456,
+      "learning_rate": 1.999716884063367e-05,
+      "loss": 1.1044,
+      "step": 244
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.981439821590262,
+      "learning_rate": 1.9997049652719655e-05,
+      "loss": 1.1347,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 298.8321545901624,
+      "learning_rate": 1.999692800781121e-05,
+      "loss": 0.9508,
+      "step": 246
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.030060046557593,
+      "learning_rate": 1.9996803905938237e-05,
+      "loss": 1.1103,
+      "step": 247
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.209425242514626,
+      "learning_rate": 1.9996677347131237e-05,
+      "loss": 1.0899,
+      "step": 248
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.786140168586079,
+      "learning_rate": 1.999654833142131e-05,
+      "loss": 1.2291,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.675732962915219,
+      "learning_rate": 1.9996416858840167e-05,
+      "loss": 1.1652,
+      "step": 250
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.486052375807091,
+      "learning_rate": 1.999628292942012e-05,
+      "loss": 1.1669,
+      "step": 251
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.513586744808445,
+      "learning_rate": 1.9996146543194086e-05,
+      "loss": 1.2232,
+      "step": 252
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.482304872859721,
+      "learning_rate": 1.9996007700195583e-05,
+      "loss": 1.0857,
+      "step": 253
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.219060691629565,
+      "learning_rate": 1.9995866400458736e-05,
+      "loss": 1.0267,
+      "step": 254
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.629086694340683,
+      "learning_rate": 1.9995722644018275e-05,
+      "loss": 1.1984,
+      "step": 255
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.554770506006259,
+      "learning_rate": 1.9995576430909526e-05,
+      "loss": 1.1749,
+      "step": 256
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.486611373898943,
+      "learning_rate": 1.9995427761168427e-05,
+      "loss": 1.0657,
+      "step": 257
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.086051833263698,
+      "learning_rate": 1.9995276634831518e-05,
+      "loss": 1.1945,
+      "step": 258
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.784755615304702,
+      "learning_rate": 1.9995123051935938e-05,
+      "loss": 0.9962,
+      "step": 259
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.1527510227553535,
+      "learning_rate": 1.9994967012519436e-05,
+      "loss": 1.0401,
+      "step": 260
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.979116067663447,
+      "learning_rate": 1.9994808516620362e-05,
+      "loss": 0.9774,
+      "step": 261
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.845044200480525,
+      "learning_rate": 1.999464756427767e-05,
+      "loss": 1.0955,
+      "step": 262
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.388519751080092,
+      "learning_rate": 1.999448415553092e-05,
+      "loss": 1.0714,
+      "step": 263
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.667854781851994,
+      "learning_rate": 1.9994318290420264e-05,
+      "loss": 0.9685,
+      "step": 264
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.912765838925781,
+      "learning_rate": 1.9994149968986477e-05,
+      "loss": 1.2169,
+      "step": 265
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.323589806398268,
+      "learning_rate": 1.9993979191270923e-05,
+      "loss": 1.0595,
+      "step": 266
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.482357421422183,
+      "learning_rate": 1.9993805957315575e-05,
+      "loss": 1.1664,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.221027555569462,
+      "learning_rate": 1.999363026716301e-05,
+      "loss": 1.1076,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 16.703437262233603,
+      "learning_rate": 1.9993452120856412e-05,
+      "loss": 0.88,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.941943116606574,
+      "learning_rate": 1.9993271518439554e-05,
+      "loss": 1.0686,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.695548346288317,
+      "learning_rate": 1.999308845995683e-05,
+      "loss": 1.0999,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.326582127278831,
+      "learning_rate": 1.9992902945453227e-05,
+      "loss": 1.1071,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.5673615588610215,
+      "learning_rate": 1.9992714974974344e-05,
+      "loss": 1.1184,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.379449041896806,
+      "learning_rate": 1.9992524548566378e-05,
+      "loss": 1.1919,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.356848963709712,
+      "learning_rate": 1.9992331666276126e-05,
+      "loss": 1.0153,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.7495735533750665,
+      "learning_rate": 1.9992136328151e-05,
+      "loss": 1.1424,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.708712119538405,
+      "learning_rate": 1.9991938534238996e-05,
+      "loss": 1.0777,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.623074644106911,
+      "learning_rate": 1.9991738284588743e-05,
+      "loss": 1.1632,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.038994081163376,
+      "learning_rate": 1.9991535579249443e-05,
+      "loss": 1.1002,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.4993011643736915,
+      "learning_rate": 1.9991330418270926e-05,
+      "loss": 1.1014,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.312602735744396,
+      "learning_rate": 1.9991122801703606e-05,
+      "loss": 1.087,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 281.9379233611627,
+      "learning_rate": 1.9990912729598512e-05,
+      "loss": 1.3675,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.403901398769606,
+      "learning_rate": 1.9990700202007276e-05,
+      "loss": 0.9703,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.718875253576995,
+      "learning_rate": 1.999048521898213e-05,
+      "loss": 1.0062,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.383041046256885,
+      "learning_rate": 1.999026778057591e-05,
+      "loss": 1.0738,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.662485194619379,
+      "learning_rate": 1.999004788684206e-05,
+      "loss": 1.1438,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.353269823598559,
+      "learning_rate": 1.9989825537834623e-05,
+      "loss": 1.0382,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.508331270093196,
+      "learning_rate": 1.9989600733608235e-05,
+      "loss": 1.0688,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.046375086111272,
+      "learning_rate": 1.9989373474218163e-05,
+      "loss": 1.1265,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.7962186931442705,
+      "learning_rate": 1.998914375972025e-05,
+      "loss": 1.146,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.70566502248541,
+      "learning_rate": 1.998891159017096e-05,
+      "loss": 1.068,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 80.8458009620087,
+      "learning_rate": 1.998867696562735e-05,
+      "loss": 1.1027,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.554679405892336,
+      "learning_rate": 1.9988439886147082e-05,
+      "loss": 1.0465,
+      "step": 293
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.717994795659879,
+      "learning_rate": 1.9988200351788425e-05,
+      "loss": 1.125,
+      "step": 294
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.380389098649303,
+      "learning_rate": 1.9987958362610255e-05,
+      "loss": 0.9189,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.012117873767208,
+      "learning_rate": 1.9987713918672044e-05,
+      "loss": 1.2106,
+      "step": 296
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.827671238106491,
+      "learning_rate": 1.9987467020033865e-05,
+      "loss": 1.1088,
+      "step": 297
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.102478908570774,
+      "learning_rate": 1.99872176667564e-05,
+      "loss": 1.0611,
+      "step": 298
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.919138180496703,
+      "learning_rate": 1.9986965858900934e-05,
+      "loss": 1.0139,
+      "step": 299
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.634896848034215,
+      "learning_rate": 1.9986711596529356e-05,
+      "loss": 1.198,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.76351079430654,
+      "learning_rate": 1.9986454879704157e-05,
+      "loss": 1.208,
+      "step": 301
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.003244093064284,
+      "learning_rate": 1.9986195708488428e-05,
+      "loss": 1.1318,
+      "step": 302
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.3217012747656565,
+      "learning_rate": 1.9985934082945868e-05,
+      "loss": 1.0452,
+      "step": 303
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.800641119661177,
+      "learning_rate": 1.9985670003140777e-05,
+      "loss": 1.1481,
+      "step": 304
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.248394762455576,
+      "learning_rate": 1.9985403469138056e-05,
+      "loss": 1.098,
+      "step": 305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.557215159093832,
+      "learning_rate": 1.9985134481003213e-05,
+      "loss": 1.1635,
+      "step": 306
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.411297001088896,
+      "learning_rate": 1.9984863038802364e-05,
+      "loss": 1.0967,
+      "step": 307
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.619947266168186,
+      "learning_rate": 1.998458914260221e-05,
+      "loss": 1.1012,
+      "step": 308
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.7941902957102895,
+      "learning_rate": 1.9984312792470074e-05,
+      "loss": 1.1977,
+      "step": 309
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.78262219409935,
+      "learning_rate": 1.9984033988473874e-05,
+      "loss": 1.1142,
+      "step": 310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.6109737402679,
+      "learning_rate": 1.9983752730682132e-05,
+      "loss": 1.1662,
+      "step": 311
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.660705540197382,
+      "learning_rate": 1.9983469019163976e-05,
+      "loss": 1.1834,
+      "step": 312
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.477088184986276,
+      "learning_rate": 1.998318285398913e-05,
+      "loss": 1.1383,
+      "step": 313
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.494691763849153,
+      "learning_rate": 1.998289423522793e-05,
+      "loss": 1.1171,
+      "step": 314
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.583589314263017,
+      "learning_rate": 1.9982603162951303e-05,
+      "loss": 1.0342,
+      "step": 315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.239258176680355,
+      "learning_rate": 1.9982309637230796e-05,
+      "loss": 1.0779,
+      "step": 316
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.533301393135521,
+      "learning_rate": 1.9982013658138544e-05,
+      "loss": 1.1758,
+      "step": 317
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.160943400606343,
+      "learning_rate": 1.9981715225747287e-05,
+      "loss": 0.993,
+      "step": 318
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.611711397681244,
+      "learning_rate": 1.998141434013038e-05,
+      "loss": 1.061,
+      "step": 319
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.98935758205428,
+      "learning_rate": 1.9981111001361762e-05,
+      "loss": 1.1703,
+      "step": 320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.475805740319163,
+      "learning_rate": 1.9980805209515995e-05,
+      "loss": 1.1512,
+      "step": 321
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.976352036289246,
+      "learning_rate": 1.9980496964668228e-05,
+      "loss": 1.059,
+      "step": 322
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.779970748595774,
+      "learning_rate": 1.998018626689422e-05,
+      "loss": 1.1983,
+      "step": 323
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.610805708531752,
+      "learning_rate": 1.9979873116270333e-05,
+      "loss": 1.1178,
+      "step": 324
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.4942074564889785,
+      "learning_rate": 1.997955751287353e-05,
+      "loss": 1.0686,
+      "step": 325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.349619273852159,
+      "learning_rate": 1.997923945678138e-05,
+      "loss": 1.1128,
+      "step": 326
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 102.78795406026858,
+      "learning_rate": 1.9978918948072047e-05,
+      "loss": 1.5781,
+      "step": 327
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.167255102561554,
+      "learning_rate": 1.997859598682431e-05,
+      "loss": 1.1255,
+      "step": 328
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.462766578218212,
+      "learning_rate": 1.997827057311753e-05,
+      "loss": 1.1399,
+      "step": 329
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.196999251460958,
+      "learning_rate": 1.99779427070317e-05,
+      "loss": 1.0407,
+      "step": 330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.672166926366032,
+      "learning_rate": 1.99776123886474e-05,
+      "loss": 1.1284,
+      "step": 331
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.394708768001454,
+      "learning_rate": 1.9977279618045798e-05,
+      "loss": 1.1958,
+      "step": 332
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.585192132933774,
+      "learning_rate": 1.9976944395308696e-05,
+      "loss": 1.0985,
+      "step": 333
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.732639403301168,
+      "learning_rate": 1.9976606720518474e-05,
+      "loss": 0.9398,
+      "step": 334
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.385645024123697,
+      "learning_rate": 1.9976266593758123e-05,
+      "loss": 1.1773,
+      "step": 335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.883495168249728,
+      "learning_rate": 1.9975924015111243e-05,
+      "loss": 0.9511,
+      "step": 336
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.378627617287336,
+      "learning_rate": 1.9975578984662017e-05,
+      "loss": 0.9816,
+      "step": 337
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.272296105323741,
+      "learning_rate": 1.9975231502495255e-05,
+      "loss": 1.0512,
+      "step": 338
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.1752536572488035,
+      "learning_rate": 1.9974881568696358e-05,
+      "loss": 1.1703,
+      "step": 339
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.280722914541011,
+      "learning_rate": 1.997452918335133e-05,
+      "loss": 1.1305,
+      "step": 340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.659806010735759,
+      "learning_rate": 1.9974174346546767e-05,
+      "loss": 1.0883,
+      "step": 341
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.021467650646564,
+      "learning_rate": 1.9973817058369892e-05,
+      "loss": 1.21,
+      "step": 342
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.73234270501322,
+      "learning_rate": 1.9973457318908508e-05,
+      "loss": 0.982,
+      "step": 343
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.808261971142038,
+      "learning_rate": 1.997309512825103e-05,
+      "loss": 1.1985,
+      "step": 344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.862801194088539,
+      "learning_rate": 1.9972730486486476e-05,
+      "loss": 1.158,
+      "step": 345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.679447242215395,
+      "learning_rate": 1.9972363393704466e-05,
+      "loss": 1.1748,
+      "step": 346
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.0787881510678945,
+      "learning_rate": 1.997199384999522e-05,
+      "loss": 1.0843,
+      "step": 347
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.801558853286235,
+      "learning_rate": 1.997162185544956e-05,
+      "loss": 1.0041,
+      "step": 348
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.679054095707912,
+      "learning_rate": 1.9971247410158908e-05,
+      "loss": 0.972,
+      "step": 349
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.383388745457075,
+      "learning_rate": 1.99708705142153e-05,
+      "loss": 1.0952,
+      "step": 350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.590760880880462,
+      "learning_rate": 1.9970491167711365e-05,
+      "loss": 1.0535,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.253862279071871,
+      "learning_rate": 1.9970109370740333e-05,
+      "loss": 1.1829,
+      "step": 352
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.440583224131077,
+      "learning_rate": 1.996972512339604e-05,
+      "loss": 1.0307,
+      "step": 353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.708308805584549,
+      "learning_rate": 1.9969338425772918e-05,
+      "loss": 1.1596,
+      "step": 354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.982991930831451,
+      "learning_rate": 1.996894927796602e-05,
+      "loss": 1.0943,
+      "step": 355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.5351066336760235,
+      "learning_rate": 1.9968557680070972e-05,
+      "loss": 1.2282,
+      "step": 356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.088719223079842,
+      "learning_rate": 1.996816363218403e-05,
+      "loss": 1.018,
+      "step": 357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.403197360776069,
+      "learning_rate": 1.9967767134402033e-05,
+      "loss": 1.0434,
+      "step": 358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.3813635695825655,
+      "learning_rate": 1.9967368186822428e-05,
+      "loss": 0.9905,
+      "step": 359
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.229914743790464,
+      "learning_rate": 1.996696678954327e-05,
+      "loss": 1.1482,
+      "step": 360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.770387275611607,
+      "learning_rate": 1.996656294266321e-05,
+      "loss": 1.1022,
+      "step": 361
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.874988333761614,
+      "learning_rate": 1.9966156646281502e-05,
+      "loss": 1.1424,
+      "step": 362
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9995638240159233,
+      "learning_rate": 1.9965747900498002e-05,
+      "loss": 0.9644,
+      "step": 363
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9612943598465606,
+      "learning_rate": 1.9965336705413167e-05,
+      "loss": 1.1147,
+      "step": 364
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.956813036420114,
+      "learning_rate": 1.996492306112806e-05,
+      "loss": 1.0985,
+      "step": 365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.552523678043079,
+      "learning_rate": 1.9964506967744336e-05,
+      "loss": 1.0996,
+      "step": 366
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.438890757180122,
+      "learning_rate": 1.9964088425364267e-05,
+      "loss": 1.1752,
+      "step": 367
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.816508118861078,
+      "learning_rate": 1.9963667434090717e-05,
+      "loss": 1.104,
+      "step": 368
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.858878038488043,
+      "learning_rate": 1.9963243994027157e-05,
+      "loss": 1.1146,
+      "step": 369
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.368493036325697,
+      "learning_rate": 1.9962818105277648e-05,
+      "loss": 1.1827,
+      "step": 370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.1246671758312585,
+      "learning_rate": 1.9962389767946867e-05,
+      "loss": 1.0241,
+      "step": 371
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9237526131257954,
+      "learning_rate": 1.9961958982140088e-05,
+      "loss": 1.035,
+      "step": 372
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9771041368400177,
+      "learning_rate": 1.996152574796318e-05,
+      "loss": 1.004,
+      "step": 373
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.265715900162166,
+      "learning_rate": 1.996109006552263e-05,
+      "loss": 1.1111,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.32195360794682,
+      "learning_rate": 1.9960651934925514e-05,
+      "loss": 1.0701,
+      "step": 375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.977376902921775,
+      "learning_rate": 1.9960211356279502e-05,
+      "loss": 1.0774,
+      "step": 376
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.4335843341897005,
+      "learning_rate": 1.9959768329692884e-05,
+      "loss": 1.1743,
+      "step": 377
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.873209849255563,
+      "learning_rate": 1.995932285527455e-05,
+      "loss": 1.0233,
+      "step": 378
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.401801021300256,
+      "learning_rate": 1.995887493313397e-05,
+      "loss": 1.1243,
+      "step": 379
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.386100362770559,
+      "learning_rate": 1.9958424563381245e-05,
+      "loss": 1.1078,
+      "step": 380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.403668567877387,
+      "learning_rate": 1.9957971746127052e-05,
+      "loss": 1.0611,
+      "step": 381
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.496062950052071,
+      "learning_rate": 1.9957516481482686e-05,
+      "loss": 1.0897,
+      "step": 382
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.1495625285379605,
+      "learning_rate": 1.9957058769560042e-05,
+      "loss": 1.1743,
+      "step": 383
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.460766501417975,
+      "learning_rate": 1.9956598610471603e-05,
+      "loss": 1.1613,
+      "step": 384
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.389722614887968,
+      "learning_rate": 1.995613600433047e-05,
+      "loss": 1.0859,
+      "step": 385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.576442531385308,
+      "learning_rate": 1.995567095125034e-05,
+      "loss": 1.0435,
+      "step": 386
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.237018164006678,
+      "learning_rate": 1.9955203451345507e-05,
+      "loss": 1.0716,
+      "step": 387
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.028493746861455,
+      "learning_rate": 1.9954733504730868e-05,
+      "loss": 1.0211,
+      "step": 388
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.362387612595332,
+      "learning_rate": 1.9954261111521926e-05,
+      "loss": 1.0603,
+      "step": 389
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.280742489164971,
+      "learning_rate": 1.9953786271834777e-05,
+      "loss": 1.0198,
+      "step": 390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.343266350378013,
+      "learning_rate": 1.9953308985786132e-05,
+      "loss": 1.1851,
+      "step": 391
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.382126821153117,
+      "learning_rate": 1.9952829253493287e-05,
+      "loss": 1.1069,
+      "step": 392
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.4645047350214515,
+      "learning_rate": 1.995234707507415e-05,
+      "loss": 1.0897,
+      "step": 393
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.347303636479825,
+      "learning_rate": 1.9951862450647227e-05,
+      "loss": 1.0755,
+      "step": 394
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.687625133699968,
+      "learning_rate": 1.9951375380331622e-05,
+      "loss": 1.0315,
+      "step": 395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.909428928532073,
+      "learning_rate": 1.9950885864247043e-05,
+      "loss": 1.1643,
+      "step": 396
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.578689500603283,
+      "learning_rate": 1.9950393902513804e-05,
+      "loss": 1.0898,
+      "step": 397
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.609108223587599,
+      "learning_rate": 1.9949899495252816e-05,
+      "loss": 1.0755,
+      "step": 398
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.841143079264003,
+      "learning_rate": 1.9949402642585585e-05,
+      "loss": 1.1456,
+      "step": 399
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.629729376883061,
+      "learning_rate": 1.9948903344634223e-05,
+      "loss": 1.065,
+      "step": 400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 55.86395081501506,
+      "learning_rate": 1.994840160152145e-05,
+      "loss": 1.9193,
+      "step": 401
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6236513738541767,
+      "learning_rate": 1.9947897413370575e-05,
+      "loss": 1.0813,
+      "step": 402
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.37260215681647,
+      "learning_rate": 1.9947390780305515e-05,
+      "loss": 1.2356,
+      "step": 403
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.394105492668787,
+      "learning_rate": 1.9946881702450788e-05,
+      "loss": 1.0607,
+      "step": 404
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 89.50489669809592,
+      "learning_rate": 1.994637017993151e-05,
+      "loss": 1.5628,
+      "step": 405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.597609858089008,
+      "learning_rate": 1.9945856212873394e-05,
+      "loss": 1.0847,
+      "step": 406
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.459506128727871,
+      "learning_rate": 1.9945339801402768e-05,
+      "loss": 1.0432,
+      "step": 407
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.071685345913883,
+      "learning_rate": 1.9944820945646543e-05,
+      "loss": 1.0675,
+      "step": 408
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.231922265254759,
+      "learning_rate": 1.9944299645732242e-05,
+      "loss": 1.0479,
+      "step": 409
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.389508171925526,
+      "learning_rate": 1.9943775901787988e-05,
+      "loss": 1.0723,
+      "step": 410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.503983891850023,
+      "learning_rate": 1.99432497139425e-05,
+      "loss": 1.0749,
+      "step": 411
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.152011570105329,
+      "learning_rate": 1.9942721082325103e-05,
+      "loss": 1.0135,
+      "step": 412
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 50.68316350833893,
+      "learning_rate": 1.9942190007065715e-05,
+      "loss": 1.1019,
+      "step": 413
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.500780906339276,
+      "learning_rate": 1.9941656488294865e-05,
+      "loss": 1.0915,
+      "step": 414
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.191607605165063,
+      "learning_rate": 1.9941120526143673e-05,
+      "loss": 1.0326,
+      "step": 415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.330054099494727,
+      "learning_rate": 1.9940582120743867e-05,
+      "loss": 0.9911,
+      "step": 416
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.511469571369914,
+      "learning_rate": 1.9940041272227767e-05,
+      "loss": 1.1471,
+      "step": 417
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.329974316823594,
+      "learning_rate": 1.9939497980728305e-05,
+      "loss": 1.0064,
+      "step": 418
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7911891744849937,
+      "learning_rate": 1.9938952246378996e-05,
+      "loss": 1.0258,
+      "step": 419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.619262957851557,
+      "learning_rate": 1.993840406931398e-05,
+      "loss": 0.9322,
+      "step": 420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.866106813938292,
+      "learning_rate": 1.9937853449667975e-05,
+      "loss": 1.0814,
+      "step": 421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.454545061711105,
+      "learning_rate": 1.993730038757631e-05,
+      "loss": 1.1405,
+      "step": 422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.344621327807707,
+      "learning_rate": 1.993674488317491e-05,
+      "loss": 1.202,
+      "step": 423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.97729689010236,
+      "learning_rate": 1.9936186936600307e-05,
+      "loss": 1.1599,
+      "step": 424
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.863469621733676,
+      "learning_rate": 1.9935626547989623e-05,
+      "loss": 1.0807,
+      "step": 425
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.242651785752871,
+      "learning_rate": 1.9935063717480587e-05,
+      "loss": 1.0514,
+      "step": 426
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.455541990923886,
+      "learning_rate": 1.9934498445211533e-05,
+      "loss": 1.0635,
+      "step": 427
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 52.6839859691502,
+      "learning_rate": 1.993393073132138e-05,
+      "loss": 1.2474,
+      "step": 428
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.836687324373342,
+      "learning_rate": 1.9933360575949666e-05,
+      "loss": 0.9968,
+      "step": 429
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.629865348213475,
+      "learning_rate": 1.993278797923651e-05,
+      "loss": 1.1198,
+      "step": 430
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.358351299783933,
+      "learning_rate": 1.9932212941322647e-05,
+      "loss": 1.0978,
+      "step": 431
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9377679799485334,
+      "learning_rate": 1.99316354623494e-05,
+      "loss": 1.0485,
+      "step": 432
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.6018038205612255,
+      "learning_rate": 1.99310555424587e-05,
+      "loss": 1.0514,
+      "step": 433
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.0440373728530785,
+      "learning_rate": 1.9930473181793072e-05,
+      "loss": 0.9412,
+      "step": 434
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.2822041977740986,
+      "learning_rate": 1.992988838049565e-05,
+      "loss": 0.9845,
+      "step": 435
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.3024294123951226,
+      "learning_rate": 1.992930113871016e-05,
+      "loss": 1.0778,
+      "step": 436
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.426499951361489,
+      "learning_rate": 1.9928711456580924e-05,
+      "loss": 1.0527,
+      "step": 437
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.97926425301124,
+      "learning_rate": 1.9928119334252874e-05,
+      "loss": 1.0938,
+      "step": 438
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.494487628691639,
+      "learning_rate": 1.9927524771871537e-05,
+      "loss": 1.0805,
+      "step": 439
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.1492409727708885,
+      "learning_rate": 1.992692776958304e-05,
+      "loss": 1.1888,
+      "step": 440
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7536887124074765,
+      "learning_rate": 1.9926328327534108e-05,
+      "loss": 1.1237,
+      "step": 441
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.17207580577035,
+      "learning_rate": 1.9925726445872064e-05,
+      "loss": 1.117,
+      "step": 442
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.034264466789217,
+      "learning_rate": 1.992512212474484e-05,
+      "loss": 1.038,
+      "step": 443
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.935357931179911,
+      "learning_rate": 1.992451536430096e-05,
+      "loss": 0.9145,
+      "step": 444
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.188855247069045,
+      "learning_rate": 1.9923906164689545e-05,
+      "loss": 1.1691,
+      "step": 445
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.872423420867507,
+      "learning_rate": 1.9923294526060318e-05,
+      "loss": 0.9843,
+      "step": 446
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.043213930905514,
+      "learning_rate": 1.992268044856361e-05,
+      "loss": 1.2413,
+      "step": 447
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8760649774844254,
+      "learning_rate": 1.992206393235034e-05,
+      "loss": 1.0825,
+      "step": 448
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.574510132496088,
+      "learning_rate": 1.9921444977572026e-05,
+      "loss": 1.0488,
+      "step": 449
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.872733544881923,
+      "learning_rate": 1.9920823584380797e-05,
+      "loss": 1.212,
+      "step": 450
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.738193513731524,
+      "learning_rate": 1.992019975292937e-05,
+      "loss": 1.2698,
+      "step": 451
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.1666626790638075,
+      "learning_rate": 1.9919573483371065e-05,
+      "loss": 1.0719,
+      "step": 452
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.152054735117916,
+      "learning_rate": 1.9918944775859805e-05,
+      "loss": 1.0218,
+      "step": 453
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.0873913265028285,
+      "learning_rate": 1.9918313630550107e-05,
+      "loss": 1.1434,
+      "step": 454
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9934330107747416,
+      "learning_rate": 1.9917680047597085e-05,
+      "loss": 1.1273,
+      "step": 455
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.376873182361851,
+      "learning_rate": 1.9917044027156462e-05,
+      "loss": 1.06,
+      "step": 456
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.715172465931275,
+      "learning_rate": 1.9916405569384548e-05,
+      "loss": 1.0255,
+      "step": 457
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 68.34701161214348,
+      "learning_rate": 1.9915764674438263e-05,
+      "loss": 1.2339,
+      "step": 458
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.665587219676493,
+      "learning_rate": 1.991512134247512e-05,
+      "loss": 1.0084,
+      "step": 459
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8415418746312766,
+      "learning_rate": 1.9914475573653228e-05,
+      "loss": 1.0346,
+      "step": 460
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8246643408106213,
+      "learning_rate": 1.9913827368131303e-05,
+      "loss": 1.0004,
+      "step": 461
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.114748947891342,
+      "learning_rate": 1.991317672606866e-05,
+      "loss": 1.1898,
+      "step": 462
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6777417567463417,
+      "learning_rate": 1.9912523647625195e-05,
+      "loss": 1.1035,
+      "step": 463
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.9174622373220895,
+      "learning_rate": 1.991186813296143e-05,
+      "loss": 0.9963,
+      "step": 464
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.6798659957889805,
+      "learning_rate": 1.9911210182238462e-05,
+      "loss": 1.1222,
+      "step": 465
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.879278324637952,
+      "learning_rate": 1.9910549795618003e-05,
+      "loss": 1.0055,
+      "step": 466
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.0401268558658145,
+      "learning_rate": 1.9909886973262356e-05,
+      "loss": 1.1659,
+      "step": 467
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 20.71395567573766,
+      "learning_rate": 1.9909221715334428e-05,
+      "loss": 1.1785,
+      "step": 468
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9734760731086074,
+      "learning_rate": 1.9908554021997715e-05,
+      "loss": 1.0212,
+      "step": 469
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.3755313994964915,
+      "learning_rate": 1.990788389341632e-05,
+      "loss": 0.9178,
+      "step": 470
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9667867009878366,
+      "learning_rate": 1.990721132975494e-05,
+      "loss": 1.139,
+      "step": 471
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.996757965018768,
+      "learning_rate": 1.9906536331178873e-05,
+      "loss": 1.0152,
+      "step": 472
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.065537265184598,
+      "learning_rate": 1.9905858897854013e-05,
+      "loss": 1.0876,
+      "step": 473
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.853734590158613,
+      "learning_rate": 1.990517902994686e-05,
+      "loss": 1.0335,
+      "step": 474
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7530342271361223,
+      "learning_rate": 1.9904496727624498e-05,
+      "loss": 1.1245,
+      "step": 475
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.770253015681632,
+      "learning_rate": 1.9903811991054628e-05,
+      "loss": 1.0104,
+      "step": 476
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.957710365597872,
+      "learning_rate": 1.990312482040553e-05,
+      "loss": 1.0127,
+      "step": 477
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9657426975966645,
+      "learning_rate": 1.9902435215846096e-05,
+      "loss": 1.1692,
+      "step": 478
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.962760636201044,
+      "learning_rate": 1.9901743177545807e-05,
+      "loss": 1.0969,
+      "step": 479
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.0883879467739215,
+      "learning_rate": 1.9901048705674752e-05,
+      "loss": 1.0422,
+      "step": 480
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.118434571260297,
+      "learning_rate": 1.990035180040361e-05,
+      "loss": 1.0369,
+      "step": 481
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 25.230164702697778,
+      "learning_rate": 1.9899652461903662e-05,
+      "loss": 1.0999,
+      "step": 482
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 11.068930128392779,
+      "learning_rate": 1.9898950690346784e-05,
+      "loss": 1.0602,
+      "step": 483
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.64858207036738,
+      "learning_rate": 1.9898246485905456e-05,
+      "loss": 1.238,
+      "step": 484
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.167629424192774,
+      "learning_rate": 1.9897539848752743e-05,
+      "loss": 0.9315,
+      "step": 485
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.325684939850164,
+      "learning_rate": 1.9896830779062325e-05,
+      "loss": 1.0613,
+      "step": 486
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.591801602541418,
+      "learning_rate": 1.989611927700847e-05,
+      "loss": 1.1534,
+      "step": 487
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8731127863301777,
+      "learning_rate": 1.9895405342766044e-05,
+      "loss": 1.0965,
+      "step": 488
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9256611935348227,
+      "learning_rate": 1.989468897651051e-05,
+      "loss": 1.0757,
+      "step": 489
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.842769298855828,
+      "learning_rate": 1.9893970178417933e-05,
+      "loss": 1.0648,
+      "step": 490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.188248922725501,
+      "learning_rate": 1.989324894866497e-05,
+      "loss": 1.0819,
+      "step": 491
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.498270743406448,
+      "learning_rate": 1.9892525287428885e-05,
+      "loss": 1.0548,
+      "step": 492
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.064522194976723,
+      "learning_rate": 1.989179919488753e-05,
+      "loss": 1.177,
+      "step": 493
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.192200466678487,
+      "learning_rate": 1.9891070671219358e-05,
+      "loss": 1.0633,
+      "step": 494
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6610634108607893,
+      "learning_rate": 1.9890339716603424e-05,
+      "loss": 1.0271,
+      "step": 495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.19383464605963,
+      "learning_rate": 1.988960633121937e-05,
+      "loss": 1.1628,
+      "step": 496
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.324148856756533,
+      "learning_rate": 1.9888870515247445e-05,
+      "loss": 1.1725,
+      "step": 497
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.094678732832611,
+      "learning_rate": 1.988813226886849e-05,
+      "loss": 1.1152,
+      "step": 498
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 100.17236461049514,
+      "learning_rate": 1.9887391592263947e-05,
+      "loss": 1.713,
+      "step": 499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.200236194076793,
+      "learning_rate": 1.9886648485615852e-05,
+      "loss": 1.0963,
+      "step": 500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.654556285087075,
+      "learning_rate": 1.9885902949106842e-05,
+      "loss": 1.1179,
+      "step": 501
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.559411065207602,
+      "learning_rate": 1.988515498292015e-05,
+      "loss": 1.0149,
+      "step": 502
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.07254697409706,
+      "learning_rate": 1.9884404587239597e-05,
+      "loss": 1.0651,
+      "step": 503
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8097851753577676,
+      "learning_rate": 1.9883651762249618e-05,
+      "loss": 0.968,
+      "step": 504
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.229524204117181,
+      "learning_rate": 1.9882896508135236e-05,
+      "loss": 1.102,
+      "step": 505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8740072573172974,
+      "learning_rate": 1.9882138825082066e-05,
+      "loss": 1.1369,
+      "step": 506
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7317673445914985,
+      "learning_rate": 1.9881378713276323e-05,
+      "loss": 0.9885,
+      "step": 507
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.295999670061802,
+      "learning_rate": 1.9880616172904833e-05,
+      "loss": 1.0094,
+      "step": 508
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.281729889433247,
+      "learning_rate": 1.9879851204154996e-05,
+      "loss": 0.9946,
+      "step": 509
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.396280320607497,
+      "learning_rate": 1.9879083807214827e-05,
+      "loss": 1.0109,
+      "step": 510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.20457733433576,
+      "learning_rate": 1.9878313982272926e-05,
+      "loss": 1.179,
+      "step": 511
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0518090897332595,
+      "learning_rate": 1.9877541729518496e-05,
+      "loss": 1.1367,
+      "step": 512
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8944019034517012,
+      "learning_rate": 1.9876767049141334e-05,
+      "loss": 1.0746,
+      "step": 513
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7920745114516183,
+      "learning_rate": 1.987598994133184e-05,
+      "loss": 1.1664,
+      "step": 514
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9722247307359373,
+      "learning_rate": 1.9875210406280993e-05,
+      "loss": 1.1303,
+      "step": 515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.182258325736007,
+      "learning_rate": 1.9874428444180395e-05,
+      "loss": 0.9902,
+      "step": 516
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.592378225884718,
+      "learning_rate": 1.987364405522222e-05,
+      "loss": 1.1012,
+      "step": 517
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.105204210476816,
+      "learning_rate": 1.9872857239599254e-05,
+      "loss": 1.1137,
+      "step": 518
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.167332270422541,
+      "learning_rate": 1.9872067997504873e-05,
+      "loss": 1.1159,
+      "step": 519
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.08947117736944,
+      "learning_rate": 1.9871276329133054e-05,
+      "loss": 1.0862,
+      "step": 520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.049584156774763,
+      "learning_rate": 1.9870482234678362e-05,
+      "loss": 1.023,
+      "step": 521
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.232270233273242,
+      "learning_rate": 1.9869685714335966e-05,
+      "loss": 1.0735,
+      "step": 522
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.2805838206307625,
+      "learning_rate": 1.986888676830162e-05,
+      "loss": 1.0401,
+      "step": 523
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.004363644777837,
+      "learning_rate": 1.9868085396771696e-05,
+      "loss": 1.0112,
+      "step": 524
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.1293441807832805,
+      "learning_rate": 1.9867281599943143e-05,
+      "loss": 1.0584,
+      "step": 525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.060908520373464,
+      "learning_rate": 1.986647537801351e-05,
+      "loss": 0.9488,
+      "step": 526
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.247746303908252,
+      "learning_rate": 1.9865666731180946e-05,
+      "loss": 1.0041,
+      "step": 527
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.175237819303534,
+      "learning_rate": 1.9864855659644188e-05,
+      "loss": 1.0653,
+      "step": 528
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7167967522781247,
+      "learning_rate": 1.986404216360258e-05,
+      "loss": 1.0688,
+      "step": 529
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.819542500669309,
+      "learning_rate": 1.986322624325606e-05,
+      "loss": 1.1037,
+      "step": 530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.298779141385325,
+      "learning_rate": 1.986240789880515e-05,
+      "loss": 1.1797,
+      "step": 531
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7973224850258624,
+      "learning_rate": 1.9861587130450983e-05,
+      "loss": 1.0882,
+      "step": 532
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 14.693961767865298,
+      "learning_rate": 1.9860763938395276e-05,
+      "loss": 1.0669,
+      "step": 533
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.07838118819138,
+      "learning_rate": 1.985993832284035e-05,
+      "loss": 1.1385,
+      "step": 534
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.39069818425914,
+      "learning_rate": 1.9859110283989115e-05,
+      "loss": 1.0543,
+      "step": 535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.034497990877742,
+      "learning_rate": 1.985827982204508e-05,
+      "loss": 1.096,
+      "step": 536
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.960533593233808,
+      "learning_rate": 1.9857446937212354e-05,
+      "loss": 1.0496,
+      "step": 537
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.309808184319671,
+      "learning_rate": 1.9856611629695632e-05,
+      "loss": 0.996,
+      "step": 538
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5147331155631987,
+      "learning_rate": 1.985577389970021e-05,
+      "loss": 1.0178,
+      "step": 539
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7971287131584917,
+      "learning_rate": 1.9854933747431978e-05,
+      "loss": 1.0094,
+      "step": 540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.386234386949602,
+      "learning_rate": 1.9854091173097423e-05,
+      "loss": 1.1386,
+      "step": 541
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.000079644424622,
+      "learning_rate": 1.9853246176903628e-05,
+      "loss": 0.9622,
+      "step": 542
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.339334686890175,
+      "learning_rate": 1.9852398759058267e-05,
+      "loss": 1.0296,
+      "step": 543
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7739468015723268,
+      "learning_rate": 1.985154891976961e-05,
+      "loss": 1.0136,
+      "step": 544
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0151575509057436,
+      "learning_rate": 1.9850696659246527e-05,
+      "loss": 1.1141,
+      "step": 545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.124140334325582,
+      "learning_rate": 1.984984197769848e-05,
+      "loss": 1.068,
+      "step": 546
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.798960767229643,
+      "learning_rate": 1.984898487533552e-05,
+      "loss": 1.042,
+      "step": 547
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.88888530591446,
+      "learning_rate": 1.9848125352368304e-05,
+      "loss": 1.0092,
+      "step": 548
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.590651116963613,
+      "learning_rate": 1.984726340900808e-05,
+      "loss": 1.0699,
+      "step": 549
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.111484667737962,
+      "learning_rate": 1.9846399045466683e-05,
+      "loss": 1.1293,
+      "step": 550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.1949741389531106,
+      "learning_rate": 1.9845532261956556e-05,
+      "loss": 0.9913,
+      "step": 551
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.577889355072513,
+      "learning_rate": 1.984466305869073e-05,
+      "loss": 0.9844,
+      "step": 552
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9399427915413283,
+      "learning_rate": 1.9843791435882823e-05,
+      "loss": 1.1503,
+      "step": 553
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.8808408631804285,
+      "learning_rate": 1.9842917393747063e-05,
+      "loss": 1.0585,
+      "step": 554
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.148362947766365,
+      "learning_rate": 1.984204093249826e-05,
+      "loss": 1.0008,
+      "step": 555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.022762411797265,
+      "learning_rate": 1.984116205235183e-05,
+      "loss": 1.0576,
+      "step": 556
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7848552941502684,
+      "learning_rate": 1.984028075352377e-05,
+      "loss": 1.1343,
+      "step": 557
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.194440701612392,
+      "learning_rate": 1.9839397036230683e-05,
+      "loss": 1.1125,
+      "step": 558
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8417025422559186,
+      "learning_rate": 1.983851090068976e-05,
+      "loss": 1.0576,
+      "step": 559
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6424172482294517,
+      "learning_rate": 1.983762234711879e-05,
+      "loss": 1.0682,
+      "step": 560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8586575805191674,
+      "learning_rate": 1.9836731375736152e-05,
+      "loss": 1.14,
+      "step": 561
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7510893528582527,
+      "learning_rate": 1.9835837986760826e-05,
+      "loss": 1.0344,
+      "step": 562
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8841637756939287,
+      "learning_rate": 1.983494218041238e-05,
+      "loss": 1.0781,
+      "step": 563
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.638573793002325,
+      "learning_rate": 1.9834043956910977e-05,
+      "loss": 1.0021,
+      "step": 564
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.84261917367866,
+      "learning_rate": 1.9833143316477373e-05,
+      "loss": 1.0403,
+      "step": 565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.99726497722348,
+      "learning_rate": 1.9832240259332926e-05,
+      "loss": 1.129,
+      "step": 566
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.287995225308431,
+      "learning_rate": 1.9831334785699573e-05,
+      "loss": 1.0588,
+      "step": 567
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.517804901348297,
+      "learning_rate": 1.9830426895799863e-05,
+      "loss": 1.0125,
+      "step": 568
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.016138201292742,
+      "learning_rate": 1.9829516589856927e-05,
+      "loss": 1.1118,
+      "step": 569
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7553954379546117,
+      "learning_rate": 1.9828603868094493e-05,
+      "loss": 1.0246,
+      "step": 570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 27.1244431360396,
+      "learning_rate": 1.982768873073688e-05,
+      "loss": 1.1834,
+      "step": 571
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.972960301159113,
+      "learning_rate": 1.9826771178009004e-05,
+      "loss": 0.9803,
+      "step": 572
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.135770025339726,
+      "learning_rate": 1.9825851210136377e-05,
+      "loss": 1.0974,
+      "step": 573
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.661648194074585,
+      "learning_rate": 1.98249288273451e-05,
+      "loss": 1.0345,
+      "step": 574
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.2506882343689805,
+      "learning_rate": 1.9824004029861865e-05,
+      "loss": 1.0352,
+      "step": 575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.918479068734851,
+      "learning_rate": 1.9823076817913965e-05,
+      "loss": 1.0562,
+      "step": 576
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8531954074705776,
+      "learning_rate": 1.9822147191729282e-05,
+      "loss": 1.1355,
+      "step": 577
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.093074093090599,
+      "learning_rate": 1.9821215151536292e-05,
+      "loss": 1.0082,
+      "step": 578
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.74784655442073,
+      "learning_rate": 1.9820280697564062e-05,
+      "loss": 1.0291,
+      "step": 579
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8133856210226313,
+      "learning_rate": 1.981934383004226e-05,
+      "loss": 1.0908,
+      "step": 580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.275051791622801,
+      "learning_rate": 1.9818404549201134e-05,
+      "loss": 1.116,
+      "step": 581
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.717147775719318,
+      "learning_rate": 1.981746285527154e-05,
+      "loss": 1.0581,
+      "step": 582
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.179106753877097,
+      "learning_rate": 1.9816518748484918e-05,
+      "loss": 1.1118,
+      "step": 583
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4321594846881496,
+      "learning_rate": 1.9815572229073302e-05,
+      "loss": 1.1244,
+      "step": 584
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 44.024519397743994,
+      "learning_rate": 1.9814623297269318e-05,
+      "loss": 1.1965,
+      "step": 585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.688970794697736,
+      "learning_rate": 1.981367195330619e-05,
+      "loss": 0.9591,
+      "step": 586
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.613809556224252,
+      "learning_rate": 1.9812718197417732e-05,
+      "loss": 1.0524,
+      "step": 587
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 19.4274900048794,
+      "learning_rate": 1.981176202983835e-05,
+      "loss": 1.0762,
+      "step": 588
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.2882461921580655,
+      "learning_rate": 1.981080345080304e-05,
+      "loss": 1.1226,
+      "step": 589
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.470081133562788,
+      "learning_rate": 1.980984246054739e-05,
+      "loss": 1.0775,
+      "step": 590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.236591608309034,
+      "learning_rate": 1.98088790593076e-05,
+      "loss": 1.1254,
+      "step": 591
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.943827833626578,
+      "learning_rate": 1.9807913247320437e-05,
+      "loss": 1.105,
+      "step": 592
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.051124059877304,
+      "learning_rate": 1.9806945024823268e-05,
+      "loss": 1.0184,
+      "step": 593
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8137043251275897,
+      "learning_rate": 1.9805974392054057e-05,
+      "loss": 1.0157,
+      "step": 594
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8810926289689256,
+      "learning_rate": 1.980500134925136e-05,
+      "loss": 1.0248,
+      "step": 595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.249114811226602,
+      "learning_rate": 1.9804025896654323e-05,
+      "loss": 1.0927,
+      "step": 596
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.937037416045328,
+      "learning_rate": 1.9803048034502686e-05,
+      "loss": 0.9526,
+      "step": 597
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.785040156544341,
+      "learning_rate": 1.9802067763036777e-05,
+      "loss": 1.0215,
+      "step": 598
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8696537498290473,
+      "learning_rate": 1.9801085082497524e-05,
+      "loss": 1.0136,
+      "step": 599
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.042449573541756,
+      "learning_rate": 1.980009999312644e-05,
+      "loss": 0.9528,
+      "step": 600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.901999977345569,
+      "learning_rate": 1.979911249516563e-05,
+      "loss": 1.0932,
+      "step": 601
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9792433955931603,
+      "learning_rate": 1.9798122588857792e-05,
+      "loss": 1.0753,
+      "step": 602
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.855871910802278,
+      "learning_rate": 1.9797130274446226e-05,
+      "loss": 0.9931,
+      "step": 603
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8407523106332966,
+      "learning_rate": 1.9796135552174806e-05,
+      "loss": 1.052,
+      "step": 604
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6097273941283787,
+      "learning_rate": 1.979513842228801e-05,
+      "loss": 1.0838,
+      "step": 605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7398132484603246,
+      "learning_rate": 1.9794138885030904e-05,
+      "loss": 1.0803,
+      "step": 606
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9933304582799956,
+      "learning_rate": 1.979313694064915e-05,
+      "loss": 1.0301,
+      "step": 607
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8702263071544802,
+      "learning_rate": 1.979213258938899e-05,
+      "loss": 1.091,
+      "step": 608
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.532117228434824,
+      "learning_rate": 1.9791125831497275e-05,
+      "loss": 1.0777,
+      "step": 609
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6877373194692176,
+      "learning_rate": 1.979011666722143e-05,
+      "loss": 1.0882,
+      "step": 610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6714087433404483,
+      "learning_rate": 1.9789105096809486e-05,
+      "loss": 1.0265,
+      "step": 611
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.253830373999391,
+      "learning_rate": 1.9788091120510054e-05,
+      "loss": 1.0392,
+      "step": 612
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.383486203412253,
+      "learning_rate": 1.978707473857234e-05,
+      "loss": 0.9853,
+      "step": 613
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9207782444723676,
+      "learning_rate": 1.9786055951246145e-05,
+      "loss": 1.0359,
+      "step": 614
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.27299707690402,
+      "learning_rate": 1.978503475878186e-05,
+      "loss": 1.0598,
+      "step": 615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.611938859351543,
+      "learning_rate": 1.978401116143046e-05,
+      "loss": 1.0947,
+      "step": 616
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.110523917481957,
+      "learning_rate": 1.9782985159443516e-05,
+      "loss": 1.0986,
+      "step": 617
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9721702902162335,
+      "learning_rate": 1.97819567530732e-05,
+      "loss": 0.9685,
+      "step": 618
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6766477960601947,
+      "learning_rate": 1.9780925942572254e-05,
+      "loss": 0.9494,
+      "step": 619
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.257434563692485,
+      "learning_rate": 1.9779892728194033e-05,
+      "loss": 1.1736,
+      "step": 620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7810009161622755,
+      "learning_rate": 1.977885711019246e-05,
+      "loss": 0.9732,
+      "step": 621
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.154657757370879,
+      "learning_rate": 1.977781908882207e-05,
+      "loss": 1.0125,
+      "step": 622
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.90341959750237,
+      "learning_rate": 1.9776778664337983e-05,
+      "loss": 1.0735,
+      "step": 623
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.029745128260756,
+      "learning_rate": 1.97757358369959e-05,
+      "loss": 1.0611,
+      "step": 624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6947993759190734,
+      "learning_rate": 1.9774690607052113e-05,
+      "loss": 0.9721,
+      "step": 625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5784447673937834,
+      "learning_rate": 1.9773642974763518e-05,
+      "loss": 1.0404,
+      "step": 626
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.071737235973688,
+      "learning_rate": 1.9772592940387592e-05,
+      "loss": 0.997,
+      "step": 627
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.618782283976266,
+      "learning_rate": 1.9771540504182404e-05,
+      "loss": 1.148,
+      "step": 628
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3777647657574272,
+      "learning_rate": 1.9770485666406612e-05,
+      "loss": 1.0647,
+      "step": 629
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.811698216066833,
+      "learning_rate": 1.976942842731947e-05,
+      "loss": 1.1163,
+      "step": 630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8554878867521087,
+      "learning_rate": 1.976836878718081e-05,
+      "loss": 1.0056,
+      "step": 631
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5595599483392077,
+      "learning_rate": 1.9767306746251073e-05,
+      "loss": 1.0362,
+      "step": 632
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.355698503630512,
+      "learning_rate": 1.976624230479127e-05,
+      "loss": 1.0318,
+      "step": 633
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5800019812910033,
+      "learning_rate": 1.976517546306301e-05,
+      "loss": 1.143,
+      "step": 634
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.109303971927999,
+      "learning_rate": 1.9764106221328495e-05,
+      "loss": 1.1507,
+      "step": 635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 11.341830505198706,
+      "learning_rate": 1.976303457985052e-05,
+      "loss": 0.9271,
+      "step": 636
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.158302554726382,
+      "learning_rate": 1.9761960538892456e-05,
+      "loss": 1.0866,
+      "step": 637
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6323990269274447,
+      "learning_rate": 1.9760884098718277e-05,
+      "loss": 1.0624,
+      "step": 638
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.593221906283679,
+      "learning_rate": 1.9759805259592543e-05,
+      "loss": 0.9801,
+      "step": 639
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8310072853709722,
+      "learning_rate": 1.97587240217804e-05,
+      "loss": 1.1142,
+      "step": 640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.708340807762302,
+      "learning_rate": 1.975764038554758e-05,
+      "loss": 1.065,
+      "step": 641
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.602682523381259,
+      "learning_rate": 1.9756554351160423e-05,
+      "loss": 1.0641,
+      "step": 642
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.677081539155567,
+      "learning_rate": 1.9755465918885836e-05,
+      "loss": 1.0378,
+      "step": 643
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7668871395570487,
+      "learning_rate": 1.975437508899133e-05,
+      "loss": 1.1344,
+      "step": 644
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8359506126491563,
+      "learning_rate": 1.9753281861745e-05,
+      "loss": 0.9694,
+      "step": 645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.206408482603931,
+      "learning_rate": 1.9752186237415528e-05,
+      "loss": 1.0533,
+      "step": 646
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7597389829258656,
+      "learning_rate": 1.9751088216272186e-05,
+      "loss": 1.1411,
+      "step": 647
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7120014142519246,
+      "learning_rate": 1.9749987798584845e-05,
+      "loss": 1.0662,
+      "step": 648
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9987562421524383,
+      "learning_rate": 1.9748884984623952e-05,
+      "loss": 1.0139,
+      "step": 649
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.282466973540785,
+      "learning_rate": 1.974777977466054e-05,
+      "loss": 1.0623,
+      "step": 650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.193144935671022,
+      "learning_rate": 1.9746672168966253e-05,
+      "loss": 1.0129,
+      "step": 651
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.481908768091186,
+      "learning_rate": 1.9745562167813302e-05,
+      "loss": 0.9982,
+      "step": 652
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7044461584994797,
+      "learning_rate": 1.9744449771474496e-05,
+      "loss": 1.0352,
+      "step": 653
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.21153552187076,
+      "learning_rate": 1.9743334980223224e-05,
+      "loss": 0.9555,
+      "step": 654
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.597834834338504,
+      "learning_rate": 1.9742217794333483e-05,
+      "loss": 1.0555,
+      "step": 655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.509560249870177,
+      "learning_rate": 1.9741098214079833e-05,
+      "loss": 1.122,
+      "step": 656
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.634147646970652,
+      "learning_rate": 1.9739976239737444e-05,
+      "loss": 1.1282,
+      "step": 657
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9355859624871243,
+      "learning_rate": 1.973885187158206e-05,
+      "loss": 1.0385,
+      "step": 658
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9223226174844936,
+      "learning_rate": 1.9737725109890027e-05,
+      "loss": 1.155,
+      "step": 659
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.839110455653833,
+      "learning_rate": 1.9736595954938263e-05,
+      "loss": 0.9182,
+      "step": 660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3843046936219263,
+      "learning_rate": 1.973546440700429e-05,
+      "loss": 1.0475,
+      "step": 661
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.95335162441481,
+      "learning_rate": 1.9734330466366204e-05,
+      "loss": 1.1071,
+      "step": 662
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5083022418894405,
+      "learning_rate": 1.97331941333027e-05,
+      "loss": 1.0379,
+      "step": 663
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 12.503320132172616,
+      "learning_rate": 1.9732055408093055e-05,
+      "loss": 1.095,
+      "step": 664
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8602096098013314,
+      "learning_rate": 1.973091429101714e-05,
+      "loss": 0.9738,
+      "step": 665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.567515287128138,
+      "learning_rate": 1.9729770782355402e-05,
+      "loss": 1.0115,
+      "step": 666
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.1619498585102885,
+      "learning_rate": 1.9728624882388887e-05,
+      "loss": 1.113,
+      "step": 667
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8403129128241567,
+      "learning_rate": 1.972747659139923e-05,
+      "loss": 1.0106,
+      "step": 668
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6732666376119303,
+      "learning_rate": 1.972632590966864e-05,
+      "loss": 1.0863,
+      "step": 669
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.63342630749796,
+      "learning_rate": 1.9725172837479926e-05,
+      "loss": 0.9528,
+      "step": 670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 27.87691370932522,
+      "learning_rate": 1.9724017375116482e-05,
+      "loss": 1.176,
+      "step": 671
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.683702726110297,
+      "learning_rate": 1.972285952286229e-05,
+      "loss": 1.1573,
+      "step": 672
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.617695052292551,
+      "learning_rate": 1.972169928100191e-05,
+      "loss": 1.0399,
+      "step": 673
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.297740360065807,
+      "learning_rate": 1.9720536649820506e-05,
+      "loss": 0.9832,
+      "step": 674
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9857184734105227,
+      "learning_rate": 1.9719371629603815e-05,
+      "loss": 1.0523,
+      "step": 675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5057614162854858,
+      "learning_rate": 1.9718204220638172e-05,
+      "loss": 1.0747,
+      "step": 676
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.682742988164712,
+      "learning_rate": 1.9717034423210482e-05,
+      "loss": 1.1584,
+      "step": 677
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.903662120563867,
+      "learning_rate": 1.9715862237608257e-05,
+      "loss": 1.0778,
+      "step": 678
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.647491392542745,
+      "learning_rate": 1.9714687664119588e-05,
+      "loss": 1.0308,
+      "step": 679
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.981060186015297,
+      "learning_rate": 1.9713510703033146e-05,
+      "loss": 0.9786,
+      "step": 680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9494055759818574,
+      "learning_rate": 1.97123313546382e-05,
+      "loss": 1.1111,
+      "step": 681
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7899296812945624,
+      "learning_rate": 1.9711149619224604e-05,
+      "loss": 1.0374,
+      "step": 682
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.871215861412619,
+      "learning_rate": 1.9709965497082786e-05,
+      "loss": 1.111,
+      "step": 683
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.67068255262732,
+      "learning_rate": 1.9708778988503777e-05,
+      "loss": 1.2062,
+      "step": 684
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.657131507652763,
+      "learning_rate": 1.9707590093779185e-05,
+      "loss": 1.0664,
+      "step": 685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8816693045054795,
+      "learning_rate": 1.9706398813201207e-05,
+      "loss": 0.9103,
+      "step": 686
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9452131690989654,
+      "learning_rate": 1.970520514706263e-05,
+      "loss": 1.0652,
+      "step": 687
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8074508879557953,
+      "learning_rate": 1.9704009095656818e-05,
+      "loss": 1.0803,
+      "step": 688
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6545433112322137,
+      "learning_rate": 1.9702810659277728e-05,
+      "loss": 1.0225,
+      "step": 689
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.981129458125146,
+      "learning_rate": 1.9701609838219907e-05,
+      "loss": 0.9947,
+      "step": 690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9045146363435643,
+      "learning_rate": 1.9700406632778476e-05,
+      "loss": 1.0019,
+      "step": 691
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.32347733310266,
+      "learning_rate": 1.9699201043249155e-05,
+      "loss": 1.0846,
+      "step": 692
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.686360328821075,
+      "learning_rate": 1.969799306992824e-05,
+      "loss": 1.0297,
+      "step": 693
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.01988978049729,
+      "learning_rate": 1.9696782713112622e-05,
+      "loss": 1.0701,
+      "step": 694
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8563763177269594,
+      "learning_rate": 1.9695569973099765e-05,
+      "loss": 1.0882,
+      "step": 695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.937134848854293,
+      "learning_rate": 1.9694354850187735e-05,
+      "loss": 1.1373,
+      "step": 696
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8809806630437658,
+      "learning_rate": 1.969313734467517e-05,
+      "loss": 1.1474,
+      "step": 697
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6916313779988994,
+      "learning_rate": 1.96919174568613e-05,
+      "loss": 0.9033,
+      "step": 698
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.508260916496033,
+      "learning_rate": 1.969069518704594e-05,
+      "loss": 1.0931,
+      "step": 699
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.763611485514735,
+      "learning_rate": 1.9689470535529486e-05,
+      "loss": 0.9209,
+      "step": 700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.03664208026575,
+      "learning_rate": 1.968824350261293e-05,
+      "loss": 0.9931,
+      "step": 701
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4671068542228993,
+      "learning_rate": 1.9687014088597835e-05,
+      "loss": 1.1389,
+      "step": 702
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6583966015737657,
+      "learning_rate": 1.968578229378636e-05,
+      "loss": 1.1128,
+      "step": 703
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.10644984836611,
+      "learning_rate": 1.9684548118481243e-05,
+      "loss": 1.0528,
+      "step": 704
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.777346025347817,
+      "learning_rate": 1.9683311562985814e-05,
+      "loss": 1.1407,
+      "step": 705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.164046825464703,
+      "learning_rate": 1.9682072627603983e-05,
+      "loss": 1.0261,
+      "step": 706
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8758635415084726,
+      "learning_rate": 1.968083131264024e-05,
+      "loss": 1.1707,
+      "step": 707
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.12795773688994,
+      "learning_rate": 1.9679587618399674e-05,
+      "loss": 0.9405,
+      "step": 708
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.441454785504363,
+      "learning_rate": 1.9678341545187942e-05,
+      "loss": 0.953,
+      "step": 709
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.834429816258187,
+      "learning_rate": 1.96770930933113e-05,
+      "loss": 1.0836,
+      "step": 710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.631866186382932,
+      "learning_rate": 1.9675842263076582e-05,
+      "loss": 1.0692,
+      "step": 711
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6349093277364863,
+      "learning_rate": 1.96745890547912e-05,
+      "loss": 1.0568,
+      "step": 712
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.989233135030243,
+      "learning_rate": 1.967333346876316e-05,
+      "loss": 0.9956,
+      "step": 713
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5976223876570796,
+      "learning_rate": 1.967207550530106e-05,
+      "loss": 1.1034,
+      "step": 714
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5445659972537817,
+      "learning_rate": 1.967081516471406e-05,
+      "loss": 0.9946,
+      "step": 715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.962320605519689,
+      "learning_rate": 1.9669552447311924e-05,
+      "loss": 1.0553,
+      "step": 716
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.617516912860955,
+      "learning_rate": 1.9668287353404985e-05,
+      "loss": 1.0419,
+      "step": 717
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9464756935533716,
+      "learning_rate": 1.9667019883304174e-05,
+      "loss": 1.0305,
+      "step": 718
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.000106682444183,
+      "learning_rate": 1.9665750037320997e-05,
+      "loss": 1.0838,
+      "step": 719
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6435817212177857,
+      "learning_rate": 1.9664477815767547e-05,
+      "loss": 1.0185,
+      "step": 720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.0570616965153805,
+      "learning_rate": 1.96632032189565e-05,
+      "loss": 1.1407,
+      "step": 721
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.486195508863955,
+      "learning_rate": 1.9661926247201114e-05,
+      "loss": 1.1585,
+      "step": 722
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.229993019271557,
+      "learning_rate": 1.966064690081524e-05,
+      "loss": 1.0752,
+      "step": 723
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.189231501099848,
+      "learning_rate": 1.9659365180113297e-05,
+      "loss": 1.0476,
+      "step": 724
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7017992871809446,
+      "learning_rate": 1.96580810854103e-05,
+      "loss": 1.0926,
+      "step": 725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.44021198429482,
+      "learning_rate": 1.965679461702185e-05,
+      "loss": 0.9926,
+      "step": 726
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.803164724153642,
+      "learning_rate": 1.9655505775264112e-05,
+      "loss": 1.1952,
+      "step": 727
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7636211547252727,
+      "learning_rate": 1.965421456045385e-05,
+      "loss": 1.1308,
+      "step": 728
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.558419284935113,
+      "learning_rate": 1.9652920972908417e-05,
+      "loss": 1.2105,
+      "step": 729
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.620390642949841,
+      "learning_rate": 1.9651625012945735e-05,
+      "loss": 1.0653,
+      "step": 730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8772827825120344,
+      "learning_rate": 1.9650326680884312e-05,
+      "loss": 0.9776,
+      "step": 731
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7144554851138603,
+      "learning_rate": 1.9649025977043248e-05,
+      "loss": 0.992,
+      "step": 732
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7390114991851795,
+      "learning_rate": 1.9647722901742214e-05,
+      "loss": 0.9189,
+      "step": 733
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.052214849078998,
+      "learning_rate": 1.964641745530147e-05,
+      "loss": 1.033,
+      "step": 734
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.806759472316589,
+      "learning_rate": 1.9645109638041863e-05,
+      "loss": 1.0422,
+      "step": 735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 38.73670037211717,
+      "learning_rate": 1.964379945028481e-05,
+      "loss": 1.2231,
+      "step": 736
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5072756289465525,
+      "learning_rate": 1.9642486892352326e-05,
+      "loss": 1.0772,
+      "step": 737
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7031172727863413,
+      "learning_rate": 1.9641171964566993e-05,
+      "loss": 1.1608,
+      "step": 738
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8411867594705162,
+      "learning_rate": 1.9639854667251995e-05,
+      "loss": 1.0666,
+      "step": 739
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8576674197679814,
+      "learning_rate": 1.9638535000731075e-05,
+      "loss": 1.1506,
+      "step": 740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.2951006484863985,
+      "learning_rate": 1.9637212965328574e-05,
+      "loss": 1.0101,
+      "step": 741
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3545177240029,
+      "learning_rate": 1.9635888561369415e-05,
+      "loss": 1.0713,
+      "step": 742
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4714756904233024,
+      "learning_rate": 1.9634561789179093e-05,
+      "loss": 1.0043,
+      "step": 743
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3900929988252564,
+      "learning_rate": 1.9633232649083697e-05,
+      "loss": 0.9865,
+      "step": 744
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8653969256444016,
+      "learning_rate": 1.9631901141409888e-05,
+      "loss": 1.0011,
+      "step": 745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5304687792541385,
+      "learning_rate": 1.9630567266484918e-05,
+      "loss": 1.0675,
+      "step": 746
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.41936470088297,
+      "learning_rate": 1.962923102463661e-05,
+      "loss": 1.0268,
+      "step": 747
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9386474133604734,
+      "learning_rate": 1.962789241619338e-05,
+      "loss": 0.9598,
+      "step": 748
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7646981028998376,
+      "learning_rate": 1.962655144148422e-05,
+      "loss": 1.0279,
+      "step": 749
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.60321616972536,
+      "learning_rate": 1.9625208100838698e-05,
+      "loss": 1.0293,
+      "step": 750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.524009445596224,
+      "learning_rate": 1.9623862394586978e-05,
+      "loss": 1.043,
+      "step": 751
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.749208823784576,
+      "learning_rate": 1.9622514323059788e-05,
+      "loss": 1.0363,
+      "step": 752
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5714432297536622,
+      "learning_rate": 1.9621163886588457e-05,
+      "loss": 0.9095,
+      "step": 753
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.644928681020688,
+      "learning_rate": 1.9619811085504877e-05,
+      "loss": 1.1088,
+      "step": 754
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5243190898327037,
+      "learning_rate": 1.961845592014153e-05,
+      "loss": 1.08,
+      "step": 755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5845755162216784,
+      "learning_rate": 1.961709839083147e-05,
+      "loss": 1.0601,
+      "step": 756
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8386738684758286,
+      "learning_rate": 1.9615738497908355e-05,
+      "loss": 1.109,
+      "step": 757
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4281897633822194,
+      "learning_rate": 1.96143762417064e-05,
+      "loss": 1.0573,
+      "step": 758
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9133448707630483,
+      "learning_rate": 1.961301162256041e-05,
+      "loss": 1.0018,
+      "step": 759
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.623091036318239,
+      "learning_rate": 1.961164464080577e-05,
+      "loss": 1.0683,
+      "step": 760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.801994531879226,
+      "learning_rate": 1.9610275296778443e-05,
+      "loss": 1.0179,
+      "step": 761
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.742948802159206,
+      "learning_rate": 1.960890359081498e-05,
+      "loss": 0.9554,
+      "step": 762
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6508770514039557,
+      "learning_rate": 1.9607529523252507e-05,
+      "loss": 0.9559,
+      "step": 763
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.021960305430146,
+      "learning_rate": 1.960615309442873e-05,
+      "loss": 1.0517,
+      "step": 764
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.593832974448563,
+      "learning_rate": 1.9604774304681937e-05,
+      "loss": 1.1218,
+      "step": 765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.624834274865465,
+      "learning_rate": 1.9603393154350996e-05,
+      "loss": 0.9769,
+      "step": 766
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.1752990912282915,
+      "learning_rate": 1.960200964377535e-05,
+      "loss": 1.1958,
+      "step": 767
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8413333660780666,
+      "learning_rate": 1.960062377329504e-05,
+      "loss": 0.9992,
+      "step": 768
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.261820789196503,
+      "learning_rate": 1.9599235543250662e-05,
+      "loss": 1.0441,
+      "step": 769
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.1863418239141374,
+      "learning_rate": 1.959784495398341e-05,
+      "loss": 1.1747,
+      "step": 770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6590672543008886,
+      "learning_rate": 1.959645200583505e-05,
+      "loss": 1.0349,
+      "step": 771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.838377478601075,
+      "learning_rate": 1.959505669914793e-05,
+      "loss": 1.1176,
+      "step": 772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7519830934354257,
+      "learning_rate": 1.9593659034264973e-05,
+      "loss": 1.0497,
+      "step": 773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.980442885458292,
+      "learning_rate": 1.9592259011529694e-05,
+      "loss": 1.0741,
+      "step": 774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.041357306903657,
+      "learning_rate": 1.9590856631286173e-05,
+      "loss": 0.9991,
+      "step": 775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.719310496292942,
+      "learning_rate": 1.9589451893879075e-05,
+      "loss": 1.0681,
+      "step": 776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.648060644724543,
+      "learning_rate": 1.958804479965365e-05,
+      "loss": 1.0148,
+      "step": 777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.385002892701598,
+      "learning_rate": 1.9586635348955726e-05,
+      "loss": 1.122,
+      "step": 778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.288261943500257,
+      "learning_rate": 1.9585223542131694e-05,
+      "loss": 1.0396,
+      "step": 779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.462958609822878,
+      "learning_rate": 1.9583809379528543e-05,
+      "loss": 0.9797,
+      "step": 780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.405964285120037,
+      "learning_rate": 1.9582392861493835e-05,
+      "loss": 1.0695,
+      "step": 781
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6432920238058637,
+      "learning_rate": 1.958097398837571e-05,
+      "loss": 0.9781,
+      "step": 782
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.481847514892652,
+      "learning_rate": 1.9579552760522887e-05,
+      "loss": 1.0203,
+      "step": 783
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.578362002892191,
+      "learning_rate": 1.9578129178284664e-05,
+      "loss": 1.0154,
+      "step": 784
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.002478093346479,
+      "learning_rate": 1.9576703242010923e-05,
+      "loss": 1.1206,
+      "step": 785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9881379135003465,
+      "learning_rate": 1.9575274952052105e-05,
+      "loss": 1.0538,
+      "step": 786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.081683690181119,
+      "learning_rate": 1.957384430875926e-05,
+      "loss": 0.9941,
+      "step": 787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7175468212827343,
+      "learning_rate": 1.957241131248399e-05,
+      "loss": 1.0754,
+      "step": 788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.534881287893714,
+      "learning_rate": 1.9570975963578485e-05,
+      "loss": 0.9715,
+      "step": 789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.432903947303448,
+      "learning_rate": 1.9569538262395517e-05,
+      "loss": 0.9574,
+      "step": 790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.796886352729334,
+      "learning_rate": 1.9568098209288436e-05,
+      "loss": 1.1519,
+      "step": 791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.906061714461183,
+      "learning_rate": 1.9566655804611156e-05,
+      "loss": 1.0291,
+      "step": 792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.770470886347632,
+      "learning_rate": 1.956521104871819e-05,
+      "loss": 1.2086,
+      "step": 793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.223319458578048,
+      "learning_rate": 1.9563763941964615e-05,
+      "loss": 1.0458,
+      "step": 794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9765422553961254,
+      "learning_rate": 1.956231448470609e-05,
+      "loss": 1.0013,
+      "step": 795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7688427996518112,
+      "learning_rate": 1.9560862677298848e-05,
+      "loss": 1.0523,
+      "step": 796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.944295892478544,
+      "learning_rate": 1.9559408520099703e-05,
+      "loss": 0.9943,
+      "step": 797
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.285120031612841,
+      "learning_rate": 1.955795201346605e-05,
+      "loss": 1.07,
+      "step": 798
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.360886095298879,
+      "learning_rate": 1.955649315775585e-05,
+      "loss": 1.0184,
+      "step": 799
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.453504735902985,
+      "learning_rate": 1.955503195332766e-05,
+      "loss": 1.0194,
+      "step": 800
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7720666704980377,
+      "learning_rate": 1.9553568400540594e-05,
+      "loss": 1.1322,
+      "step": 801
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.250064309135509,
+      "learning_rate": 1.9552102499754356e-05,
+      "loss": 1.048,
+      "step": 802
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.151007308172151,
+      "learning_rate": 1.955063425132922e-05,
+      "loss": 1.0671,
+      "step": 803
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.504712656197753,
+      "learning_rate": 1.9549163655626044e-05,
+      "loss": 0.9555,
+      "step": 804
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9652043005001696,
+      "learning_rate": 1.954769071300626e-05,
+      "loss": 1.1281,
+      "step": 805
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.417589770846741,
+      "learning_rate": 1.954621542383187e-05,
+      "loss": 0.9591,
+      "step": 806
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1455662258661916,
+      "learning_rate": 1.9544737788465465e-05,
+      "loss": 0.9198,
+      "step": 807
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.371944981568343,
+      "learning_rate": 1.9543257807270204e-05,
+      "loss": 1.0036,
+      "step": 808
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6233229149796298,
+      "learning_rate": 1.9541775480609823e-05,
+      "loss": 1.0705,
+      "step": 809
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7455642763851267,
+      "learning_rate": 1.9540290808848637e-05,
+      "loss": 1.0605,
+      "step": 810
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.67169620521609,
+      "learning_rate": 1.953880379235154e-05,
+      "loss": 1.0644,
+      "step": 811
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.563725907642809,
+      "learning_rate": 1.9537314431483996e-05,
+      "loss": 1.0724,
+      "step": 812
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6587065592888997,
+      "learning_rate": 1.953582272661205e-05,
+      "loss": 1.0925,
+      "step": 813
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.1180071058625725,
+      "learning_rate": 1.9534328678102316e-05,
+      "loss": 1.1036,
+      "step": 814
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.189957751724557,
+      "learning_rate": 1.9532832286321996e-05,
+      "loss": 1.0334,
+      "step": 815
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4550391618359204,
+      "learning_rate": 1.9531333551638858e-05,
+      "loss": 1.035,
+      "step": 816
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4523104004446936,
+      "learning_rate": 1.952983247442125e-05,
+      "loss": 1.1495,
+      "step": 817
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.324773915920645,
+      "learning_rate": 1.9528329055038094e-05,
+      "loss": 1.0231,
+      "step": 818
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.779739514408404,
+      "learning_rate": 1.9526823293858888e-05,
+      "loss": 1.0396,
+      "step": 819
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5490924567379984,
+      "learning_rate": 1.9525315191253706e-05,
+      "loss": 1.0161,
+      "step": 820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.455144063503647,
+      "learning_rate": 1.95238047475932e-05,
+      "loss": 1.0069,
+      "step": 821
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.790835644184474,
+      "learning_rate": 1.9522291963248588e-05,
+      "loss": 1.0302,
+      "step": 822
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.411888232632968,
+      "learning_rate": 1.9520776838591673e-05,
+      "loss": 0.993,
+      "step": 823
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.606694666212379,
+      "learning_rate": 1.9519259373994834e-05,
+      "loss": 0.9957,
+      "step": 824
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.792601477958348,
+      "learning_rate": 1.951773956983102e-05,
+      "loss": 1.1405,
+      "step": 825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8188365519437704,
+      "learning_rate": 1.951621742647375e-05,
+      "loss": 1.2124,
+      "step": 826
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6597773196242165,
+      "learning_rate": 1.9514692944297137e-05,
+      "loss": 1.1435,
+      "step": 827
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.194410296347501,
+      "learning_rate": 1.9513166123675838e-05,
+      "loss": 1.0048,
+      "step": 828
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.582744555272188,
+      "learning_rate": 1.951163696498512e-05,
+      "loss": 0.9583,
+      "step": 829
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6207705927160245,
+      "learning_rate": 1.9510105468600797e-05,
+      "loss": 1.0335,
+      "step": 830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6278228482113617,
+      "learning_rate": 1.950857163489927e-05,
+      "loss": 1.1164,
+      "step": 831
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5307035817039605,
+      "learning_rate": 1.9507035464257515e-05,
+      "loss": 0.9872,
+      "step": 832
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.824472225302144,
+      "learning_rate": 1.9505496957053078e-05,
+      "loss": 1.0227,
+      "step": 833
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9818563480117914,
+      "learning_rate": 1.9503956113664078e-05,
+      "loss": 1.1242,
+      "step": 834
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.869625239367482,
+      "learning_rate": 1.9502412934469215e-05,
+      "loss": 1.0929,
+      "step": 835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2976612309343882,
+      "learning_rate": 1.950086741984776e-05,
+      "loss": 1.0272,
+      "step": 836
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 12.769326972793142,
+      "learning_rate": 1.949931957017955e-05,
+      "loss": 1.1102,
+      "step": 837
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8481939837470174,
+      "learning_rate": 1.9497769385845012e-05,
+      "loss": 1.025,
+      "step": 838
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3352628206516677,
+      "learning_rate": 1.9496216867225134e-05,
+      "loss": 1.0924,
+      "step": 839
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3533347409436707,
+      "learning_rate": 1.949466201470148e-05,
+      "loss": 1.0003,
+      "step": 840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.542021472164938,
+      "learning_rate": 1.9493104828656187e-05,
+      "loss": 1.1117,
+      "step": 841
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6301160736140052,
+      "learning_rate": 1.949154530947198e-05,
+      "loss": 1.0294,
+      "step": 842
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3460425083645564,
+      "learning_rate": 1.948998345753213e-05,
+      "loss": 1.0436,
+      "step": 843
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2446259615251285,
+      "learning_rate": 1.9488419273220503e-05,
+      "loss": 0.9428,
+      "step": 844
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.310632804542017,
+      "learning_rate": 1.9486852756921534e-05,
+      "loss": 1.0236,
+      "step": 845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.364801744366941,
+      "learning_rate": 1.9485283909020226e-05,
+      "loss": 1.0437,
+      "step": 846
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2934837675991124,
+      "learning_rate": 1.9483712729902152e-05,
+      "loss": 1.0043,
+      "step": 847
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6795888850025262,
+      "learning_rate": 1.9482139219953478e-05,
+      "loss": 1.0745,
+      "step": 848
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.411422407071628,
+      "learning_rate": 1.9480563379560915e-05,
+      "loss": 0.934,
+      "step": 849
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.460071140538643,
+      "learning_rate": 1.9478985209111767e-05,
+      "loss": 1.0779,
+      "step": 850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.380162168990996,
+      "learning_rate": 1.9477404708993904e-05,
+      "loss": 0.9893,
+      "step": 851
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2651632117157217,
+      "learning_rate": 1.9475821879595765e-05,
+      "loss": 0.9183,
+      "step": 852
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2452134078511325,
+      "learning_rate": 1.947423672130637e-05,
+      "loss": 1.0649,
+      "step": 853
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.754485555340307,
+      "learning_rate": 1.9472649234515303e-05,
+      "loss": 1.2328,
+      "step": 854
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3907456912187404,
+      "learning_rate": 1.9471059419612724e-05,
+      "loss": 1.1444,
+      "step": 855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.104580673184325,
+      "learning_rate": 1.9469467276989366e-05,
+      "loss": 0.8636,
+      "step": 856
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5082479856079214,
+      "learning_rate": 1.946787280703653e-05,
+      "loss": 0.9407,
+      "step": 857
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.222298799756904,
+      "learning_rate": 1.9466276010146097e-05,
+      "loss": 1.1906,
+      "step": 858
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4676251777168177,
+      "learning_rate": 1.9464676886710513e-05,
+      "loss": 0.9284,
+      "step": 859
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.748250993913235,
+      "learning_rate": 1.9463075437122792e-05,
+      "loss": 1.0374,
+      "step": 860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5471276859661747,
+      "learning_rate": 1.9461471661776536e-05,
+      "loss": 1.105,
+      "step": 861
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8740533564053607,
+      "learning_rate": 1.94598655610659e-05,
+      "loss": 0.9877,
+      "step": 862
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.680125228295212,
+      "learning_rate": 1.945825713538562e-05,
+      "loss": 1.1043,
+      "step": 863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.148577769562295,
+      "learning_rate": 1.9456646385131006e-05,
+      "loss": 1.0213,
+      "step": 864
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.574910256683985,
+      "learning_rate": 1.945503331069793e-05,
+      "loss": 1.1023,
+      "step": 865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8228224803141773,
+      "learning_rate": 1.9453417912482843e-05,
+      "loss": 1.0985,
+      "step": 866
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7000581035892046,
+      "learning_rate": 1.9451800190882767e-05,
+      "loss": 1.0697,
+      "step": 867
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.529672076487252,
+      "learning_rate": 1.9450180146295286e-05,
+      "loss": 0.8486,
+      "step": 868
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5757251881192853,
+      "learning_rate": 1.944855777911857e-05,
+      "loss": 1.1062,
+      "step": 869
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.895084514836089,
+      "learning_rate": 1.9446933089751342e-05,
+      "loss": 1.0017,
+      "step": 870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.4793166789795213,
+      "learning_rate": 1.944530607859291e-05,
+      "loss": 1.0879,
+      "step": 871
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9300137550777734,
+      "learning_rate": 1.9443676746043146e-05,
+      "loss": 0.9493,
+      "step": 872
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.730736121130802,
+      "learning_rate": 1.94420450925025e-05,
+      "loss": 1.0333,
+      "step": 873
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9223348859910074,
+      "learning_rate": 1.944041111837198e-05,
+      "loss": 1.1305,
+      "step": 874
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.418566499173944,
+      "learning_rate": 1.9438774824053174e-05,
+      "loss": 1.0247,
+      "step": 875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.710512251678389,
+      "learning_rate": 1.9437136209948235e-05,
+      "loss": 1.0405,
+      "step": 876
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6291136655677816,
+      "learning_rate": 1.9435495276459892e-05,
+      "loss": 1.0759,
+      "step": 877
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6660336507415625,
+      "learning_rate": 1.9433852023991438e-05,
+      "loss": 1.051,
+      "step": 878
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7405109288915126,
+      "learning_rate": 1.9432206452946738e-05,
+      "loss": 0.9667,
+      "step": 879
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.795044548099862,
+      "learning_rate": 1.943055856373023e-05,
+      "loss": 0.9896,
+      "step": 880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.226216202872285,
+      "learning_rate": 1.9428908356746914e-05,
+      "loss": 1.0357,
+      "step": 881
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6773692597552903,
+      "learning_rate": 1.942725583240237e-05,
+      "loss": 0.9852,
+      "step": 882
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5055818432672696,
+      "learning_rate": 1.9425600991102737e-05,
+      "loss": 1.0631,
+      "step": 883
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7887724499734183,
+      "learning_rate": 1.9423943833254736e-05,
+      "loss": 1.0482,
+      "step": 884
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.497326903888074,
+      "learning_rate": 1.9422284359265638e-05,
+      "loss": 0.9832,
+      "step": 885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5198908403193294,
+      "learning_rate": 1.942062256954331e-05,
+      "loss": 0.9818,
+      "step": 886
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6349791779928973,
+      "learning_rate": 1.941895846449616e-05,
+      "loss": 0.9803,
+      "step": 887
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.139055135898923,
+      "learning_rate": 1.9417292044533187e-05,
+      "loss": 1.1132,
+      "step": 888
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3110527123846234,
+      "learning_rate": 1.9415623310063946e-05,
+      "loss": 0.99,
+      "step": 889
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4959525311625033,
+      "learning_rate": 1.941395226149857e-05,
+      "loss": 1.0202,
+      "step": 890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1293930744554226,
+      "learning_rate": 1.9412278899247748e-05,
+      "loss": 0.9667,
+      "step": 891
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2870879037590712,
+      "learning_rate": 1.941060322372275e-05,
+      "loss": 1.0469,
+      "step": 892
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.432365023023609,
+      "learning_rate": 1.940892523533541e-05,
+      "loss": 0.9081,
+      "step": 893
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.349988942640587,
+      "learning_rate": 1.940724493449813e-05,
+      "loss": 0.9457,
+      "step": 894
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.060076170295506,
+      "learning_rate": 1.9405562321623884e-05,
+      "loss": 1.0294,
+      "step": 895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5164199715881743,
+      "learning_rate": 1.940387739712621e-05,
+      "loss": 0.9969,
+      "step": 896
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4851977251380837,
+      "learning_rate": 1.940219016141921e-05,
+      "loss": 0.9124,
+      "step": 897
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8061545721513124,
+      "learning_rate": 1.9400500614917563e-05,
+      "loss": 1.0424,
+      "step": 898
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.424299477384805,
+      "learning_rate": 1.939880875803651e-05,
+      "loss": 1.0319,
+      "step": 899
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.744122164546516,
+      "learning_rate": 1.9397114591191866e-05,
+      "loss": 1.0477,
+      "step": 900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.445323224862705,
+      "learning_rate": 1.9395418114800005e-05,
+      "loss": 1.0916,
+      "step": 901
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.93929967636297,
+      "learning_rate": 1.939371932927788e-05,
+      "loss": 0.9989,
+      "step": 902
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.544785983942014,
+      "learning_rate": 1.9392018235043e-05,
+      "loss": 1.152,
+      "step": 903
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.892545070640911,
+      "learning_rate": 1.9390314832513442e-05,
+      "loss": 1.1214,
+      "step": 904
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 21.987263122887104,
+      "learning_rate": 1.9388609122107864e-05,
+      "loss": 1.1062,
+      "step": 905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6562630231187017,
+      "learning_rate": 1.9386901104245475e-05,
+      "loss": 0.9848,
+      "step": 906
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.518496911988627,
+      "learning_rate": 1.9385190779346058e-05,
+      "loss": 1.0192,
+      "step": 907
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5625079045270605,
+      "learning_rate": 1.938347814782997e-05,
+      "loss": 1.0707,
+      "step": 908
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.528834299569891,
+      "learning_rate": 1.938176321011812e-05,
+      "loss": 0.9774,
+      "step": 909
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6018264100036257,
+      "learning_rate": 1.9380045966631994e-05,
+      "loss": 1.1302,
+      "step": 910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.312438165651826,
+      "learning_rate": 1.9378326417793646e-05,
+      "loss": 1.0168,
+      "step": 911
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2024102793678733,
+      "learning_rate": 1.9376604564025685e-05,
+      "loss": 1.0516,
+      "step": 912
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5675605420399723,
+      "learning_rate": 1.93748804057513e-05,
+      "loss": 1.1061,
+      "step": 913
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.826640315336116,
+      "learning_rate": 1.9373153943394242e-05,
+      "loss": 1.0564,
+      "step": 914
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.624789270347231,
+      "learning_rate": 1.9371425177378824e-05,
+      "loss": 1.0781,
+      "step": 915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5070096506335084,
+      "learning_rate": 1.9369694108129928e-05,
+      "loss": 1.0496,
+      "step": 916
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6300723896660076,
+      "learning_rate": 1.9367960736073002e-05,
+      "loss": 1.0479,
+      "step": 917
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.578842278567324,
+      "learning_rate": 1.9366225061634064e-05,
+      "loss": 0.9765,
+      "step": 918
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.581842751711827,
+      "learning_rate": 1.936448708523969e-05,
+      "loss": 1.1024,
+      "step": 919
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1951935078674176,
+      "learning_rate": 1.936274680731703e-05,
+      "loss": 0.9714,
+      "step": 920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.331940461952663,
+      "learning_rate": 1.9361004228293788e-05,
+      "loss": 0.94,
+      "step": 921
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8236395933927243,
+      "learning_rate": 1.935925934859825e-05,
+      "loss": 1.0488,
+      "step": 922
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.363276927878964,
+      "learning_rate": 1.935751216865925e-05,
+      "loss": 1.1099,
+      "step": 923
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.596516227911937,
+      "learning_rate": 1.9355762688906202e-05,
+      "loss": 1.116,
+      "step": 924
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.348121383472041,
+      "learning_rate": 1.935401090976908e-05,
+      "loss": 1.0218,
+      "step": 925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.733343930017785,
+      "learning_rate": 1.935225683167842e-05,
+      "loss": 1.0043,
+      "step": 926
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.758966389591936,
+      "learning_rate": 1.935050045506532e-05,
+      "loss": 0.9766,
+      "step": 927
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.580226561244626,
+      "learning_rate": 1.9348741780361455e-05,
+      "loss": 1.0552,
+      "step": 928
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.206751124430905,
+      "learning_rate": 1.9346980807999055e-05,
+      "loss": 1.0489,
+      "step": 929
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 30.23771096559672,
+      "learning_rate": 1.934521753841092e-05,
+      "loss": 1.2118,
+      "step": 930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.322973754926158,
+      "learning_rate": 1.9343451972030407e-05,
+      "loss": 1.0232,
+      "step": 931
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6437034126980783,
+      "learning_rate": 1.934168410929145e-05,
+      "loss": 1.02,
+      "step": 932
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9613253906670383,
+      "learning_rate": 1.9339913950628536e-05,
+      "loss": 0.9895,
+      "step": 933
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7446020566588074,
+      "learning_rate": 1.933814149647672e-05,
+      "loss": 0.999,
+      "step": 934
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.666157074758006,
+      "learning_rate": 1.933636674727162e-05,
+      "loss": 1.0742,
+      "step": 935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5990163747656454,
+      "learning_rate": 1.9334589703449424e-05,
+      "loss": 0.9915,
+      "step": 936
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.679207043975646,
+      "learning_rate": 1.9332810365446876e-05,
+      "loss": 1.1157,
+      "step": 937
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2096157271834107,
+      "learning_rate": 1.9331028733701292e-05,
+      "loss": 0.9847,
+      "step": 938
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7360702314930085,
+      "learning_rate": 1.932924480865054e-05,
+      "loss": 1.0495,
+      "step": 939
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.163737309143626,
+      "learning_rate": 1.9327458590733062e-05,
+      "loss": 1.1496,
+      "step": 940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2904270215911993,
+      "learning_rate": 1.9325670080387863e-05,
+      "loss": 0.9515,
+      "step": 941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6522178641809253,
+      "learning_rate": 1.932387927805451e-05,
+      "loss": 1.0595,
+      "step": 942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.46856880025099,
+      "learning_rate": 1.9322086184173122e-05,
+      "loss": 1.0325,
+      "step": 943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.287603494668713,
+      "learning_rate": 1.93202907991844e-05,
+      "loss": 0.9141,
+      "step": 944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6251112696061867,
+      "learning_rate": 1.9318493123529597e-05,
+      "loss": 0.9155,
+      "step": 945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.987137986210219,
+      "learning_rate": 1.931669315765053e-05,
+      "loss": 1.0145,
+      "step": 946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3424632848737277,
+      "learning_rate": 1.9314890901989583e-05,
+      "loss": 1.0415,
+      "step": 947
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2353948522445077,
+      "learning_rate": 1.9313086356989698e-05,
+      "loss": 0.9595,
+      "step": 948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.576702121389513,
+      "learning_rate": 1.931127952309438e-05,
+      "loss": 1.1594,
+      "step": 949
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3757575820385246,
+      "learning_rate": 1.93094704007477e-05,
+      "loss": 1.0258,
+      "step": 950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.293818922537095,
+      "learning_rate": 1.9307658990394293e-05,
+      "loss": 1.0243,
+      "step": 951
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3688391700609315,
+      "learning_rate": 1.9305845292479346e-05,
+      "loss": 1.009,
+      "step": 952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.80828691267798,
+      "learning_rate": 1.930402930744862e-05,
+      "loss": 1.1966,
+      "step": 953
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.35247059313853,
+      "learning_rate": 1.930221103574843e-05,
+      "loss": 0.9744,
+      "step": 954
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4461099267973307,
+      "learning_rate": 1.930039047782566e-05,
+      "loss": 1.0224,
+      "step": 955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.611440236247767,
+      "learning_rate": 1.9298567634127748e-05,
+      "loss": 1.0344,
+      "step": 956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.713992871427326,
+      "learning_rate": 1.92967425051027e-05,
+      "loss": 1.0251,
+      "step": 957
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.449908187883703,
+      "learning_rate": 1.929491509119908e-05,
+      "loss": 0.9121,
+      "step": 958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.062615991041237,
+      "learning_rate": 1.9293085392866016e-05,
+      "loss": 0.9828,
+      "step": 959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.42483551399482,
+      "learning_rate": 1.9291253410553198e-05,
+      "loss": 0.9705,
+      "step": 960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.945041099579405,
+      "learning_rate": 1.9289419144710874e-05,
+      "loss": 1.0338,
+      "step": 961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.364824164234715,
+      "learning_rate": 1.928758259578986e-05,
+      "loss": 1.1394,
+      "step": 962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7583163155418124,
+      "learning_rate": 1.928574376424152e-05,
+      "loss": 1.0891,
+      "step": 963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6452998336249296,
+      "learning_rate": 1.9283902650517792e-05,
+      "loss": 1.0854,
+      "step": 964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4468051684722307,
+      "learning_rate": 1.9282059255071166e-05,
+      "loss": 0.9391,
+      "step": 965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7334849226378406,
+      "learning_rate": 1.9280213578354704e-05,
+      "loss": 1.0808,
+      "step": 966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.623689601351252,
+      "learning_rate": 1.927836562082202e-05,
+      "loss": 1.0664,
+      "step": 967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1724322437995816,
+      "learning_rate": 1.9276515382927284e-05,
+      "loss": 0.9325,
+      "step": 968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5921713171702434,
+      "learning_rate": 1.9274662865125235e-05,
+      "loss": 0.9363,
+      "step": 969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2800275079762544,
+      "learning_rate": 1.9272808067871173e-05,
+      "loss": 0.9843,
+      "step": 970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5548791781879294,
+      "learning_rate": 1.9270950991620955e-05,
+      "loss": 1.0932,
+      "step": 971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4175167933953956,
+      "learning_rate": 1.9269091636830998e-05,
+      "loss": 1.0733,
+      "step": 972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3628941721131165,
+      "learning_rate": 1.9267230003958276e-05,
+      "loss": 0.9672,
+      "step": 973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6295281626924045,
+      "learning_rate": 1.926536609346033e-05,
+      "loss": 1.0338,
+      "step": 974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.648118265875055,
+      "learning_rate": 1.926349990579525e-05,
+      "loss": 0.8895,
+      "step": 975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1202643710316105,
+      "learning_rate": 1.9261631441421703e-05,
+      "loss": 0.9469,
+      "step": 976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.649872049484259,
+      "learning_rate": 1.92597607007989e-05,
+      "loss": 1.0917,
+      "step": 977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.175536441933523,
+      "learning_rate": 1.9257887684386614e-05,
+      "loss": 1.112,
+      "step": 978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1752679166085684,
+      "learning_rate": 1.925601239264518e-05,
+      "loss": 1.0251,
+      "step": 979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.681899019040447,
+      "learning_rate": 1.92541348260355e-05,
+      "loss": 1.039,
+      "step": 980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4234714701595217,
+      "learning_rate": 1.925225498501902e-05,
+      "loss": 0.964,
+      "step": 981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.71223071364033,
+      "learning_rate": 1.9250372870057754e-05,
+      "loss": 1.0858,
+      "step": 982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.212599435406759,
+      "learning_rate": 1.924848848161427e-05,
+      "loss": 0.948,
+      "step": 983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5192934057012577,
+      "learning_rate": 1.9246601820151705e-05,
+      "loss": 1.0132,
+      "step": 984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.928109835134174,
+      "learning_rate": 1.924471288613374e-05,
+      "loss": 1.0298,
+      "step": 985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7549381559022263,
+      "learning_rate": 1.9242821680024625e-05,
+      "loss": 1.0001,
+      "step": 986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.239151421371474,
+      "learning_rate": 1.9240928202289168e-05,
+      "loss": 1.061,
+      "step": 987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.54700933413855,
+      "learning_rate": 1.9239032453392733e-05,
+      "loss": 1.0863,
+      "step": 988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.670780179665936,
+      "learning_rate": 1.9237134433801235e-05,
+      "loss": 1.1231,
+      "step": 989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.289598309932846,
+      "learning_rate": 1.923523414398116e-05,
+      "loss": 1.0588,
+      "step": 990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.489414422570557,
+      "learning_rate": 1.9233331584399542e-05,
+      "loss": 0.8921,
+      "step": 991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4955835415910568,
+      "learning_rate": 1.923142675552398e-05,
+      "loss": 0.9683,
+      "step": 992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.967525951762857,
+      "learning_rate": 1.9229519657822627e-05,
+      "loss": 1.1301,
+      "step": 993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6364302200755745,
+      "learning_rate": 1.9227610291764194e-05,
+      "loss": 1.0683,
+      "step": 994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6486532832454164,
+      "learning_rate": 1.922569865781795e-05,
+      "loss": 1.005,
+      "step": 995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.081910338426013,
+      "learning_rate": 1.922378475645372e-05,
+      "loss": 1.0145,
+      "step": 996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4099413617680425,
+      "learning_rate": 1.9221868588141886e-05,
+      "loss": 0.943,
+      "step": 997
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.570163953758169,
+      "learning_rate": 1.9219950153353393e-05,
+      "loss": 1.159,
+      "step": 998
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3798061968174875,
+      "learning_rate": 1.9218029452559733e-05,
+      "loss": 1.043,
+      "step": 999
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4073060967897386,
+      "learning_rate": 1.9216106486232965e-05,
+      "loss": 1.0475,
+      "step": 1000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.943847825429316,
+      "learning_rate": 1.9214181254845696e-05,
+      "loss": 1.0153,
+      "step": 1001
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7346551319913384,
+      "learning_rate": 1.92122537588711e-05,
+      "loss": 0.957,
+      "step": 1002
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4143581157133607,
+      "learning_rate": 1.9210323998782892e-05,
+      "loss": 1.0708,
+      "step": 1003
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.551479921614783,
+      "learning_rate": 1.920839197505536e-05,
+      "loss": 1.0025,
+      "step": 1004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.542686643829843,
+      "learning_rate": 1.9206457688163338e-05,
+      "loss": 0.9833,
+      "step": 1005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5015473236779293,
+      "learning_rate": 1.9204521138582227e-05,
+      "loss": 1.0198,
+      "step": 1006
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.563350601802636,
+      "learning_rate": 1.9202582326787963e-05,
+      "loss": 1.0396,
+      "step": 1007
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.292978792419636,
+      "learning_rate": 1.9200641253257064e-05,
+      "loss": 1.0761,
+      "step": 1008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.555083711210735,
+      "learning_rate": 1.9198697918466585e-05,
+      "loss": 0.9949,
+      "step": 1009
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5581102126972795,
+      "learning_rate": 1.9196752322894144e-05,
+      "loss": 0.9737,
+      "step": 1010
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4863765960327533,
+      "learning_rate": 1.9194804467017916e-05,
+      "loss": 1.0135,
+      "step": 1011
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.583370538658214,
+      "learning_rate": 1.9192854351316627e-05,
+      "loss": 0.952,
+      "step": 1012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8863410802663925,
+      "learning_rate": 1.9190901976269564e-05,
+      "loss": 0.9946,
+      "step": 1013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.169190485865015,
+      "learning_rate": 1.9188947342356562e-05,
+      "loss": 0.9519,
+      "step": 1014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.697746603380063,
+      "learning_rate": 1.9186990450058018e-05,
+      "loss": 1.1704,
+      "step": 1015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8443610504164885,
+      "learning_rate": 1.918503129985488e-05,
+      "loss": 0.9317,
+      "step": 1016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.613406671233491,
+      "learning_rate": 1.9183069892228648e-05,
+      "loss": 1.0153,
+      "step": 1017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.750460562880933,
+      "learning_rate": 1.9181106227661394e-05,
+      "loss": 1.0646,
+      "step": 1018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.543565576301301,
+      "learning_rate": 1.9179140306635715e-05,
+      "loss": 1.0755,
+      "step": 1019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1179341803007268,
+      "learning_rate": 1.917717212963479e-05,
+      "loss": 1.0341,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.693231646113712,
+      "learning_rate": 1.917520169714234e-05,
+      "loss": 1.0512,
+      "step": 1021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6424860961845242,
+      "learning_rate": 1.917322900964264e-05,
+      "loss": 0.9632,
+      "step": 1022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.143360794092663,
+      "learning_rate": 1.9171254067620525e-05,
+      "loss": 0.9626,
+      "step": 1023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.019061378499708,
+      "learning_rate": 1.9169276871561372e-05,
+      "loss": 1.0185,
+      "step": 1024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7413802411315658,
+      "learning_rate": 1.9167297421951133e-05,
+      "loss": 0.9387,
+      "step": 1025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4166342976005435,
+      "learning_rate": 1.916531571927629e-05,
+      "loss": 0.9611,
+      "step": 1026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4654811937421526,
+      "learning_rate": 1.9163331764023893e-05,
+      "loss": 0.9905,
+      "step": 1027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3771874199837653,
+      "learning_rate": 1.9161345556681548e-05,
+      "loss": 0.9824,
+      "step": 1028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9825758010399266,
+      "learning_rate": 1.91593570977374e-05,
+      "loss": 0.9843,
+      "step": 1029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.500299091233586,
+      "learning_rate": 1.9157366387680164e-05,
+      "loss": 1.0668,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.002409073404688,
+      "learning_rate": 1.9155373426999096e-05,
+      "loss": 1.0126,
+      "step": 1031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1178497772698615,
+      "learning_rate": 1.9153378216184013e-05,
+      "loss": 0.912,
+      "step": 1032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.875971732474303,
+      "learning_rate": 1.9151380755725282e-05,
+      "loss": 1.1375,
+      "step": 1033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.194747601119983,
+      "learning_rate": 1.9149381046113817e-05,
+      "loss": 0.9627,
+      "step": 1034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.566551938326764,
+      "learning_rate": 1.91473790878411e-05,
+      "loss": 0.9576,
+      "step": 1035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7111810907560225,
+      "learning_rate": 1.9145374881399144e-05,
+      "loss": 0.8816,
+      "step": 1036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.279089276894769,
+      "learning_rate": 1.9143368427280542e-05,
+      "loss": 1.0099,
+      "step": 1037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6872390618227047,
+      "learning_rate": 1.914135972597841e-05,
+      "loss": 1.006,
+      "step": 1038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4511173229460845,
+      "learning_rate": 1.9139348777986443e-05,
+      "loss": 0.9577,
+      "step": 1039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.246134239880193,
+      "learning_rate": 1.913733558379886e-05,
+      "loss": 0.9671,
+      "step": 1040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.16485398398124,
+      "learning_rate": 1.9135320143910465e-05,
+      "loss": 1.0102,
+      "step": 1041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.498137213691302,
+      "learning_rate": 1.9133302458816586e-05,
+      "loss": 1.1029,
+      "step": 1042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9868504509756595,
+      "learning_rate": 1.9131282529013114e-05,
+      "loss": 0.9275,
+      "step": 1043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2756053563969894,
+      "learning_rate": 1.9129260354996493e-05,
+      "loss": 1.0949,
+      "step": 1044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7173141603029385,
+      "learning_rate": 1.9127235937263716e-05,
+      "loss": 1.0517,
+      "step": 1045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.085399324277153,
+      "learning_rate": 1.912520927631233e-05,
+      "loss": 1.1233,
+      "step": 1046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.118322366645487,
+      "learning_rate": 1.9123180372640428e-05,
+      "loss": 1.0455,
+      "step": 1047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.29119929398604,
+      "learning_rate": 1.912114922674666e-05,
+      "loss": 0.9441,
+      "step": 1048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.1222949030253933,
+      "learning_rate": 1.9119115839130227e-05,
+      "loss": 0.9119,
+      "step": 1049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.436195624820573,
+      "learning_rate": 1.9117080210290873e-05,
+      "loss": 1.0198,
+      "step": 1050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5871822829406406,
+      "learning_rate": 1.9115042340728904e-05,
+      "loss": 1.0135,
+      "step": 1051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3799660086390766,
+      "learning_rate": 1.9113002230945166e-05,
+      "loss": 1.0078,
+      "step": 1052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2646985796547536,
+      "learning_rate": 1.9110959881441064e-05,
+      "loss": 0.8787,
+      "step": 1053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.15046936919999,
+      "learning_rate": 1.910891529271855e-05,
+      "loss": 1.0025,
+      "step": 1054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7288170635268973,
+      "learning_rate": 1.9106868465280125e-05,
+      "loss": 0.87,
+      "step": 1055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.454722575329552,
+      "learning_rate": 1.9104819399628846e-05,
+      "loss": 1.0159,
+      "step": 1056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3362831788943454,
+      "learning_rate": 1.9102768096268312e-05,
+      "loss": 1.0137,
+      "step": 1057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6138909028367387,
+      "learning_rate": 1.9100714555702673e-05,
+      "loss": 0.8808,
+      "step": 1058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.289557942859989,
+      "learning_rate": 1.909865877843664e-05,
+      "loss": 0.9233,
+      "step": 1059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.448109506123443,
+      "learning_rate": 1.9096600764975458e-05,
+      "loss": 1.0014,
+      "step": 1060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9931691189641203,
+      "learning_rate": 1.9094540515824933e-05,
+      "loss": 0.9862,
+      "step": 1061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.22334136087793,
+      "learning_rate": 1.909247803149141e-05,
+      "loss": 1.0753,
+      "step": 1062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3609406665630064,
+      "learning_rate": 1.9090413312481803e-05,
+      "loss": 1.0485,
+      "step": 1063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3705566669272056,
+      "learning_rate": 1.908834635930355e-05,
+      "loss": 1.0542,
+      "step": 1064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5487167698301945,
+      "learning_rate": 1.9086277172464657e-05,
+      "loss": 0.9729,
+      "step": 1065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.33753925640636,
+      "learning_rate": 1.9084205752473665e-05,
+      "loss": 1.0363,
+      "step": 1066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.43946877330562,
+      "learning_rate": 1.9082132099839678e-05,
+      "loss": 1.053,
+      "step": 1067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5135991771708275,
+      "learning_rate": 1.9080056215072335e-05,
+      "loss": 1.0468,
+      "step": 1068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2571006947721535,
+      "learning_rate": 1.9077978098681838e-05,
+      "loss": 1.0529,
+      "step": 1069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.041552540240493,
+      "learning_rate": 1.9075897751178924e-05,
+      "loss": 0.9847,
+      "step": 1070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.571357087030962,
+      "learning_rate": 1.9073815173074887e-05,
+      "loss": 0.945,
+      "step": 1071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.476556745270255,
+      "learning_rate": 1.9071730364881564e-05,
+      "loss": 1.0811,
+      "step": 1072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4951504929385124,
+      "learning_rate": 1.9069643327111343e-05,
+      "loss": 0.9277,
+      "step": 1073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.124336692811566,
+      "learning_rate": 1.9067554060277163e-05,
+      "loss": 0.9597,
+      "step": 1074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.527427864527887,
+      "learning_rate": 1.90654625648925e-05,
+      "loss": 1.0146,
+      "step": 1075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.455911351310847,
+      "learning_rate": 1.9063368841471394e-05,
+      "loss": 0.9454,
+      "step": 1076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3105290669891643,
+      "learning_rate": 1.9061272890528414e-05,
+      "loss": 0.9481,
+      "step": 1077
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2685108467907966,
+      "learning_rate": 1.9059174712578692e-05,
+      "loss": 0.9618,
+      "step": 1078
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9964576477210114,
+      "learning_rate": 1.90570743081379e-05,
+      "loss": 1.0274,
+      "step": 1079
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6542742179123198,
+      "learning_rate": 1.905497167772226e-05,
+      "loss": 0.9668,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5924874165069554,
+      "learning_rate": 1.9052866821848536e-05,
+      "loss": 1.0075,
+      "step": 1081
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.276204761156857,
+      "learning_rate": 1.9050759741034043e-05,
+      "loss": 0.9846,
+      "step": 1082
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.327735856352041,
+      "learning_rate": 1.904865043579664e-05,
+      "loss": 0.9602,
+      "step": 1083
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.197330224276992,
+      "learning_rate": 1.9046538906654745e-05,
+      "loss": 1.0542,
+      "step": 1084
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.855628629594563,
+      "learning_rate": 1.904442515412731e-05,
+      "loss": 0.9439,
+      "step": 1085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.381720411209631,
+      "learning_rate": 1.9042309178733825e-05,
+      "loss": 1.1813,
+      "step": 1086
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.714974701017107,
+      "learning_rate": 1.904019098099435e-05,
+      "loss": 1.0041,
+      "step": 1087
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.882190357917079,
+      "learning_rate": 1.9038070561429468e-05,
+      "loss": 0.986,
+      "step": 1088
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4545158928815023,
+      "learning_rate": 1.9035947920560327e-05,
+      "loss": 1.0752,
+      "step": 1089
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5499121222160874,
+      "learning_rate": 1.903382305890861e-05,
+      "loss": 1.0289,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4587592105542635,
+      "learning_rate": 1.903169597699655e-05,
+      "loss": 0.9598,
+      "step": 1091
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.334344279452804,
+      "learning_rate": 1.9029566675346916e-05,
+      "loss": 0.9692,
+      "step": 1092
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3315491706075857,
+      "learning_rate": 1.9027435154483044e-05,
+      "loss": 1.0866,
+      "step": 1093
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.383403403950702,
+      "learning_rate": 1.9025301414928795e-05,
+      "loss": 1.0768,
+      "step": 1094
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.440092757443575,
+      "learning_rate": 1.9023165457208577e-05,
+      "loss": 0.9649,
+      "step": 1095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.193803361721852,
+      "learning_rate": 1.9021027281847354e-05,
+      "loss": 1.0055,
+      "step": 1096
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.809350222993406,
+      "learning_rate": 1.9018886889370636e-05,
+      "loss": 0.9537,
+      "step": 1097
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.502375873447698,
+      "learning_rate": 1.9016744280304457e-05,
+      "loss": 0.9742,
+      "step": 1098
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4995535233150092,
+      "learning_rate": 1.901459945517542e-05,
+      "loss": 0.9055,
+      "step": 1099
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4377117018648042,
+      "learning_rate": 1.9012452414510667e-05,
+      "loss": 1.0599,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.895500725772442,
+      "learning_rate": 1.901030315883787e-05,
+      "loss": 1.0256,
+      "step": 1101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.370139212744162,
+      "learning_rate": 1.9008151688685255e-05,
+      "loss": 0.9934,
+      "step": 1102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.177175252089327,
+      "learning_rate": 1.9005998004581606e-05,
+      "loss": 1.0087,
+      "step": 1103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0968416904480254,
+      "learning_rate": 1.9003842107056224e-05,
+      "loss": 0.9516,
+      "step": 1104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.683234071334093,
+      "learning_rate": 1.9001683996638978e-05,
+      "loss": 0.9738,
+      "step": 1105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7105828983358378,
+      "learning_rate": 1.8999523673860266e-05,
+      "loss": 0.9792,
+      "step": 1106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.890400305329194,
+      "learning_rate": 1.8997361139251036e-05,
+      "loss": 1.0313,
+      "step": 1107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4767511809652216,
+      "learning_rate": 1.8995196393342778e-05,
+      "loss": 1.1165,
+      "step": 1108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.2470769090829967,
+      "learning_rate": 1.899302943666753e-05,
+      "loss": 1.0464,
+      "step": 1109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.1368112228941754,
+      "learning_rate": 1.899086026975786e-05,
+      "loss": 0.9509,
+      "step": 1110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3698732139606413,
+      "learning_rate": 1.89886888931469e-05,
+      "loss": 1.0738,
+      "step": 1111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.37952733028334,
+      "learning_rate": 1.89865153073683e-05,
+      "loss": 0.8875,
+      "step": 1112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3612014588697523,
+      "learning_rate": 1.898433951295628e-05,
+      "loss": 1.0317,
+      "step": 1113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.574115698052471,
+      "learning_rate": 1.8982161510445577e-05,
+      "loss": 1.0568,
+      "step": 1114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5775869562020173,
+      "learning_rate": 1.8979981300371492e-05,
+      "loss": 1.0596,
+      "step": 1115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.464328670626159,
+      "learning_rate": 1.8977798883269858e-05,
+      "loss": 0.9411,
+      "step": 1116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.801939625366067,
+      "learning_rate": 1.897561425967705e-05,
+      "loss": 0.9531,
+      "step": 1117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.8808600285040273,
+      "learning_rate": 1.8973427430129986e-05,
+      "loss": 1.0827,
+      "step": 1118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4145714567838223,
+      "learning_rate": 1.8971238395166128e-05,
+      "loss": 0.9949,
+      "step": 1119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.097114782795992,
+      "learning_rate": 1.8969047155323484e-05,
+      "loss": 0.9522,
+      "step": 1120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3808431452573715,
+      "learning_rate": 1.896685371114059e-05,
+      "loss": 1.0558,
+      "step": 1121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.213562730595572,
+      "learning_rate": 1.8964658063156542e-05,
+      "loss": 0.9506,
+      "step": 1122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.23503323959287,
+      "learning_rate": 1.8962460211910968e-05,
+      "loss": 0.9897,
+      "step": 1123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6174475409775577,
+      "learning_rate": 1.896026015794403e-05,
+      "loss": 0.9811,
+      "step": 1124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3036656051786983,
+      "learning_rate": 1.8958057901796446e-05,
+      "loss": 1.0645,
+      "step": 1125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.317419068376816,
+      "learning_rate": 1.895585344400947e-05,
+      "loss": 1.0342,
+      "step": 1126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.558894837151238,
+      "learning_rate": 1.8953646785124895e-05,
+      "loss": 0.9451,
+      "step": 1127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.262813533733853,
+      "learning_rate": 1.8951437925685053e-05,
+      "loss": 1.0044,
+      "step": 1128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.615319625164075,
+      "learning_rate": 1.8949226866232822e-05,
+      "loss": 1.0139,
+      "step": 1129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7198693704317325,
+      "learning_rate": 1.8947013607311614e-05,
+      "loss": 1.1314,
+      "step": 1130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.179425008195092,
+      "learning_rate": 1.8944798149465394e-05,
+      "loss": 1.066,
+      "step": 1131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.3484987417758982,
+      "learning_rate": 1.8942580493238655e-05,
+      "loss": 1.0742,
+      "step": 1132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.610034457301421,
+      "learning_rate": 1.8940360639176435e-05,
+      "loss": 1.0215,
+      "step": 1133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4479854456032344,
+      "learning_rate": 1.893813858782431e-05,
+      "loss": 1.096,
+      "step": 1134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.989140158166724,
+      "learning_rate": 1.89359143397284e-05,
+      "loss": 0.9756,
+      "step": 1135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5927108470611304,
+      "learning_rate": 1.8933687895435363e-05,
+      "loss": 0.9714,
+      "step": 1136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4687364701321726,
+      "learning_rate": 1.8931459255492396e-05,
+      "loss": 1.0522,
+      "step": 1137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.355312740447251,
+      "learning_rate": 1.8929228420447234e-05,
+      "loss": 1.0474,
+      "step": 1138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.329216562737884,
+      "learning_rate": 1.8926995390848158e-05,
+      "loss": 0.9987,
+      "step": 1139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4784093304330126,
+      "learning_rate": 1.892476016724398e-05,
+      "loss": 0.9661,
+      "step": 1140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.578408602821964,
+      "learning_rate": 1.892252275018406e-05,
+      "loss": 0.9503,
+      "step": 1141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.291121071892807,
+      "learning_rate": 1.892028314021829e-05,
+      "loss": 1.0053,
+      "step": 1142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6034792441405075,
+      "learning_rate": 1.89180413378971e-05,
+      "loss": 1.0765,
+      "step": 1143
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.558401358394951,
+      "learning_rate": 1.8915797343771464e-05,
+      "loss": 1.0834,
+      "step": 1144
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.230448244024029,
+      "learning_rate": 1.89135511583929e-05,
+      "loss": 1.0834,
+      "step": 1145
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.040874118640548,
+      "learning_rate": 1.8911302782313442e-05,
+      "loss": 0.9802,
+      "step": 1146
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.330212581063196,
+      "learning_rate": 1.8909052216085695e-05,
+      "loss": 1.0496,
+      "step": 1147
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.276501255977922,
+      "learning_rate": 1.8906799460262776e-05,
+      "loss": 0.9585,
+      "step": 1148
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7767677443907792,
+      "learning_rate": 1.8904544515398348e-05,
+      "loss": 0.8529,
+      "step": 1149
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.339419253619894,
+      "learning_rate": 1.8902287382046613e-05,
+      "loss": 0.9477,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 14.351120059873713,
+      "learning_rate": 1.8900028060762317e-05,
+      "loss": 1.3391,
+      "step": 1151
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7045864230149754,
+      "learning_rate": 1.8897766552100735e-05,
+      "loss": 1.0245,
+      "step": 1152
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5387993543284884,
+      "learning_rate": 1.889550285661768e-05,
+      "loss": 1.1662,
+      "step": 1153
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.363800219601907,
+      "learning_rate": 1.8893236974869503e-05,
+      "loss": 0.9452,
+      "step": 1154
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.775347794310135,
+      "learning_rate": 1.8890968907413103e-05,
+      "loss": 1.0074,
+      "step": 1155
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.444627030455306,
+      "learning_rate": 1.88886986548059e-05,
+      "loss": 1.0151,
+      "step": 1156
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4675043582061207,
+      "learning_rate": 1.888642621760586e-05,
+      "loss": 0.9661,
+      "step": 1157
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7803298567162296,
+      "learning_rate": 1.8884151596371487e-05,
+      "loss": 1.0353,
+      "step": 1158
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3184377487122805,
+      "learning_rate": 1.888187479166182e-05,
+      "loss": 0.9759,
+      "step": 1159
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.816499168073624,
+      "learning_rate": 1.8879595804036424e-05,
+      "loss": 1.0366,
+      "step": 1160
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2112431734482842,
+      "learning_rate": 1.8877314634055418e-05,
+      "loss": 1.0806,
+      "step": 1161
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.818032573141438,
+      "learning_rate": 1.8875031282279452e-05,
+      "loss": 0.9278,
+      "step": 1162
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3892480376956096,
+      "learning_rate": 1.88727457492697e-05,
+      "loss": 1.0231,
+      "step": 1163
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.338034412494329,
+      "learning_rate": 1.8870458035587896e-05,
+      "loss": 0.9942,
+      "step": 1164
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4858890392363016,
+      "learning_rate": 1.886816814179629e-05,
+      "loss": 1.0469,
+      "step": 1165
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.932865879989182,
+      "learning_rate": 1.8865876068457663e-05,
+      "loss": 0.9496,
+      "step": 1166
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 20.12195654806366,
+      "learning_rate": 1.8863581816135355e-05,
+      "loss": 1.2516,
+      "step": 1167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3345948724808117,
+      "learning_rate": 1.8861285385393226e-05,
+      "loss": 1.0154,
+      "step": 1168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5548649180893577,
+      "learning_rate": 1.8858986776795673e-05,
+      "loss": 0.945,
+      "step": 1169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.05695734572861,
+      "learning_rate": 1.8856685990907627e-05,
+      "loss": 0.9177,
+      "step": 1170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3215030028624435,
+      "learning_rate": 1.8854383028294563e-05,
+      "loss": 0.914,
+      "step": 1171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.534699822951794,
+      "learning_rate": 1.885207788952248e-05,
+      "loss": 0.9461,
+      "step": 1172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.630422675866747,
+      "learning_rate": 1.884977057515792e-05,
+      "loss": 1.0953,
+      "step": 1173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3762286736972444,
+      "learning_rate": 1.884746108576795e-05,
+      "loss": 1.0774,
+      "step": 1174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2132032320980506,
+      "learning_rate": 1.8845149421920183e-05,
+      "loss": 1.0781,
+      "step": 1175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.140304792219557,
+      "learning_rate": 1.884283558418276e-05,
+      "loss": 0.9953,
+      "step": 1176
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.586259201431184,
+      "learning_rate": 1.8840519573124355e-05,
+      "loss": 0.9602,
+      "step": 1177
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0692544921950002,
+      "learning_rate": 1.8838201389314186e-05,
+      "loss": 0.9531,
+      "step": 1178
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.059529893306157,
+      "learning_rate": 1.8835881033321987e-05,
+      "loss": 0.9284,
+      "step": 1179
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8748227160913618,
+      "learning_rate": 1.883355850571804e-05,
+      "loss": 0.9953,
+      "step": 1180
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.547535182182359,
+      "learning_rate": 1.8831233807073162e-05,
+      "loss": 1.0275,
+      "step": 1181
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.332665967214049,
+      "learning_rate": 1.8828906937958697e-05,
+      "loss": 1.0284,
+      "step": 1182
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.997463814001846,
+      "learning_rate": 1.882657789894652e-05,
+      "loss": 1.0457,
+      "step": 1183
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.167722059482026,
+      "learning_rate": 1.882424669060905e-05,
+      "loss": 0.9619,
+      "step": 1184
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3420295940411417,
+      "learning_rate": 1.8821913313519222e-05,
+      "loss": 1.0286,
+      "step": 1185
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4323628920909,
+      "learning_rate": 1.8819577768250527e-05,
+      "loss": 0.9414,
+      "step": 1186
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.471810980202804,
+      "learning_rate": 1.8817240055376967e-05,
+      "loss": 1.0593,
+      "step": 1187
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.014072585779561,
+      "learning_rate": 1.8814900175473095e-05,
+      "loss": 1.0279,
+      "step": 1188
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5142431015633555,
+      "learning_rate": 1.881255812911398e-05,
+      "loss": 0.9126,
+      "step": 1189
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1249815251555946,
+      "learning_rate": 1.8810213916875232e-05,
+      "loss": 0.9398,
+      "step": 1190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.095309690894207,
+      "learning_rate": 1.8807867539333e-05,
+      "loss": 1.0398,
+      "step": 1191
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2039191939862124,
+      "learning_rate": 1.8805518997063947e-05,
+      "loss": 1.0105,
+      "step": 1192
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3474256293689044,
+      "learning_rate": 1.8803168290645287e-05,
+      "loss": 0.9162,
+      "step": 1193
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0466196271546186,
+      "learning_rate": 1.8800815420654758e-05,
+      "loss": 0.9939,
+      "step": 1194
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.346909179921009,
+      "learning_rate": 1.8798460387670622e-05,
+      "loss": 0.9594,
+      "step": 1195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.486888312468145,
+      "learning_rate": 1.8796103192271688e-05,
+      "loss": 1.064,
+      "step": 1196
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5432823849117385,
+      "learning_rate": 1.8793743835037287e-05,
+      "loss": 1.0371,
+      "step": 1197
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7968288431929107,
+      "learning_rate": 1.8791382316547276e-05,
+      "loss": 1.0197,
+      "step": 1198
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7397975872652305,
+      "learning_rate": 1.878901863738206e-05,
+      "loss": 1.0208,
+      "step": 1199
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.039320733733214,
+      "learning_rate": 1.8786652798122557e-05,
+      "loss": 1.0011,
+      "step": 1200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.443305156795146,
+      "learning_rate": 1.878428479935023e-05,
+      "loss": 1.0169,
+      "step": 1201
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4890962089355315,
+      "learning_rate": 1.878191464164706e-05,
+      "loss": 1.0539,
+      "step": 1202
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3572622978384468,
+      "learning_rate": 1.8779542325595572e-05,
+      "loss": 1.0377,
+      "step": 1203
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4645394707459998,
+      "learning_rate": 1.877716785177881e-05,
+      "loss": 1.0387,
+      "step": 1204
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.4755656750309,
+      "learning_rate": 1.8774791220780358e-05,
+      "loss": 1.0699,
+      "step": 1205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2676798552234088,
+      "learning_rate": 1.877241243318432e-05,
+      "loss": 0.9958,
+      "step": 1206
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.379353075378151,
+      "learning_rate": 1.8770031489575336e-05,
+      "loss": 0.9279,
+      "step": 1207
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.009969046904054,
+      "learning_rate": 1.876764839053858e-05,
+      "loss": 0.9824,
+      "step": 1208
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.195814747484976,
+      "learning_rate": 1.8765263136659747e-05,
+      "loss": 1.0223,
+      "step": 1209
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.98702869224199,
+      "learning_rate": 1.8762875728525063e-05,
+      "loss": 0.9445,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5172367808354177,
+      "learning_rate": 1.876048616672129e-05,
+      "loss": 0.8926,
+      "step": 1211
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4871172798691954,
+      "learning_rate": 1.875809445183572e-05,
+      "loss": 1.0432,
+      "step": 1212
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5064858806387686,
+      "learning_rate": 1.8755700584456156e-05,
+      "loss": 1.0091,
+      "step": 1213
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1161036668435775,
+      "learning_rate": 1.8753304565170953e-05,
+      "loss": 0.987,
+      "step": 1214
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4399247964110287,
+      "learning_rate": 1.8750906394568984e-05,
+      "loss": 1.132,
+      "step": 1215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.439032864890778,
+      "learning_rate": 1.8748506073239654e-05,
+      "loss": 0.8858,
+      "step": 1216
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9288136576207817,
+      "learning_rate": 1.874610360177289e-05,
+      "loss": 0.9878,
+      "step": 1217
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.412264557712734,
+      "learning_rate": 1.8743698980759154e-05,
+      "loss": 1.0101,
+      "step": 1218
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0818804914527385,
+      "learning_rate": 1.8741292210789435e-05,
+      "loss": 0.9747,
+      "step": 1219
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3379575642167576,
+      "learning_rate": 1.873888329245525e-05,
+      "loss": 1.0409,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.108719942681841,
+      "learning_rate": 1.8736472226348643e-05,
+      "loss": 0.8671,
+      "step": 1221
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.433163389985319,
+      "learning_rate": 1.873405901306219e-05,
+      "loss": 1.0531,
+      "step": 1222
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.367416135930543,
+      "learning_rate": 1.873164365318898e-05,
+      "loss": 1.115,
+      "step": 1223
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1944459496476703,
+      "learning_rate": 1.8729226147322653e-05,
+      "loss": 1.0592,
+      "step": 1224
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3534934567247827,
+      "learning_rate": 1.8726806496057356e-05,
+      "loss": 0.9829,
+      "step": 1225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4050693643260876,
+      "learning_rate": 1.8724384699987776e-05,
+      "loss": 1.0231,
+      "step": 1226
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2027668780710354,
+      "learning_rate": 1.872196075970912e-05,
+      "loss": 0.9412,
+      "step": 1227
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.225637604293947,
+      "learning_rate": 1.871953467581713e-05,
+      "loss": 1.0382,
+      "step": 1228
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0737196348462033,
+      "learning_rate": 1.8717106448908065e-05,
+      "loss": 1.0373,
+      "step": 1229
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.091074286664361,
+      "learning_rate": 1.871467607957871e-05,
+      "loss": 1.0822,
+      "step": 1230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1267167205958373,
+      "learning_rate": 1.871224356842639e-05,
+      "loss": 0.9143,
+      "step": 1231
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.697620074634605,
+      "learning_rate": 1.8709808916048948e-05,
+      "loss": 0.9823,
+      "step": 1232
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5889398544631512,
+      "learning_rate": 1.8707372123044746e-05,
+      "loss": 0.9262,
+      "step": 1233
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5183700822735426,
+      "learning_rate": 1.8704933190012683e-05,
+      "loss": 1.0104,
+      "step": 1234
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0718014289597306,
+      "learning_rate": 1.8702492117552185e-05,
+      "loss": 0.9683,
+      "step": 1235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6196911699554666,
+      "learning_rate": 1.870004890626319e-05,
+      "loss": 0.8832,
+      "step": 1236
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.072246372090567,
+      "learning_rate": 1.869760355674618e-05,
+      "loss": 0.8994,
+      "step": 1237
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1411597515237917,
+      "learning_rate": 1.869515606960215e-05,
+      "loss": 0.9609,
+      "step": 1238
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5541454860928097,
+      "learning_rate": 1.869270644543262e-05,
+      "loss": 1.1036,
+      "step": 1239
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4779650247855365,
+      "learning_rate": 1.869025468483964e-05,
+      "loss": 1.0456,
+      "step": 1240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.2139546208099206,
+      "learning_rate": 1.868780078842579e-05,
+      "loss": 1.021,
+      "step": 1241
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.225726212496146,
+      "learning_rate": 1.8685344756794163e-05,
+      "loss": 0.9912,
+      "step": 1242
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.186249770125133,
+      "learning_rate": 1.8682886590548385e-05,
+      "loss": 0.9572,
+      "step": 1243
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.422670951947607,
+      "learning_rate": 1.8680426290292603e-05,
+      "loss": 0.9215,
+      "step": 1244
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.374702483528041,
+      "learning_rate": 1.867796385663149e-05,
+      "loss": 1.0541,
+      "step": 1245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7650617765604655,
+      "learning_rate": 1.8675499290170243e-05,
+      "loss": 0.964,
+      "step": 1246
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.366569229191678,
+      "learning_rate": 1.8673032591514586e-05,
+      "loss": 0.9945,
+      "step": 1247
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3031430086247657,
+      "learning_rate": 1.8670563761270762e-05,
+      "loss": 1.0205,
+      "step": 1248
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.004205593212754,
+      "learning_rate": 1.866809280004554e-05,
+      "loss": 0.8999,
+      "step": 1249
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9436629690025033,
+      "learning_rate": 1.8665619708446216e-05,
+      "loss": 0.9898,
+      "step": 1250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.263914489945494,
+      "learning_rate": 1.86631444870806e-05,
+      "loss": 0.9416,
+      "step": 1251
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.388649773428647,
+      "learning_rate": 1.8660667136557038e-05,
+      "loss": 1.0842,
+      "step": 1252
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.284605163224415,
+      "learning_rate": 1.865818765748439e-05,
+      "loss": 0.9032,
+      "step": 1253
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.36238307880044,
+      "learning_rate": 1.8655706050472045e-05,
+      "loss": 0.9571,
+      "step": 1254
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3955441759934106,
+      "learning_rate": 1.8653222316129914e-05,
+      "loss": 0.9705,
+      "step": 1255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4215491117960726,
+      "learning_rate": 1.865073645506842e-05,
+      "loss": 0.96,
+      "step": 1256
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0799731990389914,
+      "learning_rate": 1.864824846789853e-05,
+      "loss": 0.9825,
+      "step": 1257
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.16621852615012,
+      "learning_rate": 1.8645758355231712e-05,
+      "loss": 0.9223,
+      "step": 1258
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9485309307936913,
+      "learning_rate": 1.8643266117679973e-05,
+      "loss": 0.9407,
+      "step": 1259
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0035964787739,
+      "learning_rate": 1.864077175585583e-05,
+      "loss": 1.0411,
+      "step": 1260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.4098162622882207,
+      "learning_rate": 1.863827527037233e-05,
+      "loss": 0.9541,
+      "step": 1261
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.11021479917164,
+      "learning_rate": 1.8635776661843037e-05,
+      "loss": 0.9688,
+      "step": 1262
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.319618014901573,
+      "learning_rate": 1.863327593088204e-05,
+      "loss": 1.1829,
+      "step": 1263
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3061764646598077,
+      "learning_rate": 1.8630773078103947e-05,
+      "loss": 0.8817,
+      "step": 1264
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.189631081214496,
+      "learning_rate": 1.8628268104123895e-05,
+      "loss": 0.9856,
+      "step": 1265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.2252370788231,
+      "learning_rate": 1.8625761009557527e-05,
+      "loss": 1.0615,
+      "step": 1266
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.112050914940334,
+      "learning_rate": 1.8623251795021026e-05,
+      "loss": 1.0279,
+      "step": 1267
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.472255495162265,
+      "learning_rate": 1.8620740461131078e-05,
+      "loss": 1.02,
+      "step": 1268
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3616466067796567,
+      "learning_rate": 1.8618227008504903e-05,
+      "loss": 1.0231,
+      "step": 1269
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.135356926756423,
+      "learning_rate": 1.861571143776024e-05,
+      "loss": 0.9841,
+      "step": 1270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1447769298644737,
+      "learning_rate": 1.8613193749515336e-05,
+      "loss": 0.9787,
+      "step": 1271
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9614186561247378,
+      "learning_rate": 1.861067394438898e-05,
+      "loss": 1.0221,
+      "step": 1272
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.629189000456825,
+      "learning_rate": 1.860815202300046e-05,
+      "loss": 1.0493,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5198156291852416,
+      "learning_rate": 1.86056279859696e-05,
+      "loss": 1.1185,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.228813921955044,
+      "learning_rate": 1.860310183391673e-05,
+      "loss": 1.0301,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.294294729255725,
+      "learning_rate": 1.8600573567462722e-05,
+      "loss": 1.0082,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.191062319084121,
+      "learning_rate": 1.8598043187228937e-05,
+      "loss": 1.1555,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2943202829958955,
+      "learning_rate": 1.8595510693837277e-05,
+      "loss": 0.9557,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3228994938865206,
+      "learning_rate": 1.8592976087910162e-05,
+      "loss": 1.0104,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4589290611274968,
+      "learning_rate": 1.8590439370070518e-05,
+      "loss": 1.0474,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.089947078240719,
+      "learning_rate": 1.858790054094181e-05,
+      "loss": 0.8952,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.223869917514333,
+      "learning_rate": 1.8585359601148005e-05,
+      "loss": 0.9582,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5241596955063565,
+      "learning_rate": 1.8582816551313594e-05,
+      "loss": 1.0233,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1393632796116306,
+      "learning_rate": 1.858027139206359e-05,
+      "loss": 0.9223,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0147943589181088,
+      "learning_rate": 1.857772412402352e-05,
+      "loss": 0.9559,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.76710874031879,
+      "learning_rate": 1.857517474781943e-05,
+      "loss": 1.0139,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.012356243143458,
+      "learning_rate": 1.857262326407789e-05,
+      "loss": 0.9859,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.214110568695533,
+      "learning_rate": 1.8570069673425978e-05,
+      "loss": 1.0213,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4750350660104727,
+      "learning_rate": 1.8567513976491298e-05,
+      "loss": 0.988,
+      "step": 1289
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.760410578608612,
+      "learning_rate": 1.856495617390197e-05,
+      "loss": 1.2646,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1666923521830657,
+      "learning_rate": 1.856239626628662e-05,
+      "loss": 0.9592,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.017141775920228,
+      "learning_rate": 1.8559834254274413e-05,
+      "loss": 1.0568,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2580613936555403,
+      "learning_rate": 1.8557270138495017e-05,
+      "loss": 1.0296,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.722825184968627,
+      "learning_rate": 1.8554703919578617e-05,
+      "loss": 1.0888,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8520569991398985,
+      "learning_rate": 1.855213559815592e-05,
+      "loss": 1.0294,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4248927635388777,
+      "learning_rate": 1.8549565174858148e-05,
+      "loss": 1.0494,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.382377772814438,
+      "learning_rate": 1.8546992650317035e-05,
+      "loss": 0.9951,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.3623590404197627,
+      "learning_rate": 1.854441802516484e-05,
+      "loss": 1.0029,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.938670771717353,
+      "learning_rate": 1.854184130003433e-05,
+      "loss": 1.1985,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.177523092609845,
+      "learning_rate": 1.8539262475558794e-05,
+      "loss": 1.0393,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.142874464959742,
+      "learning_rate": 1.8536681552372035e-05,
+      "loss": 1.0258,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1804710183341443,
+      "learning_rate": 1.8534098531108374e-05,
+      "loss": 0.8893,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.385721670884511,
+      "learning_rate": 1.8531513412402635e-05,
+      "loss": 1.1044,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1355109726054384,
+      "learning_rate": 1.852892619689018e-05,
+      "loss": 1.0309,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2700105399947113,
+      "learning_rate": 1.8526336885206863e-05,
+      "loss": 1.0062,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.83603218845176,
+      "learning_rate": 1.8523745477989074e-05,
+      "loss": 0.9573,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.274591564653697,
+      "learning_rate": 1.85211519758737e-05,
+      "loss": 0.9857,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.738378402447023,
+      "learning_rate": 1.851855637949816e-05,
+      "loss": 0.9904,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4655143401997606,
+      "learning_rate": 1.851595868950037e-05,
+      "loss": 1.032,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.267631524880856,
+      "learning_rate": 1.8513358906518773e-05,
+      "loss": 0.9451,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1241374456914595,
+      "learning_rate": 1.8510757031192325e-05,
+      "loss": 0.9012,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1106665906055904,
+      "learning_rate": 1.850815306416049e-05,
+      "loss": 0.9004,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.57024670362329,
+      "learning_rate": 1.8505547006063254e-05,
+      "loss": 1.0498,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1600728276232752,
+      "learning_rate": 1.8502938857541112e-05,
+      "loss": 1.0602,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.301046353011309,
+      "learning_rate": 1.850032861923507e-05,
+      "loss": 1.1654,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.238340991080974,
+      "learning_rate": 1.8497716291786653e-05,
+      "loss": 0.9199,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.824416857467904,
+      "learning_rate": 1.8495101875837903e-05,
+      "loss": 1.0123,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.962611707799049,
+      "learning_rate": 1.8492485372031363e-05,
+      "loss": 0.9427,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.2661569512729365,
+      "learning_rate": 1.84898667810101e-05,
+      "loss": 1.0405,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0660251603874067,
+      "learning_rate": 1.8487246103417692e-05,
+      "loss": 0.8864,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.321859969901298,
+      "learning_rate": 1.848462333989822e-05,
+      "loss": 1.0434,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5085205805452113,
+      "learning_rate": 1.8481998491096294e-05,
+      "loss": 0.992,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.729147238125545,
+      "learning_rate": 1.8479371557657028e-05,
+      "loss": 1.0681,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.314751970939591,
+      "learning_rate": 1.847674254022604e-05,
+      "loss": 0.9572,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0172757269056842,
+      "learning_rate": 1.847411143944948e-05,
+      "loss": 0.9826,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.285621040400702,
+      "learning_rate": 1.847147825597399e-05,
+      "loss": 1.1254,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.448844738758746,
+      "learning_rate": 1.8468842990446738e-05,
+      "loss": 1.052,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.495073379772717,
+      "learning_rate": 1.84662056435154e-05,
+      "loss": 1.0485,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.588068668936893,
+      "learning_rate": 1.8463566215828153e-05,
+      "loss": 1.1313,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5046696564398703,
+      "learning_rate": 1.8460924708033703e-05,
+      "loss": 1.0228,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.311822770954448,
+      "learning_rate": 1.845828112078125e-05,
+      "loss": 0.9789,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6065329287172045,
+      "learning_rate": 1.8455635454720523e-05,
+      "loss": 0.952,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.337957980232742,
+      "learning_rate": 1.845298771050175e-05,
+      "loss": 1.039,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.1850080920372354,
+      "learning_rate": 1.845033788877567e-05,
+      "loss": 0.933,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.230879748045431,
+      "learning_rate": 1.8447685990193535e-05,
+      "loss": 1.0526,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.85396414990987,
+      "learning_rate": 1.8445032015407107e-05,
+      "loss": 1.0069,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.9378848623087728,
+      "learning_rate": 1.844237596506866e-05,
+      "loss": 0.9669,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.854668930026993,
+      "learning_rate": 1.8439717839830974e-05,
+      "loss": 1.0146,
+      "step": 1338
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.099600996688908,
+      "learning_rate": 1.843705764034735e-05,
+      "loss": 1.0386,
+      "step": 1339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2855360587889026,
+      "learning_rate": 1.843439536727158e-05,
+      "loss": 0.9487,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4579980415822007,
+      "learning_rate": 1.843173102125798e-05,
+      "loss": 0.9574,
+      "step": 1341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7273952322801116,
+      "learning_rate": 1.842906460296137e-05,
+      "loss": 0.9395,
+      "step": 1342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.275683451454534,
+      "learning_rate": 1.8426396113037085e-05,
+      "loss": 0.9334,
+      "step": 1343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.210881842639474,
+      "learning_rate": 1.8423725552140964e-05,
+      "loss": 0.986,
+      "step": 1344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 34.68569610373674,
+      "learning_rate": 1.842105292092935e-05,
+      "loss": 1.1653,
+      "step": 1345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.802505121467872,
+      "learning_rate": 1.841837822005911e-05,
+      "loss": 0.9074,
+      "step": 1346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1070745294033553,
+      "learning_rate": 1.8415701450187603e-05,
+      "loss": 1.0476,
+      "step": 1347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.262380215220285,
+      "learning_rate": 1.8413022611972707e-05,
+      "loss": 1.0297,
+      "step": 1348
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.033952372865329,
+      "learning_rate": 1.8410341706072805e-05,
+      "loss": 1.015,
+      "step": 1349
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.323263789902725,
+      "learning_rate": 1.8407658733146788e-05,
+      "loss": 0.9336,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3046967974580457,
+      "learning_rate": 1.8404973693854056e-05,
+      "loss": 0.9665,
+      "step": 1351
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2788674106321425,
+      "learning_rate": 1.840228658885452e-05,
+      "loss": 1.001,
+      "step": 1352
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1826543622998025,
+      "learning_rate": 1.8399597418808588e-05,
+      "loss": 0.9061,
+      "step": 1353
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.232063653180614,
+      "learning_rate": 1.8396906184377185e-05,
+      "loss": 0.9828,
+      "step": 1354
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1539792259539645,
+      "learning_rate": 1.8394212886221743e-05,
+      "loss": 1.0339,
+      "step": 1355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2276005454751435,
+      "learning_rate": 1.8391517525004202e-05,
+      "loss": 0.9972,
+      "step": 1356
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.933260145178024,
+      "learning_rate": 1.8388820101387e-05,
+      "loss": 0.8192,
+      "step": 1357
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.302982115720018,
+      "learning_rate": 1.8386120616033088e-05,
+      "loss": 1.0183,
+      "step": 1358
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6318377709444007,
+      "learning_rate": 1.8383419069605928e-05,
+      "loss": 1.0127,
+      "step": 1359
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0740579352046966,
+      "learning_rate": 1.838071546276948e-05,
+      "loss": 0.9438,
+      "step": 1360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7634617730211786,
+      "learning_rate": 1.837800979618822e-05,
+      "loss": 0.9507,
+      "step": 1361
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.340650030000716,
+      "learning_rate": 1.837530207052712e-05,
+      "loss": 0.9877,
+      "step": 1362
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1075200381421957,
+      "learning_rate": 1.8372592286451668e-05,
+      "loss": 1.0457,
+      "step": 1363
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.336523570898075,
+      "learning_rate": 1.836988044462785e-05,
+      "loss": 1.0177,
+      "step": 1364
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.314086950533041,
+      "learning_rate": 1.836716654572216e-05,
+      "loss": 1.0487,
+      "step": 1365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.6581734234286944,
+      "learning_rate": 1.83644505904016e-05,
+      "loss": 0.9876,
+      "step": 1366
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.224387145206634,
+      "learning_rate": 1.836173257933367e-05,
+      "loss": 0.9458,
+      "step": 1367
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.140033745322203,
+      "learning_rate": 1.835901251318639e-05,
+      "loss": 0.9106,
+      "step": 1368
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.875829759401323,
+      "learning_rate": 1.835629039262827e-05,
+      "loss": 0.9946,
+      "step": 1369
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.450143819458855,
+      "learning_rate": 1.8353566218328333e-05,
+      "loss": 1.1555,
+      "step": 1370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.442123035895518,
+      "learning_rate": 1.8350839990956104e-05,
+      "loss": 0.9642,
+      "step": 1371
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1978975027659757,
+      "learning_rate": 1.8348111711181614e-05,
+      "loss": 0.8898,
+      "step": 1372
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.259845525408301,
+      "learning_rate": 1.8345381379675397e-05,
+      "loss": 1.0292,
+      "step": 1373
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.165231681661386,
+      "learning_rate": 1.834264899710849e-05,
+      "loss": 0.9839,
+      "step": 1374
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.895179041315747,
+      "learning_rate": 1.8339914564152442e-05,
+      "loss": 0.912,
+      "step": 1375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.33805355430569,
+      "learning_rate": 1.8337178081479292e-05,
+      "loss": 1.0233,
+      "step": 1376
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.270798423226357,
+      "learning_rate": 1.8334439549761596e-05,
+      "loss": 1.0438,
+      "step": 1377
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.386659692187951,
+      "learning_rate": 1.8331698969672405e-05,
+      "loss": 1.0603,
+      "step": 1378
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.29273059399376,
+      "learning_rate": 1.832895634188528e-05,
+      "loss": 0.9176,
+      "step": 1379
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2139646813203813,
+      "learning_rate": 1.832621166707428e-05,
+      "loss": 1.0601,
+      "step": 1380
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.591974463038327,
+      "learning_rate": 1.8323464945913967e-05,
+      "loss": 0.9984,
+      "step": 1381
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0970045925072758,
+      "learning_rate": 1.8320716179079414e-05,
+      "loss": 0.9418,
+      "step": 1382
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3020292537179117,
+      "learning_rate": 1.831796536724619e-05,
+      "loss": 0.977,
+      "step": 1383
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1556072875509744,
+      "learning_rate": 1.831521251109036e-05,
+      "loss": 0.9942,
+      "step": 1384
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4902091848804098,
+      "learning_rate": 1.8312457611288506e-05,
+      "loss": 1.038,
+      "step": 1385
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.785598451488263,
+      "learning_rate": 1.8309700668517703e-05,
+      "loss": 0.9436,
+      "step": 1386
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0068519029089527,
+      "learning_rate": 1.8306941683455528e-05,
+      "loss": 1.0645,
+      "step": 1387
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.220487909077531,
+      "learning_rate": 1.8304180656780062e-05,
+      "loss": 0.9674,
+      "step": 1388
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.163243004696232,
+      "learning_rate": 1.8301417589169898e-05,
+      "loss": 1.0972,
+      "step": 1389
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.4175119770718325,
+      "learning_rate": 1.8298652481304104e-05,
+      "loss": 0.8804,
+      "step": 1390
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.820628451840013,
+      "learning_rate": 1.829588533386228e-05,
+      "loss": 1.0035,
+      "step": 1391
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0343947926471664,
+      "learning_rate": 1.8293116147524506e-05,
+      "loss": 0.9804,
+      "step": 1392
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.48329283864859,
+      "learning_rate": 1.829034492297137e-05,
+      "loss": 1.0318,
+      "step": 1393
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.2509945928878103,
+      "learning_rate": 1.8287571660883967e-05,
+      "loss": 0.9648,
+      "step": 1394
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3114408715390753,
+      "learning_rate": 1.828479636194388e-05,
+      "loss": 1.0345,
+      "step": 1395
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.026796233762158,
+      "learning_rate": 1.8282019026833205e-05,
+      "loss": 1.0983,
+      "step": 1396
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3602743957091996,
+      "learning_rate": 1.8279239656234537e-05,
+      "loss": 0.9512,
+      "step": 1397
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9546018486657926,
+      "learning_rate": 1.8276458250830954e-05,
+      "loss": 0.9391,
+      "step": 1398
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8325729704271416,
+      "learning_rate": 1.8273674811306056e-05,
+      "loss": 1.0347,
+      "step": 1399
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.0823581307912495,
+      "learning_rate": 1.8270889338343934e-05,
+      "loss": 0.9359,
+      "step": 1400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1512300975836927,
+      "learning_rate": 1.8268101832629177e-05,
+      "loss": 0.9626,
+      "step": 1401
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3099205822582833,
+      "learning_rate": 1.826531229484688e-05,
+      "loss": 0.8663,
+      "step": 1402
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.183353819847706,
+      "learning_rate": 1.8262520725682628e-05,
+      "loss": 1.008,
+      "step": 1403
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.123487813760808,
+      "learning_rate": 1.8259727125822514e-05,
+      "loss": 0.9991,
+      "step": 1404
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1737631440311107,
+      "learning_rate": 1.825693149595312e-05,
+      "loss": 1.0189,
+      "step": 1405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2698968208545995,
+      "learning_rate": 1.8254133836761547e-05,
+      "loss": 1.0219,
+      "step": 1406
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.993678348890785,
+      "learning_rate": 1.825133414893537e-05,
+      "loss": 0.978,
+      "step": 1407
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.8830861585649417,
+      "learning_rate": 1.8248532433162672e-05,
+      "loss": 1.0617,
+      "step": 1408
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5798528339110396,
+      "learning_rate": 1.8245728690132044e-05,
+      "loss": 1.0044,
+      "step": 1409
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.211449496027972,
+      "learning_rate": 1.8242922920532566e-05,
+      "loss": 0.9857,
+      "step": 1410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.243617650417417,
+      "learning_rate": 1.8240115125053816e-05,
+      "loss": 0.9601,
+      "step": 1411
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.259960752667389,
+      "learning_rate": 1.823730530438587e-05,
+      "loss": 1.0182,
+      "step": 1412
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3528873993989965,
+      "learning_rate": 1.82344934592193e-05,
+      "loss": 0.9254,
+      "step": 1413
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3073550478896805,
+      "learning_rate": 1.8231679590245185e-05,
+      "loss": 0.9833,
+      "step": 1414
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3731207113327057,
+      "learning_rate": 1.8228863698155095e-05,
+      "loss": 1.0441,
+      "step": 1415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4125271371405015,
+      "learning_rate": 1.8226045783641094e-05,
+      "loss": 0.9709,
+      "step": 1416
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4114700891629424,
+      "learning_rate": 1.8223225847395745e-05,
+      "loss": 1.1116,
+      "step": 1417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.302800754545404,
+      "learning_rate": 1.822040389011212e-05,
+      "loss": 1.0117,
+      "step": 1418
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.472788092538862,
+      "learning_rate": 1.8217579912483758e-05,
+      "loss": 0.9642,
+      "step": 1419
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4919990970322954,
+      "learning_rate": 1.821475391520473e-05,
+      "loss": 0.9361,
+      "step": 1420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.578796053182178,
+      "learning_rate": 1.821192589896958e-05,
+      "loss": 0.9699,
+      "step": 1421
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.943372444254199,
+      "learning_rate": 1.8209095864473357e-05,
+      "loss": 0.9645,
+      "step": 1422
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.957107692774212,
+      "learning_rate": 1.82062638124116e-05,
+      "loss": 1.1112,
+      "step": 1423
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3304564555728913,
+      "learning_rate": 1.8203429743480356e-05,
+      "loss": 0.9292,
+      "step": 1424
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.4338118492304623,
+      "learning_rate": 1.8200593658376152e-05,
+      "loss": 1.0365,
+      "step": 1425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.6715752225401705,
+      "learning_rate": 1.819775555779602e-05,
+      "loss": 0.9023,
+      "step": 1426
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9897434792770268,
+      "learning_rate": 1.8194915442437487e-05,
+      "loss": 0.9809,
+      "step": 1427
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2133451138736357,
+      "learning_rate": 1.8192073312998574e-05,
+      "loss": 1.0375,
+      "step": 1428
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2079646767332197,
+      "learning_rate": 1.8189229170177797e-05,
+      "loss": 0.9501,
+      "step": 1429
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5101854117888256,
+      "learning_rate": 1.8186383014674164e-05,
+      "loss": 0.9791,
+      "step": 1430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.195330863399731,
+      "learning_rate": 1.8183534847187184e-05,
+      "loss": 0.9585,
+      "step": 1431
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.942786231263773,
+      "learning_rate": 1.818068466841685e-05,
+      "loss": 0.9436,
+      "step": 1432
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.380362306882057,
+      "learning_rate": 1.8177832479063663e-05,
+      "loss": 1.0471,
+      "step": 1433
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9912397523219942,
+      "learning_rate": 1.817497827982861e-05,
+      "loss": 0.9627,
+      "step": 1434
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.131458645693922,
+      "learning_rate": 1.817212207141317e-05,
+      "loss": 1.0078,
+      "step": 1435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5044622984176814,
+      "learning_rate": 1.816926385451932e-05,
+      "loss": 0.9924,
+      "step": 1436
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2764286835338834,
+      "learning_rate": 1.8166403629849533e-05,
+      "loss": 0.9193,
+      "step": 1437
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.955062497191095,
+      "learning_rate": 1.8163541398106765e-05,
+      "loss": 0.967,
+      "step": 1438
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2063986435021348,
+      "learning_rate": 1.8160677159994482e-05,
+      "loss": 0.9361,
+      "step": 1439
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3253753539110837,
+      "learning_rate": 1.815781091621662e-05,
+      "loss": 0.9221,
+      "step": 1440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.147461882152192,
+      "learning_rate": 1.815494266747764e-05,
+      "loss": 1.0155,
+      "step": 1441
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.05539303881804,
+      "learning_rate": 1.8152072414482456e-05,
+      "loss": 1.026,
+      "step": 1442
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0298394868604213,
+      "learning_rate": 1.8149200157936512e-05,
+      "loss": 1.0316,
+      "step": 1443
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.056299294425606,
+      "learning_rate": 1.814632589854572e-05,
+      "loss": 1.1486,
+      "step": 1444
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9964798081837896,
+      "learning_rate": 1.8143449637016495e-05,
+      "loss": 0.9409,
+      "step": 1445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.500281476121371,
+      "learning_rate": 1.8140571374055737e-05,
+      "loss": 0.9786,
+      "step": 1446
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7525323910569925,
+      "learning_rate": 1.8137691110370852e-05,
+      "loss": 0.9384,
+      "step": 1447
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.005090782342863,
+      "learning_rate": 1.813480884666972e-05,
+      "loss": 1.0044,
+      "step": 1448
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.996105565791517,
+      "learning_rate": 1.813192458366072e-05,
+      "loss": 1.0387,
+      "step": 1449
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2121612732948535,
+      "learning_rate": 1.8129038322052727e-05,
+      "loss": 0.9612,
+      "step": 1450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0154524615068325,
+      "learning_rate": 1.8126150062555107e-05,
+      "loss": 0.9509,
+      "step": 1451
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.124974923490825,
+      "learning_rate": 1.81232598058777e-05,
+      "loss": 0.9746,
+      "step": 1452
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.264530323861853,
+      "learning_rate": 1.8120367552730865e-05,
+      "loss": 0.9343,
+      "step": 1453
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.394005356454698,
+      "learning_rate": 1.8117473303825426e-05,
+      "loss": 1.0168,
+      "step": 1454
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0760671123580106,
+      "learning_rate": 1.811457705987271e-05,
+      "loss": 1.067,
+      "step": 1455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.196152023485417,
+      "learning_rate": 1.811167882158454e-05,
+      "loss": 0.9804,
+      "step": 1456
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.5596382958999953,
+      "learning_rate": 1.8108778589673216e-05,
+      "loss": 1.0635,
+      "step": 1457
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.116134009243598,
+      "learning_rate": 1.810587636485153e-05,
+      "loss": 1.1408,
+      "step": 1458
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0966651802438867,
+      "learning_rate": 1.8102972147832775e-05,
+      "loss": 0.9698,
+      "step": 1459
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9877198835598753,
+      "learning_rate": 1.8100065939330716e-05,
+      "loss": 0.9354,
+      "step": 1460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.0404801902248355,
+      "learning_rate": 1.809715774005963e-05,
+      "loss": 1.0685,
+      "step": 1461
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.948143128537915,
+      "learning_rate": 1.8094247550734262e-05,
+      "loss": 1.0719,
+      "step": 1462
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.066447715290956,
+      "learning_rate": 1.8091335372069857e-05,
+      "loss": 0.9708,
+      "step": 1463
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.115030578189605,
+      "learning_rate": 1.8088421204782153e-05,
+      "loss": 1.0377,
+      "step": 1464
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1425095469293893,
+      "learning_rate": 1.808550504958736e-05,
+      "loss": 0.9705,
+      "step": 1465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.244217587585244,
+      "learning_rate": 1.8082586907202194e-05,
+      "loss": 0.9619,
+      "step": 1466
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.20412918400006,
+      "learning_rate": 1.8079666778343853e-05,
+      "loss": 1.0447,
+      "step": 1467
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.459745892769258,
+      "learning_rate": 1.8076744663730016e-05,
+      "loss": 0.9423,
+      "step": 1468
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.138069027731141,
+      "learning_rate": 1.8073820564078865e-05,
+      "loss": 1.0486,
+      "step": 1469
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.425622599914716,
+      "learning_rate": 1.8070894480109056e-05,
+      "loss": 0.8424,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2962585539079887,
+      "learning_rate": 1.806796641253974e-05,
+      "loss": 1.0893,
+      "step": 1471
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2303926782782053,
+      "learning_rate": 1.8065036362090555e-05,
+      "loss": 0.975,
+      "step": 1472
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.037522825764029,
+      "learning_rate": 1.806210432948163e-05,
+      "loss": 1.1881,
+      "step": 1473
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.288380690108407,
+      "learning_rate": 1.8059170315433565e-05,
+      "loss": 0.9964,
+      "step": 1474
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.920521462332691,
+      "learning_rate": 1.8056234320667465e-05,
+      "loss": 0.9707,
+      "step": 1475
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.257974512775954,
+      "learning_rate": 1.8053296345904913e-05,
+      "loss": 1.1267,
+      "step": 1476
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.307161089261329,
+      "learning_rate": 1.8050356391867988e-05,
+      "loss": 0.9739,
+      "step": 1477
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.24739860932814,
+      "learning_rate": 1.804741445927924e-05,
+      "loss": 0.8637,
+      "step": 1478
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9989659358800385,
+      "learning_rate": 1.8044470548861715e-05,
+      "loss": 0.9047,
+      "step": 1479
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.10437368249323,
+      "learning_rate": 1.8041524661338943e-05,
+      "loss": 0.9233,
+      "step": 1480
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.794739597626496,
+      "learning_rate": 1.8038576797434945e-05,
+      "loss": 0.827,
+      "step": 1481
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1128014833427837,
+      "learning_rate": 1.803562695787422e-05,
+      "loss": 0.9067,
+      "step": 1482
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9988136690702203,
+      "learning_rate": 1.8032675143381756e-05,
+      "loss": 0.9164,
+      "step": 1483
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.011789954198862,
+      "learning_rate": 1.8029721354683024e-05,
+      "loss": 0.8821,
+      "step": 1484
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1678259833469715,
+      "learning_rate": 1.802676559250399e-05,
+      "loss": 0.9442,
+      "step": 1485
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.186525158553192,
+      "learning_rate": 1.8023807857571092e-05,
+      "loss": 0.9632,
+      "step": 1486
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.397793056047527,
+      "learning_rate": 1.8020848150611257e-05,
+      "loss": 0.9299,
+      "step": 1487
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1337686121653165,
+      "learning_rate": 1.8017886472351898e-05,
+      "loss": 1.0272,
+      "step": 1488
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2023763342435103,
+      "learning_rate": 1.8014922823520918e-05,
+      "loss": 1.0095,
+      "step": 1489
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.179613140976094,
+      "learning_rate": 1.8011957204846694e-05,
+      "loss": 0.9999,
+      "step": 1490
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9569570559790943,
+      "learning_rate": 1.8008989617058088e-05,
+      "loss": 0.9663,
+      "step": 1491
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.819221986668142,
+      "learning_rate": 1.800602006088446e-05,
+      "loss": 0.9449,
+      "step": 1492
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0987259678730616,
+      "learning_rate": 1.8003048537055634e-05,
+      "loss": 0.9663,
+      "step": 1493
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3105056929517525,
+      "learning_rate": 1.8000075046301937e-05,
+      "loss": 0.971,
+      "step": 1494
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0776614929117736,
+      "learning_rate": 1.7997099589354162e-05,
+      "loss": 0.9827,
+      "step": 1495
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.808460903546118,
+      "learning_rate": 1.799412216694359e-05,
+      "loss": 0.8083,
+      "step": 1496
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.12190977184157,
+      "learning_rate": 1.7991142779801997e-05,
+      "loss": 0.9388,
+      "step": 1497
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.4668326908121068,
+      "learning_rate": 1.798816142866163e-05,
+      "loss": 1.0586,
+      "step": 1498
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0588969927516345,
+      "learning_rate": 1.798517811425522e-05,
+      "loss": 0.9958,
+      "step": 1499
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0271086429161507,
+      "learning_rate": 1.798219283731598e-05,
+      "loss": 1.0154,
+      "step": 1500
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.113264536411894,
+      "learning_rate": 1.797920559857761e-05,
+      "loss": 0.9389,
+      "step": 1501
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.181762870336013,
+      "learning_rate": 1.7976216398774292e-05,
+      "loss": 1.0066,
+      "step": 1502
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.957030913988016,
+      "learning_rate": 1.7973225238640682e-05,
+      "loss": 1.0518,
+      "step": 1503
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1225430695720218,
+      "learning_rate": 1.7970232118911927e-05,
+      "loss": 0.954,
+      "step": 1504
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8346355377657373,
+      "learning_rate": 1.796723704032365e-05,
+      "loss": 1.0051,
+      "step": 1505
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.222418799430063,
+      "learning_rate": 1.7964240003611958e-05,
+      "loss": 0.9165,
+      "step": 1506
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.943155474907022,
+      "learning_rate": 1.7961241009513436e-05,
+      "loss": 0.9589,
+      "step": 1507
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.091694549571042,
+      "learning_rate": 1.795824005876516e-05,
+      "loss": 1.2324,
+      "step": 1508
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.289114095658659,
+      "learning_rate": 1.7955237152104673e-05,
+      "loss": 0.9287,
+      "step": 1509
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.59972267883229,
+      "learning_rate": 1.7952232290270007e-05,
+      "loss": 0.9872,
+      "step": 1510
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.5090128528213906,
+      "learning_rate": 1.794922547399967e-05,
+      "loss": 0.9683,
+      "step": 1511
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1106921697068057,
+      "learning_rate": 1.7946216704032662e-05,
+      "loss": 0.9422,
+      "step": 1512
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0391493782354364,
+      "learning_rate": 1.7943205981108442e-05,
+      "loss": 0.9405,
+      "step": 1513
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8944574754838523,
+      "learning_rate": 1.794019330596697e-05,
+      "loss": 0.9658,
+      "step": 1514
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0285972606047755,
+      "learning_rate": 1.7937178679348675e-05,
+      "loss": 0.9986,
+      "step": 1515
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0896134059161215,
+      "learning_rate": 1.7934162101994464e-05,
+      "loss": 0.8707,
+      "step": 1516
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3401810543095163,
+      "learning_rate": 1.7931143574645735e-05,
+      "loss": 1.0286,
+      "step": 1517
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.711419672994701,
+      "learning_rate": 1.7928123098044353e-05,
+      "loss": 0.9151,
+      "step": 1518
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2190767791084656,
+      "learning_rate": 1.792510067293266e-05,
+      "loss": 1.0052,
+      "step": 1519
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.570504511484273,
+      "learning_rate": 1.7922076300053493e-05,
+      "loss": 0.9412,
+      "step": 1520
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.212523605322943,
+      "learning_rate": 1.7919049980150155e-05,
+      "loss": 1.0195,
+      "step": 1521
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3752111549249815,
+      "learning_rate": 1.791602171396643e-05,
+      "loss": 0.9564,
+      "step": 1522
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8989772659100908,
+      "learning_rate": 1.7912991502246578e-05,
+      "loss": 0.9777,
+      "step": 1523
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.0079429607041592,
+      "learning_rate": 1.7909959345735346e-05,
+      "loss": 0.9082,
+      "step": 1524
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.3863097419634585,
+      "learning_rate": 1.790692524517795e-05,
+      "loss": 0.9536,
+      "step": 1525
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.2678279877371126,
+      "learning_rate": 1.7903889201320083e-05,
+      "loss": 0.9782,
+      "step": 1526
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.219374266336708,
+      "learning_rate": 1.7900851214907925e-05,
+      "loss": 0.9392,
+      "step": 1527
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.04359995207335,
+      "learning_rate": 1.789781128668813e-05,
+      "loss": 1.068,
+      "step": 1528
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.190840020564337,
+      "learning_rate": 1.789476941740782e-05,
+      "loss": 0.9424,
+      "step": 1529
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.107360791628095,
+      "learning_rate": 1.7891725607814602e-05,
+      "loss": 0.9268,
+      "step": 1530
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.8123151853339428,
+      "learning_rate": 1.7888679858656566e-05,
+      "loss": 0.909,
+      "step": 1531
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9281300713797394,
+      "learning_rate": 1.7885632170682266e-05,
+      "loss": 0.9875,
+      "step": 1532
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.1581987594287066,
+      "learning_rate": 1.7882582544640736e-05,
+      "loss": 0.8797,
+      "step": 1533
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.955887862870306,
+      "learning_rate": 1.7879530981281492e-05,
+      "loss": 0.956,
+      "step": 1534
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.9509542667470274,
+      "learning_rate": 1.7876477481354527e-05,
+      "loss": 1.0239,
+      "step": 1535
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.543176942825485,
+      "learning_rate": 1.78734220456103e-05,
+      "loss": 0.8289,
+      "step": 1536
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1781665264428587,
+      "learning_rate": 1.787036467479975e-05,
+      "loss": 0.9332,
+      "step": 1537
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.009086672085074,
+      "learning_rate": 1.7867305369674297e-05,
+      "loss": 1.0283,
+      "step": 1538
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.001359683433981,
+      "learning_rate": 1.7864244130985827e-05,
+      "loss": 0.9153,
+      "step": 1539
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5693255061400704,
+      "learning_rate": 1.7861180959486718e-05,
+      "loss": 1.0207,
+      "step": 1540
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1355580568313433,
+      "learning_rate": 1.7858115855929797e-05,
+      "loss": 0.8707,
+      "step": 1541
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.535492268525197,
+      "learning_rate": 1.785504882106839e-05,
+      "loss": 1.1984,
+      "step": 1542
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2349686394576134,
+      "learning_rate": 1.7851979855656292e-05,
+      "loss": 1.023,
+      "step": 1543
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.128701067205983,
+      "learning_rate": 1.784890896044776e-05,
+      "loss": 0.931,
+      "step": 1544
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1756175385193623,
+      "learning_rate": 1.784583613619753e-05,
+      "loss": 1.0761,
+      "step": 1545
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2455472161944567,
+      "learning_rate": 1.784276138366083e-05,
+      "loss": 0.9923,
+      "step": 1546
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.372209180487672,
+      "learning_rate": 1.7839684703593333e-05,
+      "loss": 0.9046,
+      "step": 1547
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7337452784886636,
+      "learning_rate": 1.7836606096751215e-05,
+      "loss": 0.9483,
+      "step": 1548
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.047376458044384,
+      "learning_rate": 1.78335255638911e-05,
+      "loss": 0.9749,
+      "step": 1549
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9377508290351053,
+      "learning_rate": 1.7830443105770104e-05,
+      "loss": 0.9275,
+      "step": 1550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3213718081164423,
+      "learning_rate": 1.7827358723145803e-05,
+      "loss": 1.0016,
+      "step": 1551
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.125648807004082,
+      "learning_rate": 1.7824272416776255e-05,
+      "loss": 0.9499,
+      "step": 1552
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.992772454537541,
+      "learning_rate": 1.7821184187419986e-05,
+      "loss": 1.0947,
+      "step": 1553
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3498984461989743,
+      "learning_rate": 1.7818094035835997e-05,
+      "loss": 0.977,
+      "step": 1554
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.808823836859759,
+      "learning_rate": 1.781500196278376e-05,
+      "loss": 1.1493,
+      "step": 1555
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7414312834078487,
+      "learning_rate": 1.7811907969023226e-05,
+      "loss": 0.9074,
+      "step": 1556
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.096725210487838,
+      "learning_rate": 1.7808812055314803e-05,
+      "loss": 1.0496,
+      "step": 1557
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1454246260790453,
+      "learning_rate": 1.7805714222419383e-05,
+      "loss": 0.9642,
+      "step": 1558
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.39456934603716,
+      "learning_rate": 1.7802614471098325e-05,
+      "loss": 0.9703,
+      "step": 1559
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8039173071331853,
+      "learning_rate": 1.7799512802113463e-05,
+      "loss": 0.8879,
+      "step": 1560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9674655561333063,
+      "learning_rate": 1.77964092162271e-05,
+      "loss": 1.0401,
+      "step": 1561
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9232148337048396,
+      "learning_rate": 1.7793303714202012e-05,
+      "loss": 0.941,
+      "step": 1562
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9934636053586705,
+      "learning_rate": 1.7790196296801443e-05,
+      "loss": 0.9383,
+      "step": 1563
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9223907005030956,
+      "learning_rate": 1.7787086964789107e-05,
+      "loss": 0.9159,
+      "step": 1564
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.032596903774553,
+      "learning_rate": 1.7783975718929188e-05,
+      "loss": 0.8643,
+      "step": 1565
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.829783049588635,
+      "learning_rate": 1.7780862559986353e-05,
+      "loss": 0.9304,
+      "step": 1566
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.371369832138475,
+      "learning_rate": 1.777774748872572e-05,
+      "loss": 1.0539,
+      "step": 1567
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.098014467231705,
+      "learning_rate": 1.7774630505912894e-05,
+      "loss": 1.0586,
+      "step": 1568
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.854963050985832,
+      "learning_rate": 1.7771511612313932e-05,
+      "loss": 1.0926,
+      "step": 1569
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.03748855360975,
+      "learning_rate": 1.7768390808695376e-05,
+      "loss": 1.0332,
+      "step": 1570
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2395041609630377,
+      "learning_rate": 1.7765268095824234e-05,
+      "loss": 0.8884,
+      "step": 1571
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5766715300588996,
+      "learning_rate": 1.776214347446798e-05,
+      "loss": 1.0497,
+      "step": 1572
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.115277159581461,
+      "learning_rate": 1.7759016945394554e-05,
+      "loss": 0.9077,
+      "step": 1573
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9805465907515285,
+      "learning_rate": 1.7755888509372378e-05,
+      "loss": 0.9555,
+      "step": 1574
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.8237610297834492,
+      "learning_rate": 1.7752758167170327e-05,
+      "loss": 0.8722,
+      "step": 1575
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.346858502108061,
+      "learning_rate": 1.774962591955775e-05,
+      "loss": 0.9513,
+      "step": 1576
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.972990157972103,
+      "learning_rate": 1.774649176730447e-05,
+      "loss": 1.0721,
+      "step": 1577
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.475587952247821,
+      "learning_rate": 1.7743355711180772e-05,
+      "loss": 1.0924,
+      "step": 1578
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3373101780404073,
+      "learning_rate": 1.774021775195741e-05,
+      "loss": 0.9527,
+      "step": 1579
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.680026816225568,
+      "learning_rate": 1.773707789040561e-05,
+      "loss": 1.0351,
+      "step": 1580
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.20231272940428,
+      "learning_rate": 1.7733936127297053e-05,
+      "loss": 0.9025,
+      "step": 1581
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.135934758066214,
+      "learning_rate": 1.7730792463403907e-05,
+      "loss": 0.9604,
+      "step": 1582
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.048731265002102,
+      "learning_rate": 1.772764689949879e-05,
+      "loss": 0.9117,
+      "step": 1583
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.067612401660131,
+      "learning_rate": 1.7724499436354796e-05,
+      "loss": 0.9069,
+      "step": 1584
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.3240696511439345,
+      "learning_rate": 1.7721350074745483e-05,
+      "loss": 0.8682,
+      "step": 1585
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2752756682214947,
+      "learning_rate": 1.771819881544487e-05,
+      "loss": 1.0783,
+      "step": 1586
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.882160412513278,
+      "learning_rate": 1.771504565922746e-05,
+      "loss": 0.9463,
+      "step": 1587
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9978510227502273,
+      "learning_rate": 1.7711890606868202e-05,
+      "loss": 0.9297,
+      "step": 1588
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.9646358151861225,
+      "learning_rate": 1.770873365914252e-05,
+      "loss": 0.923,
+      "step": 1589
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1091547510176922,
+      "learning_rate": 1.77055748168263e-05,
+      "loss": 0.9384,
+      "step": 1590
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1877979522883124,
+      "learning_rate": 1.7702414080695907e-05,
+      "loss": 1.0714,
+      "step": 1591
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2640181587855484,
+      "learning_rate": 1.7699251451528152e-05,
+      "loss": 1.0194,
+      "step": 1592
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.925145799537389,
+      "learning_rate": 1.7696086930100323e-05,
+      "loss": 0.8909,
+      "step": 1593
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.133756814413861,
+      "learning_rate": 1.7692920517190175e-05,
+      "loss": 0.9,
+      "step": 1594
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2710171605177805,
+      "learning_rate": 1.768975221357592e-05,
+      "loss": 0.9672,
+      "step": 1595
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.068629248668608,
+      "learning_rate": 1.7686582020036234e-05,
+      "loss": 0.9769,
+      "step": 1596
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.830114814388996,
+      "learning_rate": 1.7683409937350267e-05,
+      "loss": 0.9226,
+      "step": 1597
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.052722151159689,
+      "learning_rate": 1.768023596629763e-05,
+      "loss": 0.9782,
+      "step": 1598
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.181907971299097,
+      "learning_rate": 1.7677060107658387e-05,
+      "loss": 0.94,
+      "step": 1599
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.0584100002745003,
+      "learning_rate": 1.7673882362213085e-05,
+      "loss": 0.9347,
+      "step": 1600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0730524882941292,
+      "learning_rate": 1.7670702730742722e-05,
+      "loss": 0.9766,
+      "step": 1601
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.743519182081731,
+      "learning_rate": 1.7667521214028757e-05,
+      "loss": 0.8648,
+      "step": 1602
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0055379073044355,
+      "learning_rate": 1.7664337812853122e-05,
+      "loss": 0.9404,
+      "step": 1603
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1226930811580136,
+      "learning_rate": 1.766115252799821e-05,
+      "loss": 1.0251,
+      "step": 1604
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2624488224353567,
+      "learning_rate": 1.7657965360246867e-05,
+      "loss": 0.9208,
+      "step": 1605
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0925944435585944,
+      "learning_rate": 1.7654776310382417e-05,
+      "loss": 0.8821,
+      "step": 1606
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3295149139126234,
+      "learning_rate": 1.7651585379188635e-05,
+      "loss": 1.0782,
+      "step": 1607
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.903090564857671,
+      "learning_rate": 1.7648392567449764e-05,
+      "loss": 0.9306,
+      "step": 1608
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.981070901378658,
+      "learning_rate": 1.7645197875950507e-05,
+      "loss": 1.0098,
+      "step": 1609
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2721472207784768,
+      "learning_rate": 1.764200130547603e-05,
+      "loss": 1.0286,
+      "step": 1610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0694378697061078,
+      "learning_rate": 1.763880285681196e-05,
+      "loss": 1.0652,
+      "step": 1611
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6228799725758836,
+      "learning_rate": 1.7635602530744387e-05,
+      "loss": 0.9625,
+      "step": 1612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.298349477713538,
+      "learning_rate": 1.763240032805986e-05,
+      "loss": 1.1157,
+      "step": 1613
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4453013308981912,
+      "learning_rate": 1.762919624954539e-05,
+      "loss": 1.0214,
+      "step": 1614
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.025070577653514,
+      "learning_rate": 1.762599029598845e-05,
+      "loss": 0.9589,
+      "step": 1615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.233682546527499,
+      "learning_rate": 1.7622782468176974e-05,
+      "loss": 1.036,
+      "step": 1616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.211561445588491,
+      "learning_rate": 1.761957276689936e-05,
+      "loss": 0.9992,
+      "step": 1617
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.277215642818311,
+      "learning_rate": 1.761636119294446e-05,
+      "loss": 0.9019,
+      "step": 1618
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.402280666544251,
+      "learning_rate": 1.761314774710158e-05,
+      "loss": 1.0616,
+      "step": 1619
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.736470200541887,
+      "learning_rate": 1.760993243016051e-05,
+      "loss": 0.92,
+      "step": 1620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.940226552076836,
+      "learning_rate": 1.7606715242911468e-05,
+      "loss": 0.8812,
+      "step": 1621
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1723309239716233,
+      "learning_rate": 1.7603496186145168e-05,
+      "loss": 0.8744,
+      "step": 1622
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.191071987973035,
+      "learning_rate": 1.7600275260652746e-05,
+      "loss": 0.9277,
+      "step": 1623
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.239417253255005,
+      "learning_rate": 1.7597052467225827e-05,
+      "loss": 0.9204,
+      "step": 1624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.192265372865029,
+      "learning_rate": 1.7593827806656477e-05,
+      "loss": 0.9358,
+      "step": 1625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0184333324198205,
+      "learning_rate": 1.7590601279737232e-05,
+      "loss": 0.9797,
+      "step": 1626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.279109165748821,
+      "learning_rate": 1.7587372887261077e-05,
+      "loss": 1.0752,
+      "step": 1627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6953986178070153,
+      "learning_rate": 1.7584142630021458e-05,
+      "loss": 0.9401,
+      "step": 1628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.172191903175026,
+      "learning_rate": 1.758091050881229e-05,
+      "loss": 1.0515,
+      "step": 1629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1727497820346677,
+      "learning_rate": 1.7577676524427934e-05,
+      "loss": 1.0884,
+      "step": 1630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1850780747179583,
+      "learning_rate": 1.7574440677663212e-05,
+      "loss": 1.1565,
+      "step": 1631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.02776283992973,
+      "learning_rate": 1.7571202969313403e-05,
+      "loss": 1.0804,
+      "step": 1632
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.1377205276174385,
+      "learning_rate": 1.7567963400174245e-05,
+      "loss": 0.9776,
+      "step": 1633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2465903735570545,
+      "learning_rate": 1.7564721971041937e-05,
+      "loss": 0.8512,
+      "step": 1634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.453705119144487,
+      "learning_rate": 1.756147868271313e-05,
+      "loss": 0.9562,
+      "step": 1635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1922476204360235,
+      "learning_rate": 1.7558233535984933e-05,
+      "loss": 0.9605,
+      "step": 1636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.700710967013734,
+      "learning_rate": 1.755498653165491e-05,
+      "loss": 1.0348,
+      "step": 1637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2615697733633113,
+      "learning_rate": 1.7551737670521084e-05,
+      "loss": 1.0719,
+      "step": 1638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1274414203254888,
+      "learning_rate": 1.7548486953381937e-05,
+      "loss": 0.9591,
+      "step": 1639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1152780749906666,
+      "learning_rate": 1.75452343810364e-05,
+      "loss": 0.9675,
+      "step": 1640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0035425673242537,
+      "learning_rate": 1.7541979954283864e-05,
+      "loss": 1.0109,
+      "step": 1641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.077978228296154,
+      "learning_rate": 1.7538723673924185e-05,
+      "loss": 0.922,
+      "step": 1642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.739028913696504,
+      "learning_rate": 1.7535465540757654e-05,
+      "loss": 0.8579,
+      "step": 1643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3548693985388605,
+      "learning_rate": 1.7532205555585032e-05,
+      "loss": 0.8721,
+      "step": 1644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1780558911413643,
+      "learning_rate": 1.7528943719207536e-05,
+      "loss": 0.9789,
+      "step": 1645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.872343924062486,
+      "learning_rate": 1.752568003242683e-05,
+      "loss": 1.1482,
+      "step": 1646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.722081445165806,
+      "learning_rate": 1.7522414496045037e-05,
+      "loss": 0.9274,
+      "step": 1647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.109960299502892,
+      "learning_rate": 1.7519147110864736e-05,
+      "loss": 1.0573,
+      "step": 1648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9004201318080516,
+      "learning_rate": 1.7515877877688957e-05,
+      "loss": 0.912,
+      "step": 1649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.069111610446601,
+      "learning_rate": 1.7512606797321185e-05,
+      "loss": 1.1247,
+      "step": 1650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.716118441546312,
+      "learning_rate": 1.7509333870565364e-05,
+      "loss": 0.9329,
+      "step": 1651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1907520999620735,
+      "learning_rate": 1.7506059098225884e-05,
+      "loss": 0.9952,
+      "step": 1652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0379065945927852,
+      "learning_rate": 1.750278248110759e-05,
+      "loss": 0.9498,
+      "step": 1653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.037290982411241,
+      "learning_rate": 1.749950402001579e-05,
+      "loss": 0.9708,
+      "step": 1654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.107830375806654,
+      "learning_rate": 1.749622371575623e-05,
+      "loss": 1.0052,
+      "step": 1655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.839118227383374,
+      "learning_rate": 1.749294156913512e-05,
+      "loss": 0.9887,
+      "step": 1656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.296029959912079,
+      "learning_rate": 1.748965758095912e-05,
+      "loss": 0.9337,
+      "step": 1657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0948027241708327,
+      "learning_rate": 1.7486371752035346e-05,
+      "loss": 0.9437,
+      "step": 1658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.095667626766412,
+      "learning_rate": 1.7483084083171353e-05,
+      "loss": 0.9261,
+      "step": 1659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.9205509097298443,
+      "learning_rate": 1.7479794575175167e-05,
+      "loss": 0.9924,
+      "step": 1660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.982852037709161,
+      "learning_rate": 1.7476503228855254e-05,
+      "loss": 0.9751,
+      "step": 1661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0325740886574923,
+      "learning_rate": 1.747321004502053e-05,
+      "loss": 0.9428,
+      "step": 1662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.321678349275457,
+      "learning_rate": 1.746991502448037e-05,
+      "loss": 1.2062,
+      "step": 1663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.207975724754327,
+      "learning_rate": 1.7466618168044604e-05,
+      "loss": 0.9685,
+      "step": 1664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.972145230828689,
+      "learning_rate": 1.74633194765235e-05,
+      "loss": 0.9395,
+      "step": 1665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0197396910817815,
+      "learning_rate": 1.746001895072778e-05,
+      "loss": 1.0408,
+      "step": 1666
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.391534577184481,
+      "learning_rate": 1.7456716591468632e-05,
+      "loss": 0.8894,
+      "step": 1667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6510303451802515,
+      "learning_rate": 1.7453412399557673e-05,
+      "loss": 0.9196,
+      "step": 1668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.69220613921482,
+      "learning_rate": 1.7450106375806988e-05,
+      "loss": 1.0706,
+      "step": 1669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4899235907480253,
+      "learning_rate": 1.7446798521029104e-05,
+      "loss": 0.9742,
+      "step": 1670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.403430566508585,
+      "learning_rate": 1.744348883603699e-05,
+      "loss": 1.0179,
+      "step": 1671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8503841944131776,
+      "learning_rate": 1.7440177321644085e-05,
+      "loss": 0.8921,
+      "step": 1672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.117540564950011,
+      "learning_rate": 1.7436863978664264e-05,
+      "loss": 0.8934,
+      "step": 1673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0231089470141677,
+      "learning_rate": 1.7433548807911846e-05,
+      "loss": 1.0579,
+      "step": 1674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.91070938996234,
+      "learning_rate": 1.7430231810201616e-05,
+      "loss": 0.9335,
+      "step": 1675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.499268591374772,
+      "learning_rate": 1.7426912986348797e-05,
+      "loss": 0.9894,
+      "step": 1676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.024082227265866,
+      "learning_rate": 1.742359233716906e-05,
+      "loss": 0.9656,
+      "step": 1677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9461564317791638,
+      "learning_rate": 1.742026986347853e-05,
+      "loss": 0.9275,
+      "step": 1678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.29468643253041,
+      "learning_rate": 1.7416945566093775e-05,
+      "loss": 1.1882,
+      "step": 1679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3109274899198193,
+      "learning_rate": 1.7413619445831815e-05,
+      "loss": 0.9028,
+      "step": 1680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.883852212607492,
+      "learning_rate": 1.741029150351012e-05,
+      "loss": 0.9543,
+      "step": 1681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1513321934934866,
+      "learning_rate": 1.7406961739946605e-05,
+      "loss": 1.012,
+      "step": 1682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0608795599630514,
+      "learning_rate": 1.7403630155959626e-05,
+      "loss": 0.9721,
+      "step": 1683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9213729732013003,
+      "learning_rate": 1.7400296752368e-05,
+      "loss": 1.0405,
+      "step": 1684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8757556386881253,
+      "learning_rate": 1.7396961529990978e-05,
+      "loss": 0.9275,
+      "step": 1685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0883656707818514,
+      "learning_rate": 1.739362448964827e-05,
+      "loss": 1.0599,
+      "step": 1686
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8710024780202064,
+      "learning_rate": 1.7390285632160025e-05,
+      "loss": 0.94,
+      "step": 1687
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9563153036212664,
+      "learning_rate": 1.738694495834684e-05,
+      "loss": 0.9108,
+      "step": 1688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0319132328021308,
+      "learning_rate": 1.738360246902976e-05,
+      "loss": 0.9292,
+      "step": 1689
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9543356755124495,
+      "learning_rate": 1.738025816503027e-05,
+      "loss": 0.9085,
+      "step": 1690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.127768570393247,
+      "learning_rate": 1.7376912047170312e-05,
+      "loss": 0.9394,
+      "step": 1691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.149126382527672,
+      "learning_rate": 1.7373564116272268e-05,
+      "loss": 1.0142,
+      "step": 1692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.16266300650271,
+      "learning_rate": 1.7370214373158962e-05,
+      "loss": 1.0159,
+      "step": 1693
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7613314349272127,
+      "learning_rate": 1.7366862818653668e-05,
+      "loss": 0.8128,
+      "step": 1694
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3646451143842917,
+      "learning_rate": 1.7363509453580104e-05,
+      "loss": 0.9478,
+      "step": 1695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.7744638924974123,
+      "learning_rate": 1.7360154278762437e-05,
+      "loss": 0.9192,
+      "step": 1696
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9470719493608946,
+      "learning_rate": 1.7356797295025267e-05,
+      "loss": 0.9279,
+      "step": 1697
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0824620731472856,
+      "learning_rate": 1.7353438503193657e-05,
+      "loss": 0.9528,
+      "step": 1698
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0844674185118532,
+      "learning_rate": 1.7350077904093094e-05,
+      "loss": 0.9883,
+      "step": 1699
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.093740006730346,
+      "learning_rate": 1.734671549854952e-05,
+      "loss": 0.8871,
+      "step": 1700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9962269846127003,
+      "learning_rate": 1.7343351287389328e-05,
+      "loss": 1.0069,
+      "step": 1701
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0301979248106408,
+      "learning_rate": 1.733998527143934e-05,
+      "loss": 0.9567,
+      "step": 1702
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.900471623801645,
+      "learning_rate": 1.733661745152683e-05,
+      "loss": 0.8398,
+      "step": 1703
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1403426337374203,
+      "learning_rate": 1.733324782847951e-05,
+      "loss": 1.0183,
+      "step": 1704
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4374172819757125,
+      "learning_rate": 1.7329876403125547e-05,
+      "loss": 0.8892,
+      "step": 1705
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.030795047947954,
+      "learning_rate": 1.7326503176293536e-05,
+      "loss": 0.9275,
+      "step": 1706
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8903678182970776,
+      "learning_rate": 1.7323128148812525e-05,
+      "loss": 0.8777,
+      "step": 1707
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.287031736581128,
+      "learning_rate": 1.7319751321511998e-05,
+      "loss": 0.9923,
+      "step": 1708
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0188191593876836,
+      "learning_rate": 1.7316372695221888e-05,
+      "loss": 0.8788,
+      "step": 1709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.019615610447504,
+      "learning_rate": 1.7312992270772564e-05,
+      "loss": 0.9385,
+      "step": 1710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8104085965673415,
+      "learning_rate": 1.730961004899484e-05,
+      "loss": 0.906,
+      "step": 1711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.907845279180565,
+      "learning_rate": 1.7306226030719972e-05,
+      "loss": 0.9527,
+      "step": 1712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4993671502111137,
+      "learning_rate": 1.7302840216779657e-05,
+      "loss": 1.0291,
+      "step": 1713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.106702965434131,
+      "learning_rate": 1.7299452608006034e-05,
+      "loss": 0.93,
+      "step": 1714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1101385424088748,
+      "learning_rate": 1.7296063205231676e-05,
+      "loss": 0.972,
+      "step": 1715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1293898506147744,
+      "learning_rate": 1.729267200928961e-05,
+      "loss": 1.104,
+      "step": 1716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9730924394570124,
+      "learning_rate": 1.7289279021013298e-05,
+      "loss": 1.0663,
+      "step": 1717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.931559610901426,
+      "learning_rate": 1.7285884241236636e-05,
+      "loss": 0.8909,
+      "step": 1718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.945747264549108,
+      "learning_rate": 1.7282487670793967e-05,
+      "loss": 0.9308,
+      "step": 1719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.372598482443039,
+      "learning_rate": 1.7279089310520075e-05,
+      "loss": 0.9671,
+      "step": 1720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.252864971605568,
+      "learning_rate": 1.727568916125018e-05,
+      "loss": 0.9744,
+      "step": 1721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0895113567236314,
+      "learning_rate": 1.727228722381994e-05,
+      "loss": 0.9229,
+      "step": 1722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.882664161383922,
+      "learning_rate": 1.7268883499065467e-05,
+      "loss": 0.9749,
+      "step": 1723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1085081031820945,
+      "learning_rate": 1.7265477987823287e-05,
+      "loss": 1.0048,
+      "step": 1724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0004823214772047,
+      "learning_rate": 1.7262070690930386e-05,
+      "loss": 0.8328,
+      "step": 1725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.038832818626629,
+      "learning_rate": 1.7258661609224186e-05,
+      "loss": 0.9228,
+      "step": 1726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.192020710399445,
+      "learning_rate": 1.7255250743542533e-05,
+      "loss": 1.0519,
+      "step": 1727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6668903579950114,
+      "learning_rate": 1.7251838094723732e-05,
+      "loss": 0.8243,
+      "step": 1728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0670559820655092,
+      "learning_rate": 1.7248423663606514e-05,
+      "loss": 0.8683,
+      "step": 1729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8676037134225485,
+      "learning_rate": 1.7245007451030046e-05,
+      "loss": 0.9981,
+      "step": 1730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.79497576470677,
+      "learning_rate": 1.724158945783394e-05,
+      "loss": 1.1708,
+      "step": 1731
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.99738917376387,
+      "learning_rate": 1.723816968485825e-05,
+      "loss": 1.0504,
+      "step": 1732
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2178324841637616,
+      "learning_rate": 1.7234748132943445e-05,
+      "loss": 1.0838,
+      "step": 1733
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.4115054171733132,
+      "learning_rate": 1.723132480293046e-05,
+      "loss": 0.9939,
+      "step": 1734
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7712272440729984,
+      "learning_rate": 1.7227899695660647e-05,
+      "loss": 0.8403,
+      "step": 1735
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.402421858255346,
+      "learning_rate": 1.7224472811975803e-05,
+      "loss": 0.9411,
+      "step": 1736
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0223444412321774,
+      "learning_rate": 1.722104415271816e-05,
+      "loss": 0.9427,
+      "step": 1737
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.017597258777127,
+      "learning_rate": 1.7217613718730385e-05,
+      "loss": 0.9807,
+      "step": 1738
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1635316587108906,
+      "learning_rate": 1.7214181510855582e-05,
+      "loss": 0.8457,
+      "step": 1739
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7802900453298824,
+      "learning_rate": 1.7210747529937296e-05,
+      "loss": 0.9754,
+      "step": 1740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7253571537480985,
+      "learning_rate": 1.72073117768195e-05,
+      "loss": 0.8775,
+      "step": 1741
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.86569813967416,
+      "learning_rate": 1.7203874252346607e-05,
+      "loss": 0.9749,
+      "step": 1742
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0864039605414426,
+      "learning_rate": 1.720043495736346e-05,
+      "loss": 0.8447,
+      "step": 1743
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.790503841646772,
+      "learning_rate": 1.7196993892715344e-05,
+      "loss": 0.9175,
+      "step": 1744
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9046962258179656,
+      "learning_rate": 1.719355105924798e-05,
+      "loss": 0.9491,
+      "step": 1745
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.188869353506275,
+      "learning_rate": 1.7190106457807515e-05,
+      "loss": 0.8645,
+      "step": 1746
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.886800885546785,
+      "learning_rate": 1.7186660089240535e-05,
+      "loss": 0.8278,
+      "step": 1747
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.055015385983957,
+      "learning_rate": 1.7183211954394063e-05,
+      "loss": 0.9109,
+      "step": 1748
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6597265454520023,
+      "learning_rate": 1.7179762054115553e-05,
+      "loss": 0.9243,
+      "step": 1749
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2097148429469717,
+      "learning_rate": 1.7176310389252897e-05,
+      "loss": 0.9359,
+      "step": 1750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1357290092403733,
+      "learning_rate": 1.717285696065441e-05,
+      "loss": 0.8657,
+      "step": 1751
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.542033237018684,
+      "learning_rate": 1.7169401769168855e-05,
+      "loss": 0.9101,
+      "step": 1752
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.104045813974907,
+      "learning_rate": 1.7165944815645418e-05,
+      "loss": 0.9854,
+      "step": 1753
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0470300166305817,
+      "learning_rate": 1.7162486100933725e-05,
+      "loss": 0.927,
+      "step": 1754
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.036460270383953,
+      "learning_rate": 1.7159025625883823e-05,
+      "loss": 0.8991,
+      "step": 1755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.187501332975838,
+      "learning_rate": 1.7155563391346207e-05,
+      "loss": 1.2687,
+      "step": 1756
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.127249750494202,
+      "learning_rate": 1.7152099398171796e-05,
+      "loss": 1.001,
+      "step": 1757
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.086661240260967,
+      "learning_rate": 1.7148633647211936e-05,
+      "loss": 0.9008,
+      "step": 1758
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6321929427561344,
+      "learning_rate": 1.714516613931842e-05,
+      "loss": 1.0084,
+      "step": 1759
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.03961360149895,
+      "learning_rate": 1.7141696875343463e-05,
+      "loss": 0.933,
+      "step": 1760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.351519334357102,
+      "learning_rate": 1.7138225856139707e-05,
+      "loss": 0.9046,
+      "step": 1761
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.097830329883883,
+      "learning_rate": 1.7134753082560236e-05,
+      "loss": 1.0162,
+      "step": 1762
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.101611726104539,
+      "learning_rate": 1.7131278555458558e-05,
+      "loss": 0.9609,
+      "step": 1763
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9844918351327556,
+      "learning_rate": 1.7127802275688615e-05,
+      "loss": 0.9106,
+      "step": 1764
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2691109853035045,
+      "learning_rate": 1.7124324244104782e-05,
+      "loss": 0.8839,
+      "step": 1765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.120653729928492,
+      "learning_rate": 1.7120844461561857e-05,
+      "loss": 0.8986,
+      "step": 1766
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.085688777871114,
+      "learning_rate": 1.7117362928915073e-05,
+      "loss": 0.9608,
+      "step": 1767
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.988794942793776,
+      "learning_rate": 1.7113879647020098e-05,
+      "loss": 0.9892,
+      "step": 1768
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.142175207922706,
+      "learning_rate": 1.711039461673302e-05,
+      "loss": 0.846,
+      "step": 1769
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1702848317984467,
+      "learning_rate": 1.7106907838910365e-05,
+      "loss": 0.9765,
+      "step": 1770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.480872514729015,
+      "learning_rate": 1.7103419314409084e-05,
+      "loss": 1.1757,
+      "step": 1771
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7328141943559916,
+      "learning_rate": 1.709992904408656e-05,
+      "loss": 0.8963,
+      "step": 1772
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0810657826869616,
+      "learning_rate": 1.70964370288006e-05,
+      "loss": 1.0049,
+      "step": 1773
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1060015118018565,
+      "learning_rate": 1.7092943269409442e-05,
+      "loss": 0.8715,
+      "step": 1774
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1102710293969507,
+      "learning_rate": 1.7089447766771762e-05,
+      "loss": 0.994,
+      "step": 1775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2154911009841762,
+      "learning_rate": 1.708595052174665e-05,
+      "loss": 0.9058,
+      "step": 1776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.557012466247219,
+      "learning_rate": 1.7082451535193635e-05,
+      "loss": 1.167,
+      "step": 1777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.783641336634145,
+      "learning_rate": 1.7078950807972667e-05,
+      "loss": 0.8901,
+      "step": 1778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9934657945221828,
+      "learning_rate": 1.7075448340944125e-05,
+      "loss": 0.9141,
+      "step": 1779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.150396420366335,
+      "learning_rate": 1.7071944134968817e-05,
+      "loss": 0.8861,
+      "step": 1780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3958232210218937,
+      "learning_rate": 1.7068438190907987e-05,
+      "loss": 0.9127,
+      "step": 1781
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0008540440950373,
+      "learning_rate": 1.7064930509623287e-05,
+      "loss": 0.9078,
+      "step": 1782
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.890349461668927,
+      "learning_rate": 1.706142109197681e-05,
+      "loss": 0.9415,
+      "step": 1783
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.865060978081574,
+      "learning_rate": 1.7057909938831077e-05,
+      "loss": 1.0334,
+      "step": 1784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.231093802050666,
+      "learning_rate": 1.7054397051049028e-05,
+      "loss": 0.9508,
+      "step": 1785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8203188281216938,
+      "learning_rate": 1.705088242949403e-05,
+      "loss": 0.9999,
+      "step": 1786
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0897255312247514,
+      "learning_rate": 1.704736607502988e-05,
+      "loss": 1.0371,
+      "step": 1787
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9172885478578827,
+      "learning_rate": 1.70438479885208e-05,
+      "loss": 0.9821,
+      "step": 1788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9720055648574664,
+      "learning_rate": 1.7040328170831438e-05,
+      "loss": 0.8796,
+      "step": 1789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1382099352612167,
+      "learning_rate": 1.703680662282686e-05,
+      "loss": 0.8977,
+      "step": 1790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.9843929380153864,
+      "learning_rate": 1.7033283345372577e-05,
+      "loss": 0.8788,
+      "step": 1791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7407504477935687,
+      "learning_rate": 1.7029758339334493e-05,
+      "loss": 0.9148,
+      "step": 1792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1593277458544846,
+      "learning_rate": 1.7026231605578977e-05,
+      "loss": 0.9894,
+      "step": 1793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.8811698457573636,
+      "learning_rate": 1.7022703144972783e-05,
+      "loss": 0.9389,
+      "step": 1794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.019957969047658,
+      "learning_rate": 1.7019172958383117e-05,
+      "loss": 1.1372,
+      "step": 1795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1741523508416507,
+      "learning_rate": 1.70156410466776e-05,
+      "loss": 0.9265,
+      "step": 1796
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.031657184880145,
+      "learning_rate": 1.7012107410724272e-05,
+      "loss": 1.009,
+      "step": 1797
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.046990652580771,
+      "learning_rate": 1.7008572051391605e-05,
+      "loss": 1.0,
+      "step": 1798
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0803455961577884,
+      "learning_rate": 1.7005034969548494e-05,
+      "loss": 0.9237,
+      "step": 1799
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.935693160122931,
+      "learning_rate": 1.7001496166064247e-05,
+      "loss": 1.0491,
+      "step": 1800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7763607875745375,
+      "learning_rate": 1.6997955641808607e-05,
+      "loss": 0.9119,
+      "step": 1801
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1123010538326614,
+      "learning_rate": 1.6994413397651736e-05,
+      "loss": 1.0113,
+      "step": 1802
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.130483097020963,
+      "learning_rate": 1.6990869434464217e-05,
+      "loss": 0.9616,
+      "step": 1803
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3471090454376884,
+      "learning_rate": 1.698732375311706e-05,
+      "loss": 0.939,
+      "step": 1804
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9468683554247557,
+      "learning_rate": 1.698377635448169e-05,
+      "loss": 0.8409,
+      "step": 1805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.293572112069917,
+      "learning_rate": 1.6980227239429957e-05,
+      "loss": 0.9963,
+      "step": 1806
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8948531850277086,
+      "learning_rate": 1.6976676408834137e-05,
+      "loss": 0.9784,
+      "step": 1807
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.139711415530516,
+      "learning_rate": 1.6973123863566927e-05,
+      "loss": 0.9313,
+      "step": 1808
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1516894042204884,
+      "learning_rate": 1.6969569604501437e-05,
+      "loss": 0.9585,
+      "step": 1809
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.973758728621806,
+      "learning_rate": 1.6966013632511207e-05,
+      "loss": 0.8765,
+      "step": 1810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0817783963642067,
+      "learning_rate": 1.6962455948470197e-05,
+      "loss": 0.9369,
+      "step": 1811
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.790610696099825,
+      "learning_rate": 1.6958896553252783e-05,
+      "loss": 0.893,
+      "step": 1812
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.816385602375839,
+      "learning_rate": 1.6955335447733768e-05,
+      "loss": 0.9152,
+      "step": 1813
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8898860914694873,
+      "learning_rate": 1.6951772632788366e-05,
+      "loss": 1.0058,
+      "step": 1814
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9058069000027467,
+      "learning_rate": 1.6948208109292224e-05,
+      "loss": 0.9882,
+      "step": 1815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0880966053571,
+      "learning_rate": 1.6944641878121397e-05,
+      "loss": 0.9876,
+      "step": 1816
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.077793953569075,
+      "learning_rate": 1.6941073940152367e-05,
+      "loss": 0.905,
+      "step": 1817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.197647594173069,
+      "learning_rate": 1.693750429626203e-05,
+      "loss": 0.9701,
+      "step": 1818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.87426005324195,
+      "learning_rate": 1.693393294732771e-05,
+      "loss": 0.8155,
+      "step": 1819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8098810857134957,
+      "learning_rate": 1.6930359894227137e-05,
+      "loss": 1.0062,
+      "step": 1820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.236471667352018,
+      "learning_rate": 1.6926785137838475e-05,
+      "loss": 1.1021,
+      "step": 1821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7811463349506473,
+      "learning_rate": 1.6923208679040292e-05,
+      "loss": 0.9271,
+      "step": 1822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.119477770391288,
+      "learning_rate": 1.6919630518711588e-05,
+      "loss": 1.0113,
+      "step": 1823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7893762696875952,
+      "learning_rate": 1.691605065773177e-05,
+      "loss": 1.0375,
+      "step": 1824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6966638468986184,
+      "learning_rate": 1.6912469096980664e-05,
+      "loss": 0.9959,
+      "step": 1825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.271609100047215,
+      "learning_rate": 1.6908885837338525e-05,
+      "loss": 0.9358,
+      "step": 1826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.6535814458895586,
+      "learning_rate": 1.6905300879686012e-05,
+      "loss": 0.9314,
+      "step": 1827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0781102996355814,
+      "learning_rate": 1.6901714224904215e-05,
+      "loss": 0.9368,
+      "step": 1828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8514555295039825,
+      "learning_rate": 1.689812587387462e-05,
+      "loss": 0.9585,
+      "step": 1829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1607038231576516,
+      "learning_rate": 1.6894535827479152e-05,
+      "loss": 1.0184,
+      "step": 1830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7494442796069913,
+      "learning_rate": 1.6890944086600145e-05,
+      "loss": 0.9984,
+      "step": 1831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2687465123361275,
+      "learning_rate": 1.6887350652120346e-05,
+      "loss": 0.9707,
+      "step": 1832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.078713456184605,
+      "learning_rate": 1.688375552492292e-05,
+      "loss": 0.9926,
+      "step": 1833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7217043180958616,
+      "learning_rate": 1.688015870589144e-05,
+      "loss": 0.9223,
+      "step": 1834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.806529041950445,
+      "learning_rate": 1.6876560195909916e-05,
+      "loss": 0.9107,
+      "step": 1835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 15.422810929651483,
+      "learning_rate": 1.687295999586276e-05,
+      "loss": 1.2869,
+      "step": 1836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.194812363041196,
+      "learning_rate": 1.6869358106634794e-05,
+      "loss": 1.0234,
+      "step": 1837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0263600740187413,
+      "learning_rate": 1.686575452911126e-05,
+      "loss": 0.9931,
+      "step": 1838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.217364912389536,
+      "learning_rate": 1.6862149264177826e-05,
+      "loss": 0.9718,
+      "step": 1839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.864268872192165,
+      "learning_rate": 1.6858542312720555e-05,
+      "loss": 0.971,
+      "step": 1840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9868935967262358,
+      "learning_rate": 1.685493367562594e-05,
+      "loss": 0.9497,
+      "step": 1841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.02682450886872,
+      "learning_rate": 1.6851323353780883e-05,
+      "loss": 1.0055,
+      "step": 1842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.985454978585505,
+      "learning_rate": 1.6847711348072694e-05,
+      "loss": 0.9894,
+      "step": 1843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.062161794366507,
+      "learning_rate": 1.684409765938911e-05,
+      "loss": 0.9619,
+      "step": 1844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.920527000454507,
+      "learning_rate": 1.684048228861827e-05,
+      "loss": 0.9533,
+      "step": 1845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8508734491132364,
+      "learning_rate": 1.6836865236648736e-05,
+      "loss": 0.8948,
+      "step": 1846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.109326001566116,
+      "learning_rate": 1.683324650436947e-05,
+      "loss": 0.866,
+      "step": 1847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9223301296017796,
+      "learning_rate": 1.682962609266986e-05,
+      "loss": 0.9989,
+      "step": 1848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9451164618486563,
+      "learning_rate": 1.68260040024397e-05,
+      "loss": 0.9472,
+      "step": 1849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.06203765888581,
+      "learning_rate": 1.68223802345692e-05,
+      "loss": 0.9352,
+      "step": 1850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.947449793617199,
+      "learning_rate": 1.6818754789948974e-05,
+      "loss": 0.9157,
+      "step": 1851
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2220571907558564,
+      "learning_rate": 1.6815127669470066e-05,
+      "loss": 1.0802,
+      "step": 1852
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9745392570414517,
+      "learning_rate": 1.6811498874023914e-05,
+      "loss": 0.9811,
+      "step": 1853
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9466446307315883,
+      "learning_rate": 1.680786840450237e-05,
+      "loss": 0.8016,
+      "step": 1854
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.221723283817632,
+      "learning_rate": 1.6804236261797707e-05,
+      "loss": 0.9568,
+      "step": 1855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.8783807082840136,
+      "learning_rate": 1.6800602446802604e-05,
+      "loss": 1.0294,
+      "step": 1856
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3234673136431776,
+      "learning_rate": 1.679696696041015e-05,
+      "loss": 1.0243,
+      "step": 1857
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1457564055469462,
+      "learning_rate": 1.6793329803513845e-05,
+      "loss": 1.0493,
+      "step": 1858
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.833612508072633,
+      "learning_rate": 1.67896909770076e-05,
+      "loss": 0.9366,
+      "step": 1859
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0762053615072427,
+      "learning_rate": 1.678605048178574e-05,
+      "loss": 0.9819,
+      "step": 1860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2781884938178423,
+      "learning_rate": 1.678240831874299e-05,
+      "loss": 0.9688,
+      "step": 1861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3692602382750247,
+      "learning_rate": 1.677876448877449e-05,
+      "loss": 0.8875,
+      "step": 1862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.013326890826228,
+      "learning_rate": 1.6775118992775805e-05,
+      "loss": 0.8707,
+      "step": 1863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.459040885528723,
+      "learning_rate": 1.6771471831642885e-05,
+      "loss": 1.0118,
+      "step": 1864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.901135398632171,
+      "learning_rate": 1.67678230062721e-05,
+      "loss": 0.9964,
+      "step": 1865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.146941247952338,
+      "learning_rate": 1.6764172517560232e-05,
+      "loss": 1.0776,
+      "step": 1866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.239123497963529,
+      "learning_rate": 1.6760520366404465e-05,
+      "loss": 0.935,
+      "step": 1867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1929933652649822,
+      "learning_rate": 1.67568665537024e-05,
+      "loss": 0.8425,
+      "step": 1868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.033169365361725,
+      "learning_rate": 1.675321108035204e-05,
+      "loss": 0.9379,
+      "step": 1869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0333266108265007,
+      "learning_rate": 1.6749553947251796e-05,
+      "loss": 0.8782,
+      "step": 1870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.876592953022209,
+      "learning_rate": 1.674589515530049e-05,
+      "loss": 0.8542,
+      "step": 1871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.302964058161183,
+      "learning_rate": 1.6742234705397353e-05,
+      "loss": 0.9858,
+      "step": 1872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.124949533715328,
+      "learning_rate": 1.6738572598442017e-05,
+      "loss": 0.9404,
+      "step": 1873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0780370387166376,
+      "learning_rate": 1.6734908835334528e-05,
+      "loss": 0.9963,
+      "step": 1874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8498009806918034,
+      "learning_rate": 1.673124341697533e-05,
+      "loss": 0.9845,
+      "step": 1875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.816079342172047,
+      "learning_rate": 1.672757634426529e-05,
+      "loss": 0.9985,
+      "step": 1876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7436541873730276,
+      "learning_rate": 1.6723907618105664e-05,
+      "loss": 0.9205,
+      "step": 1877
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.95311268150872,
+      "learning_rate": 1.6720237239398125e-05,
+      "loss": 0.8629,
+      "step": 1878
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 19.916438316118874,
+      "learning_rate": 1.671656520904475e-05,
+      "loss": 1.22,
+      "step": 1879
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0861564067907,
+      "learning_rate": 1.671289152794802e-05,
+      "loss": 0.9189,
+      "step": 1880
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.089133404397287,
+      "learning_rate": 1.670921619701082e-05,
+      "loss": 0.9411,
+      "step": 1881
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.221747175011406,
+      "learning_rate": 1.6705539217136447e-05,
+      "loss": 0.9864,
+      "step": 1882
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.996888161089025,
+      "learning_rate": 1.6701860589228597e-05,
+      "loss": 1.0954,
+      "step": 1883
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.534999187383088,
+      "learning_rate": 1.6698180314191375e-05,
+      "loss": 1.1416,
+      "step": 1884
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.250891562903199,
+      "learning_rate": 1.6694498392929293e-05,
+      "loss": 1.0242,
+      "step": 1885
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1721803755932974,
+      "learning_rate": 1.669081482634726e-05,
+      "loss": 1.0573,
+      "step": 1886
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0931656989393654,
+      "learning_rate": 1.668712961535059e-05,
+      "loss": 0.9245,
+      "step": 1887
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7676923458266494,
+      "learning_rate": 1.668344276084501e-05,
+      "loss": 0.9846,
+      "step": 1888
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2333026082841565,
+      "learning_rate": 1.6679754263736644e-05,
+      "loss": 0.9653,
+      "step": 1889
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9318513606737286,
+      "learning_rate": 1.6676064124932016e-05,
+      "loss": 0.942,
+      "step": 1890
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.804796108438422,
+      "learning_rate": 1.6672372345338067e-05,
+      "loss": 1.0292,
+      "step": 1891
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7689755284678292,
+      "learning_rate": 1.666867892586213e-05,
+      "loss": 0.9791,
+      "step": 1892
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.730471857480553,
+      "learning_rate": 1.6664983867411947e-05,
+      "loss": 0.9998,
+      "step": 1893
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9574138452567618,
+      "learning_rate": 1.6661287170895647e-05,
+      "loss": 1.0932,
+      "step": 1894
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9343442651527223,
+      "learning_rate": 1.665758883722179e-05,
+      "loss": 0.9637,
+      "step": 1895
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.928698965252209,
+      "learning_rate": 1.6653888867299312e-05,
+      "loss": 0.9635,
+      "step": 1896
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0528707428961575,
+      "learning_rate": 1.6650187262037567e-05,
+      "loss": 1.0008,
+      "step": 1897
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0335541906489136,
+      "learning_rate": 1.6646484022346305e-05,
+      "loss": 0.8778,
+      "step": 1898
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2308635269804653,
+      "learning_rate": 1.6642779149135677e-05,
+      "loss": 0.9607,
+      "step": 1899
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.0332734710472575,
+      "learning_rate": 1.663907264331624e-05,
+      "loss": 0.8383,
+      "step": 1900
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.840711811673916,
+      "learning_rate": 1.6635364505798946e-05,
+      "loss": 0.9219,
+      "step": 1901
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 16.051039006193275,
+      "learning_rate": 1.663165473749515e-05,
+      "loss": 1.1543,
+      "step": 1902
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.045344183831331,
+      "learning_rate": 1.6627943339316616e-05,
+      "loss": 0.9771,
+      "step": 1903
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9859000516392347,
+      "learning_rate": 1.662423031217549e-05,
+      "loss": 0.9071,
+      "step": 1904
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.274059140274478,
+      "learning_rate": 1.6620515656984343e-05,
+      "loss": 0.8979,
+      "step": 1905
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9111963302746187,
+      "learning_rate": 1.6616799374656124e-05,
+      "loss": 0.8854,
+      "step": 1906
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6802138181802873,
+      "learning_rate": 1.6613081466104196e-05,
+      "loss": 0.9194,
+      "step": 1907
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.973806688301168,
+      "learning_rate": 1.660936193224231e-05,
+      "loss": 0.9397,
+      "step": 1908
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9352989064745683,
+      "learning_rate": 1.660564077398463e-05,
+      "loss": 0.9442,
+      "step": 1909
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9344758621160834,
+      "learning_rate": 1.6601917992245712e-05,
+      "loss": 0.9395,
+      "step": 1910
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8050275279211845,
+      "learning_rate": 1.6598193587940508e-05,
+      "loss": 1.0315,
+      "step": 1911
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.226670654091148,
+      "learning_rate": 1.659446756198437e-05,
+      "loss": 0.9383,
+      "step": 1912
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.112674249923903,
+      "learning_rate": 1.6590739915293056e-05,
+      "loss": 1.0231,
+      "step": 1913
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.665704117888447,
+      "learning_rate": 1.6587010648782717e-05,
+      "loss": 0.9799,
+      "step": 1914
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.465238430343106,
+      "learning_rate": 1.65832797633699e-05,
+      "loss": 1.1079,
+      "step": 1915
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.911361683645425,
+      "learning_rate": 1.657954725997155e-05,
+      "loss": 0.9573,
+      "step": 1916
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.927834883238085,
+      "learning_rate": 1.6575813139505016e-05,
+      "loss": 0.9304,
+      "step": 1917
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.880364399662898,
+      "learning_rate": 1.6572077402888037e-05,
+      "loss": 0.8702,
+      "step": 1918
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9971621634760734,
+      "learning_rate": 1.6568340051038754e-05,
+      "loss": 0.963,
+      "step": 1919
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.7330086419109856,
+      "learning_rate": 1.6564601084875703e-05,
+      "loss": 0.9331,
+      "step": 1920
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2160871480825253,
+      "learning_rate": 1.6560860505317813e-05,
+      "loss": 1.0767,
+      "step": 1921
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6222550750639892,
+      "learning_rate": 1.655711831328442e-05,
+      "loss": 0.8794,
+      "step": 1922
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4903844136454523,
+      "learning_rate": 1.6553374509695244e-05,
+      "loss": 0.8383,
+      "step": 1923
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.8607341149397123,
+      "learning_rate": 1.6549629095470413e-05,
+      "loss": 0.9757,
+      "step": 1924
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.501136761886263,
+      "learning_rate": 1.6545882071530443e-05,
+      "loss": 1.1234,
+      "step": 1925
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.003942035104415,
+      "learning_rate": 1.654213343879624e-05,
+      "loss": 1.0772,
+      "step": 1926
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.455659166595698,
+      "learning_rate": 1.6538383198189122e-05,
+      "loss": 1.0843,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.885447760378492,
+      "learning_rate": 1.653463135063079e-05,
+      "loss": 0.9104,
+      "step": 1928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.144702023071639,
+      "learning_rate": 1.6530877897043343e-05,
+      "loss": 1.0566,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1471032152462115,
+      "learning_rate": 1.6527122838349274e-05,
+      "loss": 1.0278,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.890264832007854,
+      "learning_rate": 1.652336617547147e-05,
+      "loss": 0.9163,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9317766549252466,
+      "learning_rate": 1.6519607909333216e-05,
+      "loss": 0.8429,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4212716262094105,
+      "learning_rate": 1.6515848040858186e-05,
+      "loss": 0.9769,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.248556388386718,
+      "learning_rate": 1.6512086570970455e-05,
+      "loss": 0.8718,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.903299186840235,
+      "learning_rate": 1.650832350059448e-05,
+      "loss": 1.0638,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8786177507895068,
+      "learning_rate": 1.650455883065512e-05,
+      "loss": 0.991,
+      "step": 1936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8504920475696833,
+      "learning_rate": 1.650079256207763e-05,
+      "loss": 0.9286,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.108477746343841,
+      "learning_rate": 1.6497024695787646e-05,
+      "loss": 0.9307,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7398926753706903,
+      "learning_rate": 1.649325523271121e-05,
+      "loss": 0.9631,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.05048561325502,
+      "learning_rate": 1.6489484173774747e-05,
+      "loss": 0.934,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1390413170814235,
+      "learning_rate": 1.648571151990508e-05,
+      "loss": 0.9479,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.797072943904557,
+      "learning_rate": 1.648193727202942e-05,
+      "loss": 1.1066,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.838565494232607,
+      "learning_rate": 1.647816143107537e-05,
+      "loss": 1.0361,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0395818003900956,
+      "learning_rate": 1.6474383997970928e-05,
+      "loss": 0.9884,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9470028853067123,
+      "learning_rate": 1.6470604973644483e-05,
+      "loss": 0.9513,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7727680957102336,
+      "learning_rate": 1.6466824359024803e-05,
+      "loss": 0.966,
+      "step": 1946
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.713985789430561,
+      "learning_rate": 1.646304215504107e-05,
+      "loss": 1.1203,
+      "step": 1947
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0833960862474434,
+      "learning_rate": 1.645925836262284e-05,
+      "loss": 0.9877,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7057696057868297,
+      "learning_rate": 1.6455472982700055e-05,
+      "loss": 0.8799,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.884971897595675,
+      "learning_rate": 1.6451686016203065e-05,
+      "loss": 0.9695,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.159371915076138,
+      "learning_rate": 1.6447897464062593e-05,
+      "loss": 0.9409,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0243298134150836,
+      "learning_rate": 1.644410732720977e-05,
+      "loss": 0.904,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8079014057457963,
+      "learning_rate": 1.644031560657609e-05,
+      "loss": 0.9069,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.102887289670685,
+      "learning_rate": 1.6436522303093462e-05,
+      "loss": 0.8864,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8903116065538312,
+      "learning_rate": 1.6432727417694172e-05,
+      "loss": 0.9538,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2832123605271866,
+      "learning_rate": 1.6428930951310895e-05,
+      "loss": 0.9247,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2273969880571656,
+      "learning_rate": 1.6425132904876696e-05,
+      "loss": 0.9876,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.013714545246212,
+      "learning_rate": 1.642133327932503e-05,
+      "loss": 0.9537,
+      "step": 1958
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7463430437492122,
+      "learning_rate": 1.6417532075589733e-05,
+      "loss": 0.9166,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.844332090271417,
+      "learning_rate": 1.6413729294605043e-05,
+      "loss": 0.9816,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.087235521306631,
+      "learning_rate": 1.6409924937305567e-05,
+      "loss": 1.0865,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.752478659387522,
+      "learning_rate": 1.640611900462632e-05,
+      "loss": 0.9745,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8605402043191,
+      "learning_rate": 1.6402311497502685e-05,
+      "loss": 0.9028,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8371013619172736,
+      "learning_rate": 1.6398502416870444e-05,
+      "loss": 0.943,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1775931381327656,
+      "learning_rate": 1.6394691763665762e-05,
+      "loss": 0.8965,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1302370672192414,
+      "learning_rate": 1.6390879538825188e-05,
+      "loss": 1.0056,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0799939612715064,
+      "learning_rate": 1.6387065743285667e-05,
+      "loss": 0.9704,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6241877817898187,
+      "learning_rate": 1.6383250377984515e-05,
+      "loss": 0.9161,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0591703288395715,
+      "learning_rate": 1.637943344385944e-05,
+      "loss": 0.9756,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.780749719595347,
+      "learning_rate": 1.6375614941848553e-05,
+      "loss": 0.9215,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.78679062321962,
+      "learning_rate": 1.637179487289032e-05,
+      "loss": 0.9447,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.916558730513451,
+      "learning_rate": 1.6367973237923606e-05,
+      "loss": 1.017,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8714051172595014,
+      "learning_rate": 1.636415003788767e-05,
+      "loss": 0.9574,
+      "step": 1973
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.424562176046629,
+      "learning_rate": 1.6360325273722148e-05,
+      "loss": 0.7982,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7338308340261883,
+      "learning_rate": 1.6356498946367052e-05,
+      "loss": 0.9081,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.27037857350359,
+      "learning_rate": 1.635267105676279e-05,
+      "loss": 1.0226,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.1300143332121175,
+      "learning_rate": 1.634884160585015e-05,
+      "loss": 0.9527,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.106799123943485,
+      "learning_rate": 1.6345010594570303e-05,
+      "loss": 0.9228,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9397095069102734,
+      "learning_rate": 1.6341178023864803e-05,
+      "loss": 0.9816,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0107007075493217,
+      "learning_rate": 1.6337343894675594e-05,
+      "loss": 0.8376,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.238057799122706,
+      "learning_rate": 1.6333508207944987e-05,
+      "loss": 0.9929,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.767232716037717,
+      "learning_rate": 1.6329670964615698e-05,
+      "loss": 0.9193,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.068081714153525,
+      "learning_rate": 1.6325832165630804e-05,
+      "loss": 0.9252,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.403815270134738,
+      "learning_rate": 1.6321991811933778e-05,
+      "loss": 1.1198,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.99459133747253,
+      "learning_rate": 1.6318149904468476e-05,
+      "loss": 1.2209,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3344886744220963,
+      "learning_rate": 1.6314306444179124e-05,
+      "loss": 0.9521,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.507952799865198,
+      "learning_rate": 1.631046143201034e-05,
+      "loss": 1.1653,
+      "step": 1987
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.0183257520527236,
+      "learning_rate": 1.6306614868907118e-05,
+      "loss": 0.9675,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2917705295673354,
+      "learning_rate": 1.6302766755814837e-05,
+      "loss": 0.9624,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6996186995769325,
+      "learning_rate": 1.6298917093679256e-05,
+      "loss": 0.9873,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.92846853659815,
+      "learning_rate": 1.6295065883446514e-05,
+      "loss": 1.0206,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.6153606760314583,
+      "learning_rate": 1.6291213126063127e-05,
+      "loss": 0.8827,
+      "step": 1992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0482692285445476,
+      "learning_rate": 1.6287358822476003e-05,
+      "loss": 0.9584,
+      "step": 1993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.945888562890866,
+      "learning_rate": 1.628350297363241e-05,
+      "loss": 1.0334,
+      "step": 1994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9809895023289705,
+      "learning_rate": 1.6279645580480016e-05,
+      "loss": 0.8799,
+      "step": 1995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.040358876025292,
+      "learning_rate": 1.6275786643966857e-05,
+      "loss": 0.9606,
+      "step": 1996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.120304648204235,
+      "learning_rate": 1.627192616504135e-05,
+      "loss": 0.9562,
+      "step": 1997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8891192132781764,
+      "learning_rate": 1.6268064144652298e-05,
+      "loss": 0.9539,
+      "step": 1998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.618369413745189,
+      "learning_rate": 1.6264200583748872e-05,
+      "loss": 1.1742,
+      "step": 1999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.562754164287327,
+      "learning_rate": 1.6260335483280628e-05,
+      "loss": 0.9873,
+      "step": 2000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7923306530773155,
+      "learning_rate": 1.6256468844197503e-05,
+      "loss": 0.906,
+      "step": 2001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7896490661006172,
+      "learning_rate": 1.6252600667449803e-05,
+      "loss": 1.0057,
+      "step": 2002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0281024080777756,
+      "learning_rate": 1.624873095398822e-05,
+      "loss": 0.9022,
+      "step": 2003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.881684994302156,
+      "learning_rate": 1.6244859704763822e-05,
+      "loss": 0.9479,
+      "step": 2004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9444445700411137,
+      "learning_rate": 1.6240986920728047e-05,
+      "loss": 0.8036,
+      "step": 2005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7270622289384363,
+      "learning_rate": 1.6237112602832725e-05,
+      "loss": 0.8149,
+      "step": 2006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.932690790849735,
+      "learning_rate": 1.6233236752030055e-05,
+      "loss": 0.894,
+      "step": 2007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9384491189874713,
+      "learning_rate": 1.6229359369272604e-05,
+      "loss": 0.9274,
+      "step": 2008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2749757856588406,
+      "learning_rate": 1.622548045551333e-05,
+      "loss": 1.0205,
+      "step": 2009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2986011656446235,
+      "learning_rate": 1.6221600011705562e-05,
+      "loss": 0.9656,
+      "step": 2010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.899694894395896,
+      "learning_rate": 1.6217718038803004e-05,
+      "loss": 0.9747,
+      "step": 2011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.805963080021518,
+      "learning_rate": 1.621383453775973e-05,
+      "loss": 0.9258,
+      "step": 2012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.079907285065247,
+      "learning_rate": 1.6209949509530206e-05,
+      "loss": 1.0381,
+      "step": 2013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.85919475322715,
+      "learning_rate": 1.6206062955069252e-05,
+      "loss": 0.9665,
+      "step": 2014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.009738760174339,
+      "learning_rate": 1.6202174875332082e-05,
+      "loss": 0.9749,
+      "step": 2015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7899818371791323,
+      "learning_rate": 1.619828527127427e-05,
+      "loss": 0.8872,
+      "step": 2016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9113651910730733,
+      "learning_rate": 1.619439414385178e-05,
+      "loss": 0.93,
+      "step": 2017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1385972777593194,
+      "learning_rate": 1.6190501494020938e-05,
+      "loss": 0.878,
+      "step": 2018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.902655972971181,
+      "learning_rate": 1.6186607322738446e-05,
+      "loss": 0.9324,
+      "step": 2019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 36.255687160999884,
+      "learning_rate": 1.6182711630961385e-05,
+      "loss": 1.2675,
+      "step": 2020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8813953340320717,
+      "learning_rate": 1.6178814419647207e-05,
+      "loss": 0.9405,
+      "step": 2021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.146214173693879,
+      "learning_rate": 1.6174915689753733e-05,
+      "loss": 0.9028,
+      "step": 2022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.295327922343884,
+      "learning_rate": 1.6171015442239167e-05,
+      "loss": 0.8791,
+      "step": 2023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7436426736490667,
+      "learning_rate": 1.6167113678062074e-05,
+      "loss": 0.8547,
+      "step": 2024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.983002667352024,
+      "learning_rate": 1.6163210398181405e-05,
+      "loss": 0.9734,
+      "step": 2025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.21413114511768,
+      "learning_rate": 1.6159305603556474e-05,
+      "loss": 0.9882,
+      "step": 2026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8282094732335454,
+      "learning_rate": 1.615539929514697e-05,
+      "loss": 0.921,
+      "step": 2027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1604216542592587,
+      "learning_rate": 1.615149147391295e-05,
+      "loss": 0.9523,
+      "step": 2028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.856666507616215,
+      "learning_rate": 1.614758214081486e-05,
+      "loss": 0.936,
+      "step": 2029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.991532447251548,
+      "learning_rate": 1.6143671296813487e-05,
+      "loss": 0.9874,
+      "step": 2030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.840642279858671,
+      "learning_rate": 1.6139758942870015e-05,
+      "loss": 0.9078,
+      "step": 2031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9496055825239855,
+      "learning_rate": 1.6135845079945994e-05,
+      "loss": 0.9009,
+      "step": 2032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.4997572003360498,
+      "learning_rate": 1.6131929709003338e-05,
+      "loss": 0.8366,
+      "step": 2033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.854286295482546,
+      "learning_rate": 1.6128012831004334e-05,
+      "loss": 0.9281,
+      "step": 2034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0047445073134647,
+      "learning_rate": 1.6124094446911643e-05,
+      "loss": 0.9288,
+      "step": 2035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0747711973099534,
+      "learning_rate": 1.6120174557688296e-05,
+      "loss": 0.9777,
+      "step": 2036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.836496039165514,
+      "learning_rate": 1.6116253164297688e-05,
+      "loss": 1.0003,
+      "step": 2037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9495829885617866,
+      "learning_rate": 1.611233026770359e-05,
+      "loss": 0.958,
+      "step": 2038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.040177556037205,
+      "learning_rate": 1.6108405868870138e-05,
+      "loss": 1.0066,
+      "step": 2039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8392312040901073,
+      "learning_rate": 1.610447996876184e-05,
+      "loss": 0.9305,
+      "step": 2040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.858555557096998,
+      "learning_rate": 1.6100552568343575e-05,
+      "loss": 0.9561,
+      "step": 2041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.925244748188534,
+      "learning_rate": 1.609662366858058e-05,
+      "loss": 0.9858,
+      "step": 2042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 18.775907505689492,
+      "learning_rate": 1.6092693270438477e-05,
+      "loss": 1.1733,
+      "step": 2043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8831230007137116,
+      "learning_rate": 1.6088761374883244e-05,
+      "loss": 0.9558,
+      "step": 2044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0027686259492707,
+      "learning_rate": 1.608482798288123e-05,
+      "loss": 0.9855,
+      "step": 2045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3576253702197185,
+      "learning_rate": 1.6080893095399154e-05,
+      "loss": 0.996,
+      "step": 2046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.9658837524613344,
+      "learning_rate": 1.6076956713404096e-05,
+      "loss": 0.9021,
+      "step": 2047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.1626601985594966,
+      "learning_rate": 1.607301883786352e-05,
+      "loss": 1.0554,
+      "step": 2048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0553939427426196,
+      "learning_rate": 1.6069079469745232e-05,
+      "loss": 0.7729,
+      "step": 2049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7411280804515896,
+      "learning_rate": 1.6065138610017425e-05,
+      "loss": 0.975,
+      "step": 2050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.774742963619658,
+      "learning_rate": 1.606119625964865e-05,
+      "loss": 0.7768,
+      "step": 2051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 38.98687845629915,
+      "learning_rate": 1.605725241960783e-05,
+      "loss": 1.1693,
+      "step": 2052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7458443144950424,
+      "learning_rate": 1.6053307090864247e-05,
+      "loss": 0.9069,
+      "step": 2053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.919021481663924,
+      "learning_rate": 1.604936027438755e-05,
+      "loss": 0.9223,
+      "step": 2054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.948893360420509,
+      "learning_rate": 1.604541197114776e-05,
+      "loss": 0.94,
+      "step": 2055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.8116682169195295,
+      "learning_rate": 1.6041462182115257e-05,
+      "loss": 0.9397,
+      "step": 2056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.701273752635258,
+      "learning_rate": 1.603751090826079e-05,
+      "loss": 0.9303,
+      "step": 2057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9204076878350715,
+      "learning_rate": 1.6033558150555468e-05,
+      "loss": 0.9636,
+      "step": 2058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9249074674731808,
+      "learning_rate": 1.602960390997077e-05,
+      "loss": 0.8089,
+      "step": 2059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8604826058914865,
+      "learning_rate": 1.602564818747854e-05,
+      "loss": 0.9502,
+      "step": 2060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.642150556596377,
+      "learning_rate": 1.602169098405098e-05,
+      "loss": 0.8838,
+      "step": 2061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.671658746125172,
+      "learning_rate": 1.601773230066066e-05,
+      "loss": 0.8714,
+      "step": 2062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0325050799154147,
+      "learning_rate": 1.6013772138280516e-05,
+      "loss": 0.945,
+      "step": 2063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0000570968848552,
+      "learning_rate": 1.600981049788384e-05,
+      "loss": 1.0131,
+      "step": 2064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.904729128627338,
+      "learning_rate": 1.6005847380444296e-05,
+      "loss": 0.9075,
+      "step": 2065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.716502377018324,
+      "learning_rate": 1.6001882786935906e-05,
+      "loss": 0.8516,
+      "step": 2066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0878573330752905,
+      "learning_rate": 1.599791671833306e-05,
+      "loss": 0.8876,
+      "step": 2067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0105827423255227,
+      "learning_rate": 1.5993949175610496e-05,
+      "loss": 0.9043,
+      "step": 2068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8610460528157593,
+      "learning_rate": 1.5989980159743336e-05,
+      "loss": 0.9194,
+      "step": 2069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2183249053149567,
+      "learning_rate": 1.5986009671707048e-05,
+      "loss": 1.097,
+      "step": 2070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.866509114415062,
+      "learning_rate": 1.5982037712477466e-05,
+      "loss": 0.9069,
+      "step": 2071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9793432999723133,
+      "learning_rate": 1.5978064283030784e-05,
+      "loss": 1.0063,
+      "step": 2072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.366780686052711,
+      "learning_rate": 1.597408938434356e-05,
+      "loss": 0.9817,
+      "step": 2073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7600076399826334,
+      "learning_rate": 1.5970113017392724e-05,
+      "loss": 0.9071,
+      "step": 2074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1655593506091955,
+      "learning_rate": 1.5966135183155542e-05,
+      "loss": 0.8892,
+      "step": 2075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.04836429836989,
+      "learning_rate": 1.5962155882609657e-05,
+      "loss": 0.9741,
+      "step": 2076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.977526763468514,
+      "learning_rate": 1.5958175116733074e-05,
+      "loss": 1.2192,
+      "step": 2077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.726548999880068,
+      "learning_rate": 1.595419288650415e-05,
+      "loss": 0.9317,
+      "step": 2078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.011217259674851,
+      "learning_rate": 1.5950209192901603e-05,
+      "loss": 0.9827,
+      "step": 2079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.29525534444165,
+      "learning_rate": 1.594622403690452e-05,
+      "loss": 0.9244,
+      "step": 2080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.229445003686063,
+      "learning_rate": 1.5942237419492334e-05,
+      "loss": 0.9362,
+      "step": 2081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9204486236688245,
+      "learning_rate": 1.5938249341644847e-05,
+      "loss": 0.9227,
+      "step": 2082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.927607260082721,
+      "learning_rate": 1.5934259804342218e-05,
+      "loss": 0.9349,
+      "step": 2083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.837295266420575,
+      "learning_rate": 1.593026880856496e-05,
+      "loss": 0.8942,
+      "step": 2084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.651081252949924,
+      "learning_rate": 1.5926276355293948e-05,
+      "loss": 0.8587,
+      "step": 2085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.857579788362404,
+      "learning_rate": 1.5922282445510422e-05,
+      "loss": 1.1825,
+      "step": 2086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.641208085873871,
+      "learning_rate": 1.5918287080195962e-05,
+      "loss": 0.9027,
+      "step": 2087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.80746351383302,
+      "learning_rate": 1.5914290260332524e-05,
+      "loss": 0.9674,
+      "step": 2088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0202119253982875,
+      "learning_rate": 1.5910291986902415e-05,
+      "loss": 1.0477,
+      "step": 2089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7121329518000694,
+      "learning_rate": 1.590629226088829e-05,
+      "loss": 0.9407,
+      "step": 2090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8600057410215216,
+      "learning_rate": 1.5902291083273182e-05,
+      "loss": 0.812,
+      "step": 2091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.95858748031997,
+      "learning_rate": 1.589828845504046e-05,
+      "loss": 0.8556,
+      "step": 2092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.746810792832964,
+      "learning_rate": 1.589428437717386e-05,
+      "loss": 1.1038,
+      "step": 2093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.917450104375745,
+      "learning_rate": 1.5890278850657468e-05,
+      "loss": 0.963,
+      "step": 2094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.148069985975481,
+      "learning_rate": 1.5886271876475733e-05,
+      "loss": 1.0287,
+      "step": 2095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.919608954475873,
+      "learning_rate": 1.588226345561346e-05,
+      "loss": 1.0119,
+      "step": 2096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0119482294019226,
+      "learning_rate": 1.5878253589055807e-05,
+      "loss": 0.9433,
+      "step": 2097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.745844362970861,
+      "learning_rate": 1.587424227778828e-05,
+      "loss": 0.9186,
+      "step": 2098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7795582081388304,
+      "learning_rate": 1.587022952279675e-05,
+      "loss": 0.939,
+      "step": 2099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9849086913750154,
+      "learning_rate": 1.5866215325067442e-05,
+      "loss": 0.9054,
+      "step": 2100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.830979510479411,
+      "learning_rate": 1.586219968558693e-05,
+      "loss": 0.9651,
+      "step": 2101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.921574916421984,
+      "learning_rate": 1.5858182605342146e-05,
+      "loss": 0.8421,
+      "step": 2102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8984836166465087,
+      "learning_rate": 1.5854164085320375e-05,
+      "loss": 0.9944,
+      "step": 2103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.96415686502195,
+      "learning_rate": 1.5850144126509263e-05,
+      "loss": 0.8441,
+      "step": 2104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.399398202622699,
+      "learning_rate": 1.5846122729896796e-05,
+      "loss": 1.0563,
+      "step": 2105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8074715800805694,
+      "learning_rate": 1.5842099896471322e-05,
+      "loss": 0.8831,
+      "step": 2106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.036041778722464,
+      "learning_rate": 1.583807562722154e-05,
+      "loss": 0.9816,
+      "step": 2107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0284113302543822,
+      "learning_rate": 1.5834049923136508e-05,
+      "loss": 1.0098,
+      "step": 2108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7496789939771475,
+      "learning_rate": 1.5830022785205623e-05,
+      "loss": 0.8986,
+      "step": 2109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.953041494056494,
+      "learning_rate": 1.5825994214418646e-05,
+      "loss": 0.8537,
+      "step": 2110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.659977164215324,
+      "learning_rate": 1.582196421176569e-05,
+      "loss": 0.9785,
+      "step": 2111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.204140063929111,
+      "learning_rate": 1.5817932778237217e-05,
+      "loss": 0.9483,
+      "step": 2112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.6887868024183654,
+      "learning_rate": 1.581389991482403e-05,
+      "loss": 0.9564,
+      "step": 2113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7755813582969027,
+      "learning_rate": 1.5809865622517303e-05,
+      "loss": 0.9934,
+      "step": 2114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.845892550045217,
+      "learning_rate": 1.5805829902308555e-05,
+      "loss": 0.9452,
+      "step": 2115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.853050815785061,
+      "learning_rate": 1.580179275518964e-05,
+      "loss": 0.93,
+      "step": 2116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.2733220558775726,
+      "learning_rate": 1.5797754182152786e-05,
+      "loss": 1.0112,
+      "step": 2117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.979523192746589,
+      "learning_rate": 1.579371418419056e-05,
+      "loss": 0.8896,
+      "step": 2118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.4450783956006923,
+      "learning_rate": 1.578967276229588e-05,
+      "loss": 1.1442,
+      "step": 2119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.864541500830383,
+      "learning_rate": 1.5785629917462005e-05,
+      "loss": 0.8686,
+      "step": 2120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8665462714350847,
+      "learning_rate": 1.5781585650682565e-05,
+      "loss": 0.8522,
+      "step": 2121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.2402800359131625,
+      "learning_rate": 1.5777539962951526e-05,
+      "loss": 1.0849,
+      "step": 2122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.9401554666052516,
+      "learning_rate": 1.5773492855263196e-05,
+      "loss": 0.9797,
+      "step": 2123
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.826087551554652,
+      "learning_rate": 1.5769444328612248e-05,
+      "loss": 0.9366,
+      "step": 2124
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9748307430738388,
+      "learning_rate": 1.5765394383993693e-05,
+      "loss": 0.9461,
+      "step": 2125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.917612348057046,
+      "learning_rate": 1.57613430224029e-05,
+      "loss": 0.9599,
+      "step": 2126
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9492782151559545,
+      "learning_rate": 1.5757290244835566e-05,
+      "loss": 0.9607,
+      "step": 2127
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.104337592341858,
+      "learning_rate": 1.5753236052287766e-05,
+      "loss": 0.9517,
+      "step": 2128
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8612025981659652,
+      "learning_rate": 1.57491804457559e-05,
+      "loss": 1.0167,
+      "step": 2129
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8123553472731198,
+      "learning_rate": 1.5745123426236716e-05,
+      "loss": 0.9264,
+      "step": 2130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.830018363337432,
+      "learning_rate": 1.5741064994727327e-05,
+      "loss": 0.8245,
+      "step": 2131
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8110191476042683,
+      "learning_rate": 1.5737005152225176e-05,
+      "loss": 0.8709,
+      "step": 2132
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8502055155928923,
+      "learning_rate": 1.573294389972806e-05,
+      "loss": 1.0445,
+      "step": 2133
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0421641138681763,
+      "learning_rate": 1.5728881238234118e-05,
+      "loss": 1.0078,
+      "step": 2134
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.80236736793243,
+      "learning_rate": 1.572481716874184e-05,
+      "loss": 0.9258,
+      "step": 2135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1396454859019043,
+      "learning_rate": 1.572075169225006e-05,
+      "loss": 1.0275,
+      "step": 2136
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.01442584800464,
+      "learning_rate": 1.571668480975796e-05,
+      "loss": 0.9114,
+      "step": 2137
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9439289256223495,
+      "learning_rate": 1.5712616522265062e-05,
+      "loss": 0.9387,
+      "step": 2138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.86630479136464,
+      "learning_rate": 1.5708546830771242e-05,
+      "loss": 0.8924,
+      "step": 2139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9053380514019573,
+      "learning_rate": 1.5704475736276708e-05,
+      "loss": 0.9631,
+      "step": 2140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.368103684245844,
+      "learning_rate": 1.5700403239782035e-05,
+      "loss": 0.913,
+      "step": 2141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.857579536349821,
+      "learning_rate": 1.569632934228811e-05,
+      "loss": 1.0226,
+      "step": 2142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0564847208325254,
+      "learning_rate": 1.56922540447962e-05,
+      "loss": 0.9765,
+      "step": 2143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9287560858591966,
+      "learning_rate": 1.568817734830789e-05,
+      "loss": 0.8979,
+      "step": 2144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8933213409616583,
+      "learning_rate": 1.5684099253825117e-05,
+      "loss": 0.9059,
+      "step": 2145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.639794473491297,
+      "learning_rate": 1.5680019762350162e-05,
+      "loss": 0.9327,
+      "step": 2146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.891538719719891,
+      "learning_rate": 1.567593887488565e-05,
+      "loss": 1.0255,
+      "step": 2147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8097333302096015,
+      "learning_rate": 1.5671856592434557e-05,
+      "loss": 0.9987,
+      "step": 2148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.788275914343878,
+      "learning_rate": 1.5667772916000182e-05,
+      "loss": 0.8716,
+      "step": 2149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.968827801790018,
+      "learning_rate": 1.5663687846586183e-05,
+      "loss": 0.8922,
+      "step": 2150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9789373637634182,
+      "learning_rate": 1.5659601385196555e-05,
+      "loss": 0.7631,
+      "step": 2151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.641680662522926,
+      "learning_rate": 1.565551353283564e-05,
+      "loss": 0.8783,
+      "step": 2152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.992765042257369,
+      "learning_rate": 1.565142429050811e-05,
+      "loss": 0.9534,
+      "step": 2153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.912789857674366,
+      "learning_rate": 1.564733365921899e-05,
+      "loss": 0.9623,
+      "step": 2154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1939374684533623,
+      "learning_rate": 1.564324163997364e-05,
+      "loss": 0.9484,
+      "step": 2155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.910399280383399,
+      "learning_rate": 1.563914823377777e-05,
+      "loss": 1.05,
+      "step": 2156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7829633029832346,
+      "learning_rate": 1.5635053441637416e-05,
+      "loss": 0.899,
+      "step": 2157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8483703755801475,
+      "learning_rate": 1.563095726455897e-05,
+      "loss": 0.9353,
+      "step": 2158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1605453107869805,
+      "learning_rate": 1.5626859703549153e-05,
+      "loss": 0.9756,
+      "step": 2159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6579454798900675,
+      "learning_rate": 1.5622760759615033e-05,
+      "loss": 0.9316,
+      "step": 2160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.895988425803534,
+      "learning_rate": 1.561866043376401e-05,
+      "loss": 0.8316,
+      "step": 2161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1091023855621067,
+      "learning_rate": 1.5614558727003838e-05,
+      "loss": 0.8095,
+      "step": 2162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.723661965840028,
+      "learning_rate": 1.56104556403426e-05,
+      "loss": 0.9086,
+      "step": 2163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1270182582775274,
+      "learning_rate": 1.560635117478871e-05,
+      "loss": 0.9554,
+      "step": 2164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9598489071966627,
+      "learning_rate": 1.560224533135094e-05,
+      "loss": 0.864,
+      "step": 2165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.912429840064479,
+      "learning_rate": 1.559813811103839e-05,
+      "loss": 0.9377,
+      "step": 2166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8075984318631653,
+      "learning_rate": 1.55940295148605e-05,
+      "loss": 0.9434,
+      "step": 2167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9689893677610657,
+      "learning_rate": 1.5589919543827038e-05,
+      "loss": 0.8091,
+      "step": 2168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7527341533964913,
+      "learning_rate": 1.5585808198948135e-05,
+      "loss": 0.9535,
+      "step": 2169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.240281727452343,
+      "learning_rate": 1.5581695481234234e-05,
+      "loss": 0.9207,
+      "step": 2170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9346697305866134,
+      "learning_rate": 1.5577581391696125e-05,
+      "loss": 0.9847,
+      "step": 2171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.2363181840874184,
+      "learning_rate": 1.5573465931344943e-05,
+      "loss": 0.9466,
+      "step": 2172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1311202216405905,
+      "learning_rate": 1.5569349101192147e-05,
+      "loss": 0.7614,
+      "step": 2173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9897802189566147,
+      "learning_rate": 1.556523090224954e-05,
+      "loss": 0.9406,
+      "step": 2174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.866293002770438,
+      "learning_rate": 1.5561111335529262e-05,
+      "loss": 0.856,
+      "step": 2175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7666487683268266,
+      "learning_rate": 1.5556990402043785e-05,
+      "loss": 0.8607,
+      "step": 2176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.206477614388165,
+      "learning_rate": 1.5552868102805914e-05,
+      "loss": 0.9272,
+      "step": 2177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.703502589931452,
+      "learning_rate": 1.5548744438828806e-05,
+      "loss": 0.8436,
+      "step": 2178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9045504062867398,
+      "learning_rate": 1.5544619411125932e-05,
+      "loss": 0.9542,
+      "step": 2179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4383119642842557,
+      "learning_rate": 1.554049302071111e-05,
+      "loss": 0.9043,
+      "step": 2180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.749910787471238,
+      "learning_rate": 1.5536365268598495e-05,
+      "loss": 0.9428,
+      "step": 2181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.0887495819218884,
+      "learning_rate": 1.5532236155802568e-05,
+      "loss": 0.8238,
+      "step": 2182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7762753591598526,
+      "learning_rate": 1.5528105683338153e-05,
+      "loss": 0.9144,
+      "step": 2183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.8725126638936267,
+      "learning_rate": 1.5523973852220403e-05,
+      "loss": 0.8462,
+      "step": 2184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9885596304544615,
+      "learning_rate": 1.5519840663464803e-05,
+      "loss": 0.8969,
+      "step": 2185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.473266378506294,
+      "learning_rate": 1.5515706118087178e-05,
+      "loss": 1.0156,
+      "step": 2186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.417240186483622,
+      "learning_rate": 1.5511570217103686e-05,
+      "loss": 1.1446,
+      "step": 2187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.084298206467197,
+      "learning_rate": 1.550743296153081e-05,
+      "loss": 0.9202,
+      "step": 2188
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9819144441801195,
+      "learning_rate": 1.5503294352385376e-05,
+      "loss": 0.9559,
+      "step": 2189
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.69663067192699,
+      "learning_rate": 1.5499154390684534e-05,
+      "loss": 1.1093,
+      "step": 2190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9797669448476896,
+      "learning_rate": 1.5495013077445773e-05,
+      "loss": 0.9774,
+      "step": 2191
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.1250248267064844,
+      "learning_rate": 1.5490870413686913e-05,
+      "loss": 0.8908,
+      "step": 2192
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.978205398929475,
+      "learning_rate": 1.54867264004261e-05,
+      "loss": 1.0935,
+      "step": 2193
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3254960105618383,
+      "learning_rate": 1.5482581038681817e-05,
+      "loss": 0.9489,
+      "step": 2194
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.423818665589391,
+      "learning_rate": 1.5478434329472883e-05,
+      "loss": 0.9161,
+      "step": 2195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.997633325042334,
+      "learning_rate": 1.5474286273818437e-05,
+      "loss": 1.0483,
+      "step": 2196
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.901737431846852,
+      "learning_rate": 1.547013687273796e-05,
+      "loss": 0.8797,
+      "step": 2197
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9155767889936506,
+      "learning_rate": 1.5465986127251253e-05,
+      "loss": 1.0351,
+      "step": 2198
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.893278694107767,
+      "learning_rate": 1.546183403837845e-05,
+      "loss": 0.9258,
+      "step": 2199
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.995767480013683,
+      "learning_rate": 1.545768060714003e-05,
+      "loss": 0.8585,
+      "step": 2200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9454418666606514,
+      "learning_rate": 1.5453525834556784e-05,
+      "loss": 0.9562,
+      "step": 2201
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.743360233087499,
+      "learning_rate": 1.5449369721649835e-05,
+      "loss": 0.8992,
+      "step": 2202
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0252718804825993,
+      "learning_rate": 1.5445212269440644e-05,
+      "loss": 1.0418,
+      "step": 2203
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8837846598545855,
+      "learning_rate": 1.544105347895099e-05,
+      "loss": 0.9565,
+      "step": 2204
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2789960918949608,
+      "learning_rate": 1.5436893351202993e-05,
+      "loss": 0.9387,
+      "step": 2205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.385976907084884,
+      "learning_rate": 1.5432731887219094e-05,
+      "loss": 1.1257,
+      "step": 2206
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8727045249703735,
+      "learning_rate": 1.5428569088022067e-05,
+      "loss": 0.9819,
+      "step": 2207
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9735127238030588,
+      "learning_rate": 1.5424404954635e-05,
+      "loss": 0.9494,
+      "step": 2208
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0445408913692784,
+      "learning_rate": 1.5420239488081335e-05,
+      "loss": 0.8285,
+      "step": 2209
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.763826104396605,
+      "learning_rate": 1.5416072689384818e-05,
+      "loss": 1.1142,
+      "step": 2210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.061026010845337,
+      "learning_rate": 1.5411904559569536e-05,
+      "loss": 0.8954,
+      "step": 2211
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.084396939692579,
+      "learning_rate": 1.540773509965989e-05,
+      "loss": 0.9391,
+      "step": 2212
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8981176045057295,
+      "learning_rate": 1.5403564310680627e-05,
+      "loss": 0.841,
+      "step": 2213
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7909013188405085,
+      "learning_rate": 1.5399392193656802e-05,
+      "loss": 0.9828,
+      "step": 2214
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.957934625908896,
+      "learning_rate": 1.5395218749613808e-05,
+      "loss": 0.9562,
+      "step": 2215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.752532249725703,
+      "learning_rate": 1.5391043979577364e-05,
+      "loss": 0.9208,
+      "step": 2216
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9615314376475936,
+      "learning_rate": 1.5386867884573505e-05,
+      "loss": 0.9258,
+      "step": 2217
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.010242558942028,
+      "learning_rate": 1.53826904656286e-05,
+      "loss": 0.9155,
+      "step": 2218
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2902012119762647,
+      "learning_rate": 1.537851172376934e-05,
+      "loss": 0.9412,
+      "step": 2219
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9225946084002095,
+      "learning_rate": 1.5374331660022744e-05,
+      "loss": 0.9045,
+      "step": 2220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.939000118464009,
+      "learning_rate": 1.537015027541616e-05,
+      "loss": 0.9167,
+      "step": 2221
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8952624644975042,
+      "learning_rate": 1.5365967570977244e-05,
+      "loss": 0.9866,
+      "step": 2222
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9487056282258512,
+      "learning_rate": 1.5361783547733997e-05,
+      "loss": 0.9799,
+      "step": 2223
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0272505909958296,
+      "learning_rate": 1.5357598206714726e-05,
+      "loss": 0.8923,
+      "step": 2224
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9581269346386927,
+      "learning_rate": 1.535341154894808e-05,
+      "loss": 0.9712,
+      "step": 2225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6448689391213436,
+      "learning_rate": 1.5349223575463015e-05,
+      "loss": 0.9218,
+      "step": 2226
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7704869063269126,
+      "learning_rate": 1.534503428728882e-05,
+      "loss": 0.9657,
+      "step": 2227
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.663305732820632,
+      "learning_rate": 1.5340843685455105e-05,
+      "loss": 0.8638,
+      "step": 2228
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.90419193223228,
+      "learning_rate": 1.5336651770991806e-05,
+      "loss": 0.9674,
+      "step": 2229
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.888348711422464,
+      "learning_rate": 1.533245854492917e-05,
+      "loss": 0.8858,
+      "step": 2230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 28.443076215648286,
+      "learning_rate": 1.5328264008297774e-05,
+      "loss": 1.1028,
+      "step": 2231
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.67306035825745,
+      "learning_rate": 1.5324068162128524e-05,
+      "loss": 0.9006,
+      "step": 2232
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.883306393381524,
+      "learning_rate": 1.5319871007452643e-05,
+      "loss": 0.9638,
+      "step": 2233
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9956980307536196,
+      "learning_rate": 1.5315672545301668e-05,
+      "loss": 0.8938,
+      "step": 2234
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8465615901796655,
+      "learning_rate": 1.531147277670746e-05,
+      "loss": 0.9024,
+      "step": 2235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.0133716494370693,
+      "learning_rate": 1.5307271702702215e-05,
+      "loss": 0.8917,
+      "step": 2236
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7472224187281222,
+      "learning_rate": 1.530306932431843e-05,
+      "loss": 0.8313,
+      "step": 2237
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.9205674411797533,
+      "learning_rate": 1.529886564258894e-05,
+      "loss": 0.9551,
+      "step": 2238
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8579615936508076,
+      "learning_rate": 1.529466065854688e-05,
+      "loss": 0.9034,
+      "step": 2239
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.077793196274853,
+      "learning_rate": 1.5290454373225728e-05,
+      "loss": 0.9184,
+      "step": 2240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8593478786344995,
+      "learning_rate": 1.528624678765927e-05,
+      "loss": 1.0188,
+      "step": 2241
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.786932569982691,
+      "learning_rate": 1.528203790288161e-05,
+      "loss": 0.8707,
+      "step": 2242
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.986711280532268,
+      "learning_rate": 1.5277827719927168e-05,
+      "loss": 1.0005,
+      "step": 2243
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8440787062164885,
+      "learning_rate": 1.5273616239830697e-05,
+      "loss": 0.9213,
+      "step": 2244
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.778194347417419,
+      "learning_rate": 1.5269403463627263e-05,
+      "loss": 1.0728,
+      "step": 2245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2806136694782992,
+      "learning_rate": 1.5265189392352238e-05,
+      "loss": 0.9671,
+      "step": 2246
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2636437257793274,
+      "learning_rate": 1.5260974027041328e-05,
+      "loss": 0.9756,
+      "step": 2247
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7949362496071135,
+      "learning_rate": 1.5256757368730557e-05,
+      "loss": 1.0169,
+      "step": 2248
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.883127148241771,
+      "learning_rate": 1.5252539418456253e-05,
+      "loss": 0.9109,
+      "step": 2249
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.079927358289439,
+      "learning_rate": 1.5248320177255074e-05,
+      "loss": 0.9541,
+      "step": 2250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.066797245053238,
+      "learning_rate": 1.5244099646163987e-05,
+      "loss": 1.009,
+      "step": 2251
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.196708284476865,
+      "learning_rate": 1.5239877826220286e-05,
+      "loss": 0.9755,
+      "step": 2252
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.972529111217515,
+      "learning_rate": 1.5235654718461572e-05,
+      "loss": 0.8432,
+      "step": 2253
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6898726536996396,
+      "learning_rate": 1.5231430323925768e-05,
+      "loss": 0.8175,
+      "step": 2254
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.052173706002016,
+      "learning_rate": 1.5227204643651112e-05,
+      "loss": 0.979,
+      "step": 2255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.370484652099961,
+      "learning_rate": 1.5222977678676159e-05,
+      "loss": 0.955,
+      "step": 2256
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.857037218718541,
+      "learning_rate": 1.5218749430039772e-05,
+      "loss": 0.9497,
+      "step": 2257
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.686950220905987,
+      "learning_rate": 1.5214519898781141e-05,
+      "loss": 0.8466,
+      "step": 2258
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.023274578834779,
+      "learning_rate": 1.5210289085939769e-05,
+      "loss": 1.0174,
+      "step": 2259
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.017154767031705,
+      "learning_rate": 1.5206056992555465e-05,
+      "loss": 0.9454,
+      "step": 2260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.693819407507513,
+      "learning_rate": 1.5201823619668362e-05,
+      "loss": 0.9532,
+      "step": 2261
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7126130445896273,
+      "learning_rate": 1.5197588968318904e-05,
+      "loss": 0.957,
+      "step": 2262
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.585559358851023,
+      "learning_rate": 1.519335303954785e-05,
+      "loss": 0.9274,
+      "step": 2263
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9563349968700394,
+      "learning_rate": 1.518911583439627e-05,
+      "loss": 0.9926,
+      "step": 2264
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3004301639856544,
+      "learning_rate": 1.5184877353905556e-05,
+      "loss": 0.946,
+      "step": 2265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7952021511451006,
+      "learning_rate": 1.5180637599117401e-05,
+      "loss": 0.942,
+      "step": 2266
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0215029152891892,
+      "learning_rate": 1.5176396571073821e-05,
+      "loss": 0.9685,
+      "step": 2267
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7965263658600503,
+      "learning_rate": 1.517215427081714e-05,
+      "loss": 0.9508,
+      "step": 2268
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.917787388458425,
+      "learning_rate": 1.516791069939e-05,
+      "loss": 0.9753,
+      "step": 2269
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7820103402362046,
+      "learning_rate": 1.5163665857835348e-05,
+      "loss": 0.9348,
+      "step": 2270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0049793733461914,
+      "learning_rate": 1.515941974719645e-05,
+      "loss": 0.9459,
+      "step": 2271
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9106205504262777,
+      "learning_rate": 1.515517236851688e-05,
+      "loss": 0.8804,
+      "step": 2272
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9958370089998407,
+      "learning_rate": 1.5150923722840523e-05,
+      "loss": 0.9294,
+      "step": 2273
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.861271074958744,
+      "learning_rate": 1.5146673811211576e-05,
+      "loss": 0.9511,
+      "step": 2274
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8958602073341666,
+      "learning_rate": 1.5142422634674551e-05,
+      "loss": 0.9407,
+      "step": 2275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.847941228108326,
+      "learning_rate": 1.5138170194274269e-05,
+      "loss": 1.0268,
+      "step": 2276
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.902937511503127,
+      "learning_rate": 1.5133916491055858e-05,
+      "loss": 0.8619,
+      "step": 2277
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.941672149117542,
+      "learning_rate": 1.512966152606476e-05,
+      "loss": 1.0764,
+      "step": 2278
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.1262853598966927,
+      "learning_rate": 1.5125405300346722e-05,
+      "loss": 0.9714,
+      "step": 2279
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.502183056519954,
+      "learning_rate": 1.5121147814947812e-05,
+      "loss": 0.8329,
+      "step": 2280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.779765611110997,
+      "learning_rate": 1.5116889070914397e-05,
+      "loss": 0.9328,
+      "step": 2281
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.347802675409558,
+      "learning_rate": 1.5112629069293156e-05,
+      "loss": 0.9771,
+      "step": 2282
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7310814686797293,
+      "learning_rate": 1.5108367811131079e-05,
+      "loss": 0.8361,
+      "step": 2283
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9784430820973946,
+      "learning_rate": 1.5104105297475462e-05,
+      "loss": 0.9534,
+      "step": 2284
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.90244730992472,
+      "learning_rate": 1.5099841529373918e-05,
+      "loss": 1.0294,
+      "step": 2285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.741448217823855,
+      "learning_rate": 1.5095576507874353e-05,
+      "loss": 1.0374,
+      "step": 2286
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.84138712585894,
+      "learning_rate": 1.5091310234024991e-05,
+      "loss": 0.8247,
+      "step": 2287
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.113982915053296,
+      "learning_rate": 1.5087042708874368e-05,
+      "loss": 0.8825,
+      "step": 2288
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9396202859487635,
+      "learning_rate": 1.508277393347132e-05,
+      "loss": 0.8591,
+      "step": 2289
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.804939430938009,
+      "learning_rate": 1.5078503908864985e-05,
+      "loss": 0.9577,
+      "step": 2290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.106725332329504,
+      "learning_rate": 1.5074232636104824e-05,
+      "loss": 0.9256,
+      "step": 2291
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8731329023678387,
+      "learning_rate": 1.5069960116240597e-05,
+      "loss": 0.8268,
+      "step": 2292
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.561527056009055,
+      "learning_rate": 1.506568635032236e-05,
+      "loss": 0.8696,
+      "step": 2293
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.293742479742836,
+      "learning_rate": 1.5061411339400494e-05,
+      "loss": 0.9943,
+      "step": 2294
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2364700641093647,
+      "learning_rate": 1.5057135084525671e-05,
+      "loss": 1.0094,
+      "step": 2295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8922317115255423,
+      "learning_rate": 1.5052857586748881e-05,
+      "loss": 0.9652,
+      "step": 2296
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7080024377532146,
+      "learning_rate": 1.5048578847121405e-05,
+      "loss": 0.8256,
+      "step": 2297
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.6647130744588314,
+      "learning_rate": 1.5044298866694842e-05,
+      "loss": 1.0209,
+      "step": 2298
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7470063263170243,
+      "learning_rate": 1.504001764652109e-05,
+      "loss": 0.8917,
+      "step": 2299
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8360347324676005,
+      "learning_rate": 1.5035735187652353e-05,
+      "loss": 0.9454,
+      "step": 2300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.900545216220249,
+      "learning_rate": 1.503145149114114e-05,
+      "loss": 0.9046,
+      "step": 2301
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8178594587932593,
+      "learning_rate": 1.5027166558040262e-05,
+      "loss": 1.0504,
+      "step": 2302
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.01761000475965,
+      "learning_rate": 1.502288038940284e-05,
+      "loss": 0.9777,
+      "step": 2303
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.979517837174316,
+      "learning_rate": 1.5018592986282286e-05,
+      "loss": 0.848,
+      "step": 2304
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9140866574800333,
+      "learning_rate": 1.5014304349732327e-05,
+      "loss": 0.8894,
+      "step": 2305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8764694466178775,
+      "learning_rate": 1.5010014480806994e-05,
+      "loss": 1.0152,
+      "step": 2306
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.081285640863487,
+      "learning_rate": 1.5005723380560613e-05,
+      "loss": 0.7847,
+      "step": 2307
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.636984699828142,
+      "learning_rate": 1.5001431050047814e-05,
+      "loss": 0.9056,
+      "step": 2308
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.833963909345822,
+      "learning_rate": 1.499713749032353e-05,
+      "loss": 0.9704,
+      "step": 2309
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0114563448994174,
+      "learning_rate": 1.4992842702443005e-05,
+      "loss": 0.982,
+      "step": 2310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7306229723513553,
+      "learning_rate": 1.4988546687461774e-05,
+      "loss": 0.8364,
+      "step": 2311
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0296755610921164,
+      "learning_rate": 1.4984249446435674e-05,
+      "loss": 0.9381,
+      "step": 2312
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.991527609025026,
+      "learning_rate": 1.4979950980420847e-05,
+      "loss": 0.9938,
+      "step": 2313
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9612173589174624,
+      "learning_rate": 1.4975651290473741e-05,
+      "loss": 0.7781,
+      "step": 2314
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8810874165072233,
+      "learning_rate": 1.4971350377651093e-05,
+      "loss": 1.0572,
+      "step": 2315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.8538670159087545,
+      "learning_rate": 1.496704824300995e-05,
+      "loss": 1.0219,
+      "step": 2316
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0232911034956933,
+      "learning_rate": 1.4962744887607654e-05,
+      "loss": 0.9658,
+      "step": 2317
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.7966552102771116,
+      "learning_rate": 1.4958440312501852e-05,
+      "loss": 0.9018,
+      "step": 2318
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.008138253863501,
+      "learning_rate": 1.4954134518750483e-05,
+      "loss": 0.8631,
+      "step": 2319
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.99847234940024,
+      "learning_rate": 1.4949827507411791e-05,
+      "loss": 1.0202,
+      "step": 2320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.105716574195463,
+      "learning_rate": 1.4945519279544325e-05,
+      "loss": 0.8817,
+      "step": 2321
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.613668461496793,
+      "learning_rate": 1.4941209836206922e-05,
+      "loss": 0.9624,
+      "step": 2322
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8192085782602696,
+      "learning_rate": 1.4936899178458724e-05,
+      "loss": 0.976,
+      "step": 2323
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.931181999122721,
+      "learning_rate": 1.4932587307359165e-05,
+      "loss": 0.9704,
+      "step": 2324
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.767933016701002,
+      "learning_rate": 1.4928274223967986e-05,
+      "loss": 0.8961,
+      "step": 2325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7663837703593295,
+      "learning_rate": 1.4923959929345225e-05,
+      "loss": 0.9543,
+      "step": 2326
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8819865289259616,
+      "learning_rate": 1.4919644424551205e-05,
+      "loss": 0.92,
+      "step": 2327
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.935705325811252,
+      "learning_rate": 1.4915327710646568e-05,
+      "loss": 0.8172,
+      "step": 2328
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8807778391035526,
+      "learning_rate": 1.4911009788692235e-05,
+      "loss": 0.8079,
+      "step": 2329
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7340974992817775,
+      "learning_rate": 1.4906690659749426e-05,
+      "loss": 0.9884,
+      "step": 2330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8010092611333235,
+      "learning_rate": 1.4902370324879668e-05,
+      "loss": 0.8415,
+      "step": 2331
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.805620053742415,
+      "learning_rate": 1.4898048785144775e-05,
+      "loss": 0.9433,
+      "step": 2332
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.806019089524397,
+      "learning_rate": 1.4893726041606864e-05,
+      "loss": 0.9695,
+      "step": 2333
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8069677773537647,
+      "learning_rate": 1.4889402095328343e-05,
+      "loss": 0.8722,
+      "step": 2334
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.906550234653562,
+      "learning_rate": 1.488507694737191e-05,
+      "loss": 0.9039,
+      "step": 2335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1084072393256705,
+      "learning_rate": 1.4880750598800574e-05,
+      "loss": 0.9259,
+      "step": 2336
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.749123472026514,
+      "learning_rate": 1.4876423050677627e-05,
+      "loss": 0.9346,
+      "step": 2337
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6330518662858595,
+      "learning_rate": 1.4872094304066656e-05,
+      "loss": 0.987,
+      "step": 2338
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.788749903404174,
+      "learning_rate": 1.486776436003155e-05,
+      "loss": 0.9606,
+      "step": 2339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7976341491384056,
+      "learning_rate": 1.4863433219636488e-05,
+      "loss": 0.9753,
+      "step": 2340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.665942700415535,
+      "learning_rate": 1.4859100883945936e-05,
+      "loss": 0.9599,
+      "step": 2341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.930986893091736,
+      "learning_rate": 1.4854767354024668e-05,
+      "loss": 0.8953,
+      "step": 2342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0580837922604696,
+      "learning_rate": 1.4850432630937741e-05,
+      "loss": 1.0139,
+      "step": 2343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8370525920262972,
+      "learning_rate": 1.4846096715750509e-05,
+      "loss": 0.898,
+      "step": 2344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0053150475311012,
+      "learning_rate": 1.4841759609528619e-05,
+      "loss": 0.9479,
+      "step": 2345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7670959422581003,
+      "learning_rate": 1.4837421313338008e-05,
+      "loss": 0.915,
+      "step": 2346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9217310890737367,
+      "learning_rate": 1.4833081828244908e-05,
+      "loss": 0.8599,
+      "step": 2347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.490086661578173,
+      "learning_rate": 1.4828741155315844e-05,
+      "loss": 0.9828,
+      "step": 2348
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.849816628823393,
+      "learning_rate": 1.4824399295617631e-05,
+      "loss": 0.8926,
+      "step": 2349
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.999178133361301,
+      "learning_rate": 1.4820056250217377e-05,
+      "loss": 0.9283,
+      "step": 2350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2252969975357586,
+      "learning_rate": 1.4815712020182482e-05,
+      "loss": 0.8567,
+      "step": 2351
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6421881777254463,
+      "learning_rate": 1.4811366606580633e-05,
+      "loss": 0.9335,
+      "step": 2352
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.859144819762544,
+      "learning_rate": 1.480702001047981e-05,
+      "loss": 0.9672,
+      "step": 2353
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.149151970612346,
+      "learning_rate": 1.4802672232948287e-05,
+      "loss": 0.9672,
+      "step": 2354
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.961148024285571,
+      "learning_rate": 1.4798323275054627e-05,
+      "loss": 0.8123,
+      "step": 2355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0213153897706237,
+      "learning_rate": 1.4793973137867679e-05,
+      "loss": 0.9199,
+      "step": 2356
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6555091707488243,
+      "learning_rate": 1.4789621822456585e-05,
+      "loss": 0.8879,
+      "step": 2357
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1922995389939093,
+      "learning_rate": 1.4785269329890779e-05,
+      "loss": 0.8563,
+      "step": 2358
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.877845027283451,
+      "learning_rate": 1.4780915661239979e-05,
+      "loss": 0.9435,
+      "step": 2359
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9264636830785347,
+      "learning_rate": 1.4776560817574192e-05,
+      "loss": 0.9215,
+      "step": 2360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8938324963130237,
+      "learning_rate": 1.4772204799963723e-05,
+      "loss": 0.9374,
+      "step": 2361
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.039456953253997,
+      "learning_rate": 1.4767847609479155e-05,
+      "loss": 0.9589,
+      "step": 2362
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.699834713722586,
+      "learning_rate": 1.4763489247191367e-05,
+      "loss": 0.9505,
+      "step": 2363
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.894943910673886,
+      "learning_rate": 1.4759129714171515e-05,
+      "loss": 1.0437,
+      "step": 2364
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7951532789142117,
+      "learning_rate": 1.4754769011491052e-05,
+      "loss": 0.8761,
+      "step": 2365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8583742591953425,
+      "learning_rate": 1.4750407140221723e-05,
+      "loss": 0.8651,
+      "step": 2366
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8018366439100553,
+      "learning_rate": 1.4746044101435546e-05,
+      "loss": 0.9273,
+      "step": 2367
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4891138367062005,
+      "learning_rate": 1.4741679896204842e-05,
+      "loss": 0.935,
+      "step": 2368
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.806867292338262,
+      "learning_rate": 1.47373145256022e-05,
+      "loss": 0.851,
+      "step": 2369
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.9532302608146273,
+      "learning_rate": 1.4732947990700512e-05,
+      "loss": 0.9913,
+      "step": 2370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.777258410564699,
+      "learning_rate": 1.4728580292572947e-05,
+      "loss": 0.9188,
+      "step": 2371
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8586548046452687,
+      "learning_rate": 1.4724211432292965e-05,
+      "loss": 1.0016,
+      "step": 2372
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7528671137789376,
+      "learning_rate": 1.4719841410934307e-05,
+      "loss": 0.9907,
+      "step": 2373
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.711606229508736,
+      "learning_rate": 1.4715470229571007e-05,
+      "loss": 0.938,
+      "step": 2374
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6358126001386037,
+      "learning_rate": 1.4711097889277373e-05,
+      "loss": 0.9436,
+      "step": 2375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3690839599101254,
+      "learning_rate": 1.4706724391128004e-05,
+      "loss": 0.8859,
+      "step": 2376
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.857511482655727,
+      "learning_rate": 1.4702349736197787e-05,
+      "loss": 0.995,
+      "step": 2377
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.657782775593164,
+      "learning_rate": 1.4697973925561885e-05,
+      "loss": 0.9377,
+      "step": 2378
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.0674842520445136,
+      "learning_rate": 1.4693596960295754e-05,
+      "loss": 0.8398,
+      "step": 2379
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6013761676243523,
+      "learning_rate": 1.4689218841475126e-05,
+      "loss": 0.8126,
+      "step": 2380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.864805727482045,
+      "learning_rate": 1.468483957017602e-05,
+      "loss": 0.8925,
+      "step": 2381
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.800285854028283,
+      "learning_rate": 1.4680459147474739e-05,
+      "loss": 0.8958,
+      "step": 2382
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2943055652064226,
+      "learning_rate": 1.4676077574447867e-05,
+      "loss": 1.0121,
+      "step": 2383
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8279058966790025,
+      "learning_rate": 1.4671694852172276e-05,
+      "loss": 0.8836,
+      "step": 2384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5618607083364355,
+      "learning_rate": 1.4667310981725113e-05,
+      "loss": 0.8828,
+      "step": 2385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.919796341251823,
+      "learning_rate": 1.4662925964183807e-05,
+      "loss": 0.8746,
+      "step": 2386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5276980612176696,
+      "learning_rate": 1.4658539800626078e-05,
+      "loss": 0.8823,
+      "step": 2387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6098839300692678,
+      "learning_rate": 1.4654152492129918e-05,
+      "loss": 0.948,
+      "step": 2388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5115034465656607,
+      "learning_rate": 1.4649764039773606e-05,
+      "loss": 0.9205,
+      "step": 2389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.731287358860139,
+      "learning_rate": 1.4645374444635703e-05,
+      "loss": 0.9242,
+      "step": 2390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1344035395663186,
+      "learning_rate": 1.4640983707795042e-05,
+      "loss": 0.8341,
+      "step": 2391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8115383914664194,
+      "learning_rate": 1.463659183033075e-05,
+      "loss": 0.8833,
+      "step": 2392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7697574743430575,
+      "learning_rate": 1.4632198813322223e-05,
+      "loss": 0.9018,
+      "step": 2393
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.83330791104321,
+      "learning_rate": 1.4627804657849143e-05,
+      "loss": 0.9454,
+      "step": 2394
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.775273132429049,
+      "learning_rate": 1.462340936499147e-05,
+      "loss": 0.8233,
+      "step": 2395
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.5958681049130883,
+      "learning_rate": 1.4619012935829444e-05,
+      "loss": 0.8597,
+      "step": 2396
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8992675561202574,
+      "learning_rate": 1.4614615371443583e-05,
+      "loss": 0.9323,
+      "step": 2397
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.2927009271512806,
+      "learning_rate": 1.4610216672914683e-05,
+      "loss": 0.8577,
+      "step": 2398
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9704436448904388,
+      "learning_rate": 1.4605816841323827e-05,
+      "loss": 0.997,
+      "step": 2399
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.353962376662028,
+      "learning_rate": 1.4601415877752362e-05,
+      "loss": 1.1038,
+      "step": 2400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7215183764031874,
+      "learning_rate": 1.459701378328193e-05,
+      "loss": 0.9097,
+      "step": 2401
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.933424502380309,
+      "learning_rate": 1.4592610558994436e-05,
+      "loss": 0.7999,
+      "step": 2402
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.926601447869054,
+      "learning_rate": 1.4588206205972074e-05,
+      "loss": 0.9358,
+      "step": 2403
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9496117509119233,
+      "learning_rate": 1.4583800725297303e-05,
+      "loss": 0.9057,
+      "step": 2404
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.254611314774993,
+      "learning_rate": 1.4579394118052874e-05,
+      "loss": 0.7935,
+      "step": 2405
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.851822266069201,
+      "learning_rate": 1.4574986385321803e-05,
+      "loss": 1.0653,
+      "step": 2406
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1348344649384035,
+      "learning_rate": 1.457057752818739e-05,
+      "loss": 0.8486,
+      "step": 2407
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.579387784667669,
+      "learning_rate": 1.456616754773321e-05,
+      "loss": 0.8526,
+      "step": 2408
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0468680359639984,
+      "learning_rate": 1.4561756445043104e-05,
+      "loss": 0.9041,
+      "step": 2409
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0982245467165415,
+      "learning_rate": 1.4557344221201206e-05,
+      "loss": 0.9566,
+      "step": 2410
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9565652107469536,
+      "learning_rate": 1.4552930877291915e-05,
+      "loss": 1.0297,
+      "step": 2411
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8739467747522607,
+      "learning_rate": 1.4548516414399904e-05,
+      "loss": 0.8787,
+      "step": 2412
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6437720748691778,
+      "learning_rate": 1.4544100833610132e-05,
+      "loss": 0.9208,
+      "step": 2413
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9485877441613635,
+      "learning_rate": 1.4539684136007815e-05,
+      "loss": 0.987,
+      "step": 2414
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7206452418675786,
+      "learning_rate": 1.4535266322678455e-05,
+      "loss": 0.9639,
+      "step": 2415
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0018829599804486,
+      "learning_rate": 1.453084739470783e-05,
+      "loss": 0.8887,
+      "step": 2416
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9505090712062456,
+      "learning_rate": 1.452642735318199e-05,
+      "loss": 0.9423,
+      "step": 2417
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9778830127369367,
+      "learning_rate": 1.4522006199187254e-05,
+      "loss": 0.9154,
+      "step": 2418
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.657643161124756,
+      "learning_rate": 1.4517583933810219e-05,
+      "loss": 0.9314,
+      "step": 2419
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8830635226181176,
+      "learning_rate": 1.4513160558137753e-05,
+      "loss": 0.9039,
+      "step": 2420
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.050561549706014,
+      "learning_rate": 1.4508736073256997e-05,
+      "loss": 0.873,
+      "step": 2421
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3887870469835066,
+      "learning_rate": 1.450431048025537e-05,
+      "loss": 0.8954,
+      "step": 2422
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9279258595634317,
+      "learning_rate": 1.4499883780220552e-05,
+      "loss": 0.9408,
+      "step": 2423
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0884637941438986,
+      "learning_rate": 1.4495455974240507e-05,
+      "loss": 0.8688,
+      "step": 2424
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.855867004287512,
+      "learning_rate": 1.4491027063403462e-05,
+      "loss": 0.9584,
+      "step": 2425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8528079362086025,
+      "learning_rate": 1.4486597048797922e-05,
+      "loss": 0.9052,
+      "step": 2426
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8168958102750477,
+      "learning_rate": 1.4482165931512655e-05,
+      "loss": 0.9273,
+      "step": 2427
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.879624150643087,
+      "learning_rate": 1.447773371263671e-05,
+      "loss": 0.9153,
+      "step": 2428
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6327227672284925,
+      "learning_rate": 1.4473300393259404e-05,
+      "loss": 0.9121,
+      "step": 2429
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7857463283637314,
+      "learning_rate": 1.4468865974470319e-05,
+      "loss": 0.9462,
+      "step": 2430
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8478943189794337,
+      "learning_rate": 1.4464430457359308e-05,
+      "loss": 0.9284,
+      "step": 2431
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0260805043516834,
+      "learning_rate": 1.4459993843016501e-05,
+      "loss": 0.9206,
+      "step": 2432
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.77957086923648,
+      "learning_rate": 1.4455556132532298e-05,
+      "loss": 0.8637,
+      "step": 2433
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.844634762005502,
+      "learning_rate": 1.4451117326997355e-05,
+      "loss": 0.9007,
+      "step": 2434
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7510660636079756,
+      "learning_rate": 1.444667742750261e-05,
+      "loss": 0.9752,
+      "step": 2435
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.729279351837006,
+      "learning_rate": 1.4442236435139265e-05,
+      "loss": 0.9481,
+      "step": 2436
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.791939618896989,
+      "learning_rate": 1.4437794350998791e-05,
+      "loss": 0.9578,
+      "step": 2437
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8060592460027998,
+      "learning_rate": 1.4433351176172925e-05,
+      "loss": 0.8578,
+      "step": 2438
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.666012032688352,
+      "learning_rate": 1.4428906911753679e-05,
+      "loss": 1.0896,
+      "step": 2439
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8237416644883946,
+      "learning_rate": 1.4424461558833332e-05,
+      "loss": 0.8734,
+      "step": 2440
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6397829663448653,
+      "learning_rate": 1.442001511850442e-05,
+      "loss": 0.8617,
+      "step": 2441
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6334892262776126,
+      "learning_rate": 1.4415567591859753e-05,
+      "loss": 0.7519,
+      "step": 2442
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1957681156517204,
+      "learning_rate": 1.4411118979992416e-05,
+      "loss": 1.0172,
+      "step": 2443
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.69562700633565,
+      "learning_rate": 1.4406669283995747e-05,
+      "loss": 1.2042,
+      "step": 2444
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.060077822738249,
+      "learning_rate": 1.4402218504963355e-05,
+      "loss": 0.9895,
+      "step": 2445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7709821109960866,
+      "learning_rate": 1.4397766643989123e-05,
+      "loss": 0.8656,
+      "step": 2446
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6858104251237664,
+      "learning_rate": 1.4393313702167185e-05,
+      "loss": 0.9585,
+      "step": 2447
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9475283693916645,
+      "learning_rate": 1.4388859680591957e-05,
+      "loss": 0.7974,
+      "step": 2448
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.7918155152676922,
+      "learning_rate": 1.4384404580358112e-05,
+      "loss": 0.9301,
+      "step": 2449
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6422292443267072,
+      "learning_rate": 1.4379948402560581e-05,
+      "loss": 0.8223,
+      "step": 2450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.790804524918857,
+      "learning_rate": 1.4375491148294578e-05,
+      "loss": 0.9744,
+      "step": 2451
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.862406171144873,
+      "learning_rate": 1.4371032818655564e-05,
+      "loss": 0.837,
+      "step": 2452
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.772748271528916,
+      "learning_rate": 1.4366573414739273e-05,
+      "loss": 0.9354,
+      "step": 2453
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5730897546706624,
+      "learning_rate": 1.4362112937641702e-05,
+      "loss": 0.9264,
+      "step": 2454
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9648680762416584,
+      "learning_rate": 1.435765138845911e-05,
+      "loss": 0.9132,
+      "step": 2455
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.96891879680553,
+      "learning_rate": 1.4353188768288022e-05,
+      "loss": 0.771,
+      "step": 2456
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.027731011106476,
+      "learning_rate": 1.4348725078225228e-05,
+      "loss": 0.9532,
+      "step": 2457
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9217523567988866,
+      "learning_rate": 1.434426031936777e-05,
+      "loss": 0.9065,
+      "step": 2458
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.902108117677878,
+      "learning_rate": 1.4339794492812966e-05,
+      "loss": 1.0494,
+      "step": 2459
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0654388452077606,
+      "learning_rate": 1.433532759965839e-05,
+      "loss": 0.8442,
+      "step": 2460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.934914525516883,
+      "learning_rate": 1.433085964100188e-05,
+      "loss": 0.8166,
+      "step": 2461
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7968856258452566,
+      "learning_rate": 1.4326390617941533e-05,
+      "loss": 0.9282,
+      "step": 2462
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7773284905132862,
+      "learning_rate": 1.4321920531575708e-05,
+      "loss": 0.9476,
+      "step": 2463
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8341525979924986,
+      "learning_rate": 1.4317449383003032e-05,
+      "loss": 0.8395,
+      "step": 2464
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7509817488619133,
+      "learning_rate": 1.4312977173322384e-05,
+      "loss": 1.0142,
+      "step": 2465
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0310197179879648,
+      "learning_rate": 1.430850390363291e-05,
+      "loss": 0.8766,
+      "step": 2466
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9366419645028676,
+      "learning_rate": 1.430402957503401e-05,
+      "loss": 1.0466,
+      "step": 2467
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6504441981995255,
+      "learning_rate": 1.4299554188625352e-05,
+      "loss": 0.8638,
+      "step": 2468
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9048530088304023,
+      "learning_rate": 1.4295077745506865e-05,
+      "loss": 0.9458,
+      "step": 2469
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8117932431738164,
+      "learning_rate": 1.4290600246778726e-05,
+      "loss": 0.9039,
+      "step": 2470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6908191575161906,
+      "learning_rate": 1.4286121693541378e-05,
+      "loss": 0.9676,
+      "step": 2471
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0444535921462332,
+      "learning_rate": 1.4281642086895527e-05,
+      "loss": 0.9215,
+      "step": 2472
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0096304418790654,
+      "learning_rate": 1.4277161427942137e-05,
+      "loss": 0.9357,
+      "step": 2473
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9660334572945746,
+      "learning_rate": 1.427267971778242e-05,
+      "loss": 0.9491,
+      "step": 2474
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6068280607700136,
+      "learning_rate": 1.4268196957517866e-05,
+      "loss": 0.973,
+      "step": 2475
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.076600308765351,
+      "learning_rate": 1.4263713148250203e-05,
+      "loss": 0.85,
+      "step": 2476
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.127527029548337,
+      "learning_rate": 1.4259228291081431e-05,
+      "loss": 1.0009,
+      "step": 2477
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.796536067560663,
+      "learning_rate": 1.4254742387113795e-05,
+      "loss": 0.8847,
+      "step": 2478
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7476099402360243,
+      "learning_rate": 1.4250255437449812e-05,
+      "loss": 0.9103,
+      "step": 2479
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4654239989935136,
+      "learning_rate": 1.4245767443192246e-05,
+      "loss": 0.986,
+      "step": 2480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0750075287655423,
+      "learning_rate": 1.424127840544412e-05,
+      "loss": 0.924,
+      "step": 2481
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.649710861424403,
+      "learning_rate": 1.423678832530871e-05,
+      "loss": 0.8283,
+      "step": 2482
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.986231643973584,
+      "learning_rate": 1.4232297203889556e-05,
+      "loss": 1.1026,
+      "step": 2483
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6902403214536084,
+      "learning_rate": 1.4227805042290447e-05,
+      "loss": 0.858,
+      "step": 2484
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.86598463124649,
+      "learning_rate": 1.4223311841615435e-05,
+      "loss": 0.9599,
+      "step": 2485
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6725105892389482,
+      "learning_rate": 1.421881760296882e-05,
+      "loss": 0.9663,
+      "step": 2486
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5386181291260144,
+      "learning_rate": 1.4214322327455157e-05,
+      "loss": 0.9455,
+      "step": 2487
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9625059832666265,
+      "learning_rate": 1.4209826016179263e-05,
+      "loss": 1.0126,
+      "step": 2488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.761055038983844,
+      "learning_rate": 1.42053286702462e-05,
+      "loss": 0.8517,
+      "step": 2489
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8698649570185073,
+      "learning_rate": 1.4200830290761295e-05,
+      "loss": 0.9186,
+      "step": 2490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.176874951282763,
+      "learning_rate": 1.419633087883012e-05,
+      "loss": 0.9243,
+      "step": 2491
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7541805305928913,
+      "learning_rate": 1.419183043555851e-05,
+      "loss": 0.8945,
+      "step": 2492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7154264957294645,
+      "learning_rate": 1.4187328962052536e-05,
+      "loss": 0.9352,
+      "step": 2493
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.769275581735851,
+      "learning_rate": 1.4182826459418543e-05,
+      "loss": 0.9409,
+      "step": 2494
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.54796386871129,
+      "learning_rate": 1.417832292876312e-05,
+      "loss": 0.9307,
+      "step": 2495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.869832085250058,
+      "learning_rate": 1.4173818371193106e-05,
+      "loss": 0.9986,
+      "step": 2496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.737659375631931,
+      "learning_rate": 1.41693127878156e-05,
+      "loss": 0.8104,
+      "step": 2497
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0286718411841465,
+      "learning_rate": 1.4164806179737935e-05,
+      "loss": 0.8746,
+      "step": 2498
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.940764782727735,
+      "learning_rate": 1.4160298548067725e-05,
+      "loss": 0.8917,
+      "step": 2499
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.896939827544822,
+      "learning_rate": 1.4155789893912808e-05,
+      "loss": 0.8441,
+      "step": 2500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5839952314690016,
+      "learning_rate": 1.4151280218381287e-05,
+      "loss": 0.9311,
+      "step": 2501
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.845449847456862,
+      "learning_rate": 1.4146769522581519e-05,
+      "loss": 0.897,
+      "step": 2502
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6564675143300276,
+      "learning_rate": 1.4142257807622103e-05,
+      "loss": 0.8739,
+      "step": 2503
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.80321608838917,
+      "learning_rate": 1.4137745074611888e-05,
+      "loss": 0.9843,
+      "step": 2504
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.6055525554314296,
+      "learning_rate": 1.4133231324659984e-05,
+      "loss": 0.8857,
+      "step": 2505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7794862939293057,
+      "learning_rate": 1.412871655887574e-05,
+      "loss": 0.8227,
+      "step": 2506
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.860975874901178,
+      "learning_rate": 1.412420077836876e-05,
+      "loss": 0.9598,
+      "step": 2507
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.95525509948893,
+      "learning_rate": 1.4119683984248898e-05,
+      "loss": 0.9288,
+      "step": 2508
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2260253091053848,
+      "learning_rate": 1.4115166177626252e-05,
+      "loss": 1.0039,
+      "step": 2509
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.8290548924846135,
+      "learning_rate": 1.4110647359611175e-05,
+      "loss": 0.8938,
+      "step": 2510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9476608734997245,
+      "learning_rate": 1.4106127531314261e-05,
+      "loss": 0.8894,
+      "step": 2511
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.652241950773867,
+      "learning_rate": 1.4101606693846363e-05,
+      "loss": 0.9125,
+      "step": 2512
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7796707100209352,
+      "learning_rate": 1.4097084848318574e-05,
+      "loss": 0.9231,
+      "step": 2513
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.863374745844007,
+      "learning_rate": 1.4092561995842238e-05,
+      "loss": 1.0797,
+      "step": 2514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7617959711169444,
+      "learning_rate": 1.4088038137528938e-05,
+      "loss": 0.9663,
+      "step": 2515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.163921586737289,
+      "learning_rate": 1.408351327449052e-05,
+      "loss": 0.9333,
+      "step": 2516
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7792699930149816,
+      "learning_rate": 1.4078987407839066e-05,
+      "loss": 0.8374,
+      "step": 2517
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7188267540892985,
+      "learning_rate": 1.4074460538686908e-05,
+      "loss": 0.8946,
+      "step": 2518
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6558835918581876,
+      "learning_rate": 1.4069932668146619e-05,
+      "loss": 0.9045,
+      "step": 2519
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.866876242378881,
+      "learning_rate": 1.406540379733103e-05,
+      "loss": 0.8737,
+      "step": 2520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.198276090274672,
+      "learning_rate": 1.4060873927353203e-05,
+      "loss": 0.9967,
+      "step": 2521
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5484619529654458,
+      "learning_rate": 1.4056343059326458e-05,
+      "loss": 0.8782,
+      "step": 2522
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.743265614544366,
+      "learning_rate": 1.405181119436435e-05,
+      "loss": 0.8819,
+      "step": 2523
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.084943695798522,
+      "learning_rate": 1.4047278333580689e-05,
+      "loss": 0.8367,
+      "step": 2524
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6568453931938603,
+      "learning_rate": 1.4042744478089528e-05,
+      "loss": 0.9544,
+      "step": 2525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9522303052046026,
+      "learning_rate": 1.4038209629005156e-05,
+      "loss": 0.8682,
+      "step": 2526
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0082325997961306,
+      "learning_rate": 1.4033673787442108e-05,
+      "loss": 0.8472,
+      "step": 2527
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5897995504801026,
+      "learning_rate": 1.4029136954515175e-05,
+      "loss": 0.9053,
+      "step": 2528
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.056262723244552,
+      "learning_rate": 1.402459913133938e-05,
+      "loss": 0.9708,
+      "step": 2529
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.830129743021095,
+      "learning_rate": 1.4020060319029991e-05,
+      "loss": 1.0032,
+      "step": 2530
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6939067046894354,
+      "learning_rate": 1.4015520518702526e-05,
+      "loss": 0.8888,
+      "step": 2531
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.655218785250219,
+      "learning_rate": 1.4010979731472731e-05,
+      "loss": 0.8829,
+      "step": 2532
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.67010313060736,
+      "learning_rate": 1.4006437958456616e-05,
+      "loss": 0.8782,
+      "step": 2533
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.861776718201834,
+      "learning_rate": 1.4001895200770412e-05,
+      "loss": 0.9893,
+      "step": 2534
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.64690624389765,
+      "learning_rate": 1.3997351459530605e-05,
+      "loss": 0.8615,
+      "step": 2535
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.82273974702354,
+      "learning_rate": 1.3992806735853924e-05,
+      "loss": 1.0044,
+      "step": 2536
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7952746643921755,
+      "learning_rate": 1.3988261030857327e-05,
+      "loss": 0.853,
+      "step": 2537
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.0034032826719534,
+      "learning_rate": 1.3983714345658021e-05,
+      "loss": 1.0211,
+      "step": 2538
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.683248570522212,
+      "learning_rate": 1.3979166681373459e-05,
+      "loss": 0.8442,
+      "step": 2539
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8977292429823303,
+      "learning_rate": 1.3974618039121326e-05,
+      "loss": 0.8503,
+      "step": 2540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.838726270766741,
+      "learning_rate": 1.3970068420019552e-05,
+      "loss": 0.9449,
+      "step": 2541
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6119929718476276,
+      "learning_rate": 1.3965517825186306e-05,
+      "loss": 0.9552,
+      "step": 2542
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.691711091798615,
+      "learning_rate": 1.3960966255739992e-05,
+      "loss": 0.9775,
+      "step": 2543
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.886609800357804,
+      "learning_rate": 1.3956413712799263e-05,
+      "loss": 0.9099,
+      "step": 2544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.5510696879764474,
+      "learning_rate": 1.3951860197483008e-05,
+      "loss": 0.8674,
+      "step": 2545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.2071989200026714,
+      "learning_rate": 1.3947305710910346e-05,
+      "loss": 0.8818,
+      "step": 2546
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7232029556025936,
+      "learning_rate": 1.394275025420065e-05,
+      "loss": 0.8767,
+      "step": 2547
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.822174025808589,
+      "learning_rate": 1.3938193828473521e-05,
+      "loss": 1.0398,
+      "step": 2548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.91246822858765,
+      "learning_rate": 1.3933636434848797e-05,
+      "loss": 0.8926,
+      "step": 2549
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7026549086967,
+      "learning_rate": 1.3929078074446561e-05,
+      "loss": 1.0069,
+      "step": 2550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3907508424871993,
+      "learning_rate": 1.392451874838713e-05,
+      "loss": 0.8861,
+      "step": 2551
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.365881535947219,
+      "learning_rate": 1.3919958457791056e-05,
+      "loss": 0.9715,
+      "step": 2552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9741631822721617,
+      "learning_rate": 1.3915397203779138e-05,
+      "loss": 0.8546,
+      "step": 2553
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4551308205268096,
+      "learning_rate": 1.3910834987472393e-05,
+      "loss": 0.8067,
+      "step": 2554
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7640869588849273,
+      "learning_rate": 1.3906271809992093e-05,
+      "loss": 0.8744,
+      "step": 2555
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7444788971223937,
+      "learning_rate": 1.3901707672459738e-05,
+      "loss": 0.9381,
+      "step": 2556
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8966581939330647,
+      "learning_rate": 1.3897142575997062e-05,
+      "loss": 0.8446,
+      "step": 2557
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7516082432017703,
+      "learning_rate": 1.3892576521726045e-05,
+      "loss": 0.8426,
+      "step": 2558
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7161833682050105,
+      "learning_rate": 1.388800951076889e-05,
+      "loss": 0.927,
+      "step": 2559
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8069829382886025,
+      "learning_rate": 1.3883441544248037e-05,
+      "loss": 0.903,
+      "step": 2560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8060823802410213,
+      "learning_rate": 1.3878872623286169e-05,
+      "loss": 0.9407,
+      "step": 2561
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.1672323489059027,
+      "learning_rate": 1.38743027490062e-05,
+      "loss": 0.8627,
+      "step": 2562
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9899067313211556,
+      "learning_rate": 1.386973192253127e-05,
+      "loss": 0.9436,
+      "step": 2563
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.742952405237023,
+      "learning_rate": 1.3865160144984766e-05,
+      "loss": 0.7929,
+      "step": 2564
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.709277125727227,
+      "learning_rate": 1.38605874174903e-05,
+      "loss": 0.8206,
+      "step": 2565
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8380760313223443,
+      "learning_rate": 1.3856013741171723e-05,
+      "loss": 0.9272,
+      "step": 2566
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.557049466152109,
+      "learning_rate": 1.3851439117153114e-05,
+      "loss": 0.8834,
+      "step": 2567
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.674380736873026,
+      "learning_rate": 1.3846863546558783e-05,
+      "loss": 0.8271,
+      "step": 2568
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.729408030623172,
+      "learning_rate": 1.3842287030513287e-05,
+      "loss": 0.917,
+      "step": 2569
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4925015801647987,
+      "learning_rate": 1.3837709570141401e-05,
+      "loss": 0.8981,
+      "step": 2570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.881328573934881,
+      "learning_rate": 1.3833131166568132e-05,
+      "loss": 0.9229,
+      "step": 2571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6522807333490324,
+      "learning_rate": 1.3828551820918726e-05,
+      "loss": 1.0144,
+      "step": 2572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.631217876018588,
+      "learning_rate": 1.382397153431866e-05,
+      "loss": 0.8376,
+      "step": 2573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.752984398348073,
+      "learning_rate": 1.3819390307893637e-05,
+      "loss": 0.8932,
+      "step": 2574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6337997335427636,
+      "learning_rate": 1.3814808142769596e-05,
+      "loss": 0.8712,
+      "step": 2575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.767756879305521,
+      "learning_rate": 1.3810225040072702e-05,
+      "loss": 0.8711,
+      "step": 2576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.833671564740066,
+      "learning_rate": 1.380564100092936e-05,
+      "loss": 0.9122,
+      "step": 2577
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6823193648331993,
+      "learning_rate": 1.3801056026466187e-05,
+      "loss": 0.9515,
+      "step": 2578
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7052167356636985,
+      "learning_rate": 1.3796470117810047e-05,
+      "loss": 0.9358,
+      "step": 2579
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7684445261992323,
+      "learning_rate": 1.3791883276088032e-05,
+      "loss": 0.9435,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7770127368337647,
+      "learning_rate": 1.3787295502427456e-05,
+      "loss": 1.0437,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7540315264272577,
+      "learning_rate": 1.3782706797955862e-05,
+      "loss": 0.8721,
+      "step": 2582
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6708408424496337,
+      "learning_rate": 1.3778117163801026e-05,
+      "loss": 0.8377,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4750890755484996,
+      "learning_rate": 1.3773526601090953e-05,
+      "loss": 0.9632,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8539918952627445,
+      "learning_rate": 1.3768935110953876e-05,
+      "loss": 0.8852,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7756479228382283,
+      "learning_rate": 1.376434269451825e-05,
+      "loss": 0.9469,
+      "step": 2586
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8092528873183813,
+      "learning_rate": 1.3759749352912766e-05,
+      "loss": 0.8656,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7460581990366535,
+      "learning_rate": 1.3755155087266339e-05,
+      "loss": 0.9428,
+      "step": 2588
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6634448363350254,
+      "learning_rate": 1.3750559898708104e-05,
+      "loss": 0.7914,
+      "step": 2589
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3016500768582326,
+      "learning_rate": 1.3745963788367438e-05,
+      "loss": 0.9247,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8604673823201012,
+      "learning_rate": 1.3741366757373928e-05,
+      "loss": 0.7247,
+      "step": 2591
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.634064735941571,
+      "learning_rate": 1.3736768806857405e-05,
+      "loss": 0.7969,
+      "step": 2592
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6425265646373126,
+      "learning_rate": 1.373216993794791e-05,
+      "loss": 0.7945,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8211610865763577,
+      "learning_rate": 1.3727570151775716e-05,
+      "loss": 0.9039,
+      "step": 2594
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.569205222997264,
+      "learning_rate": 1.3722969449471319e-05,
+      "loss": 0.9151,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.716633530129442,
+      "learning_rate": 1.3718367832165451e-05,
+      "loss": 0.8682,
+      "step": 2596
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9394906838196153,
+      "learning_rate": 1.3713765300989053e-05,
+      "loss": 0.8901,
+      "step": 2597
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8753101369751275,
+      "learning_rate": 1.37091618570733e-05,
+      "loss": 1.0159,
+      "step": 2598
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6976133669347346,
+      "learning_rate": 1.3704557501549594e-05,
+      "loss": 0.8259,
+      "step": 2599
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9676339712488287,
+      "learning_rate": 1.3699952235549547e-05,
+      "loss": 0.8872,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.015750805388782,
+      "learning_rate": 1.369534606020501e-05,
+      "loss": 0.9446,
+      "step": 2601
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0731985649847657,
+      "learning_rate": 1.3690738976648053e-05,
+      "loss": 1.0013,
+      "step": 2602
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.163854041137835,
+      "learning_rate": 1.3686130986010965e-05,
+      "loss": 0.9987,
+      "step": 2603
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.706206243594574,
+      "learning_rate": 1.3681522089426265e-05,
+      "loss": 0.8524,
+      "step": 2604
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.575483493975654,
+      "learning_rate": 1.3676912288026685e-05,
+      "loss": 0.7937,
+      "step": 2605
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9879921457113077,
+      "learning_rate": 1.3672301582945187e-05,
+      "loss": 0.8572,
+      "step": 2606
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.853495583921802,
+      "learning_rate": 1.3667689975314955e-05,
+      "loss": 0.98,
+      "step": 2607
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.90434906178557,
+      "learning_rate": 1.366307746626939e-05,
+      "loss": 0.8291,
+      "step": 2608
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.837540695181841,
+      "learning_rate": 1.365846405694212e-05,
+      "loss": 0.8312,
+      "step": 2609
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.641247798531113,
+      "learning_rate": 1.3653849748466991e-05,
+      "loss": 1.1737,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.845182795899973,
+      "learning_rate": 1.364923454197807e-05,
+      "loss": 0.9513,
+      "step": 2611
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8796126601879175,
+      "learning_rate": 1.3644618438609643e-05,
+      "loss": 0.8865,
+      "step": 2612
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8526823560302756,
+      "learning_rate": 1.364000143949622e-05,
+      "loss": 0.9331,
+      "step": 2613
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.3528468609485875,
+      "learning_rate": 1.3635383545772534e-05,
+      "loss": 0.8366,
+      "step": 2614
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.41234900636786,
+      "learning_rate": 1.3630764758573529e-05,
+      "loss": 0.834,
+      "step": 2615
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.988943225087445,
+      "learning_rate": 1.3626145079034374e-05,
+      "loss": 0.9219,
+      "step": 2616
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9776376496628636,
+      "learning_rate": 1.3621524508290457e-05,
+      "loss": 0.9286,
+      "step": 2617
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.229730135176999,
+      "learning_rate": 1.361690304747739e-05,
+      "loss": 0.9246,
+      "step": 2618
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9490639624794284,
+      "learning_rate": 1.361228069773099e-05,
+      "loss": 0.9191,
+      "step": 2619
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9437329662922713,
+      "learning_rate": 1.3607657460187307e-05,
+      "loss": 0.9408,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5307046738228607,
+      "learning_rate": 1.36030333359826e-05,
+      "loss": 0.8944,
+      "step": 2621
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6236321071528637,
+      "learning_rate": 1.3598408326253348e-05,
+      "loss": 0.8936,
+      "step": 2622
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0871202845956263,
+      "learning_rate": 1.3593782432136251e-05,
+      "loss": 0.9609,
+      "step": 2623
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6858071831140933,
+      "learning_rate": 1.3589155654768224e-05,
+      "loss": 0.8933,
+      "step": 2624
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.045716570735623,
+      "learning_rate": 1.35845279952864e-05,
+      "loss": 1.1436,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8137361622080816,
+      "learning_rate": 1.3579899454828126e-05,
+      "loss": 0.9405,
+      "step": 2626
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8962261675877774,
+      "learning_rate": 1.3575270034530967e-05,
+      "loss": 0.8779,
+      "step": 2627
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6786738209282155,
+      "learning_rate": 1.3570639735532707e-05,
+      "loss": 0.8522,
+      "step": 2628
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8984792337350838,
+      "learning_rate": 1.3566008558971342e-05,
+      "loss": 0.8925,
+      "step": 2629
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6947101762433547,
+      "learning_rate": 1.3561376505985085e-05,
+      "loss": 0.8026,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7174665627606913,
+      "learning_rate": 1.3556743577712363e-05,
+      "loss": 0.929,
+      "step": 2631
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8030923315646157,
+      "learning_rate": 1.3552109775291828e-05,
+      "loss": 0.9614,
+      "step": 2632
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.067036667463295,
+      "learning_rate": 1.354747509986233e-05,
+      "loss": 0.8985,
+      "step": 2633
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.737467558758066,
+      "learning_rate": 1.3542839552562945e-05,
+      "loss": 0.9098,
+      "step": 2634
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9905535153970537,
+      "learning_rate": 1.353820313453296e-05,
+      "loss": 0.8895,
+      "step": 2635
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8996950341165877,
+      "learning_rate": 1.353356584691188e-05,
+      "loss": 0.9421,
+      "step": 2636
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.1401490422536424,
+      "learning_rate": 1.3528927690839414e-05,
+      "loss": 0.8608,
+      "step": 2637
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7015344706197664,
+      "learning_rate": 1.3524288667455497e-05,
+      "loss": 0.9702,
+      "step": 2638
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.580217257775127,
+      "learning_rate": 1.3519648777900264e-05,
+      "loss": 0.8723,
+      "step": 2639
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0334489963523077,
+      "learning_rate": 1.3515008023314077e-05,
+      "loss": 0.9329,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.94368863839663,
+      "learning_rate": 1.3510366404837499e-05,
+      "loss": 0.8787,
+      "step": 2641
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4523195891235248,
+      "learning_rate": 1.3505723923611309e-05,
+      "loss": 0.85,
+      "step": 2642
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.845812910882723,
+      "learning_rate": 1.3501080580776504e-05,
+      "loss": 0.9852,
+      "step": 2643
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.6528192407446602,
+      "learning_rate": 1.3496436377474282e-05,
+      "loss": 0.8928,
+      "step": 2644
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.815621799890215,
+      "learning_rate": 1.3491791314846059e-05,
+      "loss": 0.857,
+      "step": 2645
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.398570829960979,
+      "learning_rate": 1.348714539403346e-05,
+      "loss": 0.8943,
+      "step": 2646
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7354191271811326,
+      "learning_rate": 1.3482498616178329e-05,
+      "loss": 0.9616,
+      "step": 2647
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0440592594486175,
+      "learning_rate": 1.3477850982422704e-05,
+      "loss": 0.8634,
+      "step": 2648
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6829014105178377,
+      "learning_rate": 1.3473202493908847e-05,
+      "loss": 0.9402,
+      "step": 2649
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6462089340973725,
+      "learning_rate": 1.3468553151779229e-05,
+      "loss": 0.7961,
+      "step": 2650
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9826913776524546,
+      "learning_rate": 1.3463902957176526e-05,
+      "loss": 0.8488,
+      "step": 2651
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8634755106514653,
+      "learning_rate": 1.3459251911243623e-05,
+      "loss": 0.8495,
+      "step": 2652
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6810739867840625,
+      "learning_rate": 1.345460001512362e-05,
+      "loss": 1.0048,
+      "step": 2653
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8634739647852445,
+      "learning_rate": 1.344994726995982e-05,
+      "loss": 0.9138,
+      "step": 2654
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0049448169785546,
+      "learning_rate": 1.3445293676895742e-05,
+      "loss": 0.9287,
+      "step": 2655
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8368502452839635,
+      "learning_rate": 1.34406392370751e-05,
+      "loss": 0.8992,
+      "step": 2656
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7559802643598608,
+      "learning_rate": 1.3435983951641831e-05,
+      "loss": 0.844,
+      "step": 2657
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1639489232846976,
+      "learning_rate": 1.3431327821740074e-05,
+      "loss": 0.9783,
+      "step": 2658
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.5013584115803975,
+      "learning_rate": 1.3426670848514172e-05,
+      "loss": 0.8197,
+      "step": 2659
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6903599600202748,
+      "learning_rate": 1.3422013033108683e-05,
+      "loss": 0.8648,
+      "step": 2660
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.889426640643883,
+      "learning_rate": 1.341735437666836e-05,
+      "loss": 0.9477,
+      "step": 2661
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.561490050920701,
+      "learning_rate": 1.341269488033818e-05,
+      "loss": 0.8577,
+      "step": 2662
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.93139917411127,
+      "learning_rate": 1.3408034545263307e-05,
+      "loss": 0.8558,
+      "step": 2663
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8483228045095426,
+      "learning_rate": 1.3403373372589126e-05,
+      "loss": 0.8472,
+      "step": 2664
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8900801444678326,
+      "learning_rate": 1.339871136346122e-05,
+      "loss": 0.9375,
+      "step": 2665
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.761176789590491,
+      "learning_rate": 1.3394048519025385e-05,
+      "loss": 0.88,
+      "step": 2666
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.995974666022481,
+      "learning_rate": 1.3389384840427609e-05,
+      "loss": 0.9282,
+      "step": 2667
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.795462437751262,
+      "learning_rate": 1.3384720328814101e-05,
+      "loss": 0.8563,
+      "step": 2668
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8992183719809117,
+      "learning_rate": 1.338005498533126e-05,
+      "loss": 0.9361,
+      "step": 2669
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6516545078942926,
+      "learning_rate": 1.3375388811125707e-05,
+      "loss": 0.9369,
+      "step": 2670
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.369168985699152,
+      "learning_rate": 1.337072180734425e-05,
+      "loss": 0.7759,
+      "step": 2671
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.088417388011945,
+      "learning_rate": 1.3366053975133904e-05,
+      "loss": 0.8851,
+      "step": 2672
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9663036317798768,
+      "learning_rate": 1.3361385315641898e-05,
+      "loss": 0.9257,
+      "step": 2673
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1256307601041797,
+      "learning_rate": 1.3356715830015652e-05,
+      "loss": 0.9925,
+      "step": 2674
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.090405001302476,
+      "learning_rate": 1.3352045519402799e-05,
+      "loss": 1.1464,
+      "step": 2675
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.604324707239911,
+      "learning_rate": 1.3347374384951171e-05,
+      "loss": 0.8574,
+      "step": 2676
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7363852452328645,
+      "learning_rate": 1.33427024278088e-05,
+      "loss": 0.8872,
+      "step": 2677
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7479083095110153,
+      "learning_rate": 1.333802964912392e-05,
+      "loss": 0.8885,
+      "step": 2678
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7348959309497,
+      "learning_rate": 1.333335605004497e-05,
+      "loss": 0.8658,
+      "step": 2679
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.015353432771644,
+      "learning_rate": 1.332868163172059e-05,
+      "loss": 0.949,
+      "step": 2680
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.513831609319815,
+      "learning_rate": 1.3324006395299624e-05,
+      "loss": 0.8491,
+      "step": 2681
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.646123238994641,
+      "learning_rate": 1.3319330341931112e-05,
+      "loss": 0.8096,
+      "step": 2682
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.870007829546668,
+      "learning_rate": 1.3314653472764293e-05,
+      "loss": 0.8376,
+      "step": 2683
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.54424602096051,
+      "learning_rate": 1.3309975788948616e-05,
+      "loss": 1.0279,
+      "step": 2684
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.650425831800214,
+      "learning_rate": 1.330529729163372e-05,
+      "loss": 0.8417,
+      "step": 2685
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.793823761613536,
+      "learning_rate": 1.330061798196945e-05,
+      "loss": 0.858,
+      "step": 2686
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0661166656301595,
+      "learning_rate": 1.3295937861105848e-05,
+      "loss": 0.8369,
+      "step": 2687
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8365228650619096,
+      "learning_rate": 1.3291256930193164e-05,
+      "loss": 0.8982,
+      "step": 2688
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.829947055923032,
+      "learning_rate": 1.3286575190381828e-05,
+      "loss": 0.9457,
+      "step": 2689
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8245219608219343,
+      "learning_rate": 1.3281892642822488e-05,
+      "loss": 0.8713,
+      "step": 2690
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7515076639581935,
+      "learning_rate": 1.3277209288665977e-05,
+      "loss": 0.8659,
+      "step": 2691
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.039281991650958,
+      "learning_rate": 1.3272525129063339e-05,
+      "loss": 0.9434,
+      "step": 2692
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.78357479127376,
+      "learning_rate": 1.3267840165165802e-05,
+      "loss": 0.9466,
+      "step": 2693
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.708980316331865,
+      "learning_rate": 1.3263154398124807e-05,
+      "loss": 0.7815,
+      "step": 2694
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.539856934636554,
+      "learning_rate": 1.325846782909198e-05,
+      "loss": 0.8458,
+      "step": 2695
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9120051675086893,
+      "learning_rate": 1.3253780459219143e-05,
+      "loss": 0.9523,
+      "step": 2696
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6221844576287014,
+      "learning_rate": 1.3249092289658327e-05,
+      "loss": 0.8075,
+      "step": 2697
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.1776531212135213,
+      "learning_rate": 1.324440332156175e-05,
+      "loss": 0.8592,
+      "step": 2698
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8181917492879025,
+      "learning_rate": 1.323971355608183e-05,
+      "loss": 0.9806,
+      "step": 2699
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.916697538337927,
+      "learning_rate": 1.323502299437118e-05,
+      "loss": 0.9191,
+      "step": 2700
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.682544239590553,
+      "learning_rate": 1.3230331637582605e-05,
+      "loss": 0.9452,
+      "step": 2701
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.803323300095509,
+      "learning_rate": 1.3225639486869113e-05,
+      "loss": 0.8157,
+      "step": 2702
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7628897421404925,
+      "learning_rate": 1.3220946543383904e-05,
+      "loss": 0.8419,
+      "step": 2703
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9899136183471016,
+      "learning_rate": 1.3216252808280366e-05,
+      "loss": 0.9206,
+      "step": 2704
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9304013279937533,
+      "learning_rate": 1.3211558282712092e-05,
+      "loss": 0.8583,
+      "step": 2705
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.0558376279085238,
+      "learning_rate": 1.3206862967832863e-05,
+      "loss": 0.9753,
+      "step": 2706
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.049415755795687,
+      "learning_rate": 1.3202166864796659e-05,
+      "loss": 1.0676,
+      "step": 2707
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.903397325916613,
+      "learning_rate": 1.3197469974757644e-05,
+      "loss": 0.8897,
+      "step": 2708
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.927310383815381,
+      "learning_rate": 1.3192772298870187e-05,
+      "loss": 0.8699,
+      "step": 2709
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.7231277330666335,
+      "learning_rate": 1.3188073838288844e-05,
+      "loss": 0.7612,
+      "step": 2710
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.9162454164657357,
+      "learning_rate": 1.3183374594168368e-05,
+      "loss": 1.0089,
+      "step": 2711
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.759293862016098,
+      "learning_rate": 1.3178674567663692e-05,
+      "loss": 0.9146,
+      "step": 2712
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.873108983408784,
+      "learning_rate": 1.3173973759929956e-05,
+      "loss": 0.8961,
+      "step": 2713
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.900800727457516,
+      "learning_rate": 1.3169272172122493e-05,
+      "loss": 0.8252,
+      "step": 2714
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.671252646096681,
+      "learning_rate": 1.3164569805396812e-05,
+      "loss": 0.9027,
+      "step": 2715
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.4209487105746272,
+      "learning_rate": 1.3159866660908629e-05,
+      "loss": 0.818,
+      "step": 2716
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.596162460313096,
+      "learning_rate": 1.315516273981384e-05,
+      "loss": 0.8533,
+      "step": 2717
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8102198032524477,
+      "learning_rate": 1.3150458043268541e-05,
+      "loss": 0.9579,
+      "step": 2718
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7144446997601652,
+      "learning_rate": 1.3145752572429012e-05,
+      "loss": 0.9368,
+      "step": 2719
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9364685528235817,
+      "learning_rate": 1.3141046328451724e-05,
+      "loss": 0.8868,
+      "step": 2720
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.064401225713515,
+      "learning_rate": 1.3136339312493346e-05,
+      "loss": 0.9467,
+      "step": 2721
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5323842705217143,
+      "learning_rate": 1.3131631525710729e-05,
+      "loss": 0.822,
+      "step": 2722
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.936538224523595,
+      "learning_rate": 1.3126922969260908e-05,
+      "loss": 0.8762,
+      "step": 2723
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.640141372170476,
+      "learning_rate": 1.312221364430112e-05,
+      "loss": 0.9057,
+      "step": 2724
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9851422679608923,
+      "learning_rate": 1.3117503551988786e-05,
+      "loss": 0.9964,
+      "step": 2725
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.729257381358218,
+      "learning_rate": 1.311279269348151e-05,
+      "loss": 0.905,
+      "step": 2726
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8125014739544567,
+      "learning_rate": 1.3108081069937095e-05,
+      "loss": 0.8715,
+      "step": 2727
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0387568343577667,
+      "learning_rate": 1.3103368682513519e-05,
+      "loss": 0.8337,
+      "step": 2728
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0422667008518354,
+      "learning_rate": 1.309865553236896e-05,
+      "loss": 0.8908,
+      "step": 2729
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.685792667296222,
+      "learning_rate": 1.3093941620661777e-05,
+      "loss": 0.8665,
+      "step": 2730
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.864671960381527,
+      "learning_rate": 1.3089226948550513e-05,
+      "loss": 0.8474,
+      "step": 2731
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6702622460485546,
+      "learning_rate": 1.3084511517193908e-05,
+      "loss": 0.8985,
+      "step": 2732
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.943472642915905,
+      "learning_rate": 1.3079795327750882e-05,
+      "loss": 0.8237,
+      "step": 2733
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.80814879490897,
+      "learning_rate": 1.307507838138054e-05,
+      "loss": 0.9702,
+      "step": 2734
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.763516539187601,
+      "learning_rate": 1.3070360679242171e-05,
+      "loss": 0.9229,
+      "step": 2735
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.513463001939668,
+      "learning_rate": 1.3065642222495263e-05,
+      "loss": 0.8847,
+      "step": 2736
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6352679845951053,
+      "learning_rate": 1.3060923012299474e-05,
+      "loss": 0.808,
+      "step": 2737
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6534091049974515,
+      "learning_rate": 1.3056203049814657e-05,
+      "loss": 0.8816,
+      "step": 2738
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6125283309661,
+      "learning_rate": 1.3051482336200844e-05,
+      "loss": 0.8314,
+      "step": 2739
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5198187375176504,
+      "learning_rate": 1.3046760872618256e-05,
+      "loss": 0.8629,
+      "step": 2740
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.314150472453498,
+      "learning_rate": 1.3042038660227294e-05,
+      "loss": 0.9143,
+      "step": 2741
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.682097793806096,
+      "learning_rate": 1.3037315700188545e-05,
+      "loss": 0.8695,
+      "step": 2742
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.691789862579373,
+      "learning_rate": 1.3032591993662782e-05,
+      "loss": 0.9287,
+      "step": 2743
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7938866440413697,
+      "learning_rate": 1.3027867541810961e-05,
+      "loss": 0.9067,
+      "step": 2744
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.806362708709313,
+      "learning_rate": 1.3023142345794217e-05,
+      "loss": 0.925,
+      "step": 2745
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8228075732201483,
+      "learning_rate": 1.3018416406773872e-05,
+      "loss": 0.9025,
+      "step": 2746
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6855012662957063,
+      "learning_rate": 1.3013689725911429e-05,
+      "loss": 0.8737,
+      "step": 2747
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.816680124122433,
+      "learning_rate": 1.3008962304368574e-05,
+      "loss": 0.9318,
+      "step": 2748
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7972551434857094,
+      "learning_rate": 1.3004234143307173e-05,
+      "loss": 1.0027,
+      "step": 2749
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8968584915515048,
+      "learning_rate": 1.2999505243889276e-05,
+      "loss": 0.8936,
+      "step": 2750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8214777133442754,
+      "learning_rate": 1.2994775607277117e-05,
+      "loss": 0.8905,
+      "step": 2751
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7101579231098265,
+      "learning_rate": 1.2990045234633103e-05,
+      "loss": 0.9263,
+      "step": 2752
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.677902149989637,
+      "learning_rate": 1.2985314127119827e-05,
+      "loss": 0.8503,
+      "step": 2753
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7699643153619378,
+      "learning_rate": 1.2980582285900067e-05,
+      "loss": 0.9604,
+      "step": 2754
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9787744770673408,
+      "learning_rate": 1.2975849712136777e-05,
+      "loss": 0.9063,
+      "step": 2755
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6750675394118044,
+      "learning_rate": 1.2971116406993087e-05,
+      "loss": 1.0037,
+      "step": 2756
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.76107592986856,
+      "learning_rate": 1.2966382371632312e-05,
+      "loss": 0.8496,
+      "step": 2757
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.810847329041354,
+      "learning_rate": 1.2961647607217947e-05,
+      "loss": 0.9515,
+      "step": 2758
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5966067731708664,
+      "learning_rate": 1.295691211491366e-05,
+      "loss": 0.8798,
+      "step": 2759
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.62491735299009,
+      "learning_rate": 1.2952175895883305e-05,
+      "loss": 0.8362,
+      "step": 2760
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.739999232523196,
+      "learning_rate": 1.2947438951290914e-05,
+      "loss": 0.8861,
+      "step": 2761
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.794841678738877,
+      "learning_rate": 1.2942701282300694e-05,
+      "loss": 0.8742,
+      "step": 2762
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.5384183680272097,
+      "learning_rate": 1.2937962890077025e-05,
+      "loss": 0.8167,
+      "step": 2763
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.618528137650998,
+      "learning_rate": 1.2933223775784476e-05,
+      "loss": 0.9138,
+      "step": 2764
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9500922611185936,
+      "learning_rate": 1.292848394058779e-05,
+      "loss": 0.9018,
+      "step": 2765
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.842181916751234,
+      "learning_rate": 1.2923743385651886e-05,
+      "loss": 0.9543,
+      "step": 2766
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.0128918616195084,
+      "learning_rate": 1.2919002112141856e-05,
+      "loss": 1.0471,
+      "step": 2767
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.561823446799556,
+      "learning_rate": 1.2914260121222973e-05,
+      "loss": 0.9366,
+      "step": 2768
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.70083738483455,
+      "learning_rate": 1.2909517414060686e-05,
+      "loss": 0.9575,
+      "step": 2769
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.6430987035979,
+      "learning_rate": 1.2904773991820619e-05,
+      "loss": 0.9195,
+      "step": 2770
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.628729956718192,
+      "learning_rate": 1.290002985566857e-05,
+      "loss": 0.8752,
+      "step": 2771
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.1639188996877206,
+      "learning_rate": 1.2895285006770521e-05,
+      "loss": 1.1261,
+      "step": 2772
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.704694091407873,
+      "learning_rate": 1.2890539446292617e-05,
+      "loss": 0.8353,
+      "step": 2773
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7133213384391324,
+      "learning_rate": 1.2885793175401184e-05,
+      "loss": 0.8487,
+      "step": 2774
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.766965649726584,
+      "learning_rate": 1.2881046195262722e-05,
+      "loss": 0.8796,
+      "step": 2775
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.666040307272094,
+      "learning_rate": 1.2876298507043909e-05,
+      "loss": 0.9043,
+      "step": 2776
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6977785742953646,
+      "learning_rate": 1.2871550111911594e-05,
+      "loss": 0.9711,
+      "step": 2777
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7691161056205362,
+      "learning_rate": 1.2866801011032797e-05,
+      "loss": 0.9293,
+      "step": 2778
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7409823091121863,
+      "learning_rate": 1.2862051205574711e-05,
+      "loss": 0.9393,
+      "step": 2779
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.729972015716603,
+      "learning_rate": 1.2857300696704709e-05,
+      "loss": 0.7922,
+      "step": 2780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6327159311383377,
+      "learning_rate": 1.2852549485590333e-05,
+      "loss": 0.8727,
+      "step": 2781
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.780302599242152,
+      "learning_rate": 1.2847797573399295e-05,
+      "loss": 0.995,
+      "step": 2782
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8037074207099093,
+      "learning_rate": 1.2843044961299485e-05,
+      "loss": 0.9171,
+      "step": 2783
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6445660114893594,
+      "learning_rate": 1.2838291650458958e-05,
+      "loss": 0.925,
+      "step": 2784
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.837391869572842,
+      "learning_rate": 1.2833537642045946e-05,
+      "loss": 0.9195,
+      "step": 2785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.824701378485826,
+      "learning_rate": 1.282878293722885e-05,
+      "loss": 0.8747,
+      "step": 2786
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9369630126925377,
+      "learning_rate": 1.2824027537176245e-05,
+      "loss": 0.8264,
+      "step": 2787
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.979787998965054,
+      "learning_rate": 1.2819271443056875e-05,
+      "loss": 0.8666,
+      "step": 2788
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7097297050328284,
+      "learning_rate": 1.2814514656039654e-05,
+      "loss": 0.8764,
+      "step": 2789
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7239968607272536,
+      "learning_rate": 1.2809757177293665e-05,
+      "loss": 0.8569,
+      "step": 2790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.739216278795826,
+      "learning_rate": 1.2804999007988164e-05,
+      "loss": 1.076,
+      "step": 2791
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.78097932188797,
+      "learning_rate": 1.280024014929258e-05,
+      "loss": 0.9201,
+      "step": 2792
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7816801537901346,
+      "learning_rate": 1.2795480602376498e-05,
+      "loss": 0.8948,
+      "step": 2793
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8022224490668406,
+      "learning_rate": 1.279072036840969e-05,
+      "loss": 0.9382,
+      "step": 2794
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0675784329509876,
+      "learning_rate": 1.2785959448562085e-05,
+      "loss": 0.969,
+      "step": 2795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6091188213405028,
+      "learning_rate": 1.2781197844003779e-05,
+      "loss": 0.7114,
+      "step": 2796
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.9346912134531844,
+      "learning_rate": 1.2776435555905044e-05,
+      "loss": 0.9375,
+      "step": 2797
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.720573376646091,
+      "learning_rate": 1.2771672585436319e-05,
+      "loss": 0.9549,
+      "step": 2798
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.816531795887586,
+      "learning_rate": 1.2766908933768208e-05,
+      "loss": 0.8821,
+      "step": 2799
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8434065002551825,
+      "learning_rate": 1.276214460207148e-05,
+      "loss": 0.9636,
+      "step": 2800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.513410606594475,
+      "learning_rate": 1.2757379591517078e-05,
+      "loss": 0.8945,
+      "step": 2801
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.572942288837356,
+      "learning_rate": 1.2752613903276105e-05,
+      "loss": 0.8388,
+      "step": 2802
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3877301322816327,
+      "learning_rate": 1.2747847538519835e-05,
+      "loss": 0.8968,
+      "step": 2803
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8973856675083773,
+      "learning_rate": 1.2743080498419706e-05,
+      "loss": 0.8581,
+      "step": 2804
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6479081480567146,
+      "learning_rate": 1.2738312784147321e-05,
+      "loss": 0.9422,
+      "step": 2805
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.658536179740867,
+      "learning_rate": 1.2733544396874458e-05,
+      "loss": 0.9756,
+      "step": 2806
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7879058284524008,
+      "learning_rate": 1.2728775337773045e-05,
+      "loss": 0.7959,
+      "step": 2807
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.56131826463935,
+      "learning_rate": 1.2724005608015182e-05,
+      "loss": 0.9383,
+      "step": 2808
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7028874499026885,
+      "learning_rate": 1.271923520877314e-05,
+      "loss": 0.9949,
+      "step": 2809
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6867758439522373,
+      "learning_rate": 1.2714464141219349e-05,
+      "loss": 0.9218,
+      "step": 2810
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.52131520434168,
+      "learning_rate": 1.2709692406526402e-05,
+      "loss": 0.8991,
+      "step": 2811
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6569271969175845,
+      "learning_rate": 1.2704920005867056e-05,
+      "loss": 0.9066,
+      "step": 2812
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8660174545559434,
+      "learning_rate": 1.2700146940414235e-05,
+      "loss": 0.992,
+      "step": 2813
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8443981988384084,
+      "learning_rate": 1.2695373211341027e-05,
+      "loss": 0.9334,
+      "step": 2814
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8875230537426404,
+      "learning_rate": 1.2690598819820673e-05,
+      "loss": 0.9484,
+      "step": 2815
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4775901913816125,
+      "learning_rate": 1.2685823767026595e-05,
+      "loss": 0.891,
+      "step": 2816
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7206299402662686,
+      "learning_rate": 1.2681048054132362e-05,
+      "loss": 0.9514,
+      "step": 2817
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.167692193895392,
+      "learning_rate": 1.2676271682311707e-05,
+      "loss": 0.8848,
+      "step": 2818
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.190369872857878,
+      "learning_rate": 1.2671494652738532e-05,
+      "loss": 0.9674,
+      "step": 2819
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7075328177894096,
+      "learning_rate": 1.2666716966586897e-05,
+      "loss": 0.9712,
+      "step": 2820
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.624690499566204,
+      "learning_rate": 1.2661938625031023e-05,
+      "loss": 0.8567,
+      "step": 2821
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8968353468717307,
+      "learning_rate": 1.2657159629245289e-05,
+      "loss": 0.8855,
+      "step": 2822
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.654379274046832,
+      "learning_rate": 1.2652379980404243e-05,
+      "loss": 0.864,
+      "step": 2823
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6926644329777125,
+      "learning_rate": 1.2647599679682587e-05,
+      "loss": 0.8662,
+      "step": 2824
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.082786408851719,
+      "learning_rate": 1.2642818728255187e-05,
+      "loss": 0.8975,
+      "step": 2825
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.738555091909095,
+      "learning_rate": 1.2638037127297057e-05,
+      "loss": 0.846,
+      "step": 2826
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.674779193194505,
+      "learning_rate": 1.2633254877983391e-05,
+      "loss": 0.8139,
+      "step": 2827
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6945356029820298,
+      "learning_rate": 1.2628471981489531e-05,
+      "loss": 1.0136,
+      "step": 2828
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.752427401318561,
+      "learning_rate": 1.2623688438990977e-05,
+      "loss": 0.8937,
+      "step": 2829
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.520523775277756,
+      "learning_rate": 1.2618904251663383e-05,
+      "loss": 0.8781,
+      "step": 2830
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7637733636446513,
+      "learning_rate": 1.2614119420682578e-05,
+      "loss": 0.8954,
+      "step": 2831
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.618376381488524,
+      "learning_rate": 1.2609333947224536e-05,
+      "loss": 0.8622,
+      "step": 2832
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6964185788176973,
+      "learning_rate": 1.2604547832465388e-05,
+      "loss": 0.9027,
+      "step": 2833
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4961769332020256,
+      "learning_rate": 1.2599761077581432e-05,
+      "loss": 0.896,
+      "step": 2834
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1354249286666676,
+      "learning_rate": 1.2594973683749117e-05,
+      "loss": 0.917,
+      "step": 2835
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.499449814627182,
+      "learning_rate": 1.2590185652145048e-05,
+      "loss": 0.876,
+      "step": 2836
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.6758994644461054,
+      "learning_rate": 1.258539698394599e-05,
+      "loss": 0.9023,
+      "step": 2837
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.721132926239213,
+      "learning_rate": 1.2580607680328862e-05,
+      "loss": 0.8959,
+      "step": 2838
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.012454865319753,
+      "learning_rate": 1.2575817742470744e-05,
+      "loss": 0.856,
+      "step": 2839
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.704714719021651,
+      "learning_rate": 1.2571027171548869e-05,
+      "loss": 0.8977,
+      "step": 2840
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.84919928232391,
+      "learning_rate": 1.2566235968740617e-05,
+      "loss": 0.9952,
+      "step": 2841
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6574199172892525,
+      "learning_rate": 1.256144413522354e-05,
+      "loss": 0.9264,
+      "step": 2842
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.614026383064716,
+      "learning_rate": 1.2556651672175334e-05,
+      "loss": 0.814,
+      "step": 2843
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.701109692270557,
+      "learning_rate": 1.2551858580773849e-05,
+      "loss": 0.9096,
+      "step": 2844
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.238141437418869,
+      "learning_rate": 1.2547064862197094e-05,
+      "loss": 0.9762,
+      "step": 2845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6739228366592105,
+      "learning_rate": 1.254227051762323e-05,
+      "loss": 0.8861,
+      "step": 2846
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.939568840529801,
+      "learning_rate": 1.2537475548230576e-05,
+      "loss": 0.897,
+      "step": 2847
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.158877445042974,
+      "learning_rate": 1.2532679955197598e-05,
+      "loss": 1.1339,
+      "step": 2848
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7311145656490585,
+      "learning_rate": 1.2527883739702915e-05,
+      "loss": 0.7805,
+      "step": 2849
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.887246109250409,
+      "learning_rate": 1.252308690292531e-05,
+      "loss": 0.9028,
+      "step": 2850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1492961703914863,
+      "learning_rate": 1.2518289446043708e-05,
+      "loss": 0.8904,
+      "step": 2851
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.86800072338607,
+      "learning_rate": 1.2513491370237185e-05,
+      "loss": 0.9,
+      "step": 2852
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0258328334393756,
+      "learning_rate": 1.2508692676684976e-05,
+      "loss": 0.9429,
+      "step": 2853
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7665456325819693,
+      "learning_rate": 1.250389336656647e-05,
+      "loss": 0.9439,
+      "step": 2854
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5646168892460377,
+      "learning_rate": 1.2499093441061197e-05,
+      "loss": 0.9178,
+      "step": 2855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6287532728333125,
+      "learning_rate": 1.2494292901348843e-05,
+      "loss": 0.8059,
+      "step": 2856
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.958320890926117,
+      "learning_rate": 1.2489491748609252e-05,
+      "loss": 0.8785,
+      "step": 2857
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2348670122620264,
+      "learning_rate": 1.2484689984022411e-05,
+      "loss": 0.8628,
+      "step": 2858
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.980007281557108,
+      "learning_rate": 1.2479887608768456e-05,
+      "loss": 0.9237,
+      "step": 2859
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1141166290087186,
+      "learning_rate": 1.2475084624027676e-05,
+      "loss": 0.9854,
+      "step": 2860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6165244904164173,
+      "learning_rate": 1.2470281030980514e-05,
+      "loss": 0.8273,
+      "step": 2861
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6790421711796113,
+      "learning_rate": 1.2465476830807554e-05,
+      "loss": 0.8995,
+      "step": 2862
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6940258268094537,
+      "learning_rate": 1.246067202468954e-05,
+      "loss": 0.7782,
+      "step": 2863
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7195876883706096,
+      "learning_rate": 1.245586661380735e-05,
+      "loss": 0.9042,
+      "step": 2864
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6665709592021134,
+      "learning_rate": 1.2451060599342027e-05,
+      "loss": 0.8503,
+      "step": 2865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.780535821992812,
+      "learning_rate": 1.244625398247475e-05,
+      "loss": 0.8841,
+      "step": 2866
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3861406054156697,
+      "learning_rate": 1.2441446764386852e-05,
+      "loss": 0.8027,
+      "step": 2867
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1004373602009925,
+      "learning_rate": 1.2436638946259812e-05,
+      "loss": 0.8828,
+      "step": 2868
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6726937897135388,
+      "learning_rate": 1.2431830529275258e-05,
+      "loss": 0.8096,
+      "step": 2869
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.699417394780327,
+      "learning_rate": 1.242702151461496e-05,
+      "loss": 0.8516,
+      "step": 2870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8101062223473368,
+      "learning_rate": 1.2422211903460845e-05,
+      "loss": 0.9078,
+      "step": 2871
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.152195380073992,
+      "learning_rate": 1.2417401696994976e-05,
+      "loss": 1.0162,
+      "step": 2872
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8336300027831007,
+      "learning_rate": 1.241259089639957e-05,
+      "loss": 0.799,
+      "step": 2873
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.050238017683165,
+      "learning_rate": 1.2407779502856987e-05,
+      "loss": 0.8961,
+      "step": 2874
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.808485721043457,
+      "learning_rate": 1.2402967517549727e-05,
+      "loss": 0.8932,
+      "step": 2875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7191297854715617,
+      "learning_rate": 1.2398154941660444e-05,
+      "loss": 0.9282,
+      "step": 2876
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.724204028606329,
+      "learning_rate": 1.2393341776371938e-05,
+      "loss": 0.9891,
+      "step": 2877
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7601559575612145,
+      "learning_rate": 1.2388528022867149e-05,
+      "loss": 0.8699,
+      "step": 2878
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.000915940449102,
+      "learning_rate": 1.238371368232916e-05,
+      "loss": 0.8998,
+      "step": 2879
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.492961193488793,
+      "learning_rate": 1.23788987559412e-05,
+      "loss": 0.8853,
+      "step": 2880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8430538524255873,
+      "learning_rate": 1.2374083244886643e-05,
+      "loss": 0.9793,
+      "step": 2881
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6505794293919682,
+      "learning_rate": 1.2369267150349009e-05,
+      "loss": 0.9107,
+      "step": 2882
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7726946812456537,
+      "learning_rate": 1.2364450473511958e-05,
+      "loss": 0.8269,
+      "step": 2883
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5356256975086793,
+      "learning_rate": 1.2359633215559297e-05,
+      "loss": 0.856,
+      "step": 2884
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.705532891404338,
+      "learning_rate": 1.235481537767497e-05,
+      "loss": 0.8162,
+      "step": 2885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5881663782552518,
+      "learning_rate": 1.2349996961043063e-05,
+      "loss": 0.8658,
+      "step": 2886
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7240151945541533,
+      "learning_rate": 1.234517796684781e-05,
+      "loss": 0.8341,
+      "step": 2887
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8736602833296305,
+      "learning_rate": 1.234035839627359e-05,
+      "loss": 0.8701,
+      "step": 2888
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.496571636405041,
+      "learning_rate": 1.233553825050491e-05,
+      "loss": 0.8708,
+      "step": 2889
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7426844394459926,
+      "learning_rate": 1.2330717530726435e-05,
+      "loss": 0.8606,
+      "step": 2890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 20.411245162121194,
+      "learning_rate": 1.2325896238122958e-05,
+      "loss": 0.9352,
+      "step": 2891
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.088228446918177,
+      "learning_rate": 1.2321074373879416e-05,
+      "loss": 0.8612,
+      "step": 2892
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.92235944442913,
+      "learning_rate": 1.2316251939180888e-05,
+      "loss": 0.8304,
+      "step": 2893
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9263552390861607,
+      "learning_rate": 1.2311428935212598e-05,
+      "loss": 0.7888,
+      "step": 2894
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9573729539265527,
+      "learning_rate": 1.23066053631599e-05,
+      "loss": 0.881,
+      "step": 2895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.592117602341213,
+      "learning_rate": 1.2301781224208297e-05,
+      "loss": 0.8186,
+      "step": 2896
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.804394996969601,
+      "learning_rate": 1.2296956519543424e-05,
+      "loss": 0.9044,
+      "step": 2897
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6173110797940975,
+      "learning_rate": 1.2292131250351059e-05,
+      "loss": 0.8435,
+      "step": 2898
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.744848357135238,
+      "learning_rate": 1.228730541781712e-05,
+      "loss": 0.8331,
+      "step": 2899
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.8600823953486634,
+      "learning_rate": 1.2282479023127656e-05,
+      "loss": 0.9479,
+      "step": 2900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.128079937009162,
+      "learning_rate": 1.2277652067468864e-05,
+      "loss": 0.881,
+      "step": 2901
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6693888567881565,
+      "learning_rate": 1.2272824552027072e-05,
+      "loss": 0.8039,
+      "step": 2902
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9620619172529166,
+      "learning_rate": 1.226799647798875e-05,
+      "loss": 0.8716,
+      "step": 2903
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7313416619549824,
+      "learning_rate": 1.2263167846540502e-05,
+      "loss": 0.9502,
+      "step": 2904
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.497128092116304,
+      "learning_rate": 1.2258338658869069e-05,
+      "loss": 1.0453,
+      "step": 2905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.5403130711331765,
+      "learning_rate": 1.2253508916161331e-05,
+      "loss": 0.792,
+      "step": 2906
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.122401354526183,
+      "learning_rate": 1.2248678619604308e-05,
+      "loss": 0.8809,
+      "step": 2907
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.86419136824606,
+      "learning_rate": 1.2243847770385142e-05,
+      "loss": 1.0589,
+      "step": 2908
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8875754346845306,
+      "learning_rate": 1.2239016369691126e-05,
+      "loss": 0.835,
+      "step": 2909
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.028081918992309,
+      "learning_rate": 1.2234184418709685e-05,
+      "loss": 0.8556,
+      "step": 2910
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6065156965874188,
+      "learning_rate": 1.222935191862837e-05,
+      "loss": 0.9404,
+      "step": 2911
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1134873620451864,
+      "learning_rate": 1.2224518870634879e-05,
+      "loss": 0.8696,
+      "step": 2912
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5175705354858375,
+      "learning_rate": 1.2219685275917039e-05,
+      "loss": 0.8818,
+      "step": 2913
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.810887240705469,
+      "learning_rate": 1.2214851135662813e-05,
+      "loss": 0.8863,
+      "step": 2914
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7008514488137108,
+      "learning_rate": 1.2210016451060291e-05,
+      "loss": 0.8713,
+      "step": 2915
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9587433703312325,
+      "learning_rate": 1.2205181223297712e-05,
+      "loss": 0.9416,
+      "step": 2916
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.448777441858399,
+      "learning_rate": 1.2200345453563433e-05,
+      "loss": 1.0207,
+      "step": 2917
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.714595830544037,
+      "learning_rate": 1.2195509143045953e-05,
+      "loss": 0.8795,
+      "step": 2918
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9810252209169565,
+      "learning_rate": 1.2190672292933902e-05,
+      "loss": 0.9086,
+      "step": 2919
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.750782059673263,
+      "learning_rate": 1.218583490441604e-05,
+      "loss": 0.9838,
+      "step": 2920
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.588070443866364,
+      "learning_rate": 1.2180996978681262e-05,
+      "loss": 0.8068,
+      "step": 2921
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5146551808157622,
+      "learning_rate": 1.2176158516918597e-05,
+      "loss": 0.7791,
+      "step": 2922
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.9319161515472665,
+      "learning_rate": 1.21713195203172e-05,
+      "loss": 0.8309,
+      "step": 2923
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6399633926053623,
+      "learning_rate": 1.2166479990066362e-05,
+      "loss": 0.7922,
+      "step": 2924
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5698088581181113,
+      "learning_rate": 1.2161639927355503e-05,
+      "loss": 0.838,
+      "step": 2925
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.56267127959082,
+      "learning_rate": 1.2156799333374176e-05,
+      "loss": 0.9022,
+      "step": 2926
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6103619596786913,
+      "learning_rate": 1.2151958209312063e-05,
+      "loss": 0.9504,
+      "step": 2927
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.754608561725593,
+      "learning_rate": 1.2147116556358975e-05,
+      "loss": 0.8826,
+      "step": 2928
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.05394639932344,
+      "learning_rate": 1.2142274375704855e-05,
+      "loss": 0.9252,
+      "step": 2929
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.758527341831582,
+      "learning_rate": 1.2137431668539778e-05,
+      "loss": 0.8559,
+      "step": 2930
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.650067159518404,
+      "learning_rate": 1.2132588436053942e-05,
+      "loss": 0.8278,
+      "step": 2931
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2212624060518076,
+      "learning_rate": 1.2127744679437681e-05,
+      "loss": 0.795,
+      "step": 2932
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.568855356171706,
+      "learning_rate": 1.212290039988145e-05,
+      "loss": 0.8856,
+      "step": 2933
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.895450975084306,
+      "learning_rate": 1.211805559857584e-05,
+      "loss": 1.0471,
+      "step": 2934
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.602246628619692,
+      "learning_rate": 1.211321027671157e-05,
+      "loss": 0.8483,
+      "step": 2935
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6483466494779972,
+      "learning_rate": 1.2108364435479478e-05,
+      "loss": 0.9137,
+      "step": 2936
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.708496704441042,
+      "learning_rate": 1.210351807607054e-05,
+      "loss": 0.9075,
+      "step": 2937
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2501209914236693,
+      "learning_rate": 1.2098671199675851e-05,
+      "loss": 0.9656,
+      "step": 2938
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7307672357696626,
+      "learning_rate": 1.2093823807486645e-05,
+      "loss": 0.8287,
+      "step": 2939
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.715745559891006,
+      "learning_rate": 1.2088975900694269e-05,
+      "loss": 0.7912,
+      "step": 2940
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7187018995015113,
+      "learning_rate": 1.2084127480490206e-05,
+      "loss": 0.8247,
+      "step": 2941
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5691527149286326,
+      "learning_rate": 1.2079278548066058e-05,
+      "loss": 0.8684,
+      "step": 2942
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7598411907507274,
+      "learning_rate": 1.2074429104613558e-05,
+      "loss": 0.8464,
+      "step": 2943
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.124777456564711,
+      "learning_rate": 1.2069579151324563e-05,
+      "loss": 1.0324,
+      "step": 2944
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.553503614208449,
+      "learning_rate": 1.2064728689391059e-05,
+      "loss": 0.8478,
+      "step": 2945
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.142861541542434,
+      "learning_rate": 1.2059877720005149e-05,
+      "loss": 1.0479,
+      "step": 2946
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7464581018539866,
+      "learning_rate": 1.205502624435907e-05,
+      "loss": 0.8466,
+      "step": 2947
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.437733125582104,
+      "learning_rate": 1.2050174263645169e-05,
+      "loss": 0.8697,
+      "step": 2948
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.590483344547154,
+      "learning_rate": 1.2045321779055936e-05,
+      "loss": 1.0495,
+      "step": 2949
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.379367692712861,
+      "learning_rate": 1.2040468791783973e-05,
+      "loss": 1.0306,
+      "step": 2950
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3149611334682025,
+      "learning_rate": 1.203561530302201e-05,
+      "loss": 1.001,
+      "step": 2951
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.087327985322694,
+      "learning_rate": 1.2030761313962898e-05,
+      "loss": 1.0608,
+      "step": 2952
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.516713422731289,
+      "learning_rate": 1.2025906825799604e-05,
+      "loss": 0.9792,
+      "step": 2953
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.870450403349401,
+      "learning_rate": 1.2021051839725235e-05,
+      "loss": 0.8342,
+      "step": 2954
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.623934475340532,
+      "learning_rate": 1.2016196356933005e-05,
+      "loss": 0.8126,
+      "step": 2955
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.757354876528018,
+      "learning_rate": 1.2011340378616256e-05,
+      "loss": 0.9232,
+      "step": 2956
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6140798539319494,
+      "learning_rate": 1.2006483905968456e-05,
+      "loss": 0.8552,
+      "step": 2957
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6166843878359862,
+      "learning_rate": 1.2001626940183185e-05,
+      "loss": 0.926,
+      "step": 2958
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6471979880923047,
+      "learning_rate": 1.1996769482454144e-05,
+      "loss": 0.8624,
+      "step": 2959
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6230161949992663,
+      "learning_rate": 1.1991911533975172e-05,
+      "loss": 0.7624,
+      "step": 2960
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7014344511734807,
+      "learning_rate": 1.1987053095940204e-05,
+      "loss": 0.86,
+      "step": 2961
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.649651028227363,
+      "learning_rate": 1.198219416954332e-05,
+      "loss": 0.9156,
+      "step": 2962
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5928321617079844,
+      "learning_rate": 1.19773347559787e-05,
+      "loss": 0.8112,
+      "step": 2963
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7488508110619754,
+      "learning_rate": 1.1972474856440654e-05,
+      "loss": 0.9098,
+      "step": 2964
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7198204374543224,
+      "learning_rate": 1.1967614472123607e-05,
+      "loss": 0.8645,
+      "step": 2965
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.05265967983416,
+      "learning_rate": 1.1962753604222108e-05,
+      "loss": 0.9255,
+      "step": 2966
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6759537172100645,
+      "learning_rate": 1.1957892253930819e-05,
+      "loss": 0.8436,
+      "step": 2967
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2144844492572755,
+      "learning_rate": 1.1953030422444526e-05,
+      "loss": 0.914,
+      "step": 2968
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0346285503049057,
+      "learning_rate": 1.1948168110958132e-05,
+      "loss": 0.9222,
+      "step": 2969
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.919974504371189,
+      "learning_rate": 1.194330532066665e-05,
+      "loss": 0.9076,
+      "step": 2970
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.916037722150218,
+      "learning_rate": 1.1938442052765225e-05,
+      "loss": 0.9252,
+      "step": 2971
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0399673977308037,
+      "learning_rate": 1.1933578308449108e-05,
+      "loss": 0.7823,
+      "step": 2972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.554756201579733,
+      "learning_rate": 1.1928714088913673e-05,
+      "loss": 0.828,
+      "step": 2973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7425865424645615,
+      "learning_rate": 1.1923849395354407e-05,
+      "loss": 0.8717,
+      "step": 2974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.375571545693288,
+      "learning_rate": 1.1918984228966917e-05,
+      "loss": 0.8431,
+      "step": 2975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9360801702086463,
+      "learning_rate": 1.1914118590946924e-05,
+      "loss": 0.8792,
+      "step": 2976
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9626264065606702,
+      "learning_rate": 1.1909252482490263e-05,
+      "loss": 0.8869,
+      "step": 2977
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.890042514095932,
+      "learning_rate": 1.190438590479289e-05,
+      "loss": 0.9294,
+      "step": 2978
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.981870606613395,
+      "learning_rate": 1.1899518859050869e-05,
+      "loss": 0.9017,
+      "step": 2979
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.715493009051848,
+      "learning_rate": 1.1894651346460391e-05,
+      "loss": 0.8221,
+      "step": 2980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 10.855410286020899,
+      "learning_rate": 1.188978336821775e-05,
+      "loss": 1.0945,
+      "step": 2981
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7232678577599536,
+      "learning_rate": 1.1884914925519356e-05,
+      "loss": 0.8186,
+      "step": 2982
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.482457923899972,
+      "learning_rate": 1.1880046019561735e-05,
+      "loss": 0.8432,
+      "step": 2983
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.639829647704218,
+      "learning_rate": 1.1875176651541533e-05,
+      "loss": 0.8063,
+      "step": 2984
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.107733287783824,
+      "learning_rate": 1.1870306822655502e-05,
+      "loss": 0.9191,
+      "step": 2985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7323130504400144,
+      "learning_rate": 1.1865436534100508e-05,
+      "loss": 0.9172,
+      "step": 2986
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7044458896629044,
+      "learning_rate": 1.186056578707353e-05,
+      "loss": 0.9815,
+      "step": 2987
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8592139872621245,
+      "learning_rate": 1.1855694582771666e-05,
+      "loss": 0.9378,
+      "step": 2988
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.565066517688624,
+      "learning_rate": 1.1850822922392119e-05,
+      "loss": 0.8113,
+      "step": 2989
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6779479847097685,
+      "learning_rate": 1.1845950807132203e-05,
+      "loss": 0.8412,
+      "step": 2990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5545351126336584,
+      "learning_rate": 1.1841078238189352e-05,
+      "loss": 0.9046,
+      "step": 2991
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6450426222819354,
+      "learning_rate": 1.1836205216761105e-05,
+      "loss": 0.8617,
+      "step": 2992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6275419597565746,
+      "learning_rate": 1.1831331744045114e-05,
+      "loss": 0.8616,
+      "step": 2993
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7510452230302787,
+      "learning_rate": 1.182645782123914e-05,
+      "loss": 0.8931,
+      "step": 2994
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6643225382375206,
+      "learning_rate": 1.182158344954106e-05,
+      "loss": 0.8337,
+      "step": 2995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.618555830855707,
+      "learning_rate": 1.1816708630148857e-05,
+      "loss": 0.9202,
+      "step": 2996
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8455305353008713,
+      "learning_rate": 1.1811833364260625e-05,
+      "loss": 0.9023,
+      "step": 2997
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7928103126274375,
+      "learning_rate": 1.1806957653074564e-05,
+      "loss": 0.8542,
+      "step": 2998
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7731698718633604,
+      "learning_rate": 1.1802081497788993e-05,
+      "loss": 0.7557,
+      "step": 2999
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.814943969739714,
+      "learning_rate": 1.1797204899602328e-05,
+      "loss": 0.8808,
+      "step": 3000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8355012116827134,
+      "learning_rate": 1.1792327859713104e-05,
+      "loss": 0.8385,
+      "step": 3001
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1537111761470373,
+      "learning_rate": 1.1787450379319963e-05,
+      "loss": 0.8303,
+      "step": 3002
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.744178078536589,
+      "learning_rate": 1.1782572459621646e-05,
+      "loss": 0.9255,
+      "step": 3003
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5395370642311623,
+      "learning_rate": 1.1777694101817014e-05,
+      "loss": 0.8489,
+      "step": 3004
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9295376938033324,
+      "learning_rate": 1.1772815307105027e-05,
+      "loss": 0.9934,
+      "step": 3005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8432809249138042,
+      "learning_rate": 1.176793607668476e-05,
+      "loss": 0.9256,
+      "step": 3006
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6080502082997423,
+      "learning_rate": 1.176305641175539e-05,
+      "loss": 0.8401,
+      "step": 3007
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6194753528697103,
+      "learning_rate": 1.17581763135162e-05,
+      "loss": 0.8568,
+      "step": 3008
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.518047586237937,
+      "learning_rate": 1.1753295783166581e-05,
+      "loss": 0.8328,
+      "step": 3009
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4931889524384694,
+      "learning_rate": 1.1748414821906034e-05,
+      "loss": 0.8611,
+      "step": 3010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.540801850644384,
+      "learning_rate": 1.1743533430934155e-05,
+      "loss": 0.9411,
+      "step": 3011
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7099336287100435,
+      "learning_rate": 1.173865161145066e-05,
+      "loss": 0.7831,
+      "step": 3012
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.11702898724108,
+      "learning_rate": 1.1733769364655363e-05,
+      "loss": 0.9089,
+      "step": 3013
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.378925417302546,
+      "learning_rate": 1.1728886691748183e-05,
+      "loss": 1.0161,
+      "step": 3014
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7509304800419336,
+      "learning_rate": 1.1724003593929138e-05,
+      "loss": 0.9121,
+      "step": 3015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.68915934169062,
+      "learning_rate": 1.1719120072398361e-05,
+      "loss": 0.8014,
+      "step": 3016
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7399760762830856,
+      "learning_rate": 1.1714236128356092e-05,
+      "loss": 0.8724,
+      "step": 3017
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.030162280916283,
+      "learning_rate": 1.1709351763002652e-05,
+      "loss": 0.8158,
+      "step": 3018
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.585101360406029,
+      "learning_rate": 1.1704466977538496e-05,
+      "loss": 0.9078,
+      "step": 3019
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7771498640982206,
+      "learning_rate": 1.1699581773164155e-05,
+      "loss": 0.8736,
+      "step": 3020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9162027035422398,
+      "learning_rate": 1.1694696151080282e-05,
+      "loss": 0.9148,
+      "step": 3021
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.674802039440239,
+      "learning_rate": 1.1689810112487626e-05,
+      "loss": 0.8939,
+      "step": 3022
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.9827837301978692,
+      "learning_rate": 1.1684923658587036e-05,
+      "loss": 0.7724,
+      "step": 3023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.446858036060685,
+      "learning_rate": 1.1680036790579465e-05,
+      "loss": 0.8349,
+      "step": 3024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7490258161168355,
+      "learning_rate": 1.1675149509665972e-05,
+      "loss": 0.8318,
+      "step": 3025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.758105078173259,
+      "learning_rate": 1.167026181704771e-05,
+      "loss": 0.8289,
+      "step": 3026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.353607851225301,
+      "learning_rate": 1.1665373713925936e-05,
+      "loss": 1.0972,
+      "step": 3027
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3166270913512403,
+      "learning_rate": 1.1660485201502011e-05,
+      "loss": 0.8928,
+      "step": 3028
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5484423724642578,
+      "learning_rate": 1.1655596280977395e-05,
+      "loss": 0.8378,
+      "step": 3029
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7828491941027105,
+      "learning_rate": 1.1650706953553644e-05,
+      "loss": 0.8506,
+      "step": 3030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.222252561223075,
+      "learning_rate": 1.1645817220432421e-05,
+      "loss": 0.7596,
+      "step": 3031
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8353519578274593,
+      "learning_rate": 1.1640927082815485e-05,
+      "loss": 0.8111,
+      "step": 3032
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.712517208964851,
+      "learning_rate": 1.1636036541904692e-05,
+      "loss": 0.9596,
+      "step": 3033
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.02873034434939,
+      "learning_rate": 1.1631145598901999e-05,
+      "loss": 0.934,
+      "step": 3034
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7491870247460337,
+      "learning_rate": 1.1626254255009465e-05,
+      "loss": 0.8391,
+      "step": 3035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.522093212738503,
+      "learning_rate": 1.162136251142925e-05,
+      "loss": 0.8147,
+      "step": 3036
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.03519444835517,
+      "learning_rate": 1.1616470369363602e-05,
+      "loss": 0.892,
+      "step": 3037
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.769866454847948,
+      "learning_rate": 1.161157783001487e-05,
+      "loss": 0.9277,
+      "step": 3038
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8005906197669908,
+      "learning_rate": 1.1606684894585507e-05,
+      "loss": 0.9335,
+      "step": 3039
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6771495503120724,
+      "learning_rate": 1.1601791564278057e-05,
+      "loss": 0.8585,
+      "step": 3040
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7458170630301493,
+      "learning_rate": 1.1596897840295165e-05,
+      "loss": 0.9842,
+      "step": 3041
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.42988492519934,
+      "learning_rate": 1.159200372383957e-05,
+      "loss": 0.8292,
+      "step": 3042
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5103081399212375,
+      "learning_rate": 1.1587109216114111e-05,
+      "loss": 0.9228,
+      "step": 3043
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.589221115239344,
+      "learning_rate": 1.1582214318321718e-05,
+      "loss": 0.8647,
+      "step": 3044
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.811962497245336,
+      "learning_rate": 1.1577319031665419e-05,
+      "loss": 1.0228,
+      "step": 3045
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6507114110052887,
+      "learning_rate": 1.157242335734834e-05,
+      "loss": 0.845,
+      "step": 3046
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8573693138518306,
+      "learning_rate": 1.1567527296573702e-05,
+      "loss": 0.87,
+      "step": 3047
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.457127461940877,
+      "learning_rate": 1.1562630850544816e-05,
+      "loss": 0.86,
+      "step": 3048
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.556657489478441,
+      "learning_rate": 1.1557734020465093e-05,
+      "loss": 0.7787,
+      "step": 3049
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.76817538447232,
+      "learning_rate": 1.1552836807538034e-05,
+      "loss": 1.0745,
+      "step": 3050
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.828019544724989,
+      "learning_rate": 1.154793921296724e-05,
+      "loss": 0.7933,
+      "step": 3051
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9926852113478217,
+      "learning_rate": 1.1543041237956403e-05,
+      "loss": 0.9776,
+      "step": 3052
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7029689770823206,
+      "learning_rate": 1.1538142883709305e-05,
+      "loss": 0.9321,
+      "step": 3053
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.819858750783005,
+      "learning_rate": 1.1533244151429825e-05,
+      "loss": 0.8614,
+      "step": 3054
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5754083770125895,
+      "learning_rate": 1.1528345042321933e-05,
+      "loss": 0.8534,
+      "step": 3055
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.808109444366165,
+      "learning_rate": 1.1523445557589692e-05,
+      "loss": 0.8464,
+      "step": 3056
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.642707924896552,
+      "learning_rate": 1.1518545698437262e-05,
+      "loss": 0.8235,
+      "step": 3057
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.793040715733123,
+      "learning_rate": 1.1513645466068887e-05,
+      "loss": 0.7877,
+      "step": 3058
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2252924252687065,
+      "learning_rate": 1.1508744861688912e-05,
+      "loss": 0.7785,
+      "step": 3059
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9322617811791116,
+      "learning_rate": 1.150384388650176e-05,
+      "loss": 0.8891,
+      "step": 3060
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3079267028920536,
+      "learning_rate": 1.149894254171196e-05,
+      "loss": 0.8218,
+      "step": 3061
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.826653081264617,
+      "learning_rate": 1.1494040828524122e-05,
+      "loss": 0.8896,
+      "step": 3062
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.499685351473289,
+      "learning_rate": 1.1489138748142949e-05,
+      "loss": 0.8141,
+      "step": 3063
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0839843943225453,
+      "learning_rate": 1.1484236301773239e-05,
+      "loss": 0.797,
+      "step": 3064
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9392104478446295,
+      "learning_rate": 1.1479333490619873e-05,
+      "loss": 0.8057,
+      "step": 3065
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.544480005618585,
+      "learning_rate": 1.147443031588782e-05,
+      "loss": 0.8691,
+      "step": 3066
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5467653566420805,
+      "learning_rate": 1.146952677878215e-05,
+      "loss": 0.7424,
+      "step": 3067
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.62107278620751,
+      "learning_rate": 1.1464622880508012e-05,
+      "loss": 0.967,
+      "step": 3068
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.543148916423793,
+      "learning_rate": 1.1459718622270648e-05,
+      "loss": 1.0334,
+      "step": 3069
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4324678600471663,
+      "learning_rate": 1.1454814005275388e-05,
+      "loss": 0.8146,
+      "step": 3070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8998473892388326,
+      "learning_rate": 1.1449909030727641e-05,
+      "loss": 0.8289,
+      "step": 3071
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7750117258591485,
+      "learning_rate": 1.1445003699832922e-05,
+      "loss": 1.0167,
+      "step": 3072
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8018037931812034,
+      "learning_rate": 1.144009801379682e-05,
+      "loss": 0.9312,
+      "step": 3073
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.789138018988291,
+      "learning_rate": 1.1435191973825015e-05,
+      "loss": 0.9369,
+      "step": 3074
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.692334241154601,
+      "learning_rate": 1.1430285581123278e-05,
+      "loss": 0.911,
+      "step": 3075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5145506813955163,
+      "learning_rate": 1.1425378836897457e-05,
+      "loss": 0.8242,
+      "step": 3076
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4979403035686945,
+      "learning_rate": 1.1420471742353491e-05,
+      "loss": 0.9209,
+      "step": 3077
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.502264436899139,
+      "learning_rate": 1.141556429869741e-05,
+      "loss": 0.806,
+      "step": 3078
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.734742905570955,
+      "learning_rate": 1.1410656507135328e-05,
+      "loss": 0.8528,
+      "step": 3079
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.722677435806108,
+      "learning_rate": 1.1405748368873438e-05,
+      "loss": 0.8267,
+      "step": 3080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6633849719505664,
+      "learning_rate": 1.1400839885118026e-05,
+      "loss": 0.8519,
+      "step": 3081
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.831404582758861,
+      "learning_rate": 1.1395931057075455e-05,
+      "loss": 0.839,
+      "step": 3082
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7737992784725893,
+      "learning_rate": 1.1391021885952182e-05,
+      "loss": 0.8881,
+      "step": 3083
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.287806551319888,
+      "learning_rate": 1.1386112372954745e-05,
+      "loss": 0.9836,
+      "step": 3084
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.687326202298055,
+      "learning_rate": 1.138120251928976e-05,
+      "loss": 1.0422,
+      "step": 3085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5190937625243697,
+      "learning_rate": 1.137629232616393e-05,
+      "loss": 0.8285,
+      "step": 3086
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6728531205610087,
+      "learning_rate": 1.1371381794784051e-05,
+      "loss": 0.8709,
+      "step": 3087
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1875883834182965,
+      "learning_rate": 1.1366470926356986e-05,
+      "loss": 0.8351,
+      "step": 3088
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6343183413218236,
+      "learning_rate": 1.1361559722089691e-05,
+      "loss": 0.9105,
+      "step": 3089
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.273978479754304,
+      "learning_rate": 1.1356648183189203e-05,
+      "loss": 1.0424,
+      "step": 3090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8158695606170863,
+      "learning_rate": 1.1351736310862642e-05,
+      "loss": 0.8838,
+      "step": 3091
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.808634999693899,
+      "learning_rate": 1.1346824106317204e-05,
+      "loss": 0.7946,
+      "step": 3092
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9091587714578644,
+      "learning_rate": 1.1341911570760176e-05,
+      "loss": 0.8636,
+      "step": 3093
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.481053395723128,
+      "learning_rate": 1.1336998705398918e-05,
+      "loss": 0.8706,
+      "step": 3094
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7924999881066985,
+      "learning_rate": 1.1332085511440877e-05,
+      "loss": 0.9168,
+      "step": 3095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7065320524178937,
+      "learning_rate": 1.1327171990093574e-05,
+      "loss": 0.8494,
+      "step": 3096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.798614283817104,
+      "learning_rate": 1.1322258142564619e-05,
+      "loss": 0.8357,
+      "step": 3097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9673905386874635,
+      "learning_rate": 1.13173439700617e-05,
+      "loss": 0.9729,
+      "step": 3098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.669600900005444,
+      "learning_rate": 1.1312429473792576e-05,
+      "loss": 0.8486,
+      "step": 3099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.820936598288706,
+      "learning_rate": 1.1307514654965097e-05,
+      "loss": 0.8991,
+      "step": 3100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.6125223944829936,
+      "learning_rate": 1.1302599514787186e-05,
+      "loss": 0.8566,
+      "step": 3101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.739922897046682,
+      "learning_rate": 1.129768405446685e-05,
+      "loss": 0.9179,
+      "step": 3102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7433379542441787,
+      "learning_rate": 1.1292768275212162e-05,
+      "loss": 0.8841,
+      "step": 3103
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.670627406032204,
+      "learning_rate": 1.1287852178231295e-05,
+      "loss": 0.8187,
+      "step": 3104
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9365550668107816,
+      "learning_rate": 1.1282935764732477e-05,
+      "loss": 0.9242,
+      "step": 3105
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.755246641824546,
+      "learning_rate": 1.1278019035924032e-05,
+      "loss": 0.9124,
+      "step": 3106
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7649728926786143,
+      "learning_rate": 1.1273101993014351e-05,
+      "loss": 0.782,
+      "step": 3107
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1261831995171527,
+      "learning_rate": 1.1268184637211905e-05,
+      "loss": 0.8278,
+      "step": 3108
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6353217053491713,
+      "learning_rate": 1.1263266969725244e-05,
+      "loss": 0.893,
+      "step": 3109
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8645760552772144,
+      "learning_rate": 1.1258348991762994e-05,
+      "loss": 0.7923,
+      "step": 3110
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.891997003857544,
+      "learning_rate": 1.1253430704533847e-05,
+      "loss": 0.7817,
+      "step": 3111
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.624875915617395,
+      "learning_rate": 1.124851210924659e-05,
+      "loss": 0.8476,
+      "step": 3112
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7688792539306153,
+      "learning_rate": 1.1243593207110073e-05,
+      "loss": 0.8592,
+      "step": 3113
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7742296596297518,
+      "learning_rate": 1.1238673999333223e-05,
+      "loss": 0.8803,
+      "step": 3114
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.793489256791983,
+      "learning_rate": 1.1233754487125043e-05,
+      "loss": 0.9501,
+      "step": 3115
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.813680144743365,
+      "learning_rate": 1.1228834671694613e-05,
+      "loss": 0.839,
+      "step": 3116
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.916078405373159,
+      "learning_rate": 1.1223914554251085e-05,
+      "loss": 1.0373,
+      "step": 3117
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.610000965016514,
+      "learning_rate": 1.1218994136003685e-05,
+      "loss": 0.8629,
+      "step": 3118
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7625346058741904,
+      "learning_rate": 1.1214073418161712e-05,
+      "loss": 0.9373,
+      "step": 3119
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6943411479104284,
+      "learning_rate": 1.1209152401934546e-05,
+      "loss": 0.8694,
+      "step": 3120
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.861777936032375,
+      "learning_rate": 1.1204231088531631e-05,
+      "loss": 0.7413,
+      "step": 3121
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5235310218243376,
+      "learning_rate": 1.1199309479162489e-05,
+      "loss": 0.8644,
+      "step": 3122
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.497572177741026,
+      "learning_rate": 1.119438757503671e-05,
+      "loss": 0.8023,
+      "step": 3123
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6581250101609863,
+      "learning_rate": 1.1189465377363964e-05,
+      "loss": 0.8173,
+      "step": 3124
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7763787999992227,
+      "learning_rate": 1.118454288735399e-05,
+      "loss": 0.9271,
+      "step": 3125
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5574303042021675,
+      "learning_rate": 1.1179620106216597e-05,
+      "loss": 0.9474,
+      "step": 3126
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5624182985693587,
+      "learning_rate": 1.117469703516166e-05,
+      "loss": 0.8598,
+      "step": 3127
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3757353182751983,
+      "learning_rate": 1.1169773675399144e-05,
+      "loss": 0.8019,
+      "step": 3128
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.690347384996987,
+      "learning_rate": 1.1164850028139063e-05,
+      "loss": 0.8593,
+      "step": 3129
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4945781694298197,
+      "learning_rate": 1.1159926094591514e-05,
+      "loss": 0.8506,
+      "step": 3130
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.655067345590805,
+      "learning_rate": 1.1155001875966663e-05,
+      "loss": 0.894,
+      "step": 3131
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5126207449004463,
+      "learning_rate": 1.1150077373474745e-05,
+      "loss": 0.7716,
+      "step": 3132
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3273236917279556,
+      "learning_rate": 1.1145152588326063e-05,
+      "loss": 0.75,
+      "step": 3133
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7025550289889817,
+      "learning_rate": 1.1140227521730988e-05,
+      "loss": 0.7937,
+      "step": 3134
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.742824070165681,
+      "learning_rate": 1.1135302174899971e-05,
+      "loss": 0.8877,
+      "step": 3135
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.237256074327542,
+      "learning_rate": 1.113037654904352e-05,
+      "loss": 0.8987,
+      "step": 3136
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9438627721310087,
+      "learning_rate": 1.1125450645372218e-05,
+      "loss": 0.8522,
+      "step": 3137
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.886165850380071,
+      "learning_rate": 1.1120524465096706e-05,
+      "loss": 0.8938,
+      "step": 3138
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.656321602028388,
+      "learning_rate": 1.1115598009427712e-05,
+      "loss": 0.8773,
+      "step": 3139
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6816869315729543,
+      "learning_rate": 1.1110671279576014e-05,
+      "loss": 0.8806,
+      "step": 3140
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7861332807618555,
+      "learning_rate": 1.1105744276752464e-05,
+      "loss": 0.8534,
+      "step": 3141
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7415889420632036,
+      "learning_rate": 1.1100817002167983e-05,
+      "loss": 0.8204,
+      "step": 3142
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.853777558312234,
+      "learning_rate": 1.1095889457033557e-05,
+      "loss": 0.8472,
+      "step": 3143
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.679333044964318,
+      "learning_rate": 1.1090961642560238e-05,
+      "loss": 0.8814,
+      "step": 3144
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.712063692602172,
+      "learning_rate": 1.1086033559959143e-05,
+      "loss": 0.8651,
+      "step": 3145
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.901893289890703,
+      "learning_rate": 1.1081105210441458e-05,
+      "loss": 0.9463,
+      "step": 3146
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.622942948873725,
+      "learning_rate": 1.1076176595218438e-05,
+      "loss": 0.8437,
+      "step": 3147
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.784225144697118,
+      "learning_rate": 1.1071247715501387e-05,
+      "loss": 0.9422,
+      "step": 3148
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.784905531348584,
+      "learning_rate": 1.1066318572501695e-05,
+      "loss": 0.857,
+      "step": 3149
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.614670266049605,
+      "learning_rate": 1.1061389167430804e-05,
+      "loss": 0.899,
+      "step": 3150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.786304158975111,
+      "learning_rate": 1.1056459501500223e-05,
+      "loss": 0.9484,
+      "step": 3151
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7894077314963335,
+      "learning_rate": 1.1051529575921525e-05,
+      "loss": 0.9577,
+      "step": 3152
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.621899779821202,
+      "learning_rate": 1.1046599391906347e-05,
+      "loss": 0.7899,
+      "step": 3153
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7178037657407033,
+      "learning_rate": 1.1041668950666395e-05,
+      "loss": 0.7923,
+      "step": 3154
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.387761205410121,
+      "learning_rate": 1.1036738253413431e-05,
+      "loss": 0.7847,
+      "step": 3155
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6422778659156134,
+      "learning_rate": 1.1031807301359273e-05,
+      "loss": 0.9249,
+      "step": 3156
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9058694607373634,
+      "learning_rate": 1.1026876095715825e-05,
+      "loss": 0.8803,
+      "step": 3157
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.628653013302643,
+      "learning_rate": 1.1021944637695032e-05,
+      "loss": 0.9861,
+      "step": 3158
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.639775493358449,
+      "learning_rate": 1.1017012928508905e-05,
+      "loss": 0.8768,
+      "step": 3159
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6560270557217853,
+      "learning_rate": 1.1012080969369527e-05,
+      "loss": 0.8731,
+      "step": 3160
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9183755912857343,
+      "learning_rate": 1.1007148761489031e-05,
+      "loss": 0.8827,
+      "step": 3161
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6162305768841576,
+      "learning_rate": 1.1002216306079616e-05,
+      "loss": 0.8166,
+      "step": 3162
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.580769173010447,
+      "learning_rate": 1.099728360435354e-05,
+      "loss": 0.9338,
+      "step": 3163
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.548212107677192,
+      "learning_rate": 1.0992350657523123e-05,
+      "loss": 0.7659,
+      "step": 3164
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.048637681072861,
+      "learning_rate": 1.0987417466800749e-05,
+      "loss": 0.9659,
+      "step": 3165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6843483103040544,
+      "learning_rate": 1.0982484033398855e-05,
+      "loss": 0.8631,
+      "step": 3166
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6588237759332394,
+      "learning_rate": 1.0977550358529935e-05,
+      "loss": 0.8909,
+      "step": 3167
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8436832837871253,
+      "learning_rate": 1.0972616443406558e-05,
+      "loss": 0.834,
+      "step": 3168
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8045928596793903,
+      "learning_rate": 1.0967682289241337e-05,
+      "loss": 0.8808,
+      "step": 3169
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7498769351854677,
+      "learning_rate": 1.0962747897246949e-05,
+      "loss": 0.8705,
+      "step": 3170
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.709629558405918,
+      "learning_rate": 1.0957813268636127e-05,
+      "loss": 0.8877,
+      "step": 3171
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6672565570216595,
+      "learning_rate": 1.0952878404621667e-05,
+      "loss": 0.9211,
+      "step": 3172
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.855523920099627,
+      "learning_rate": 1.0947943306416422e-05,
+      "loss": 0.834,
+      "step": 3173
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8752440568494277,
+      "learning_rate": 1.0943007975233296e-05,
+      "loss": 0.8943,
+      "step": 3174
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7796180474907732,
+      "learning_rate": 1.0938072412285257e-05,
+      "loss": 0.9756,
+      "step": 3175
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.637847864052865,
+      "learning_rate": 1.093313661878533e-05,
+      "loss": 0.843,
+      "step": 3176
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.516870968925292,
+      "learning_rate": 1.0928200595946594e-05,
+      "loss": 0.8225,
+      "step": 3177
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8040107938246868,
+      "learning_rate": 1.092326434498218e-05,
+      "loss": 0.8971,
+      "step": 3178
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.312917984262526,
+      "learning_rate": 1.0918327867105284e-05,
+      "loss": 1.0743,
+      "step": 3179
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6661378147021946,
+      "learning_rate": 1.0913391163529158e-05,
+      "loss": 0.8194,
+      "step": 3180
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.923249564548845,
+      "learning_rate": 1.0908454235467099e-05,
+      "loss": 0.9235,
+      "step": 3181
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.124631933822971,
+      "learning_rate": 1.0903517084132469e-05,
+      "loss": 0.7906,
+      "step": 3182
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6608958546046586,
+      "learning_rate": 1.0898579710738675e-05,
+      "loss": 0.9203,
+      "step": 3183
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.737613006287347,
+      "learning_rate": 1.0893642116499194e-05,
+      "loss": 0.95,
+      "step": 3184
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.804888270338999,
+      "learning_rate": 1.0888704302627542e-05,
+      "loss": 0.8623,
+      "step": 3185
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.710517749370958,
+      "learning_rate": 1.0883766270337297e-05,
+      "loss": 0.7938,
+      "step": 3186
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.846565399790556,
+      "learning_rate": 1.0878828020842091e-05,
+      "loss": 0.8408,
+      "step": 3187
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2049410040769426,
+      "learning_rate": 1.0873889555355606e-05,
+      "loss": 0.845,
+      "step": 3188
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9466179850540923,
+      "learning_rate": 1.0868950875091573e-05,
+      "loss": 0.8936,
+      "step": 3189
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6737815841660186,
+      "learning_rate": 1.0864011981263786e-05,
+      "loss": 0.9218,
+      "step": 3190
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.620826306048308,
+      "learning_rate": 1.085907287508609e-05,
+      "loss": 0.9462,
+      "step": 3191
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.778997897619199,
+      "learning_rate": 1.0854133557772373e-05,
+      "loss": 0.9397,
+      "step": 3192
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5247511860554677,
+      "learning_rate": 1.0849194030536583e-05,
+      "loss": 0.9028,
+      "step": 3193
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4814349353822736,
+      "learning_rate": 1.0844254294592716e-05,
+      "loss": 0.7551,
+      "step": 3194
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5717710565669205,
+      "learning_rate": 1.0839314351154821e-05,
+      "loss": 0.8836,
+      "step": 3195
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.061981208422528,
+      "learning_rate": 1.0834374201436996e-05,
+      "loss": 0.8326,
+      "step": 3196
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.710682198564873,
+      "learning_rate": 1.0829433846653397e-05,
+      "loss": 0.9966,
+      "step": 3197
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.631961606425523,
+      "learning_rate": 1.082449328801822e-05,
+      "loss": 0.9213,
+      "step": 3198
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.718721534981861,
+      "learning_rate": 1.0819552526745716e-05,
+      "loss": 0.7187,
+      "step": 3199
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6937766875932736,
+      "learning_rate": 1.0814611564050186e-05,
+      "loss": 0.8894,
+      "step": 3200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8889819332593887,
+      "learning_rate": 1.080967040114598e-05,
+      "loss": 0.8874,
+      "step": 3201
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7937136356082792,
+      "learning_rate": 1.08047290392475e-05,
+      "loss": 0.8424,
+      "step": 3202
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.644207184621602,
+      "learning_rate": 1.0799787479569188e-05,
+      "loss": 0.8399,
+      "step": 3203
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3400959435695334,
+      "learning_rate": 1.0794845723325544e-05,
+      "loss": 0.8781,
+      "step": 3204
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9983667609443976,
+      "learning_rate": 1.0789903771731118e-05,
+      "loss": 0.8109,
+      "step": 3205
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.617913805844195,
+      "learning_rate": 1.0784961626000497e-05,
+      "loss": 0.831,
+      "step": 3206
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.676909320940646,
+      "learning_rate": 1.0780019287348321e-05,
+      "loss": 0.9348,
+      "step": 3207
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4113877770088146,
+      "learning_rate": 1.0775076756989281e-05,
+      "loss": 0.7715,
+      "step": 3208
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.873965751785654,
+      "learning_rate": 1.0770134036138114e-05,
+      "loss": 0.863,
+      "step": 3209
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.234018267681649,
+      "learning_rate": 1.07651911260096e-05,
+      "loss": 1.0551,
+      "step": 3210
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.601399561591246,
+      "learning_rate": 1.0760248027818566e-05,
+      "loss": 0.8906,
+      "step": 3211
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5780137439885213,
+      "learning_rate": 1.0755304742779891e-05,
+      "loss": 0.8582,
+      "step": 3212
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7272475315349687,
+      "learning_rate": 1.0750361272108492e-05,
+      "loss": 0.9181,
+      "step": 3213
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0256202385156685,
+      "learning_rate": 1.0745417617019336e-05,
+      "loss": 0.9568,
+      "step": 3214
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.515863450716289,
+      "learning_rate": 1.0740473778727436e-05,
+      "loss": 0.7738,
+      "step": 3215
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.787365827288811,
+      "learning_rate": 1.0735529758447851e-05,
+      "loss": 0.9147,
+      "step": 3216
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.687163583453802,
+      "learning_rate": 1.0730585557395682e-05,
+      "loss": 0.9083,
+      "step": 3217
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.498617602623131,
+      "learning_rate": 1.0725641176786066e-05,
+      "loss": 0.8464,
+      "step": 3218
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.669957505128777,
+      "learning_rate": 1.0720696617834203e-05,
+      "loss": 0.8533,
+      "step": 3219
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7615180189226725,
+      "learning_rate": 1.0715751881755322e-05,
+      "loss": 0.8023,
+      "step": 3220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.84954465106346,
+      "learning_rate": 1.0710806969764708e-05,
+      "loss": 0.7787,
+      "step": 3221
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.761029530875673,
+      "learning_rate": 1.0705861883077676e-05,
+      "loss": 0.8775,
+      "step": 3222
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7200088736275156,
+      "learning_rate": 1.0700916622909584e-05,
+      "loss": 0.8918,
+      "step": 3223
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.580459203264464,
+      "learning_rate": 1.069597119047585e-05,
+      "loss": 0.979,
+      "step": 3224
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.784799576334272,
+      "learning_rate": 1.0691025586991913e-05,
+      "loss": 0.8962,
+      "step": 3225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.716320206142077,
+      "learning_rate": 1.0686079813673266e-05,
+      "loss": 0.8454,
+      "step": 3226
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.78252057186699,
+      "learning_rate": 1.0681133871735447e-05,
+      "loss": 0.7963,
+      "step": 3227
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.797761111042213,
+      "learning_rate": 1.0676187762394024e-05,
+      "loss": 0.8693,
+      "step": 3228
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6328034403208016,
+      "learning_rate": 1.0671241486864612e-05,
+      "loss": 0.831,
+      "step": 3229
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.509790499903019,
+      "learning_rate": 1.0666295046362866e-05,
+      "loss": 0.8735,
+      "step": 3230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.559415932905526,
+      "learning_rate": 1.0661348442104488e-05,
+      "loss": 0.8074,
+      "step": 3231
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7346927684237206,
+      "learning_rate": 1.0656401675305213e-05,
+      "loss": 0.8779,
+      "step": 3232
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.790203708689786,
+      "learning_rate": 1.0651454747180814e-05,
+      "loss": 0.8612,
+      "step": 3233
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.570354502414417,
+      "learning_rate": 1.0646507658947107e-05,
+      "loss": 0.7987,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.864626432776611,
+      "learning_rate": 1.0641560411819949e-05,
+      "loss": 0.8281,
+      "step": 3235
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.703277856066407,
+      "learning_rate": 1.0636613007015237e-05,
+      "loss": 0.8933,
+      "step": 3236
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5346887053544522,
+      "learning_rate": 1.0631665445748903e-05,
+      "loss": 0.8432,
+      "step": 3237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6720590127577113,
+      "learning_rate": 1.0626717729236916e-05,
+      "loss": 0.8069,
+      "step": 3238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.696786972098412,
+      "learning_rate": 1.062176985869529e-05,
+      "loss": 0.8801,
+      "step": 3239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.697229729382102,
+      "learning_rate": 1.061682183534007e-05,
+      "loss": 0.9402,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.948636608691428,
+      "learning_rate": 1.0611873660387342e-05,
+      "loss": 0.8494,
+      "step": 3241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8446098507092406,
+      "learning_rate": 1.0606925335053227e-05,
+      "loss": 0.8413,
+      "step": 3242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.728881060896302,
+      "learning_rate": 1.060197686055389e-05,
+      "loss": 0.9709,
+      "step": 3243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7895855154963085,
+      "learning_rate": 1.0597028238105524e-05,
+      "loss": 0.7959,
+      "step": 3244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6499197839840463,
+      "learning_rate": 1.0592079468924359e-05,
+      "loss": 0.7914,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.911507768646488,
+      "learning_rate": 1.0587130554226665e-05,
+      "loss": 0.9737,
+      "step": 3246
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.900046966280705,
+      "learning_rate": 1.0582181495228751e-05,
+      "loss": 0.9709,
+      "step": 3247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.590115515221143,
+      "learning_rate": 1.0577232293146951e-05,
+      "loss": 0.9294,
+      "step": 3248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5526063817207385,
+      "learning_rate": 1.0572282949197646e-05,
+      "loss": 0.8423,
+      "step": 3249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.53629469601658,
+      "learning_rate": 1.0567333464597238e-05,
+      "loss": 0.9556,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.980707695112363,
+      "learning_rate": 1.0562383840562179e-05,
+      "loss": 0.9552,
+      "step": 3251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.495283589020335,
+      "learning_rate": 1.0557434078308941e-05,
+      "loss": 0.7907,
+      "step": 3252
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.518120801724238,
+      "learning_rate": 1.0552484179054041e-05,
+      "loss": 0.8885,
+      "step": 3253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7614692883101126,
+      "learning_rate": 1.0547534144014027e-05,
+      "loss": 0.8517,
+      "step": 3254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8019226736605187,
+      "learning_rate": 1.0542583974405476e-05,
+      "loss": 0.9192,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.79675855043032,
+      "learning_rate": 1.0537633671445002e-05,
+      "loss": 0.8881,
+      "step": 3256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.265933060436887,
+      "learning_rate": 1.0532683236349248e-05,
+      "loss": 0.7393,
+      "step": 3257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.526929479088268,
+      "learning_rate": 1.0527732670334897e-05,
+      "loss": 0.8796,
+      "step": 3258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.520864599696683,
+      "learning_rate": 1.0522781974618652e-05,
+      "loss": 0.8469,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7165796819432604,
+      "learning_rate": 1.0517831150417264e-05,
+      "loss": 0.8224,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6270000648683722,
+      "learning_rate": 1.0512880198947501e-05,
+      "loss": 0.938,
+      "step": 3261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.79444465784981,
+      "learning_rate": 1.050792912142617e-05,
+      "loss": 0.7442,
+      "step": 3262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5324615943017106,
+      "learning_rate": 1.0502977919070106e-05,
+      "loss": 0.8856,
+      "step": 3263
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.533648583070449,
+      "learning_rate": 1.0498026593096174e-05,
+      "loss": 0.8138,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.483789225555165,
+      "learning_rate": 1.0493075144721274e-05,
+      "loss": 0.7558,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8208179904872566,
+      "learning_rate": 1.0488123575162332e-05,
+      "loss": 0.8044,
+      "step": 3266
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.618894893257103,
+      "learning_rate": 1.0483171885636307e-05,
+      "loss": 0.8982,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8409111924604673,
+      "learning_rate": 1.0478220077360184e-05,
+      "loss": 0.7764,
+      "step": 3268
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6090076484711986,
+      "learning_rate": 1.0473268151550977e-05,
+      "loss": 0.9362,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9134663249316786,
+      "learning_rate": 1.0468316109425732e-05,
+      "loss": 0.8487,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5389919301057446,
+      "learning_rate": 1.046336395220152e-05,
+      "loss": 0.8479,
+      "step": 3271
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4828818279246376,
+      "learning_rate": 1.0458411681095444e-05,
+      "loss": 0.8475,
+      "step": 3272
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8052594263529405,
+      "learning_rate": 1.0453459297324638e-05,
+      "loss": 0.7965,
+      "step": 3273
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.018704108526814,
+      "learning_rate": 1.0448506802106248e-05,
+      "loss": 1.0924,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7527598038819567,
+      "learning_rate": 1.0443554196657468e-05,
+      "loss": 0.908,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1159386946264753,
+      "learning_rate": 1.0438601482195507e-05,
+      "loss": 0.9319,
+      "step": 3276
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.705038646126299,
+      "learning_rate": 1.0433648659937604e-05,
+      "loss": 0.813,
+      "step": 3277
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5206348247401427,
+      "learning_rate": 1.042869573110102e-05,
+      "loss": 0.7875,
+      "step": 3278
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7917041949509933,
+      "learning_rate": 1.0423742696903047e-05,
+      "loss": 0.9462,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6746082283942822,
+      "learning_rate": 1.0418789558561009e-05,
+      "loss": 0.9436,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5656588357619428,
+      "learning_rate": 1.0413836317292237e-05,
+      "loss": 0.9148,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6118866244099777,
+      "learning_rate": 1.0408882974314107e-05,
+      "loss": 0.9189,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6684671051310556,
+      "learning_rate": 1.040392953084401e-05,
+      "loss": 0.8033,
+      "step": 3283
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.501298470832651,
+      "learning_rate": 1.0398975988099364e-05,
+      "loss": 0.8298,
+      "step": 3284
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5541503907350083,
+      "learning_rate": 1.0394022347297607e-05,
+      "loss": 0.8786,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5594560871040892,
+      "learning_rate": 1.038906860965621e-05,
+      "loss": 0.8968,
+      "step": 3286
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.5151461823220593,
+      "learning_rate": 1.038411477639266e-05,
+      "loss": 0.9096,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.636485823466242,
+      "learning_rate": 1.037916084872447e-05,
+      "loss": 0.8021,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.72865389979779,
+      "learning_rate": 1.0374206827869177e-05,
+      "loss": 0.907,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.818026224530828,
+      "learning_rate": 1.0369252715044343e-05,
+      "loss": 0.8863,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.7254428761840406,
+      "learning_rate": 1.0364298511467548e-05,
+      "loss": 0.88,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.728241616850513,
+      "learning_rate": 1.0359344218356393e-05,
+      "loss": 0.8195,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8484344836013813,
+      "learning_rate": 1.0354389836928507e-05,
+      "loss": 0.9687,
+      "step": 3293
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.096231522927287,
+      "learning_rate": 1.0349435368401541e-05,
+      "loss": 0.8959,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6238474886319425,
+      "learning_rate": 1.0344480813993163e-05,
+      "loss": 0.8663,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.785385869115943,
+      "learning_rate": 1.033952617492106e-05,
+      "loss": 0.9265,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4841009237509755,
+      "learning_rate": 1.0334571452402943e-05,
+      "loss": 0.8481,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6596124256370652,
+      "learning_rate": 1.032961664765655e-05,
+      "loss": 0.8183,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4437143863687227,
+      "learning_rate": 1.0324661761899629e-05,
+      "loss": 0.8173,
+      "step": 3299
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.29109457912766,
+      "learning_rate": 1.0319706796349954e-05,
+      "loss": 0.9644,
+      "step": 3300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.797304793482052,
+      "learning_rate": 1.0314751752225311e-05,
+      "loss": 0.8918,
+      "step": 3301
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7597018438887027,
+      "learning_rate": 1.0309796630743518e-05,
+      "loss": 0.819,
+      "step": 3302
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3860829479408974,
+      "learning_rate": 1.0304841433122399e-05,
+      "loss": 0.9725,
+      "step": 3303
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6324701321499036,
+      "learning_rate": 1.0299886160579806e-05,
+      "loss": 0.8367,
+      "step": 3304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.501387059647868,
+      "learning_rate": 1.0294930814333605e-05,
+      "loss": 0.7974,
+      "step": 3305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.525680580078367,
+      "learning_rate": 1.0289975395601686e-05,
+      "loss": 0.7649,
+      "step": 3306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.686467602375458,
+      "learning_rate": 1.0285019905601943e-05,
+      "loss": 0.8886,
+      "step": 3307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1918554685564837,
+      "learning_rate": 1.02800643455523e-05,
+      "loss": 0.7807,
+      "step": 3308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.636561913205046,
+      "learning_rate": 1.0275108716670698e-05,
+      "loss": 0.7962,
+      "step": 3309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.719254863507505,
+      "learning_rate": 1.0270153020175092e-05,
+      "loss": 1.0061,
+      "step": 3310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.801691612655321,
+      "learning_rate": 1.0265197257283444e-05,
+      "loss": 0.8736,
+      "step": 3311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.590671703849368,
+      "learning_rate": 1.0260241429213754e-05,
+      "loss": 0.9129,
+      "step": 3312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8841881720350333,
+      "learning_rate": 1.0255285537184016e-05,
+      "loss": 0.8225,
+      "step": 3313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4984456373558555,
+      "learning_rate": 1.0250329582412253e-05,
+      "loss": 0.8641,
+      "step": 3314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.596338224358061,
+      "learning_rate": 1.0245373566116496e-05,
+      "loss": 0.8932,
+      "step": 3315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7568003722125938,
+      "learning_rate": 1.0240417489514802e-05,
+      "loss": 0.8756,
+      "step": 3316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.866798445477397,
+      "learning_rate": 1.0235461353825234e-05,
+      "loss": 0.8654,
+      "step": 3317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5433943971852275,
+      "learning_rate": 1.0230505160265867e-05,
+      "loss": 0.8851,
+      "step": 3318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9431924246520564,
+      "learning_rate": 1.0225548910054794e-05,
+      "loss": 0.8239,
+      "step": 3319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.745923605151662,
+      "learning_rate": 1.0220592604410127e-05,
+      "loss": 0.8627,
+      "step": 3320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.277014146147613,
+      "learning_rate": 1.0215636244549985e-05,
+      "loss": 0.9909,
+      "step": 3321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9053879625025125,
+      "learning_rate": 1.02106798316925e-05,
+      "loss": 0.8434,
+      "step": 3322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.584237482121651,
+      "learning_rate": 1.0205723367055821e-05,
+      "loss": 0.842,
+      "step": 3323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8707112582385177,
+      "learning_rate": 1.0200766851858112e-05,
+      "loss": 0.935,
+      "step": 3324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.447159641817951,
+      "learning_rate": 1.0195810287317539e-05,
+      "loss": 0.8213,
+      "step": 3325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.651398982692408,
+      "learning_rate": 1.0190853674652289e-05,
+      "loss": 0.9345,
+      "step": 3326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.411584643818777,
+      "learning_rate": 1.0185897015080555e-05,
+      "loss": 0.8791,
+      "step": 3327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.728514157055375,
+      "learning_rate": 1.0180940309820553e-05,
+      "loss": 0.8527,
+      "step": 3328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.3547298609445724,
+      "learning_rate": 1.0175983560090496e-05,
+      "loss": 0.8094,
+      "step": 3329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.439730352013072,
+      "learning_rate": 1.0171026767108617e-05,
+      "loss": 0.7593,
+      "step": 3330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0534143692632862,
+      "learning_rate": 1.0166069932093152e-05,
+      "loss": 0.8036,
+      "step": 3331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.683965074654162,
+      "learning_rate": 1.016111305626236e-05,
+      "loss": 0.8081,
+      "step": 3332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.57505920815486,
+      "learning_rate": 1.0156156140834492e-05,
+      "loss": 0.753,
+      "step": 3333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.283608619024065,
+      "learning_rate": 1.0151199187027828e-05,
+      "loss": 0.9922,
+      "step": 3334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7688976110505235,
+      "learning_rate": 1.0146242196060646e-05,
+      "loss": 0.8464,
+      "step": 3335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.68577175813711,
+      "learning_rate": 1.0141285169151229e-05,
+      "loss": 0.7952,
+      "step": 3336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.562544645505176,
+      "learning_rate": 1.0136328107517881e-05,
+      "loss": 0.9107,
+      "step": 3337
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6052512511645904,
+      "learning_rate": 1.0131371012378907e-05,
+      "loss": 0.7702,
+      "step": 3338
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0419214898098548,
+      "learning_rate": 1.0126413884952626e-05,
+      "loss": 0.7645,
+      "step": 3339
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.671771314701333,
+      "learning_rate": 1.0121456726457357e-05,
+      "loss": 0.8947,
+      "step": 3340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.650762861555099,
+      "learning_rate": 1.0116499538111428e-05,
+      "loss": 0.8634,
+      "step": 3341
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.634411025008088,
+      "learning_rate": 1.0111542321133182e-05,
+      "loss": 0.8083,
+      "step": 3342
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.41310547661801,
+      "learning_rate": 1.010658507674096e-05,
+      "loss": 0.9767,
+      "step": 3343
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.568605596401079,
+      "learning_rate": 1.0101627806153117e-05,
+      "loss": 0.8827,
+      "step": 3344
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7319944186757406,
+      "learning_rate": 1.0096670510588009e-05,
+      "loss": 0.8224,
+      "step": 3345
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.283551340145881,
+      "learning_rate": 1.0091713191264001e-05,
+      "loss": 0.7559,
+      "step": 3346
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.012626492995786,
+      "learning_rate": 1.0086755849399464e-05,
+      "loss": 0.986,
+      "step": 3347
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.686251388531931,
+      "learning_rate": 1.008179848621277e-05,
+      "loss": 0.8781,
+      "step": 3348
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.4624808391304565,
+      "learning_rate": 1.0076841102922301e-05,
+      "loss": 0.8124,
+      "step": 3349
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6461052882685365,
+      "learning_rate": 1.0071883700746448e-05,
+      "loss": 0.8288,
+      "step": 3350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.712048135899838,
+      "learning_rate": 1.0066926280903598e-05,
+      "loss": 0.8369,
+      "step": 3351
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5896984185944394,
+      "learning_rate": 1.0061968844612143e-05,
+      "loss": 0.8532,
+      "step": 3352
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8039264113702904,
+      "learning_rate": 1.0057011393090481e-05,
+      "loss": 0.8418,
+      "step": 3353
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.930467198579045,
+      "learning_rate": 1.0052053927557022e-05,
+      "loss": 0.8305,
+      "step": 3354
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.02410853561508,
+      "learning_rate": 1.0047096449230164e-05,
+      "loss": 0.8893,
+      "step": 3355
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6144880865805216,
+      "learning_rate": 1.0042138959328322e-05,
+      "loss": 0.7572,
+      "step": 3356
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7977719436946544,
+      "learning_rate": 1.0037181459069905e-05,
+      "loss": 0.8333,
+      "step": 3357
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5403140670311344,
+      "learning_rate": 1.003222394967333e-05,
+      "loss": 0.7677,
+      "step": 3358
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6608801644930726,
+      "learning_rate": 1.0027266432357007e-05,
+      "loss": 0.924,
+      "step": 3359
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.704651179422154,
+      "learning_rate": 1.0022308908339365e-05,
+      "loss": 0.7022,
+      "step": 3360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.679460239629454,
+      "learning_rate": 1.0017351378838817e-05,
+      "loss": 0.8915,
+      "step": 3361
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.498793390857166,
+      "learning_rate": 1.0012393845073787e-05,
+      "loss": 0.7512,
+      "step": 3362
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.6430015156102837,
+      "learning_rate": 1.0007436308262696e-05,
+      "loss": 0.8217,
+      "step": 3363
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.735772542706487,
+      "learning_rate": 1.000247876962397e-05,
+      "loss": 0.8334,
+      "step": 3364
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.515510191667527,
+      "learning_rate": 9.997521230376032e-06,
+      "loss": 0.8312,
+      "step": 3365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8658668233468925,
+      "learning_rate": 9.992563691737304e-06,
+      "loss": 0.9067,
+      "step": 3366
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.837582352635219,
+      "learning_rate": 9.987606154926214e-06,
+      "loss": 0.7849,
+      "step": 3367
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7723264928964144,
+      "learning_rate": 9.982648621161188e-06,
+      "loss": 0.8556,
+      "step": 3368
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7701920524217827,
+      "learning_rate": 9.977691091660637e-06,
+      "loss": 0.7989,
+      "step": 3369
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7434900183215976,
+      "learning_rate": 9.972733567642994e-06,
+      "loss": 0.8515,
+      "step": 3370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8262306070225014,
+      "learning_rate": 9.967776050326675e-06,
+      "loss": 0.768,
+      "step": 3371
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.630803167653262,
+      "learning_rate": 9.962818540930095e-06,
+      "loss": 0.8,
+      "step": 3372
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7316271034593185,
+      "learning_rate": 9.95786104067168e-06,
+      "loss": 0.8565,
+      "step": 3373
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5927251961593463,
+      "learning_rate": 9.952903550769837e-06,
+      "loss": 0.843,
+      "step": 3374
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.682418900705461,
+      "learning_rate": 9.947946072442982e-06,
+      "loss": 0.9163,
+      "step": 3375
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4891777647767173,
+      "learning_rate": 9.942988606909522e-06,
+      "loss": 0.7926,
+      "step": 3376
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7278232479343947,
+      "learning_rate": 9.938031155387859e-06,
+      "loss": 0.8517,
+      "step": 3377
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6922445372108488,
+      "learning_rate": 9.933073719096406e-06,
+      "loss": 0.916,
+      "step": 3378
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.562382246736542,
+      "learning_rate": 9.928116299253553e-06,
+      "loss": 0.8839,
+      "step": 3379
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4537029707439313,
+      "learning_rate": 9.923158897077699e-06,
+      "loss": 0.8054,
+      "step": 3380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.3236126145506613,
+      "learning_rate": 9.918201513787233e-06,
+      "loss": 0.7271,
+      "step": 3381
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.055213047874377,
+      "learning_rate": 9.913244150600541e-06,
+      "loss": 0.9046,
+      "step": 3382
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.553739553514491,
+      "learning_rate": 9.908286808735999e-06,
+      "loss": 0.8065,
+      "step": 3383
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4245662109666766,
+      "learning_rate": 9.903329489411993e-06,
+      "loss": 0.8129,
+      "step": 3384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.660906783044096,
+      "learning_rate": 9.898372193846887e-06,
+      "loss": 0.8541,
+      "step": 3385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.578729689514895,
+      "learning_rate": 9.893414923259042e-06,
+      "loss": 0.8421,
+      "step": 3386
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5078486683767416,
+      "learning_rate": 9.888457678866823e-06,
+      "loss": 0.8087,
+      "step": 3387
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.521637554958834,
+      "learning_rate": 9.883500461888573e-06,
+      "loss": 0.8557,
+      "step": 3388
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6795766617286967,
+      "learning_rate": 9.878543273542648e-06,
+      "loss": 0.7181,
+      "step": 3389
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.498689771309547,
+      "learning_rate": 9.873586115047377e-06,
+      "loss": 0.8819,
+      "step": 3390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.577184398412776,
+      "learning_rate": 9.868628987621095e-06,
+      "loss": 0.8566,
+      "step": 3391
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.793311417654129,
+      "learning_rate": 9.863671892482122e-06,
+      "loss": 0.8169,
+      "step": 3392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.620997389101896,
+      "learning_rate": 9.858714830848776e-06,
+      "loss": 0.9171,
+      "step": 3393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.778207542389812,
+      "learning_rate": 9.853757803939358e-06,
+      "loss": 0.8429,
+      "step": 3394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9562677095301786,
+      "learning_rate": 9.848800812972175e-06,
+      "loss": 0.8696,
+      "step": 3395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.599642093006926,
+      "learning_rate": 9.84384385916551e-06,
+      "loss": 0.7961,
+      "step": 3396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0071841120945737,
+      "learning_rate": 9.838886943737645e-06,
+      "loss": 0.8398,
+      "step": 3397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.696431412561743,
+      "learning_rate": 9.833930067906851e-06,
+      "loss": 0.8396,
+      "step": 3398
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.430701029582184,
+      "learning_rate": 9.828973232891385e-06,
+      "loss": 0.7366,
+      "step": 3399
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8167533231043294,
+      "learning_rate": 9.824016439909505e-06,
+      "loss": 0.8484,
+      "step": 3400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.687160062145975,
+      "learning_rate": 9.81905969017945e-06,
+      "loss": 0.8009,
+      "step": 3401
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4484114257757734,
+      "learning_rate": 9.814102984919445e-06,
+      "loss": 0.7761,
+      "step": 3402
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6496492282970485,
+      "learning_rate": 9.809146325347716e-06,
+      "loss": 0.7687,
+      "step": 3403
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.864677266471541,
+      "learning_rate": 9.804189712682466e-06,
+      "loss": 1.023,
+      "step": 3404
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7989202592935984,
+      "learning_rate": 9.79923314814189e-06,
+      "loss": 0.8775,
+      "step": 3405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.586169913513614,
+      "learning_rate": 9.79427663294418e-06,
+      "loss": 0.7965,
+      "step": 3406
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8333347061539187,
+      "learning_rate": 9.7893201683075e-06,
+      "loss": 0.9712,
+      "step": 3407
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.719996695889003,
+      "learning_rate": 9.784363755450018e-06,
+      "loss": 0.9177,
+      "step": 3408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4773553077282444,
+      "learning_rate": 9.779407395589876e-06,
+      "loss": 0.7608,
+      "step": 3409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5475236772585563,
+      "learning_rate": 9.774451089945206e-06,
+      "loss": 0.7846,
+      "step": 3410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0188556226064445,
+      "learning_rate": 9.769494839734136e-06,
+      "loss": 0.8426,
+      "step": 3411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.254714206692356,
+      "learning_rate": 9.764538646174771e-06,
+      "loss": 0.9804,
+      "step": 3412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.827330030766981,
+      "learning_rate": 9.759582510485198e-06,
+      "loss": 0.857,
+      "step": 3413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4734544365088404,
+      "learning_rate": 9.754626433883506e-06,
+      "loss": 0.8474,
+      "step": 3414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.704908642099613,
+      "learning_rate": 9.749670417587753e-06,
+      "loss": 0.9602,
+      "step": 3415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6888669635440077,
+      "learning_rate": 9.744714462815987e-06,
+      "loss": 0.8605,
+      "step": 3416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.709281617887562,
+      "learning_rate": 9.739758570786251e-06,
+      "loss": 0.8596,
+      "step": 3417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7199414147145395,
+      "learning_rate": 9.734802742716556e-06,
+      "loss": 0.7955,
+      "step": 3418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.509186081480435,
+      "learning_rate": 9.729846979824913e-06,
+      "loss": 0.8612,
+      "step": 3419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9166514156562298,
+      "learning_rate": 9.724891283329305e-06,
+      "loss": 0.9359,
+      "step": 3420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.496906473025068,
+      "learning_rate": 9.7199356544477e-06,
+      "loss": 0.7673,
+      "step": 3421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5023038128266286,
+      "learning_rate": 9.714980094398059e-06,
+      "loss": 0.8332,
+      "step": 3422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.75657816335714,
+      "learning_rate": 9.710024604398317e-06,
+      "loss": 0.8968,
+      "step": 3423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5561294321392594,
+      "learning_rate": 9.705069185666396e-06,
+      "loss": 0.7787,
+      "step": 3424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1037707547457294,
+      "learning_rate": 9.700113839420197e-06,
+      "loss": 0.8882,
+      "step": 3425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4952606751734274,
+      "learning_rate": 9.695158566877606e-06,
+      "loss": 0.9268,
+      "step": 3426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.026747964805295,
+      "learning_rate": 9.690203369256486e-06,
+      "loss": 0.9835,
+      "step": 3427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.5480816986475294,
+      "learning_rate": 9.685248247774692e-06,
+      "loss": 0.7946,
+      "step": 3428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.604296116527149,
+      "learning_rate": 9.68029320365005e-06,
+      "loss": 0.8441,
+      "step": 3429
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.510085849702647,
+      "learning_rate": 9.675338238100375e-06,
+      "loss": 0.7499,
+      "step": 3430
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9373262344980984,
+      "learning_rate": 9.670383352343454e-06,
+      "loss": 0.9004,
+      "step": 3431
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6469345532827955,
+      "learning_rate": 9.665428547597057e-06,
+      "loss": 0.8273,
+      "step": 3432
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.761930481958474,
+      "learning_rate": 9.660473825078944e-06,
+      "loss": 0.819,
+      "step": 3433
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4498718953719414,
+      "learning_rate": 9.655519186006842e-06,
+      "loss": 0.7963,
+      "step": 3434
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.786495910453177,
+      "learning_rate": 9.65056463159846e-06,
+      "loss": 0.913,
+      "step": 3435
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.53808454970735,
+      "learning_rate": 9.645610163071495e-06,
+      "loss": 0.82,
+      "step": 3436
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5015219428230586,
+      "learning_rate": 9.640655781643612e-06,
+      "loss": 0.7981,
+      "step": 3437
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4680609160616767,
+      "learning_rate": 9.635701488532455e-06,
+      "loss": 0.7857,
+      "step": 3438
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.990846500890265,
+      "learning_rate": 9.63074728495566e-06,
+      "loss": 0.9372,
+      "step": 3439
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9446715414792775,
+      "learning_rate": 9.625793172130825e-06,
+      "loss": 0.7747,
+      "step": 3440
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.483984216486534,
+      "learning_rate": 9.620839151275534e-06,
+      "loss": 0.8843,
+      "step": 3441
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7169965735793693,
+      "learning_rate": 9.615885223607345e-06,
+      "loss": 0.8942,
+      "step": 3442
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.544915450697256,
+      "learning_rate": 9.610931390343792e-06,
+      "loss": 0.8571,
+      "step": 3443
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6189570146790913,
+      "learning_rate": 9.605977652702394e-06,
+      "loss": 0.8119,
+      "step": 3444
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.771286909908849,
+      "learning_rate": 9.60102401190064e-06,
+      "loss": 0.8558,
+      "step": 3445
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.627183298351705,
+      "learning_rate": 9.596070469155992e-06,
+      "loss": 0.8693,
+      "step": 3446
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.602025072939709,
+      "learning_rate": 9.591117025685897e-06,
+      "loss": 0.8599,
+      "step": 3447
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7213601358474806,
+      "learning_rate": 9.586163682707768e-06,
+      "loss": 0.754,
+      "step": 3448
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7410310874423023,
+      "learning_rate": 9.581210441438994e-06,
+      "loss": 0.8447,
+      "step": 3449
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.44157798300347,
+      "learning_rate": 9.576257303096955e-06,
+      "loss": 0.8248,
+      "step": 3450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6606986031824222,
+      "learning_rate": 9.571304268898983e-06,
+      "loss": 0.99,
+      "step": 3451
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.703515781558748,
+      "learning_rate": 9.566351340062401e-06,
+      "loss": 0.7711,
+      "step": 3452
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.727709716865192,
+      "learning_rate": 9.561398517804498e-06,
+      "loss": 0.7937,
+      "step": 3453
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.634863441280792,
+      "learning_rate": 9.556445803342532e-06,
+      "loss": 0.836,
+      "step": 3454
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.148624164260505,
+      "learning_rate": 9.551493197893755e-06,
+      "loss": 0.9873,
+      "step": 3455
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8217903157217106,
+      "learning_rate": 9.546540702675369e-06,
+      "loss": 0.841,
+      "step": 3456
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.667756818733098,
+      "learning_rate": 9.541588318904558e-06,
+      "loss": 0.8399,
+      "step": 3457
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.8915142027475165,
+      "learning_rate": 9.536636047798484e-06,
+      "loss": 0.8674,
+      "step": 3458
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5080794197051826,
+      "learning_rate": 9.531683890574275e-06,
+      "loss": 0.7672,
+      "step": 3459
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8576179806465003,
+      "learning_rate": 9.526731848449025e-06,
+      "loss": 0.8458,
+      "step": 3460
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7890894029086764,
+      "learning_rate": 9.52177992263982e-06,
+      "loss": 0.8804,
+      "step": 3461
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6054412774153697,
+      "learning_rate": 9.516828114363695e-06,
+      "loss": 0.7928,
+      "step": 3462
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.510783295864765,
+      "learning_rate": 9.51187642483767e-06,
+      "loss": 0.9286,
+      "step": 3463
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.600019031689135,
+      "learning_rate": 9.50692485527873e-06,
+      "loss": 0.7392,
+      "step": 3464
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6241823466708336,
+      "learning_rate": 9.501973406903827e-06,
+      "loss": 0.8606,
+      "step": 3465
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.190637646959537,
+      "learning_rate": 9.497022080929898e-06,
+      "loss": 0.9424,
+      "step": 3466
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.564623714774583,
+      "learning_rate": 9.492070878573835e-06,
+      "loss": 0.8429,
+      "step": 3467
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.668379840486614,
+      "learning_rate": 9.487119801052502e-06,
+      "loss": 0.8477,
+      "step": 3468
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.569726689880901,
+      "learning_rate": 9.48216884958274e-06,
+      "loss": 0.7003,
+      "step": 3469
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5774518981436936,
+      "learning_rate": 9.477218025381351e-06,
+      "loss": 0.7733,
+      "step": 3470
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8184807477848963,
+      "learning_rate": 9.472267329665107e-06,
+      "loss": 0.7704,
+      "step": 3471
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.489332514227551,
+      "learning_rate": 9.467316763650755e-06,
+      "loss": 0.8022,
+      "step": 3472
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.741348982843612,
+      "learning_rate": 9.462366328555e-06,
+      "loss": 0.8633,
+      "step": 3473
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8905647730408908,
+      "learning_rate": 9.457416025594528e-06,
+      "loss": 0.9357,
+      "step": 3474
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.951366616887644,
+      "learning_rate": 9.452465855985978e-06,
+      "loss": 0.8145,
+      "step": 3475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4285873693290116,
+      "learning_rate": 9.447515820945959e-06,
+      "loss": 0.7857,
+      "step": 3476
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.815876716660041,
+      "learning_rate": 9.442565921691062e-06,
+      "loss": 0.7452,
+      "step": 3477
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5491022711639046,
+      "learning_rate": 9.437616159437828e-06,
+      "loss": 0.8215,
+      "step": 3478
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.5599892020044384,
+      "learning_rate": 9.432666535402764e-06,
+      "loss": 0.7299,
+      "step": 3479
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7040859098807526,
+      "learning_rate": 9.427717050802359e-06,
+      "loss": 0.811,
+      "step": 3480
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9344148037173694,
+      "learning_rate": 9.42276770685305e-06,
+      "loss": 0.9246,
+      "step": 3481
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.906786834299944,
+      "learning_rate": 9.41781850477125e-06,
+      "loss": 0.9404,
+      "step": 3482
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.566431527013668,
+      "learning_rate": 9.412869445773338e-06,
+      "loss": 0.8971,
+      "step": 3483
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8506725057017905,
+      "learning_rate": 9.407920531075641e-06,
+      "loss": 0.8048,
+      "step": 3484
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.215499929654718,
+      "learning_rate": 9.40297176189448e-06,
+      "loss": 0.9818,
+      "step": 3485
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8119702972456717,
+      "learning_rate": 9.398023139446113e-06,
+      "loss": 0.8954,
+      "step": 3486
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8743448207774582,
+      "learning_rate": 9.393074664946773e-06,
+      "loss": 0.9275,
+      "step": 3487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6264244278472773,
+      "learning_rate": 9.388126339612661e-06,
+      "loss": 0.8055,
+      "step": 3488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.786429666020273,
+      "learning_rate": 9.383178164659935e-06,
+      "loss": 0.7912,
+      "step": 3489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.96515087237739,
+      "learning_rate": 9.378230141304711e-06,
+      "loss": 0.9173,
+      "step": 3490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.685269953318423,
+      "learning_rate": 9.373282270763087e-06,
+      "loss": 0.8394,
+      "step": 3491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8070326631129587,
+      "learning_rate": 9.368334554251099e-06,
+      "loss": 0.9085,
+      "step": 3492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.79038954283755,
+      "learning_rate": 9.363386992984765e-06,
+      "loss": 0.8801,
+      "step": 3493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9430808628170833,
+      "learning_rate": 9.358439588180053e-06,
+      "loss": 0.8405,
+      "step": 3494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6887706037474057,
+      "learning_rate": 9.353492341052894e-06,
+      "loss": 0.7708,
+      "step": 3495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.506287273436747,
+      "learning_rate": 9.34854525281919e-06,
+      "loss": 0.8898,
+      "step": 3496
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.798243828822919,
+      "learning_rate": 9.34359832469479e-06,
+      "loss": 0.8658,
+      "step": 3497
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8401541321532227,
+      "learning_rate": 9.338651557895513e-06,
+      "loss": 0.8637,
+      "step": 3498
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4414548026642198,
+      "learning_rate": 9.333704953637135e-06,
+      "loss": 0.8658,
+      "step": 3499
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6143105878878985,
+      "learning_rate": 9.328758513135393e-06,
+      "loss": 0.7415,
+      "step": 3500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.509734906422458,
+      "learning_rate": 9.323812237605977e-06,
+      "loss": 0.8148,
+      "step": 3501
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.55941722209386,
+      "learning_rate": 9.318866128264556e-06,
+      "loss": 0.7698,
+      "step": 3502
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6536660992690133,
+      "learning_rate": 9.313920186326734e-06,
+      "loss": 0.7173,
+      "step": 3503
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5321252514402866,
+      "learning_rate": 9.30897441300809e-06,
+      "loss": 0.8794,
+      "step": 3504
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9748037974758677,
+      "learning_rate": 9.304028809524154e-06,
+      "loss": 0.9203,
+      "step": 3505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.852371008150815,
+      "learning_rate": 9.299083377090415e-06,
+      "loss": 0.811,
+      "step": 3506
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8329209503983774,
+      "learning_rate": 9.294138116922328e-06,
+      "loss": 0.8725,
+      "step": 3507
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.556917873688633,
+      "learning_rate": 9.289193030235293e-06,
+      "loss": 0.8832,
+      "step": 3508
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7490077769808483,
+      "learning_rate": 9.284248118244676e-06,
+      "loss": 0.8685,
+      "step": 3509
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.821699231197627,
+      "learning_rate": 9.2793033821658e-06,
+      "loss": 0.8822,
+      "step": 3510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.62568208135973,
+      "learning_rate": 9.274358823213938e-06,
+      "loss": 0.8229,
+      "step": 3511
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3221816368186157,
+      "learning_rate": 9.269414442604324e-06,
+      "loss": 0.9029,
+      "step": 3512
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5729726401385427,
+      "learning_rate": 9.264470241552152e-06,
+      "loss": 0.7594,
+      "step": 3513
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.613962094609536,
+      "learning_rate": 9.259526221272564e-06,
+      "loss": 0.8595,
+      "step": 3514
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7485831111727084,
+      "learning_rate": 9.254582382980667e-06,
+      "loss": 0.7865,
+      "step": 3515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.622037998037098,
+      "learning_rate": 9.249638727891513e-06,
+      "loss": 0.9312,
+      "step": 3516
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4585507034727847,
+      "learning_rate": 9.24469525722011e-06,
+      "loss": 0.7725,
+      "step": 3517
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.491230089556136,
+      "learning_rate": 9.239751972181435e-06,
+      "loss": 0.8587,
+      "step": 3518
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6129819112402437,
+      "learning_rate": 9.234808873990405e-06,
+      "loss": 0.8502,
+      "step": 3519
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5048705499962733,
+      "learning_rate": 9.229865963861888e-06,
+      "loss": 0.7968,
+      "step": 3520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6486800302992606,
+      "learning_rate": 9.224923243010722e-06,
+      "loss": 0.8275,
+      "step": 3521
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7180029217975923,
+      "learning_rate": 9.219980712651684e-06,
+      "loss": 0.7334,
+      "step": 3522
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.600823388647472,
+      "learning_rate": 9.215038373999507e-06,
+      "loss": 0.8259,
+      "step": 3523
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.842235272446629,
+      "learning_rate": 9.210096228268885e-06,
+      "loss": 0.8671,
+      "step": 3524
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.692287898296294,
+      "learning_rate": 9.205154276674456e-06,
+      "loss": 0.9264,
+      "step": 3525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.682372073940632,
+      "learning_rate": 9.200212520430814e-06,
+      "loss": 0.8054,
+      "step": 3526
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.928798676688956,
+      "learning_rate": 9.195270960752505e-06,
+      "loss": 0.998,
+      "step": 3527
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5156680957421016,
+      "learning_rate": 9.19032959885402e-06,
+      "loss": 0.7873,
+      "step": 3528
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.643942704387546,
+      "learning_rate": 9.185388435949815e-06,
+      "loss": 0.7548,
+      "step": 3529
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.65169439065476,
+      "learning_rate": 9.180447473254289e-06,
+      "loss": 0.8838,
+      "step": 3530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6752622485002098,
+      "learning_rate": 9.175506711981782e-06,
+      "loss": 0.8896,
+      "step": 3531
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7589924845247444,
+      "learning_rate": 9.170566153346606e-06,
+      "loss": 0.8725,
+      "step": 3532
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.506078803726627,
+      "learning_rate": 9.165625798563007e-06,
+      "loss": 0.7382,
+      "step": 3533
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.53520520564263,
+      "learning_rate": 9.160685648845182e-06,
+      "loss": 0.8571,
+      "step": 3534
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.607123422142902,
+      "learning_rate": 9.155745705407288e-06,
+      "loss": 0.9119,
+      "step": 3535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5822010527861434,
+      "learning_rate": 9.15080596946342e-06,
+      "loss": 0.7983,
+      "step": 3536
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5685810562247813,
+      "learning_rate": 9.145866442227632e-06,
+      "loss": 0.8518,
+      "step": 3537
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.535089245635708,
+      "learning_rate": 9.140927124913915e-06,
+      "loss": 0.6993,
+      "step": 3538
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7257069900543813,
+      "learning_rate": 9.135988018736214e-06,
+      "loss": 0.8749,
+      "step": 3539
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.711857668052274,
+      "learning_rate": 9.13104912490843e-06,
+      "loss": 0.833,
+      "step": 3540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5288441724978767,
+      "learning_rate": 9.1261104446444e-06,
+      "loss": 0.806,
+      "step": 3541
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9774721456587407,
+      "learning_rate": 9.121171979157912e-06,
+      "loss": 0.9085,
+      "step": 3542
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5387614093810194,
+      "learning_rate": 9.116233729662705e-06,
+      "loss": 0.784,
+      "step": 3543
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9467484789677543,
+      "learning_rate": 9.111295697372463e-06,
+      "loss": 0.9633,
+      "step": 3544
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9757437755654497,
+      "learning_rate": 9.106357883500808e-06,
+      "loss": 0.8032,
+      "step": 3545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5998054749396022,
+      "learning_rate": 9.101420289261327e-06,
+      "loss": 0.8322,
+      "step": 3546
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7217799386307995,
+      "learning_rate": 9.096482915867535e-06,
+      "loss": 0.9016,
+      "step": 3547
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7475901842215342,
+      "learning_rate": 9.091545764532905e-06,
+      "loss": 0.8422,
+      "step": 3548
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.904223424761576,
+      "learning_rate": 9.086608836470847e-06,
+      "loss": 0.8544,
+      "step": 3549
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8385256278455,
+      "learning_rate": 9.081672132894716e-06,
+      "loss": 0.8998,
+      "step": 3550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.538160224033064,
+      "learning_rate": 9.076735655017822e-06,
+      "loss": 1.0688,
+      "step": 3551
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6527168081088623,
+      "learning_rate": 9.071799404053412e-06,
+      "loss": 0.8223,
+      "step": 3552
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9676891853864023,
+      "learning_rate": 9.066863381214672e-06,
+      "loss": 0.8399,
+      "step": 3553
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.685845974353761,
+      "learning_rate": 9.061927587714747e-06,
+      "loss": 0.9156,
+      "step": 3554
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.724001222133191,
+      "learning_rate": 9.056992024766706e-06,
+      "loss": 0.8723,
+      "step": 3555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.4897993153457327,
+      "learning_rate": 9.05205669358358e-06,
+      "loss": 0.7244,
+      "step": 3556
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.6163874330228825,
+      "learning_rate": 9.047121595378335e-06,
+      "loss": 0.9575,
+      "step": 3557
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.471137288002054,
+      "learning_rate": 9.042186731363876e-06,
+      "loss": 0.8606,
+      "step": 3558
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8295381734389005,
+      "learning_rate": 9.037252102753056e-06,
+      "loss": 0.866,
+      "step": 3559
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5518594589083032,
+      "learning_rate": 9.032317710758668e-06,
+      "loss": 0.7769,
+      "step": 3560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6965298501742114,
+      "learning_rate": 9.027383556593443e-06,
+      "loss": 0.818,
+      "step": 3561
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.536115670826026,
+      "learning_rate": 9.022449641470066e-06,
+      "loss": 0.7124,
+      "step": 3562
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.588450203478518,
+      "learning_rate": 9.017515966601152e-06,
+      "loss": 0.972,
+      "step": 3563
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7909023955013934,
+      "learning_rate": 9.012582533199254e-06,
+      "loss": 0.9889,
+      "step": 3564
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6482008686874416,
+      "learning_rate": 9.00764934247688e-06,
+      "loss": 0.8659,
+      "step": 3565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8874019494064798,
+      "learning_rate": 9.002716395646462e-06,
+      "loss": 0.8602,
+      "step": 3566
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6378105278298722,
+      "learning_rate": 8.997783693920387e-06,
+      "loss": 0.8591,
+      "step": 3567
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6620268667766327,
+      "learning_rate": 8.992851238510972e-06,
+      "loss": 0.8449,
+      "step": 3568
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.633541484655814,
+      "learning_rate": 8.987919030630474e-06,
+      "loss": 0.8495,
+      "step": 3569
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.397040476582493,
+      "learning_rate": 8.982987071491097e-06,
+      "loss": 0.8061,
+      "step": 3570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5219829545734362,
+      "learning_rate": 8.978055362304974e-06,
+      "loss": 0.8393,
+      "step": 3571
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.586610187878668,
+      "learning_rate": 8.973123904284175e-06,
+      "loss": 0.7636,
+      "step": 3572
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7685865321112684,
+      "learning_rate": 8.968192698640728e-06,
+      "loss": 0.8546,
+      "step": 3573
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8152562626440734,
+      "learning_rate": 8.963261746586576e-06,
+      "loss": 0.8472,
+      "step": 3574
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.470179452309494,
+      "learning_rate": 8.958331049333608e-06,
+      "loss": 0.8264,
+      "step": 3575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5236760418062016,
+      "learning_rate": 8.953400608093655e-06,
+      "loss": 0.8112,
+      "step": 3576
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1871526471365152,
+      "learning_rate": 8.948470424078477e-06,
+      "loss": 0.8764,
+      "step": 3577
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.390788855245002,
+      "learning_rate": 8.94354049849978e-06,
+      "loss": 1.0057,
+      "step": 3578
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7210093186803923,
+      "learning_rate": 8.9386108325692e-06,
+      "loss": 0.9305,
+      "step": 3579
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9855834421533496,
+      "learning_rate": 8.933681427498308e-06,
+      "loss": 0.9511,
+      "step": 3580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7666441197229914,
+      "learning_rate": 8.928752284498616e-06,
+      "loss": 0.8362,
+      "step": 3581
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.751965748661896,
+      "learning_rate": 8.923823404781569e-06,
+      "loss": 0.8808,
+      "step": 3582
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.576050038453813,
+      "learning_rate": 8.91889478955854e-06,
+      "loss": 0.7936,
+      "step": 3583
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8137578824285945,
+      "learning_rate": 8.913966440040858e-06,
+      "loss": 0.8241,
+      "step": 3584
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6778875907500406,
+      "learning_rate": 8.909038357439767e-06,
+      "loss": 0.7843,
+      "step": 3585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.668733297014157,
+      "learning_rate": 8.904110542966446e-06,
+      "loss": 0.7914,
+      "step": 3586
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6923499450062893,
+      "learning_rate": 8.89918299783202e-06,
+      "loss": 0.8458,
+      "step": 3587
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.632803767921514,
+      "learning_rate": 8.894255723247536e-06,
+      "loss": 0.8983,
+      "step": 3588
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5401591385553277,
+      "learning_rate": 8.88932872042399e-06,
+      "loss": 0.7862,
+      "step": 3589
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7332741048875278,
+      "learning_rate": 8.884401990572293e-06,
+      "loss": 0.9888,
+      "step": 3590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7439162578949094,
+      "learning_rate": 8.879475534903292e-06,
+      "loss": 0.8091,
+      "step": 3591
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2872806339100715,
+      "learning_rate": 8.874549354627786e-06,
+      "loss": 0.9352,
+      "step": 3592
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7685688420362307,
+      "learning_rate": 8.869623450956484e-06,
+      "loss": 0.8611,
+      "step": 3593
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.736941628903649,
+      "learning_rate": 8.86469782510003e-06,
+      "loss": 0.7929,
+      "step": 3594
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.58569671282389,
+      "learning_rate": 8.859772478269013e-06,
+      "loss": 0.8211,
+      "step": 3595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.025888336572811,
+      "learning_rate": 8.854847411673944e-06,
+      "loss": 0.8808,
+      "step": 3596
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.774503519864151,
+      "learning_rate": 8.849922626525258e-06,
+      "loss": 0.8471,
+      "step": 3597
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.652018012156596,
+      "learning_rate": 8.844998124033339e-06,
+      "loss": 0.8309,
+      "step": 3598
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7737368141511816,
+      "learning_rate": 8.840073905408488e-06,
+      "loss": 0.797,
+      "step": 3599
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.631941056407336,
+      "learning_rate": 8.83514997186094e-06,
+      "loss": 0.7881,
+      "step": 3600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2951018055617,
+      "learning_rate": 8.83022632460086e-06,
+      "loss": 0.808,
+      "step": 3601
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.754968213283891,
+      "learning_rate": 8.825302964838337e-06,
+      "loss": 0.8158,
+      "step": 3602
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7676551447009947,
+      "learning_rate": 8.820379893783406e-06,
+      "loss": 0.7856,
+      "step": 3603
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3558550948620947,
+      "learning_rate": 8.815457112646012e-06,
+      "loss": 0.8038,
+      "step": 3604
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9201439057338137,
+      "learning_rate": 8.810534622636035e-06,
+      "loss": 0.8423,
+      "step": 3605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.485876965646072,
+      "learning_rate": 8.805612424963293e-06,
+      "loss": 0.9396,
+      "step": 3606
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.05468597846747,
+      "learning_rate": 8.800690520837516e-06,
+      "loss": 0.8926,
+      "step": 3607
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6860278468379533,
+      "learning_rate": 8.79576891146837e-06,
+      "loss": 0.8396,
+      "step": 3608
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.4171544778269194,
+      "learning_rate": 8.790847598065457e-06,
+      "loss": 0.8286,
+      "step": 3609
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.84210486687714,
+      "learning_rate": 8.785926581838288e-06,
+      "loss": 0.9437,
+      "step": 3610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.015621670490844,
+      "learning_rate": 8.781005863996318e-06,
+      "loss": 0.9002,
+      "step": 3611
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6779526247346066,
+      "learning_rate": 8.77608544574892e-06,
+      "loss": 0.8747,
+      "step": 3612
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.650536560154308,
+      "learning_rate": 8.771165328305387e-06,
+      "loss": 0.8957,
+      "step": 3613
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6917203323536785,
+      "learning_rate": 8.766245512874959e-06,
+      "loss": 0.886,
+      "step": 3614
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6663991698263834,
+      "learning_rate": 8.76132600066678e-06,
+      "loss": 0.7626,
+      "step": 3615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.620435279245936,
+      "learning_rate": 8.75640679288993e-06,
+      "loss": 0.7998,
+      "step": 3616
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.506585621530046,
+      "learning_rate": 8.751487890753414e-06,
+      "loss": 0.8461,
+      "step": 3617
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.427248465159573,
+      "learning_rate": 8.746569295466158e-06,
+      "loss": 0.854,
+      "step": 3618
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.886733992698023,
+      "learning_rate": 8.741651008237012e-06,
+      "loss": 0.9957,
+      "step": 3619
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5762624317434146,
+      "learning_rate": 8.73673303027476e-06,
+      "loss": 0.893,
+      "step": 3620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0155415002945727,
+      "learning_rate": 8.731815362788097e-06,
+      "loss": 0.8848,
+      "step": 3621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.077168323117745,
+      "learning_rate": 8.72689800698565e-06,
+      "loss": 0.8263,
+      "step": 3622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7555288356242706,
+      "learning_rate": 8.721980964075971e-06,
+      "loss": 0.7863,
+      "step": 3623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5669626707976776,
+      "learning_rate": 8.717064235267523e-06,
+      "loss": 0.8099,
+      "step": 3624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9058792524652364,
+      "learning_rate": 8.712147821768708e-06,
+      "loss": 0.8589,
+      "step": 3625
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5965076505939697,
+      "learning_rate": 8.70723172478784e-06,
+      "loss": 0.8579,
+      "step": 3626
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8274905558414374,
+      "learning_rate": 8.702315945533156e-06,
+      "loss": 0.8621,
+      "step": 3627
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.630215743420575,
+      "learning_rate": 8.697400485212816e-06,
+      "loss": 0.9039,
+      "step": 3628
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6932115439415667,
+      "learning_rate": 8.692485345034903e-06,
+      "loss": 0.8897,
+      "step": 3629
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9440595153618623,
+      "learning_rate": 8.687570526207425e-06,
+      "loss": 0.8714,
+      "step": 3630
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.603288212077659,
+      "learning_rate": 8.682656029938304e-06,
+      "loss": 0.8508,
+      "step": 3631
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.677093356005969,
+      "learning_rate": 8.67774185743538e-06,
+      "loss": 0.8141,
+      "step": 3632
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3662069721673933,
+      "learning_rate": 8.672828009906427e-06,
+      "loss": 0.8724,
+      "step": 3633
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.53071609090638,
+      "learning_rate": 8.667914488559128e-06,
+      "loss": 0.8556,
+      "step": 3634
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7326135410952004,
+      "learning_rate": 8.663001294601082e-06,
+      "loss": 0.789,
+      "step": 3635
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.10132310788602,
+      "learning_rate": 8.658088429239826e-06,
+      "loss": 1.0455,
+      "step": 3636
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.87349936884276,
+      "learning_rate": 8.653175893682798e-06,
+      "loss": 0.8204,
+      "step": 3637
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.763286521179671,
+      "learning_rate": 8.648263689137361e-06,
+      "loss": 0.8436,
+      "step": 3638
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6273730030188003,
+      "learning_rate": 8.643351816810798e-06,
+      "loss": 0.7951,
+      "step": 3639
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.917254214247303,
+      "learning_rate": 8.638440277910308e-06,
+      "loss": 0.8058,
+      "step": 3640
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3002616303651755,
+      "learning_rate": 8.633529073643015e-06,
+      "loss": 0.7621,
+      "step": 3641
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.413059539401685,
+      "learning_rate": 8.628618205215952e-06,
+      "loss": 0.8709,
+      "step": 3642
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5414190420762237,
+      "learning_rate": 8.62370767383607e-06,
+      "loss": 0.8097,
+      "step": 3643
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.481571621621111,
+      "learning_rate": 8.618797480710244e-06,
+      "loss": 0.8251,
+      "step": 3644
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5172550773810687,
+      "learning_rate": 8.613887627045259e-06,
+      "loss": 0.9102,
+      "step": 3645
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.670664180733619,
+      "learning_rate": 8.608978114047818e-06,
+      "loss": 0.8178,
+      "step": 3646
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.606139471469613,
+      "learning_rate": 8.604068942924546e-06,
+      "loss": 0.8516,
+      "step": 3647
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.82664466148121,
+      "learning_rate": 8.599160114881979e-06,
+      "loss": 0.9226,
+      "step": 3648
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.523005606895564,
+      "learning_rate": 8.594251631126566e-06,
+      "loss": 0.8452,
+      "step": 3649
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.5205110535151665,
+      "learning_rate": 8.589343492864677e-06,
+      "loss": 0.9898,
+      "step": 3650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9766710161932304,
+      "learning_rate": 8.58443570130259e-06,
+      "loss": 0.7819,
+      "step": 3651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6300474432151995,
+      "learning_rate": 8.579528257646512e-06,
+      "loss": 0.8238,
+      "step": 3652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6781706303509085,
+      "learning_rate": 8.574621163102548e-06,
+      "loss": 0.8303,
+      "step": 3653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8478976294309364,
+      "learning_rate": 8.569714418876726e-06,
+      "loss": 0.8185,
+      "step": 3654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.675348904783975,
+      "learning_rate": 8.564808026174987e-06,
+      "loss": 0.8694,
+      "step": 3655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.645107465382493,
+      "learning_rate": 8.559901986203183e-06,
+      "loss": 0.8213,
+      "step": 3656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7247460217826753,
+      "learning_rate": 8.554996300167078e-06,
+      "loss": 0.7998,
+      "step": 3657
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5975898264618977,
+      "learning_rate": 8.55009096927236e-06,
+      "loss": 0.8683,
+      "step": 3658
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.768347542686713,
+      "learning_rate": 8.545185994724619e-06,
+      "loss": 0.8646,
+      "step": 3659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.524499498032838,
+      "learning_rate": 8.540281377729355e-06,
+      "loss": 0.9274,
+      "step": 3660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5152003835677985,
+      "learning_rate": 8.535377119491992e-06,
+      "loss": 0.7848,
+      "step": 3661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 11.956498767697864,
+      "learning_rate": 8.53047322121785e-06,
+      "loss": 0.9453,
+      "step": 3662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6256844278947464,
+      "learning_rate": 8.525569684112181e-06,
+      "loss": 0.6936,
+      "step": 3663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.80453341383359,
+      "learning_rate": 8.520666509380132e-06,
+      "loss": 0.9192,
+      "step": 3664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6025782439130096,
+      "learning_rate": 8.515763698226764e-06,
+      "loss": 0.9011,
+      "step": 3665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6216848423824293,
+      "learning_rate": 8.510861251857053e-06,
+      "loss": 0.7981,
+      "step": 3666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.526482661731994,
+      "learning_rate": 8.505959171475883e-06,
+      "loss": 0.7283,
+      "step": 3667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.840409081076734,
+      "learning_rate": 8.501057458288041e-06,
+      "loss": 0.8971,
+      "step": 3668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6405701512694044,
+      "learning_rate": 8.496156113498242e-06,
+      "loss": 0.8458,
+      "step": 3669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.647772783988904,
+      "learning_rate": 8.491255138311093e-06,
+      "loss": 0.8854,
+      "step": 3670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5549512183080996,
+      "learning_rate": 8.486354533931114e-06,
+      "loss": 0.7682,
+      "step": 3671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.539104496539669,
+      "learning_rate": 8.481454301562741e-06,
+      "loss": 0.7779,
+      "step": 3672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0747522525924325,
+      "learning_rate": 8.476554442410308e-06,
+      "loss": 0.8221,
+      "step": 3673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.833827224853693,
+      "learning_rate": 8.47165495767807e-06,
+      "loss": 0.8259,
+      "step": 3674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.3111799161124615,
+      "learning_rate": 8.46675584857018e-06,
+      "loss": 0.9928,
+      "step": 3675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6904811253228873,
+      "learning_rate": 8.461857116290698e-06,
+      "loss": 0.8276,
+      "step": 3676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.789560243636047,
+      "learning_rate": 8.456958762043602e-06,
+      "loss": 0.7791,
+      "step": 3677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7711953502100717,
+      "learning_rate": 8.452060787032762e-06,
+      "loss": 0.9128,
+      "step": 3678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7597117529170894,
+      "learning_rate": 8.447163192461967e-06,
+      "loss": 0.7911,
+      "step": 3679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.758978197325863,
+      "learning_rate": 8.44226597953491e-06,
+      "loss": 0.8669,
+      "step": 3680
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.561596737247688,
+      "learning_rate": 8.437369149455188e-06,
+      "loss": 0.8433,
+      "step": 3681
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.7707246140287785,
+      "learning_rate": 8.432472703426301e-06,
+      "loss": 1.0088,
+      "step": 3682
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5503950734906424,
+      "learning_rate": 8.427576642651664e-06,
+      "loss": 0.8231,
+      "step": 3683
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6727623912158047,
+      "learning_rate": 8.422680968334583e-06,
+      "loss": 0.8313,
+      "step": 3684
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.737454750508601,
+      "learning_rate": 8.417785681678286e-06,
+      "loss": 0.8674,
+      "step": 3685
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6719477754679586,
+      "learning_rate": 8.412890783885894e-06,
+      "loss": 0.9723,
+      "step": 3686
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.642320424254053,
+      "learning_rate": 8.40799627616043e-06,
+      "loss": 0.8246,
+      "step": 3687
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.709922300257362,
+      "learning_rate": 8.40310215970484e-06,
+      "loss": 0.8561,
+      "step": 3688
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.666770069476004,
+      "learning_rate": 8.398208435721948e-06,
+      "loss": 0.7958,
+      "step": 3689
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7526148496402754,
+      "learning_rate": 8.393315105414496e-06,
+      "loss": 0.8502,
+      "step": 3690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.6659138491305767,
+      "learning_rate": 8.388422169985133e-06,
+      "loss": 0.7236,
+      "step": 3691
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5439661145398333,
+      "learning_rate": 8.383529630636403e-06,
+      "loss": 0.7353,
+      "step": 3692
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.724577989752663,
+      "learning_rate": 8.378637488570753e-06,
+      "loss": 0.8559,
+      "step": 3693
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5940339598271605,
+      "learning_rate": 8.373745744990536e-06,
+      "loss": 0.9183,
+      "step": 3694
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4445109767260633,
+      "learning_rate": 8.368854401098001e-06,
+      "loss": 0.8698,
+      "step": 3695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5851633527893316,
+      "learning_rate": 8.363963458095312e-06,
+      "loss": 0.9235,
+      "step": 3696
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.58465589391059,
+      "learning_rate": 8.35907291718452e-06,
+      "loss": 0.8486,
+      "step": 3697
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4882994084186425,
+      "learning_rate": 8.35418277956758e-06,
+      "loss": 0.7139,
+      "step": 3698
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8000076559062457,
+      "learning_rate": 8.34929304644636e-06,
+      "loss": 0.8835,
+      "step": 3699
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.770086107774786,
+      "learning_rate": 8.34440371902261e-06,
+      "loss": 0.8634,
+      "step": 3700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4079058389435675,
+      "learning_rate": 8.33951479849799e-06,
+      "loss": 0.7542,
+      "step": 3701
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.848969415151501,
+      "learning_rate": 8.334626286074068e-06,
+      "loss": 0.8225,
+      "step": 3702
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7592930972688183,
+      "learning_rate": 8.329738182952294e-06,
+      "loss": 0.8484,
+      "step": 3703
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6074480161638633,
+      "learning_rate": 8.32485049033403e-06,
+      "loss": 0.7946,
+      "step": 3704
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7716119180727934,
+      "learning_rate": 8.319963209420538e-06,
+      "loss": 0.944,
+      "step": 3705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8029282681731065,
+      "learning_rate": 8.315076341412966e-06,
+      "loss": 0.7907,
+      "step": 3706
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6892629916864452,
+      "learning_rate": 8.310189887512376e-06,
+      "loss": 0.9146,
+      "step": 3707
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6734084490871024,
+      "learning_rate": 8.305303848919721e-06,
+      "loss": 0.7844,
+      "step": 3708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.539063798010986,
+      "learning_rate": 8.300418226835845e-06,
+      "loss": 0.8068,
+      "step": 3709
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9846429225063944,
+      "learning_rate": 8.295533022461509e-06,
+      "loss": 0.9278,
+      "step": 3710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6602592369524305,
+      "learning_rate": 8.29064823699735e-06,
+      "loss": 0.8223,
+      "step": 3711
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.416254247890502,
+      "learning_rate": 8.285763871643913e-06,
+      "loss": 0.9225,
+      "step": 3712
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9016898319075652,
+      "learning_rate": 8.28087992760164e-06,
+      "loss": 0.8301,
+      "step": 3713
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7287937427591302,
+      "learning_rate": 8.275996406070862e-06,
+      "loss": 0.8018,
+      "step": 3714
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5050043115033396,
+      "learning_rate": 8.27111330825182e-06,
+      "loss": 0.8171,
+      "step": 3715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6532678696623444,
+      "learning_rate": 8.266230635344639e-06,
+      "loss": 0.8022,
+      "step": 3716
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.561582659958214,
+      "learning_rate": 8.261348388549339e-06,
+      "loss": 0.8913,
+      "step": 3717
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.179169065907958,
+      "learning_rate": 8.256466569065848e-06,
+      "loss": 0.8957,
+      "step": 3718
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7464665614377495,
+      "learning_rate": 8.251585178093973e-06,
+      "loss": 0.869,
+      "step": 3719
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5740469266074557,
+      "learning_rate": 8.24670421683342e-06,
+      "loss": 0.7636,
+      "step": 3720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.625686248280815,
+      "learning_rate": 8.241823686483803e-06,
+      "loss": 0.9046,
+      "step": 3721
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.662924210522219,
+      "learning_rate": 8.236943588244614e-06,
+      "loss": 0.7854,
+      "step": 3722
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.002836458193394,
+      "learning_rate": 8.232063923315241e-06,
+      "loss": 0.7868,
+      "step": 3723
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.045339883343718,
+      "learning_rate": 8.227184692894975e-06,
+      "loss": 0.7397,
+      "step": 3724
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.628061506312594,
+      "learning_rate": 8.222305898182988e-06,
+      "loss": 0.9184,
+      "step": 3725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.741478080492712,
+      "learning_rate": 8.217427540378356e-06,
+      "loss": 0.7048,
+      "step": 3726
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.432295343245831,
+      "learning_rate": 8.212549620680042e-06,
+      "loss": 0.8698,
+      "step": 3727
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6309359477330854,
+      "learning_rate": 8.207672140286897e-06,
+      "loss": 0.801,
+      "step": 3728
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5958707826515677,
+      "learning_rate": 8.202795100397675e-06,
+      "loss": 0.8624,
+      "step": 3729
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.463907049171007,
+      "learning_rate": 8.197918502211012e-06,
+      "loss": 0.7411,
+      "step": 3730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7184611770503855,
+      "learning_rate": 8.193042346925436e-06,
+      "loss": 0.8144,
+      "step": 3731
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.597451160371377,
+      "learning_rate": 8.188166635739378e-06,
+      "loss": 0.8072,
+      "step": 3732
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4920600266531894,
+      "learning_rate": 8.183291369851147e-06,
+      "loss": 0.7994,
+      "step": 3733
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.836702439590227,
+      "learning_rate": 8.178416550458942e-06,
+      "loss": 0.8462,
+      "step": 3734
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4212101587552,
+      "learning_rate": 8.173542178760863e-06,
+      "loss": 0.853,
+      "step": 3735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.477224609844448,
+      "learning_rate": 8.168668255954887e-06,
+      "loss": 0.8548,
+      "step": 3736
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.78458703917749,
+      "learning_rate": 8.163794783238896e-06,
+      "loss": 0.8783,
+      "step": 3737
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.066389736990901,
+      "learning_rate": 8.158921761810652e-06,
+      "loss": 0.9896,
+      "step": 3738
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6268166088866214,
+      "learning_rate": 8.154049192867799e-06,
+      "loss": 0.8734,
+      "step": 3739
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6637608901708685,
+      "learning_rate": 8.149177077607885e-06,
+      "loss": 0.9008,
+      "step": 3740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8074823775308575,
+      "learning_rate": 8.144305417228338e-06,
+      "loss": 0.9087,
+      "step": 3741
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5991639822857464,
+      "learning_rate": 8.13943421292647e-06,
+      "loss": 0.8284,
+      "step": 3742
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3865845077404044,
+      "learning_rate": 8.134563465899494e-06,
+      "loss": 0.5926,
+      "step": 3743
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0143104439623296,
+      "learning_rate": 8.129693177344501e-06,
+      "loss": 0.7634,
+      "step": 3744
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.311530155343317,
+      "learning_rate": 8.124823348458469e-06,
+      "loss": 0.7781,
+      "step": 3745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8112785294356226,
+      "learning_rate": 8.119953980438268e-06,
+      "loss": 0.8852,
+      "step": 3746
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.7930743638010975,
+      "learning_rate": 8.115085074480646e-06,
+      "loss": 0.9101,
+      "step": 3747
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4819200007879494,
+      "learning_rate": 8.110216631782252e-06,
+      "loss": 0.7507,
+      "step": 3748
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.5773649983789944,
+      "learning_rate": 8.10534865353961e-06,
+      "loss": 0.9403,
+      "step": 3749
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9531625641283825,
+      "learning_rate": 8.10048114094913e-06,
+      "loss": 0.8998,
+      "step": 3750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3980166625417954,
+      "learning_rate": 8.095614095207114e-06,
+      "loss": 0.8242,
+      "step": 3751
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6056645544160584,
+      "learning_rate": 8.09074751750974e-06,
+      "loss": 0.798,
+      "step": 3752
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.463826485778903,
+      "learning_rate": 8.085881409053077e-06,
+      "loss": 0.7636,
+      "step": 3753
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9734517602914123,
+      "learning_rate": 8.081015771033084e-06,
+      "loss": 0.6716,
+      "step": 3754
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6689993987748344,
+      "learning_rate": 8.076150604645596e-06,
+      "loss": 0.7986,
+      "step": 3755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8901860718142007,
+      "learning_rate": 8.07128591108633e-06,
+      "loss": 0.9522,
+      "step": 3756
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.691387601662654,
+      "learning_rate": 8.066421691550895e-06,
+      "loss": 0.8825,
+      "step": 3757
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.521901884415322,
+      "learning_rate": 8.061557947234776e-06,
+      "loss": 0.8862,
+      "step": 3758
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.492459607271112,
+      "learning_rate": 8.056694679333352e-06,
+      "loss": 0.7939,
+      "step": 3759
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.661612562114557,
+      "learning_rate": 8.051831889041874e-06,
+      "loss": 0.9144,
+      "step": 3760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.706362129998038,
+      "learning_rate": 8.046969577555476e-06,
+      "loss": 0.8429,
+      "step": 3761
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.472977941116828,
+      "learning_rate": 8.042107746069186e-06,
+      "loss": 0.8602,
+      "step": 3762
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.754451211879156,
+      "learning_rate": 8.037246395777899e-06,
+      "loss": 0.9238,
+      "step": 3763
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.481373688892135,
+      "learning_rate": 8.032385527876395e-06,
+      "loss": 0.8886,
+      "step": 3764
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5611895981930513,
+      "learning_rate": 8.02752514355935e-06,
+      "loss": 0.8657,
+      "step": 3765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8296881650687014,
+      "learning_rate": 8.022665244021305e-06,
+      "loss": 0.8609,
+      "step": 3766
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.661495351257953,
+      "learning_rate": 8.017805830456682e-06,
+      "loss": 0.9261,
+      "step": 3767
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.661809405239875,
+      "learning_rate": 8.012946904059798e-06,
+      "loss": 0.7899,
+      "step": 3768
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.558394657590361,
+      "learning_rate": 8.00808846602483e-06,
+      "loss": 0.8469,
+      "step": 3769
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8327158347851484,
+      "learning_rate": 8.003230517545857e-06,
+      "loss": 0.7845,
+      "step": 3770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4525659365371824,
+      "learning_rate": 7.998373059816822e-06,
+      "loss": 0.7724,
+      "step": 3771
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.603495304223118,
+      "learning_rate": 7.993516094031548e-06,
+      "loss": 0.9337,
+      "step": 3772
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6672418683358936,
+      "learning_rate": 7.988659621383746e-06,
+      "loss": 0.842,
+      "step": 3773
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.877866056470712,
+      "learning_rate": 7.983803643067e-06,
+      "loss": 0.8995,
+      "step": 3774
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6378115325206313,
+      "learning_rate": 7.978948160274767e-06,
+      "loss": 0.8989,
+      "step": 3775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9815579099601783,
+      "learning_rate": 7.974093174200397e-06,
+      "loss": 0.8495,
+      "step": 3776
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.132213870518248,
+      "learning_rate": 7.96923868603711e-06,
+      "loss": 0.8843,
+      "step": 3777
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.412788545789461,
+      "learning_rate": 7.964384696977992e-06,
+      "loss": 0.7961,
+      "step": 3778
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.370076785003066,
+      "learning_rate": 7.959531208216028e-06,
+      "loss": 0.7651,
+      "step": 3779
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5746154436580557,
+      "learning_rate": 7.954678220944064e-06,
+      "loss": 0.878,
+      "step": 3780
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6759180343235527,
+      "learning_rate": 7.949825736354833e-06,
+      "loss": 0.8155,
+      "step": 3781
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6891425243501357,
+      "learning_rate": 7.944973755640937e-06,
+      "loss": 0.9401,
+      "step": 3782
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.516111779906683,
+      "learning_rate": 7.940122279994853e-06,
+      "loss": 0.9423,
+      "step": 3783
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7621339373286307,
+      "learning_rate": 7.935271310608946e-06,
+      "loss": 0.9382,
+      "step": 3784
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.619415577360085,
+      "learning_rate": 7.930420848675442e-06,
+      "loss": 0.8197,
+      "step": 3785
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.245032498999451,
+      "learning_rate": 7.925570895386443e-06,
+      "loss": 0.9284,
+      "step": 3786
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.700494938164556,
+      "learning_rate": 7.920721451933946e-06,
+      "loss": 0.952,
+      "step": 3787
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.534211144265974,
+      "learning_rate": 7.915872519509798e-06,
+      "loss": 0.9005,
+      "step": 3788
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.960644175712867,
+      "learning_rate": 7.911024099305733e-06,
+      "loss": 0.8329,
+      "step": 3789
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.742678404898385,
+      "learning_rate": 7.90617619251336e-06,
+      "loss": 0.8715,
+      "step": 3790
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.627606307215524,
+      "learning_rate": 7.901328800324149e-06,
+      "loss": 0.7791,
+      "step": 3791
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4158599573343573,
+      "learning_rate": 7.896481923929464e-06,
+      "loss": 0.7315,
+      "step": 3792
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4810470882863247,
+      "learning_rate": 7.891635564520527e-06,
+      "loss": 0.7382,
+      "step": 3793
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.510165859368989,
+      "learning_rate": 7.886789723288436e-06,
+      "loss": 0.7769,
+      "step": 3794
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5818045931789033,
+      "learning_rate": 7.881944401424164e-06,
+      "loss": 0.7249,
+      "step": 3795
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7965746120569066,
+      "learning_rate": 7.877099600118556e-06,
+      "loss": 0.8596,
+      "step": 3796
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.868160776338824,
+      "learning_rate": 7.872255320562324e-06,
+      "loss": 0.8343,
+      "step": 3797
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8908538497696297,
+      "learning_rate": 7.867411563946061e-06,
+      "loss": 0.9137,
+      "step": 3798
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.675812256480429,
+      "learning_rate": 7.862568331460224e-06,
+      "loss": 0.8901,
+      "step": 3799
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7616648182460937,
+      "learning_rate": 7.857725624295147e-06,
+      "loss": 0.8093,
+      "step": 3800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7560563302006402,
+      "learning_rate": 7.85288344364103e-06,
+      "loss": 0.7587,
+      "step": 3801
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.902987278257522,
+      "learning_rate": 7.848041790687938e-06,
+      "loss": 0.809,
+      "step": 3802
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9658013555750475,
+      "learning_rate": 7.843200666625827e-06,
+      "loss": 0.9063,
+      "step": 3803
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7647583871656245,
+      "learning_rate": 7.838360072644502e-06,
+      "loss": 0.7302,
+      "step": 3804
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6531480332408846,
+      "learning_rate": 7.833520009933642e-06,
+      "loss": 0.8479,
+      "step": 3805
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4018411002880904,
+      "learning_rate": 7.828680479682804e-06,
+      "loss": 0.7875,
+      "step": 3806
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6037494271984802,
+      "learning_rate": 7.823841483081408e-06,
+      "loss": 0.8453,
+      "step": 3807
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5441498531562448,
+      "learning_rate": 7.81900302131874e-06,
+      "loss": 0.8459,
+      "step": 3808
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4488141970955777,
+      "learning_rate": 7.814165095583963e-06,
+      "loss": 0.7218,
+      "step": 3809
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.697496931767143,
+      "learning_rate": 7.809327707066098e-06,
+      "loss": 0.875,
+      "step": 3810
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7787096557710593,
+      "learning_rate": 7.80449085695405e-06,
+      "loss": 0.9643,
+      "step": 3811
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.997625552294042,
+      "learning_rate": 7.799654546436572e-06,
+      "loss": 0.9041,
+      "step": 3812
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5589065401606454,
+      "learning_rate": 7.79481877670229e-06,
+      "loss": 0.7814,
+      "step": 3813
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7670004774038652,
+      "learning_rate": 7.78998354893971e-06,
+      "loss": 0.7938,
+      "step": 3814
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5168899500141206,
+      "learning_rate": 7.785148864337192e-06,
+      "loss": 0.8931,
+      "step": 3815
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.812391141474085,
+      "learning_rate": 7.780314724082963e-06,
+      "loss": 0.9063,
+      "step": 3816
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.719624742077618,
+      "learning_rate": 7.775481129365126e-06,
+      "loss": 0.7571,
+      "step": 3817
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.641408766865852,
+      "learning_rate": 7.770648081371636e-06,
+      "loss": 0.8086,
+      "step": 3818
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.5413481785443994,
+      "learning_rate": 7.76581558129032e-06,
+      "loss": 0.9068,
+      "step": 3819
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.567326515183025,
+      "learning_rate": 7.760983630308877e-06,
+      "loss": 0.8058,
+      "step": 3820
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.895640457109586,
+      "learning_rate": 7.756152229614858e-06,
+      "loss": 0.7857,
+      "step": 3821
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.705430270171675,
+      "learning_rate": 7.751321380395696e-06,
+      "loss": 0.7997,
+      "step": 3822
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3761036254213566,
+      "learning_rate": 7.74649108383867e-06,
+      "loss": 0.7441,
+      "step": 3823
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7108377611394814,
+      "learning_rate": 7.741661341130931e-06,
+      "loss": 0.8042,
+      "step": 3824
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.76401109437764,
+      "learning_rate": 7.736832153459501e-06,
+      "loss": 0.7993,
+      "step": 3825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6007382840432243,
+      "learning_rate": 7.732003522011255e-06,
+      "loss": 0.9539,
+      "step": 3826
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8750454249231185,
+      "learning_rate": 7.727175447972928e-06,
+      "loss": 0.7493,
+      "step": 3827
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7782634958747026,
+      "learning_rate": 7.72234793253114e-06,
+      "loss": 0.8482,
+      "step": 3828
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4242192072212174,
+      "learning_rate": 7.717520976872347e-06,
+      "loss": 0.7813,
+      "step": 3829
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8572436094490086,
+      "learning_rate": 7.712694582182885e-06,
+      "loss": 0.9205,
+      "step": 3830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8229708482287035,
+      "learning_rate": 7.707868749648945e-06,
+      "loss": 0.8363,
+      "step": 3831
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5074469640257666,
+      "learning_rate": 7.703043480456576e-06,
+      "loss": 0.807,
+      "step": 3832
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6569233219377844,
+      "learning_rate": 7.698218775791704e-06,
+      "loss": 0.7093,
+      "step": 3833
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.63041037066242,
+      "learning_rate": 7.693394636840102e-06,
+      "loss": 0.9002,
+      "step": 3834
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.1206808271085595,
+      "learning_rate": 7.688571064787404e-06,
+      "loss": 0.9588,
+      "step": 3835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6908652608398955,
+      "learning_rate": 7.683748060819116e-06,
+      "loss": 0.853,
+      "step": 3836
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.64015780770857,
+      "learning_rate": 7.67892562612059e-06,
+      "loss": 0.7764,
+      "step": 3837
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.161301991487249,
+      "learning_rate": 7.674103761877045e-06,
+      "loss": 0.8251,
+      "step": 3838
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.627704700289266,
+      "learning_rate": 7.669282469273566e-06,
+      "loss": 0.8344,
+      "step": 3839
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7454896666097164,
+      "learning_rate": 7.664461749495092e-06,
+      "loss": 0.8359,
+      "step": 3840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.714463524406809,
+      "learning_rate": 7.659641603726413e-06,
+      "loss": 0.748,
+      "step": 3841
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.788609811735884,
+      "learning_rate": 7.654822033152192e-06,
+      "loss": 0.8835,
+      "step": 3842
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.72491491081617,
+      "learning_rate": 7.650003038956939e-06,
+      "loss": 0.9077,
+      "step": 3843
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6085472328025125,
+      "learning_rate": 7.645184622325034e-06,
+      "loss": 0.872,
+      "step": 3844
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.813389893202622,
+      "learning_rate": 7.640366784440707e-06,
+      "loss": 0.8672,
+      "step": 3845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.326141636602756,
+      "learning_rate": 7.635549526488042e-06,
+      "loss": 0.7961,
+      "step": 3846
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6334060331516205,
+      "learning_rate": 7.630732849650993e-06,
+      "loss": 0.8555,
+      "step": 3847
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5144279864015444,
+      "learning_rate": 7.625916755113361e-06,
+      "loss": 0.8252,
+      "step": 3848
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.674159071621899,
+      "learning_rate": 7.621101244058803e-06,
+      "loss": 0.8819,
+      "step": 3849
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7674109736473715,
+      "learning_rate": 7.616286317670845e-06,
+      "loss": 0.7824,
+      "step": 3850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7562266107062277,
+      "learning_rate": 7.611471977132855e-06,
+      "loss": 0.888,
+      "step": 3851
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5461987223959257,
+      "learning_rate": 7.606658223628063e-06,
+      "loss": 0.8751,
+      "step": 3852
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4483598091432293,
+      "learning_rate": 7.601845058339557e-06,
+      "loss": 0.8321,
+      "step": 3853
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4859420768204363,
+      "learning_rate": 7.5970324824502736e-06,
+      "loss": 0.8101,
+      "step": 3854
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.020109170851338,
+      "learning_rate": 7.5922204971430165e-06,
+      "loss": 0.8826,
+      "step": 3855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.692579456461489,
+      "learning_rate": 7.587409103600433e-06,
+      "loss": 0.7842,
+      "step": 3856
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8599508800622364,
+      "learning_rate": 7.582598303005026e-06,
+      "loss": 0.8273,
+      "step": 3857
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3605072679986843,
+      "learning_rate": 7.577788096539158e-06,
+      "loss": 0.8066,
+      "step": 3858
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.609277078943832,
+      "learning_rate": 7.572978485385044e-06,
+      "loss": 0.8033,
+      "step": 3859
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.195124199704877,
+      "learning_rate": 7.5681694707247445e-06,
+      "loss": 0.9971,
+      "step": 3860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.482502952468572,
+      "learning_rate": 7.563361053740191e-06,
+      "loss": 0.7986,
+      "step": 3861
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.574590700574185,
+      "learning_rate": 7.558553235613151e-06,
+      "loss": 0.825,
+      "step": 3862
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5775304906123413,
+      "learning_rate": 7.553746017525253e-06,
+      "loss": 0.8978,
+      "step": 3863
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9492898868537,
+      "learning_rate": 7.548939400657977e-06,
+      "loss": 0.8328,
+      "step": 3864
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.71613382329485,
+      "learning_rate": 7.544133386192649e-06,
+      "loss": 0.7306,
+      "step": 3865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7268270443678637,
+      "learning_rate": 7.539327975310463e-06,
+      "loss": 0.8832,
+      "step": 3866
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5607710626551627,
+      "learning_rate": 7.534523169192447e-06,
+      "loss": 0.7932,
+      "step": 3867
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.699562397322179,
+      "learning_rate": 7.529718969019488e-06,
+      "loss": 0.7144,
+      "step": 3868
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.570287154377048,
+      "learning_rate": 7.524915375972327e-06,
+      "loss": 0.9042,
+      "step": 3869
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7691816343881577,
+      "learning_rate": 7.520112391231549e-06,
+      "loss": 0.9343,
+      "step": 3870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6067248356622845,
+      "learning_rate": 7.515310015977591e-06,
+      "loss": 0.7428,
+      "step": 3871
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6809416644963053,
+      "learning_rate": 7.510508251390749e-06,
+      "loss": 0.7999,
+      "step": 3872
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6413952753527137,
+      "learning_rate": 7.505707098651157e-06,
+      "loss": 0.7502,
+      "step": 3873
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3759380993681276,
+      "learning_rate": 7.500906558938806e-06,
+      "loss": 0.7932,
+      "step": 3874
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.605686384220005,
+      "learning_rate": 7.496106633433535e-06,
+      "loss": 0.9078,
+      "step": 3875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4652523513354687,
+      "learning_rate": 7.4913073233150236e-06,
+      "loss": 0.8745,
+      "step": 3876
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7138063748947827,
+      "learning_rate": 7.486508629762818e-06,
+      "loss": 0.848,
+      "step": 3877
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.508036618474075,
+      "learning_rate": 7.481710553956296e-06,
+      "loss": 0.7674,
+      "step": 3878
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.770740350268557,
+      "learning_rate": 7.476913097074692e-06,
+      "loss": 0.8912,
+      "step": 3879
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6922695290496668,
+      "learning_rate": 7.472116260297087e-06,
+      "loss": 0.8769,
+      "step": 3880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.57309161479025,
+      "learning_rate": 7.467320044802409e-06,
+      "loss": 0.796,
+      "step": 3881
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.512958056132519,
+      "learning_rate": 7.462524451769426e-06,
+      "loss": 0.9029,
+      "step": 3882
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.6277879501170536,
+      "learning_rate": 7.457729482376772e-06,
+      "loss": 0.7873,
+      "step": 3883
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.765976568443896,
+      "learning_rate": 7.4529351378029094e-06,
+      "loss": 0.9712,
+      "step": 3884
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4778448264754562,
+      "learning_rate": 7.4481414192261556e-06,
+      "loss": 0.7683,
+      "step": 3885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.607179107319063,
+      "learning_rate": 7.443348327824673e-06,
+      "loss": 0.8287,
+      "step": 3886
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.707889718434232,
+      "learning_rate": 7.438555864776462e-06,
+      "loss": 0.808,
+      "step": 3887
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9849966740053584,
+      "learning_rate": 7.433764031259385e-06,
+      "loss": 0.9122,
+      "step": 3888
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.112460889317166,
+      "learning_rate": 7.428972828451138e-06,
+      "loss": 0.7092,
+      "step": 3889
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7988184084042227,
+      "learning_rate": 7.424182257529258e-06,
+      "loss": 0.8475,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7428963787567104,
+      "learning_rate": 7.419392319671142e-06,
+      "loss": 0.7922,
+      "step": 3891
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.100613249299371,
+      "learning_rate": 7.414603016054016e-06,
+      "loss": 0.968,
+      "step": 3892
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6502811402975506,
+      "learning_rate": 7.4098143478549555e-06,
+      "loss": 0.9379,
+      "step": 3893
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6669734031481216,
+      "learning_rate": 7.405026316250887e-06,
+      "loss": 0.7728,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.606265333083008,
+      "learning_rate": 7.40023892241857e-06,
+      "loss": 0.7656,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4323996483300623,
+      "learning_rate": 7.395452167534615e-06,
+      "loss": 0.8158,
+      "step": 3896
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.649985215785405,
+      "learning_rate": 7.39066605277547e-06,
+      "loss": 0.8632,
+      "step": 3897
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6220263882415966,
+      "learning_rate": 7.385880579317424e-06,
+      "loss": 0.7514,
+      "step": 3898
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9149752455005213,
+      "learning_rate": 7.381095748336618e-06,
+      "loss": 0.7993,
+      "step": 3899
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6720254849645264,
+      "learning_rate": 7.3763115610090284e-06,
+      "loss": 0.9737,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.431364392785707,
+      "learning_rate": 7.371528018510472e-06,
+      "loss": 0.84,
+      "step": 3901
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6650297510276664,
+      "learning_rate": 7.3667451220166104e-06,
+      "loss": 0.791,
+      "step": 3902
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.452822733776375,
+      "learning_rate": 7.361962872702947e-06,
+      "loss": 0.7925,
+      "step": 3903
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.377872769290526,
+      "learning_rate": 7.357181271744818e-06,
+      "loss": 0.767,
+      "step": 3904
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7052139163110382,
+      "learning_rate": 7.352400320317417e-06,
+      "loss": 0.8582,
+      "step": 3905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.551578609697552,
+      "learning_rate": 7.347620019595758e-06,
+      "loss": 0.8198,
+      "step": 3906
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.491058196271869,
+      "learning_rate": 7.342840370754714e-06,
+      "loss": 0.7494,
+      "step": 3907
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3744720682912193,
+      "learning_rate": 7.338061374968984e-06,
+      "loss": 0.7923,
+      "step": 3908
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.498751007169318,
+      "learning_rate": 7.333283033413106e-06,
+      "loss": 0.863,
+      "step": 3909
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6822308660893808,
+      "learning_rate": 7.328505347261471e-06,
+      "loss": 0.8063,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.701553992183195,
+      "learning_rate": 7.323728317688296e-06,
+      "loss": 0.7296,
+      "step": 3911
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6067250965264033,
+      "learning_rate": 7.318951945867643e-06,
+      "loss": 0.8595,
+      "step": 3912
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.646105279440585,
+      "learning_rate": 7.314176232973409e-06,
+      "loss": 0.8204,
+      "step": 3913
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.472573274215426,
+      "learning_rate": 7.30940118017933e-06,
+      "loss": 0.7515,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.889346213895278,
+      "learning_rate": 7.304626788658976e-06,
+      "loss": 0.8521,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8143134429258736,
+      "learning_rate": 7.299853059585768e-06,
+      "loss": 0.8683,
+      "step": 3916
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.359919074066636,
+      "learning_rate": 7.295079994132947e-06,
+      "loss": 0.9095,
+      "step": 3917
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3986782808370823,
+      "learning_rate": 7.290307593473603e-06,
+      "loss": 0.7398,
+      "step": 3918
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7038658777691236,
+      "learning_rate": 7.2855358587806554e-06,
+      "loss": 0.8169,
+      "step": 3919
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5069386708867407,
+      "learning_rate": 7.28076479122686e-06,
+      "loss": 0.8501,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.517053657035505,
+      "learning_rate": 7.27599439198482e-06,
+      "loss": 0.7897,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8360269490317953,
+      "learning_rate": 7.2712246622269615e-06,
+      "loss": 0.8144,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.9258561023362706,
+      "learning_rate": 7.266455603125546e-06,
+      "loss": 0.8818,
+      "step": 3923
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6528039617127424,
+      "learning_rate": 7.261687215852681e-06,
+      "loss": 0.7645,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.935343617366732,
+      "learning_rate": 7.2569195015803e-06,
+      "loss": 0.8296,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4898009977984827,
+      "learning_rate": 7.252152461480166e-06,
+      "loss": 0.7976,
+      "step": 3926
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.576448374761486,
+      "learning_rate": 7.247386096723899e-06,
+      "loss": 0.9198,
+      "step": 3927
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.387055416933314,
+      "learning_rate": 7.2426204084829225e-06,
+      "loss": 0.7474,
+      "step": 3928
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6562445306220006,
+      "learning_rate": 7.237855397928522e-06,
+      "loss": 0.939,
+      "step": 3929
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4856046326947054,
+      "learning_rate": 7.233091066231796e-06,
+      "loss": 0.7783,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5341068263640323,
+      "learning_rate": 7.228327414563683e-06,
+      "loss": 0.8302,
+      "step": 3931
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6776321393981934,
+      "learning_rate": 7.223564444094959e-06,
+      "loss": 0.7617,
+      "step": 3932
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6597687482328407,
+      "learning_rate": 7.218802155996227e-06,
+      "loss": 0.8918,
+      "step": 3933
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0271167434689525,
+      "learning_rate": 7.214040551437919e-06,
+      "loss": 0.9506,
+      "step": 3934
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5823896739606136,
+      "learning_rate": 7.209279631590313e-06,
+      "loss": 0.7003,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.748517455966857,
+      "learning_rate": 7.204519397623502e-06,
+      "loss": 0.8306,
+      "step": 3936
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.612978903513092,
+      "learning_rate": 7.199759850707424e-06,
+      "loss": 0.7482,
+      "step": 3937
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6809929088899294,
+      "learning_rate": 7.195000992011838e-06,
+      "loss": 0.7985,
+      "step": 3938
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5825987471777534,
+      "learning_rate": 7.190242822706335e-06,
+      "loss": 0.7688,
+      "step": 3939
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5087445915228046,
+      "learning_rate": 7.185485343960348e-06,
+      "loss": 0.7756,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.6229532585798303,
+      "learning_rate": 7.180728556943128e-06,
+      "loss": 0.8737,
+      "step": 3941
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8277366555711994,
+      "learning_rate": 7.175972462823757e-06,
+      "loss": 0.739,
+      "step": 3942
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6299431073851434,
+      "learning_rate": 7.171217062771153e-06,
+      "loss": 0.899,
+      "step": 3943
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7129768358506614,
+      "learning_rate": 7.16646235795406e-06,
+      "loss": 0.8835,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5883181574655696,
+      "learning_rate": 7.161708349541044e-06,
+      "loss": 0.8523,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5426023684011723,
+      "learning_rate": 7.156955038700519e-06,
+      "loss": 0.7801,
+      "step": 3946
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0473689365131404,
+      "learning_rate": 7.1522024266007065e-06,
+      "loss": 0.793,
+      "step": 3947
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7815288430966674,
+      "learning_rate": 7.14745051440967e-06,
+      "loss": 0.9378,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.978236151092738,
+      "learning_rate": 7.1426993032952926e-06,
+      "loss": 0.9328,
+      "step": 3949
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7000413945650714,
+      "learning_rate": 7.137948794425289e-06,
+      "loss": 0.8611,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.851811931723268,
+      "learning_rate": 7.133198988967205e-06,
+      "loss": 0.835,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.785889191122522,
+      "learning_rate": 7.128449888088409e-06,
+      "loss": 0.7782,
+      "step": 3952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7193290149705818,
+      "learning_rate": 7.12370149295609e-06,
+      "loss": 0.8046,
+      "step": 3953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.643197344893586,
+      "learning_rate": 7.118953804737278e-06,
+      "loss": 0.888,
+      "step": 3954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.64274182536652,
+      "learning_rate": 7.1142068245988216e-06,
+      "loss": 0.8475,
+      "step": 3955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.509853365664716,
+      "learning_rate": 7.109460553707386e-06,
+      "loss": 0.8149,
+      "step": 3956
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.403526176543108,
+      "learning_rate": 7.104714993229482e-06,
+      "loss": 0.6655,
+      "step": 3957
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7434615498012906,
+      "learning_rate": 7.0999701443314295e-06,
+      "loss": 0.7385,
+      "step": 3958
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.697601809689983,
+      "learning_rate": 7.095226008179384e-06,
+      "loss": 0.7553,
+      "step": 3959
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.203386630427518,
+      "learning_rate": 7.090482585939318e-06,
+      "loss": 0.9734,
+      "step": 3960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8165695481632227,
+      "learning_rate": 7.085739878777027e-06,
+      "loss": 0.8349,
+      "step": 3961
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.609275504845507,
+      "learning_rate": 7.080997887858145e-06,
+      "loss": 0.9396,
+      "step": 3962
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8280813499025252,
+      "learning_rate": 7.076256614348116e-06,
+      "loss": 0.8813,
+      "step": 3963
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3852433919170135,
+      "learning_rate": 7.07151605941221e-06,
+      "loss": 0.741,
+      "step": 3964
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.288825295359347,
+      "learning_rate": 7.066776224215526e-06,
+      "loss": 0.6835,
+      "step": 3965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.628680125563155,
+      "learning_rate": 7.06203710992298e-06,
+      "loss": 0.8354,
+      "step": 3966
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.643593615500088,
+      "learning_rate": 7.05729871769931e-06,
+      "loss": 0.8535,
+      "step": 3967
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.676690991946433,
+      "learning_rate": 7.052561048709089e-06,
+      "loss": 0.8382,
+      "step": 3968
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4738463303615092,
+      "learning_rate": 7.0478241041166964e-06,
+      "loss": 0.8537,
+      "step": 3969
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7570416706638268,
+      "learning_rate": 7.043087885086343e-06,
+      "loss": 0.8763,
+      "step": 3970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.927495685980187,
+      "learning_rate": 7.038352392782057e-06,
+      "loss": 0.7795,
+      "step": 3971
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6282299744327693,
+      "learning_rate": 7.033617628367688e-06,
+      "loss": 0.8111,
+      "step": 3972
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4653497111160387,
+      "learning_rate": 7.028883593006914e-06,
+      "loss": 0.8788,
+      "step": 3973
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.521042858506162,
+      "learning_rate": 7.024150287863225e-06,
+      "loss": 0.8709,
+      "step": 3974
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.931058675577534,
+      "learning_rate": 7.019417714099933e-06,
+      "loss": 0.9379,
+      "step": 3975
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8664809412726044,
+      "learning_rate": 7.014685872880175e-06,
+      "loss": 0.7604,
+      "step": 3976
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.683922110588336,
+      "learning_rate": 7.009954765366902e-06,
+      "loss": 0.8586,
+      "step": 3977
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.467799967600729,
+      "learning_rate": 7.005224392722886e-06,
+      "loss": 0.7699,
+      "step": 3978
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5448159972295774,
+      "learning_rate": 7.000494756110726e-06,
+      "loss": 0.8426,
+      "step": 3979
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6894278394301554,
+      "learning_rate": 6.995765856692829e-06,
+      "loss": 0.8512,
+      "step": 3980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.53178759299369,
+      "learning_rate": 6.991037695631429e-06,
+      "loss": 0.8591,
+      "step": 3981
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.718043311800521,
+      "learning_rate": 6.986310274088574e-06,
+      "loss": 0.8791,
+      "step": 3982
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.573133728510007,
+      "learning_rate": 6.981583593226129e-06,
+      "loss": 0.773,
+      "step": 3983
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3568678578730933,
+      "learning_rate": 6.976857654205784e-06,
+      "loss": 0.7653,
+      "step": 3984
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.69196387885202,
+      "learning_rate": 6.972132458189041e-06,
+      "loss": 0.844,
+      "step": 3985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.619016705936952,
+      "learning_rate": 6.96740800633722e-06,
+      "loss": 0.8745,
+      "step": 3986
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5620443105698874,
+      "learning_rate": 6.962684299811458e-06,
+      "loss": 0.8514,
+      "step": 3987
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.693222913880473,
+      "learning_rate": 6.957961339772712e-06,
+      "loss": 0.898,
+      "step": 3988
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8891906623178865,
+      "learning_rate": 6.953239127381747e-06,
+      "loss": 0.8755,
+      "step": 3989
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.519827989426567,
+      "learning_rate": 6.948517663799159e-06,
+      "loss": 0.8378,
+      "step": 3990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6792360398650534,
+      "learning_rate": 6.943796950185344e-06,
+      "loss": 0.727,
+      "step": 3991
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.323976172597204,
+      "learning_rate": 6.939076987700528e-06,
+      "loss": 0.826,
+      "step": 3992
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8847629787119793,
+      "learning_rate": 6.934357777504741e-06,
+      "loss": 0.8186,
+      "step": 3993
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5969932789141064,
+      "learning_rate": 6.929639320757829e-06,
+      "loss": 0.7931,
+      "step": 3994
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.013839101459728,
+      "learning_rate": 6.924921618619465e-06,
+      "loss": 0.8448,
+      "step": 3995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8119454609352896,
+      "learning_rate": 6.920204672249122e-06,
+      "loss": 0.8568,
+      "step": 3996
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.555707248166815,
+      "learning_rate": 6.915488482806094e-06,
+      "loss": 0.7189,
+      "step": 3997
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.626316012005687,
+      "learning_rate": 6.9107730514494905e-06,
+      "loss": 0.891,
+      "step": 3998
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6351268057180492,
+      "learning_rate": 6.906058379338229e-06,
+      "loss": 0.8713,
+      "step": 3999
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8765134930506817,
+      "learning_rate": 6.901344467631041e-06,
+      "loss": 0.8725,
+      "step": 4000
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.423697084064312,
+      "learning_rate": 6.896631317486484e-06,
+      "loss": 0.7581,
+      "step": 4001
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.773687716759275,
+      "learning_rate": 6.891918930062908e-06,
+      "loss": 0.7716,
+      "step": 4002
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0364915828207826,
+      "learning_rate": 6.887207306518493e-06,
+      "loss": 0.753,
+      "step": 4003
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.421734376856012,
+      "learning_rate": 6.8824964480112185e-06,
+      "loss": 0.8384,
+      "step": 4004
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8184012050044376,
+      "learning_rate": 6.877786355698881e-06,
+      "loss": 0.8738,
+      "step": 4005
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7558547785075453,
+      "learning_rate": 6.873077030739095e-06,
+      "loss": 0.7818,
+      "step": 4006
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.495768634000435,
+      "learning_rate": 6.868368474289278e-06,
+      "loss": 0.8092,
+      "step": 4007
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.026842608734279,
+      "learning_rate": 6.8636606875066556e-06,
+      "loss": 0.8685,
+      "step": 4008
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.495028706880457,
+      "learning_rate": 6.8589536715482786e-06,
+      "loss": 0.7991,
+      "step": 4009
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.109487577384096,
+      "learning_rate": 6.8542474275709906e-06,
+      "loss": 0.8021,
+      "step": 4010
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5455720562557658,
+      "learning_rate": 6.8495419567314625e-06,
+      "loss": 0.8722,
+      "step": 4011
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.513268664427139,
+      "learning_rate": 6.844837260186164e-06,
+      "loss": 0.8785,
+      "step": 4012
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.050419928001916,
+      "learning_rate": 6.840133339091375e-06,
+      "loss": 0.7724,
+      "step": 4013
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6074989364288426,
+      "learning_rate": 6.835430194603191e-06,
+      "loss": 0.8731,
+      "step": 4014
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.43468203557067,
+      "learning_rate": 6.8307278278775125e-06,
+      "loss": 0.9025,
+      "step": 4015
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.824045907126482,
+      "learning_rate": 6.826026240070043e-06,
+      "loss": 0.7948,
+      "step": 4016
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.730762641642708,
+      "learning_rate": 6.821325432336311e-06,
+      "loss": 0.812,
+      "step": 4017
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.531912929558859,
+      "learning_rate": 6.816625405831638e-06,
+      "loss": 0.7354,
+      "step": 4018
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.572901191167967,
+      "learning_rate": 6.811926161711157e-06,
+      "loss": 0.8102,
+      "step": 4019
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.726521586567872,
+      "learning_rate": 6.807227701129816e-06,
+      "loss": 0.8618,
+      "step": 4020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.634553428080092,
+      "learning_rate": 6.802530025242356e-06,
+      "loss": 0.8396,
+      "step": 4021
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7548342410377984,
+      "learning_rate": 6.797833135203345e-06,
+      "loss": 0.8199,
+      "step": 4022
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6846160687305054,
+      "learning_rate": 6.79313703216714e-06,
+      "loss": 0.8959,
+      "step": 4023
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8043730901195687,
+      "learning_rate": 6.7884417172879104e-06,
+      "loss": 0.7931,
+      "step": 4024
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.545970678283305,
+      "learning_rate": 6.783747191719637e-06,
+      "loss": 0.7749,
+      "step": 4025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5451689723358326,
+      "learning_rate": 6.779053456616102e-06,
+      "loss": 0.7443,
+      "step": 4026
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.529972058083067,
+      "learning_rate": 6.774360513130888e-06,
+      "loss": 0.975,
+      "step": 4027
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.611937243351216,
+      "learning_rate": 6.769668362417397e-06,
+      "loss": 0.9137,
+      "step": 4028
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.644050252900844,
+      "learning_rate": 6.7649770056288245e-06,
+      "loss": 0.8149,
+      "step": 4029
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.504733855409512,
+      "learning_rate": 6.760286443918172e-06,
+      "loss": 0.8504,
+      "step": 4030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8987650988567344,
+      "learning_rate": 6.755596678438253e-06,
+      "loss": 0.9102,
+      "step": 4031
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6816926813216684,
+      "learning_rate": 6.750907710341674e-06,
+      "loss": 0.777,
+      "step": 4032
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.568770879329872,
+      "learning_rate": 6.746219540780859e-06,
+      "loss": 0.8053,
+      "step": 4033
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6798556604625077,
+      "learning_rate": 6.7415321709080254e-06,
+      "loss": 0.9544,
+      "step": 4034
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.558759654519992,
+      "learning_rate": 6.736845601875195e-06,
+      "loss": 0.7916,
+      "step": 4035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.85191964706347,
+      "learning_rate": 6.7321598348342e-06,
+      "loss": 0.7775,
+      "step": 4036
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.699746268096858,
+      "learning_rate": 6.727474870936667e-06,
+      "loss": 0.8829,
+      "step": 4037
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6224212365035386,
+      "learning_rate": 6.722790711334024e-06,
+      "loss": 0.7747,
+      "step": 4038
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.891059389978791,
+      "learning_rate": 6.7181073571775165e-06,
+      "loss": 0.7886,
+      "step": 4039
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.623580531520144,
+      "learning_rate": 6.713424809618176e-06,
+      "loss": 0.8556,
+      "step": 4040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6496413746498084,
+      "learning_rate": 6.708743069806839e-06,
+      "loss": 0.8466,
+      "step": 4041
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9920543365646664,
+      "learning_rate": 6.704062138894154e-06,
+      "loss": 0.7497,
+      "step": 4042
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4311107806698398,
+      "learning_rate": 6.699382018030551e-06,
+      "loss": 0.8919,
+      "step": 4043
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.668935722922524,
+      "learning_rate": 6.694702708366283e-06,
+      "loss": 0.8383,
+      "step": 4044
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6708486158887257,
+      "learning_rate": 6.690024211051389e-06,
+      "loss": 0.8418,
+      "step": 4045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.611173432896564,
+      "learning_rate": 6.685346527235707e-06,
+      "loss": 0.778,
+      "step": 4046
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.724855065721752,
+      "learning_rate": 6.68066965806889e-06,
+      "loss": 0.8278,
+      "step": 4047
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.749989817656343,
+      "learning_rate": 6.67599360470038e-06,
+      "loss": 0.7917,
+      "step": 4048
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.516759813149189,
+      "learning_rate": 6.6713183682794104e-06,
+      "loss": 0.7597,
+      "step": 4049
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.700800674394037,
+      "learning_rate": 6.666643949955033e-06,
+      "loss": 0.9328,
+      "step": 4050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.3983122261201006,
+      "learning_rate": 6.6619703508760855e-06,
+      "loss": 0.7261,
+      "step": 4051
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.748672614839632,
+      "learning_rate": 6.657297572191202e-06,
+      "loss": 0.8521,
+      "step": 4052
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6261395166600905,
+      "learning_rate": 6.652625615048831e-06,
+      "loss": 0.8024,
+      "step": 4053
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4759735826351665,
+      "learning_rate": 6.6479544805972e-06,
+      "loss": 0.776,
+      "step": 4054
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.506862194790913,
+      "learning_rate": 6.64328416998435e-06,
+      "loss": 0.7892,
+      "step": 4055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.525522899555866,
+      "learning_rate": 6.6386146843581075e-06,
+      "loss": 0.76,
+      "step": 4056
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.90535564554283,
+      "learning_rate": 6.633946024866098e-06,
+      "loss": 0.8547,
+      "step": 4057
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8295321791625,
+      "learning_rate": 6.629278192655755e-06,
+      "loss": 0.8736,
+      "step": 4058
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8648454263022014,
+      "learning_rate": 6.624611188874297e-06,
+      "loss": 0.8359,
+      "step": 4059
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.4725522208039368,
+      "learning_rate": 6.619945014668741e-06,
+      "loss": 0.829,
+      "step": 4060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.924624596128136,
+      "learning_rate": 6.615279671185904e-06,
+      "loss": 0.783,
+      "step": 4061
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.308448959924426,
+      "learning_rate": 6.6106151595723955e-06,
+      "loss": 0.6837,
+      "step": 4062
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7063686985335793,
+      "learning_rate": 6.605951480974618e-06,
+      "loss": 0.7703,
+      "step": 4063
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6589391223667045,
+      "learning_rate": 6.601288636538782e-06,
+      "loss": 0.8667,
+      "step": 4064
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.544162074879167,
+      "learning_rate": 6.596626627410876e-06,
+      "loss": 0.7775,
+      "step": 4065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7523919460415818,
+      "learning_rate": 6.591965454736695e-06,
+      "loss": 0.8147,
+      "step": 4066
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.008385669960899,
+      "learning_rate": 6.587305119661824e-06,
+      "loss": 0.9105,
+      "step": 4067
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.769938732126488,
+      "learning_rate": 6.582645623331638e-06,
+      "loss": 0.8769,
+      "step": 4068
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.67656946337048,
+      "learning_rate": 6.577986966891319e-06,
+      "loss": 0.8576,
+      "step": 4069
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.917024733414259,
+      "learning_rate": 6.57332915148583e-06,
+      "loss": 0.6818,
+      "step": 4070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5314665259462976,
+      "learning_rate": 6.568672178259927e-06,
+      "loss": 0.7612,
+      "step": 4071
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.767389303924322,
+      "learning_rate": 6.564016048358171e-06,
+      "loss": 0.7915,
+      "step": 4072
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5813872988695876,
+      "learning_rate": 6.559360762924905e-06,
+      "loss": 0.8387,
+      "step": 4073
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.626198012687205,
+      "learning_rate": 6.5547063231042616e-06,
+      "loss": 0.8293,
+      "step": 4074
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.616346063221364,
+      "learning_rate": 6.550052730040182e-06,
+      "loss": 0.904,
+      "step": 4075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.554419912580466,
+      "learning_rate": 6.545399984876382e-06,
+      "loss": 0.8675,
+      "step": 4076
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5640788502246457,
+      "learning_rate": 6.540748088756379e-06,
+      "loss": 0.7613,
+      "step": 4077
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.704161443937231,
+      "learning_rate": 6.536097042823478e-06,
+      "loss": 0.8895,
+      "step": 4078
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6358909379667366,
+      "learning_rate": 6.53144684822077e-06,
+      "loss": 0.7766,
+      "step": 4079
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.941858378301113,
+      "learning_rate": 6.5267975060911534e-06,
+      "loss": 0.8079,
+      "step": 4080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.515359174202313,
+      "learning_rate": 6.5221490175772996e-06,
+      "loss": 0.7194,
+      "step": 4081
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.586250039707266,
+      "learning_rate": 6.517501383821675e-06,
+      "loss": 0.7522,
+      "step": 4082
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6503763756279386,
+      "learning_rate": 6.512854605966542e-06,
+      "loss": 0.8073,
+      "step": 4083
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6301952442230596,
+      "learning_rate": 6.508208685153943e-06,
+      "loss": 0.8574,
+      "step": 4084
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6460918372899784,
+      "learning_rate": 6.5035636225257206e-06,
+      "loss": 0.8026,
+      "step": 4085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6467874751339475,
+      "learning_rate": 6.4989194192235e-06,
+      "loss": 0.7865,
+      "step": 4086
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.53926613183767,
+      "learning_rate": 6.4942760763886906e-06,
+      "loss": 0.8668,
+      "step": 4087
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4622835484736094,
+      "learning_rate": 6.489633595162503e-06,
+      "loss": 0.8309,
+      "step": 4088
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7848796523663166,
+      "learning_rate": 6.484991976685927e-06,
+      "loss": 0.8824,
+      "step": 4089
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.014091740841856,
+      "learning_rate": 6.4803512220997366e-06,
+      "loss": 0.9,
+      "step": 4090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4966799499927443,
+      "learning_rate": 6.475711332544506e-06,
+      "loss": 0.8105,
+      "step": 4091
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.56373881046293,
+      "learning_rate": 6.471072309160589e-06,
+      "loss": 0.8666,
+      "step": 4092
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.102359688208209,
+      "learning_rate": 6.466434153088124e-06,
+      "loss": 0.8941,
+      "step": 4093
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.506941106145849,
+      "learning_rate": 6.461796865467043e-06,
+      "loss": 0.8177,
+      "step": 4094
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.446841206056606,
+      "learning_rate": 6.457160447437055e-06,
+      "loss": 0.8387,
+      "step": 4095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.039755545522044,
+      "learning_rate": 6.452524900137671e-06,
+      "loss": 0.9508,
+      "step": 4096
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.960153109907913,
+      "learning_rate": 6.447890224708175e-06,
+      "loss": 0.8863,
+      "step": 4097
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8572295562041097,
+      "learning_rate": 6.443256422287636e-06,
+      "loss": 0.8931,
+      "step": 4098
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5975817640776717,
+      "learning_rate": 6.4386234940149175e-06,
+      "loss": 0.7323,
+      "step": 4099
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.585156872911483,
+      "learning_rate": 6.433991441028662e-06,
+      "loss": 0.8293,
+      "step": 4100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.649456417121931,
+      "learning_rate": 6.429360264467295e-06,
+      "loss": 0.8892,
+      "step": 4101
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8639255281877456,
+      "learning_rate": 6.424729965469035e-06,
+      "loss": 0.9024,
+      "step": 4102
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.780003960122886,
+      "learning_rate": 6.420100545171878e-06,
+      "loss": 0.8657,
+      "step": 4103
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.509960135382255,
+      "learning_rate": 6.4154720047136036e-06,
+      "loss": 0.7157,
+      "step": 4104
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5979157879958343,
+      "learning_rate": 6.4108443452317795e-06,
+      "loss": 0.8317,
+      "step": 4105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4868039343347768,
+      "learning_rate": 6.40621756786375e-06,
+      "loss": 0.8063,
+      "step": 4106
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2123728570860846,
+      "learning_rate": 6.401591673746654e-06,
+      "loss": 0.7474,
+      "step": 4107
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8195349629081465,
+      "learning_rate": 6.396966664017406e-06,
+      "loss": 0.8592,
+      "step": 4108
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.74887243369275,
+      "learning_rate": 6.392342539812697e-06,
+      "loss": 0.8349,
+      "step": 4109
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.414885397862542,
+      "learning_rate": 6.387719302269013e-06,
+      "loss": 0.8437,
+      "step": 4110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8044942835656865,
+      "learning_rate": 6.383096952522616e-06,
+      "loss": 0.8438,
+      "step": 4111
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4437840839634037,
+      "learning_rate": 6.378475491709543e-06,
+      "loss": 0.8465,
+      "step": 4112
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5651394312215334,
+      "learning_rate": 6.373854920965629e-06,
+      "loss": 0.8432,
+      "step": 4113
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4733310108162367,
+      "learning_rate": 6.369235241426475e-06,
+      "loss": 0.7228,
+      "step": 4114
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0439512584691006,
+      "learning_rate": 6.3646164542274705e-06,
+      "loss": 0.7518,
+      "step": 4115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9206726657813906,
+      "learning_rate": 6.359998560503784e-06,
+      "loss": 0.7668,
+      "step": 4116
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.985452574642286,
+      "learning_rate": 6.355381561390359e-06,
+      "loss": 0.7956,
+      "step": 4117
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.550473017214655,
+      "learning_rate": 6.350765458021935e-06,
+      "loss": 0.7408,
+      "step": 4118
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5381264489602375,
+      "learning_rate": 6.346150251533013e-06,
+      "loss": 0.8889,
+      "step": 4119
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.575632216288596,
+      "learning_rate": 6.341535943057882e-06,
+      "loss": 0.8932,
+      "step": 4120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4305923233576925,
+      "learning_rate": 6.336922533730611e-06,
+      "loss": 0.7149,
+      "step": 4121
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.542358361284761,
+      "learning_rate": 6.332310024685049e-06,
+      "loss": 0.7948,
+      "step": 4122
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8751522589498832,
+      "learning_rate": 6.327698417054814e-06,
+      "loss": 0.8883,
+      "step": 4123
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4987312069013172,
+      "learning_rate": 6.3230877119733184e-06,
+      "loss": 0.8417,
+      "step": 4124
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.626332516673871,
+      "learning_rate": 6.318477910573739e-06,
+      "loss": 0.8392,
+      "step": 4125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.884330720549634,
+      "learning_rate": 6.313869013989037e-06,
+      "loss": 0.8784,
+      "step": 4126
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.433399590467831,
+      "learning_rate": 6.309261023351951e-06,
+      "loss": 0.7685,
+      "step": 4127
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.234724727523963,
+      "learning_rate": 6.304653939794991e-06,
+      "loss": 0.9656,
+      "step": 4128
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7665147719725556,
+      "learning_rate": 6.300047764450456e-06,
+      "loss": 0.8758,
+      "step": 4129
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6819728229640902,
+      "learning_rate": 6.295442498450413e-06,
+      "loss": 0.8419,
+      "step": 4130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.405193789764089,
+      "learning_rate": 6.290838142926702e-06,
+      "loss": 0.7792,
+      "step": 4131
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5564651239366065,
+      "learning_rate": 6.286234699010951e-06,
+      "loss": 0.7418,
+      "step": 4132
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.547961536655596,
+      "learning_rate": 6.281632167834555e-06,
+      "loss": 0.8122,
+      "step": 4133
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.579679050497923,
+      "learning_rate": 6.277030550528681e-06,
+      "loss": 0.8219,
+      "step": 4134
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6577361052780657,
+      "learning_rate": 6.2724298482242885e-06,
+      "loss": 0.843,
+      "step": 4135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.796823419353427,
+      "learning_rate": 6.267830062052095e-06,
+      "loss": 0.8328,
+      "step": 4136
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.675985550773138,
+      "learning_rate": 6.263231193142598e-06,
+      "loss": 0.7838,
+      "step": 4137
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5707725044542515,
+      "learning_rate": 6.258633242626073e-06,
+      "loss": 0.8262,
+      "step": 4138
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.813092623388354,
+      "learning_rate": 6.254036211632563e-06,
+      "loss": 0.8425,
+      "step": 4139
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.568730615011392,
+      "learning_rate": 6.249440101291898e-06,
+      "loss": 0.7452,
+      "step": 4140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.755377671537752,
+      "learning_rate": 6.244844912733667e-06,
+      "loss": 0.8829,
+      "step": 4141
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.558947372504722,
+      "learning_rate": 6.240250647087236e-06,
+      "loss": 0.897,
+      "step": 4142
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.745447153058319,
+      "learning_rate": 6.235657305481752e-06,
+      "loss": 0.8688,
+      "step": 4143
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.647907903816932,
+      "learning_rate": 6.23106488904613e-06,
+      "loss": 0.8379,
+      "step": 4144
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.790647314327554,
+      "learning_rate": 6.2264733989090475e-06,
+      "loss": 0.8742,
+      "step": 4145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9750857386672114,
+      "learning_rate": 6.221882836198977e-06,
+      "loss": 0.8778,
+      "step": 4146
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3547014311971215,
+      "learning_rate": 6.217293202044143e-06,
+      "loss": 0.6963,
+      "step": 4147
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.4512617810217496,
+      "learning_rate": 6.212704497572548e-06,
+      "loss": 0.7892,
+      "step": 4148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5369909191742073,
+      "learning_rate": 6.208116723911971e-06,
+      "loss": 0.7219,
+      "step": 4149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.536979432046566,
+      "learning_rate": 6.203529882189951e-06,
+      "loss": 0.7174,
+      "step": 4150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.452686560261575,
+      "learning_rate": 6.198943973533816e-06,
+      "loss": 0.7855,
+      "step": 4151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6527193305604295,
+      "learning_rate": 6.194358999070646e-06,
+      "loss": 0.7499,
+      "step": 4152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3863825447074682,
+      "learning_rate": 6.189774959927297e-06,
+      "loss": 0.744,
+      "step": 4153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5502961600790366,
+      "learning_rate": 6.185191857230408e-06,
+      "loss": 0.7702,
+      "step": 4154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.777791413090332,
+      "learning_rate": 6.180609692106368e-06,
+      "loss": 0.7513,
+      "step": 4155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7429295672784546,
+      "learning_rate": 6.176028465681343e-06,
+      "loss": 0.7781,
+      "step": 4156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.583623335809288,
+      "learning_rate": 6.1714481790812765e-06,
+      "loss": 0.7991,
+      "step": 4157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5019251528214976,
+      "learning_rate": 6.166868833431869e-06,
+      "loss": 0.7802,
+      "step": 4158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.46943361864758,
+      "learning_rate": 6.162290429858602e-06,
+      "loss": 0.8392,
+      "step": 4159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.697621406379056,
+      "learning_rate": 6.157712969486716e-06,
+      "loss": 0.7135,
+      "step": 4160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5361851349014426,
+      "learning_rate": 6.153136453441216e-06,
+      "loss": 0.7468,
+      "step": 4161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.777669008265395,
+      "learning_rate": 6.14856088284689e-06,
+      "loss": 0.7782,
+      "step": 4162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.676393263557243,
+      "learning_rate": 6.14398625882828e-06,
+      "loss": 0.8292,
+      "step": 4163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.629987864288778,
+      "learning_rate": 6.1394125825097005e-06,
+      "loss": 0.7771,
+      "step": 4164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3706226588091357,
+      "learning_rate": 6.134839855015235e-06,
+      "loss": 0.7614,
+      "step": 4165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9096277892102456,
+      "learning_rate": 6.1302680774687325e-06,
+      "loss": 0.8288,
+      "step": 4166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.266472116130424,
+      "learning_rate": 6.125697250993804e-06,
+      "loss": 0.617,
+      "step": 4167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.546047394461752,
+      "learning_rate": 6.1211273767138336e-06,
+      "loss": 0.8736,
+      "step": 4168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7568097786899677,
+      "learning_rate": 6.1165584557519634e-06,
+      "loss": 0.7315,
+      "step": 4169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5725431683106645,
+      "learning_rate": 6.111990489231114e-06,
+      "loss": 0.7479,
+      "step": 4170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.418434465044092,
+      "learning_rate": 6.1074234782739576e-06,
+      "loss": 0.8278,
+      "step": 4171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6811260628803524,
+      "learning_rate": 6.102857424002937e-06,
+      "loss": 0.8043,
+      "step": 4172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7969088357041114,
+      "learning_rate": 6.098292327540265e-06,
+      "loss": 0.8,
+      "step": 4173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8678984331762813,
+      "learning_rate": 6.093728190007912e-06,
+      "loss": 0.8598,
+      "step": 4174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7141740857093932,
+      "learning_rate": 6.089165012527609e-06,
+      "loss": 0.7699,
+      "step": 4175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7721712895307693,
+      "learning_rate": 6.084602796220866e-06,
+      "loss": 0.8596,
+      "step": 4176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.534281013021705,
+      "learning_rate": 6.080041542208946e-06,
+      "loss": 0.7828,
+      "step": 4177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.551375290088987,
+      "learning_rate": 6.075481251612873e-06,
+      "loss": 0.7576,
+      "step": 4178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8120490143121186,
+      "learning_rate": 6.0709219255534424e-06,
+      "loss": 0.8315,
+      "step": 4179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2619929727898964,
+      "learning_rate": 6.066363565151203e-06,
+      "loss": 0.8645,
+      "step": 4180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5383095550716197,
+      "learning_rate": 6.061806171526482e-06,
+      "loss": 0.7716,
+      "step": 4181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4670004389518843,
+      "learning_rate": 6.0572497457993515e-06,
+      "loss": 0.7649,
+      "step": 4182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2463743452604414,
+      "learning_rate": 6.052694289089655e-06,
+      "loss": 0.7303,
+      "step": 4183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8256884332350682,
+      "learning_rate": 6.048139802516997e-06,
+      "loss": 0.7508,
+      "step": 4184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6277561858184866,
+      "learning_rate": 6.04358628720074e-06,
+      "loss": 0.7525,
+      "step": 4185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.788573749816516,
+      "learning_rate": 6.039033744260009e-06,
+      "loss": 0.8504,
+      "step": 4186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6760899491206636,
+      "learning_rate": 6.034482174813698e-06,
+      "loss": 0.8448,
+      "step": 4187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7226554524499753,
+      "learning_rate": 6.0299315799804524e-06,
+      "loss": 0.8442,
+      "step": 4188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5518099787008737,
+      "learning_rate": 6.025381960878675e-06,
+      "loss": 0.8306,
+      "step": 4189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.696138047928778,
+      "learning_rate": 6.020833318626544e-06,
+      "loss": 0.8902,
+      "step": 4190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3993432050044063,
+      "learning_rate": 6.016285654341978e-06,
+      "loss": 0.7706,
+      "step": 4191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7765750937001203,
+      "learning_rate": 6.011738969142676e-06,
+      "loss": 0.8679,
+      "step": 4192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.859266254500037,
+      "learning_rate": 6.0071932641460784e-06,
+      "loss": 0.8554,
+      "step": 4193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6019050994444024,
+      "learning_rate": 6.002648540469394e-06,
+      "loss": 0.7164,
+      "step": 4194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7344950289561343,
+      "learning_rate": 5.9981047992295895e-06,
+      "loss": 0.8349,
+      "step": 4195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.612788730546887,
+      "learning_rate": 5.993562041543388e-06,
+      "loss": 0.8315,
+      "step": 4196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.664873642699137,
+      "learning_rate": 5.989020268527268e-06,
+      "loss": 0.8662,
+      "step": 4197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7090021448831867,
+      "learning_rate": 5.9844794812974784e-06,
+      "loss": 0.8077,
+      "step": 4198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8042750369301563,
+      "learning_rate": 5.979939680970012e-06,
+      "loss": 0.8289,
+      "step": 4199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.897760226237868,
+      "learning_rate": 5.975400868660624e-06,
+      "loss": 0.845,
+      "step": 4200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8252430374625055,
+      "learning_rate": 5.970863045484829e-06,
+      "loss": 0.8849,
+      "step": 4201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4574717176968277,
+      "learning_rate": 5.966326212557892e-06,
+      "loss": 0.9231,
+      "step": 4202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5705754869443163,
+      "learning_rate": 5.9617903709948485e-06,
+      "loss": 0.8147,
+      "step": 4203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.559653301159418,
+      "learning_rate": 5.957255521910477e-06,
+      "loss": 0.8086,
+      "step": 4204
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6067710988929536,
+      "learning_rate": 5.952721666419311e-06,
+      "loss": 0.7963,
+      "step": 4205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7621380129077964,
+      "learning_rate": 5.948188805635652e-06,
+      "loss": 0.8049,
+      "step": 4206
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.652662783063158,
+      "learning_rate": 5.9436569406735475e-06,
+      "loss": 0.798,
+      "step": 4207
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.742562653994274,
+      "learning_rate": 5.939126072646798e-06,
+      "loss": 0.8252,
+      "step": 4208
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4686756050614465,
+      "learning_rate": 5.934596202668973e-06,
+      "loss": 0.7727,
+      "step": 4209
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5693938881507408,
+      "learning_rate": 5.930067331853382e-06,
+      "loss": 0.8081,
+      "step": 4210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6406554355059035,
+      "learning_rate": 5.925539461313095e-06,
+      "loss": 0.6952,
+      "step": 4211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.567357727216333,
+      "learning_rate": 5.921012592160936e-06,
+      "loss": 0.7968,
+      "step": 4212
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.549508744267877,
+      "learning_rate": 5.916486725509479e-06,
+      "loss": 0.9364,
+      "step": 4213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.653890717917521,
+      "learning_rate": 5.911961862471063e-06,
+      "loss": 0.8733,
+      "step": 4214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5635645587348996,
+      "learning_rate": 5.907438004157767e-06,
+      "loss": 0.8015,
+      "step": 4215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.901723645052626,
+      "learning_rate": 5.902915151681427e-06,
+      "loss": 0.8542,
+      "step": 4216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.630135346802989,
+      "learning_rate": 5.89839330615364e-06,
+      "loss": 0.7924,
+      "step": 4217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9543974638939074,
+      "learning_rate": 5.893872468685743e-06,
+      "loss": 0.8472,
+      "step": 4218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.608700762294027,
+      "learning_rate": 5.889352640388828e-06,
+      "loss": 0.7086,
+      "step": 4219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6845836414520825,
+      "learning_rate": 5.884833822373751e-06,
+      "loss": 0.8024,
+      "step": 4220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.712954926371659,
+      "learning_rate": 5.880316015751106e-06,
+      "loss": 0.819,
+      "step": 4221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.601280457477493,
+      "learning_rate": 5.875799221631242e-06,
+      "loss": 0.7722,
+      "step": 4222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.523625947183782,
+      "learning_rate": 5.871283441124264e-06,
+      "loss": 0.8626,
+      "step": 4223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4552347893262985,
+      "learning_rate": 5.866768675340018e-06,
+      "loss": 0.7041,
+      "step": 4224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.52178337670741,
+      "learning_rate": 5.8622549253881135e-06,
+      "loss": 0.8436,
+      "step": 4225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7843047744222957,
+      "learning_rate": 5.8577421923779025e-06,
+      "loss": 0.7545,
+      "step": 4226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7334770181269517,
+      "learning_rate": 5.853230477418483e-06,
+      "loss": 0.9143,
+      "step": 4227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.551804593077235,
+      "learning_rate": 5.8487197816187145e-06,
+      "loss": 0.7694,
+      "step": 4228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.827479143157379,
+      "learning_rate": 5.844210106087198e-06,
+      "loss": 0.8784,
+      "step": 4229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6859822414158914,
+      "learning_rate": 5.839701451932278e-06,
+      "loss": 0.7843,
+      "step": 4230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5468246556889764,
+      "learning_rate": 5.8351938202620666e-06,
+      "loss": 0.862,
+      "step": 4231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.308901391630756,
+      "learning_rate": 5.830687212184407e-06,
+      "loss": 0.8985,
+      "step": 4232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3919305019885573,
+      "learning_rate": 5.826181628806893e-06,
+      "loss": 0.7759,
+      "step": 4233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.795221305372727,
+      "learning_rate": 5.821677071236881e-06,
+      "loss": 0.8936,
+      "step": 4234
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.543397606862117,
+      "learning_rate": 5.817173540581459e-06,
+      "loss": 0.8577,
+      "step": 4235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5323267733398285,
+      "learning_rate": 5.812671037947468e-06,
+      "loss": 0.7667,
+      "step": 4236
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6574236218151848,
+      "learning_rate": 5.8081695644415e-06,
+      "loss": 0.8143,
+      "step": 4237
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.912775437447401,
+      "learning_rate": 5.803669121169883e-06,
+      "loss": 0.8918,
+      "step": 4238
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.939122520427922,
+      "learning_rate": 5.79916970923871e-06,
+      "loss": 0.858,
+      "step": 4239
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4747736151679516,
+      "learning_rate": 5.7946713297538045e-06,
+      "loss": 0.8165,
+      "step": 4240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.640788425357784,
+      "learning_rate": 5.790173983820741e-06,
+      "loss": 0.8805,
+      "step": 4241
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.806348940640807,
+      "learning_rate": 5.785677672544847e-06,
+      "loss": 0.844,
+      "step": 4242
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5707488108651746,
+      "learning_rate": 5.781182397031182e-06,
+      "loss": 0.8061,
+      "step": 4243
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.638873894062458,
+      "learning_rate": 5.776688158384565e-06,
+      "loss": 0.9218,
+      "step": 4244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8972966930258943,
+      "learning_rate": 5.772194957709553e-06,
+      "loss": 0.7464,
+      "step": 4245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.686284381777874,
+      "learning_rate": 5.767702796110448e-06,
+      "loss": 0.8202,
+      "step": 4246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.605172358461725,
+      "learning_rate": 5.763211674691296e-06,
+      "loss": 0.8269,
+      "step": 4247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.539411710086652,
+      "learning_rate": 5.758721594555887e-06,
+      "loss": 0.8579,
+      "step": 4248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4711589723884613,
+      "learning_rate": 5.7542325568077576e-06,
+      "loss": 0.7988,
+      "step": 4249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.605839636785204,
+      "learning_rate": 5.749744562550191e-06,
+      "loss": 0.8018,
+      "step": 4250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.68865790641465,
+      "learning_rate": 5.745257612886209e-06,
+      "loss": 0.8699,
+      "step": 4251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4119373086322318,
+      "learning_rate": 5.740771708918573e-06,
+      "loss": 0.7347,
+      "step": 4252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5149040499716757,
+      "learning_rate": 5.7362868517498e-06,
+      "loss": 0.8381,
+      "step": 4253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.575269057160359,
+      "learning_rate": 5.731803042482135e-06,
+      "loss": 0.8648,
+      "step": 4254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.492363871218477,
+      "learning_rate": 5.72732028221758e-06,
+      "loss": 0.7981,
+      "step": 4255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6274570081759014,
+      "learning_rate": 5.722838572057867e-06,
+      "loss": 0.6948,
+      "step": 4256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.722018640380298,
+      "learning_rate": 5.718357913104477e-06,
+      "loss": 0.7619,
+      "step": 4257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.6070857482325307,
+      "learning_rate": 5.713878306458626e-06,
+      "loss": 0.8415,
+      "step": 4258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.731178106222968,
+      "learning_rate": 5.709399753221282e-06,
+      "loss": 0.8122,
+      "step": 4259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5202536126008734,
+      "learning_rate": 5.704922254493139e-06,
+      "loss": 0.9007,
+      "step": 4260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.763827526804397,
+      "learning_rate": 5.7004458113746485e-06,
+      "loss": 0.7899,
+      "step": 4261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.9383225693445243,
+      "learning_rate": 5.695970424965993e-06,
+      "loss": 0.7951,
+      "step": 4262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.673618096203085,
+      "learning_rate": 5.691496096367093e-06,
+      "loss": 0.7874,
+      "step": 4263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4528289976047604,
+      "learning_rate": 5.687022826677619e-06,
+      "loss": 0.7333,
+      "step": 4264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.434025160199743,
+      "learning_rate": 5.682550616996968e-06,
+      "loss": 0.7751,
+      "step": 4265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5805781831193646,
+      "learning_rate": 5.678079468424293e-06,
+      "loss": 0.8292,
+      "step": 4266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.535643452437688,
+      "learning_rate": 5.67360938205847e-06,
+      "loss": 0.7846,
+      "step": 4267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7760047936406242,
+      "learning_rate": 5.66914035899812e-06,
+      "loss": 0.7979,
+      "step": 4268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3859097315091726,
+      "learning_rate": 5.664672400341614e-06,
+      "loss": 0.7801,
+      "step": 4269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8159117598444467,
+      "learning_rate": 5.6602055071870395e-06,
+      "loss": 0.7675,
+      "step": 4270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.617294328890024,
+      "learning_rate": 5.655739680632233e-06,
+      "loss": 0.7799,
+      "step": 4271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.5763635450190776,
+      "learning_rate": 5.651274921774777e-06,
+      "loss": 0.8318,
+      "step": 4272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7449717860806904,
+      "learning_rate": 5.646811231711982e-06,
+      "loss": 0.7591,
+      "step": 4273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.627656108484001,
+      "learning_rate": 5.642348611540892e-06,
+      "loss": 0.7857,
+      "step": 4274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.545063061864533,
+      "learning_rate": 5.637887062358302e-06,
+      "loss": 0.8196,
+      "step": 4275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.482595594212583,
+      "learning_rate": 5.633426585260728e-06,
+      "loss": 0.8116,
+      "step": 4276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.047445613173143,
+      "learning_rate": 5.6289671813444376e-06,
+      "loss": 0.8021,
+      "step": 4277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.422904459979712,
+      "learning_rate": 5.624508851705426e-06,
+      "loss": 0.8128,
+      "step": 4278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8670743746634257,
+      "learning_rate": 5.620051597439417e-06,
+      "loss": 0.836,
+      "step": 4279
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.473690554907657,
+      "learning_rate": 5.6155954196418905e-06,
+      "loss": 0.772,
+      "step": 4280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6703573989353697,
+      "learning_rate": 5.6111403194080435e-06,
+      "loss": 0.8059,
+      "step": 4281
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.63961123884444,
+      "learning_rate": 5.606686297832817e-06,
+      "loss": 0.7687,
+      "step": 4282
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7593795892751896,
+      "learning_rate": 5.602233356010883e-06,
+      "loss": 0.8548,
+      "step": 4283
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.516821655894118,
+      "learning_rate": 5.59778149503665e-06,
+      "loss": 0.823,
+      "step": 4284
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.034175184053461,
+      "learning_rate": 5.5933307160042575e-06,
+      "loss": 0.8375,
+      "step": 4285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.609464382963494,
+      "learning_rate": 5.588881020007588e-06,
+      "loss": 0.7616,
+      "step": 4286
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5855150647940217,
+      "learning_rate": 5.584432408140246e-06,
+      "loss": 0.7481,
+      "step": 4287
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7773608715093907,
+      "learning_rate": 5.579984881495582e-06,
+      "loss": 0.7988,
+      "step": 4288
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7223896843575988,
+      "learning_rate": 5.575538441166671e-06,
+      "loss": 0.7573,
+      "step": 4289
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.690483798746339,
+      "learning_rate": 5.5710930882463174e-06,
+      "loss": 0.8885,
+      "step": 4290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.422432484428218,
+      "learning_rate": 5.566648823827075e-06,
+      "loss": 0.7379,
+      "step": 4291
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7445270021125245,
+      "learning_rate": 5.562205649001213e-06,
+      "loss": 0.7741,
+      "step": 4292
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.699609159678072,
+      "learning_rate": 5.5577635648607396e-06,
+      "loss": 0.7712,
+      "step": 4293
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.683136191034804,
+      "learning_rate": 5.553322572497395e-06,
+      "loss": 0.8999,
+      "step": 4294
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.63819629445708,
+      "learning_rate": 5.548882673002651e-06,
+      "loss": 0.8582,
+      "step": 4295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4904520900083083,
+      "learning_rate": 5.544443867467705e-06,
+      "loss": 0.9049,
+      "step": 4296
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.677841058226375,
+      "learning_rate": 5.5400061569834995e-06,
+      "loss": 0.7495,
+      "step": 4297
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6850812833894007,
+      "learning_rate": 5.5355695426406905e-06,
+      "loss": 0.7755,
+      "step": 4298
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.718695969365637,
+      "learning_rate": 5.531134025529684e-06,
+      "loss": 0.9263,
+      "step": 4299
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6832942199376215,
+      "learning_rate": 5.5266996067405995e-06,
+      "loss": 0.7404,
+      "step": 4300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5524843916004953,
+      "learning_rate": 5.522266287363289e-06,
+      "loss": 0.7853,
+      "step": 4301
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5065546394197145,
+      "learning_rate": 5.517834068487347e-06,
+      "loss": 0.8372,
+      "step": 4302
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.544350067419898,
+      "learning_rate": 5.513402951202082e-06,
+      "loss": 0.7003,
+      "step": 4303
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7065939446365435,
+      "learning_rate": 5.508972936596542e-06,
+      "loss": 0.6783,
+      "step": 4304
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5383869242448776,
+      "learning_rate": 5.504544025759498e-06,
+      "loss": 0.7375,
+      "step": 4305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.591610208015511,
+      "learning_rate": 5.500116219779453e-06,
+      "loss": 0.8246,
+      "step": 4306
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3648144456733933,
+      "learning_rate": 5.495689519744634e-06,
+      "loss": 0.752,
+      "step": 4307
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.275339281801009,
+      "learning_rate": 5.491263926743005e-06,
+      "loss": 0.7565,
+      "step": 4308
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6419310382852195,
+      "learning_rate": 5.486839441862248e-06,
+      "loss": 0.7358,
+      "step": 4309
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5891680783499385,
+      "learning_rate": 5.482416066189783e-06,
+      "loss": 0.7734,
+      "step": 4310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8216200785004437,
+      "learning_rate": 5.477993800812749e-06,
+      "loss": 0.8772,
+      "step": 4311
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.640532241676102,
+      "learning_rate": 5.473572646818011e-06,
+      "loss": 0.8411,
+      "step": 4312
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.427155653140146,
+      "learning_rate": 5.4691526052921705e-06,
+      "loss": 0.7412,
+      "step": 4313
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4904325112756585,
+      "learning_rate": 5.464733677321548e-06,
+      "loss": 0.7864,
+      "step": 4314
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8079062192354263,
+      "learning_rate": 5.460315863992191e-06,
+      "loss": 0.7866,
+      "step": 4315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.853060485157894,
+      "learning_rate": 5.455899166389875e-06,
+      "loss": 0.8018,
+      "step": 4316
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.485773737267365,
+      "learning_rate": 5.451483585600096e-06,
+      "loss": 0.7539,
+      "step": 4317
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.797910829886508,
+      "learning_rate": 5.447069122708086e-06,
+      "loss": 0.7892,
+      "step": 4318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8726475904531275,
+      "learning_rate": 5.442655778798795e-06,
+      "loss": 0.7287,
+      "step": 4319
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.964910825888747,
+      "learning_rate": 5.438243554956895e-06,
+      "loss": 0.9582,
+      "step": 4320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6561943823597725,
+      "learning_rate": 5.433832452266794e-06,
+      "loss": 0.8632,
+      "step": 4321
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.544080014332781,
+      "learning_rate": 5.429422471812612e-06,
+      "loss": 0.8035,
+      "step": 4322
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.826157745342801,
+      "learning_rate": 5.425013614678197e-06,
+      "loss": 0.8831,
+      "step": 4323
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4112657605218684,
+      "learning_rate": 5.4206058819471276e-06,
+      "loss": 0.797,
+      "step": 4324
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4430557001274273,
+      "learning_rate": 5.416199274702699e-06,
+      "loss": 0.7924,
+      "step": 4325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4328516564445186,
+      "learning_rate": 5.411793794027931e-06,
+      "loss": 0.727,
+      "step": 4326
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8558225094747867,
+      "learning_rate": 5.407389441005569e-06,
+      "loss": 0.7773,
+      "step": 4327
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5283966333337706,
+      "learning_rate": 5.402986216718071e-06,
+      "loss": 0.77,
+      "step": 4328
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.728670654557776,
+      "learning_rate": 5.398584122247639e-06,
+      "loss": 0.8477,
+      "step": 4329
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8334669043005265,
+      "learning_rate": 5.394183158676178e-06,
+      "loss": 0.8122,
+      "step": 4330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.652568907162666,
+      "learning_rate": 5.389783327085317e-06,
+      "loss": 0.822,
+      "step": 4331
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3142620945835386,
+      "learning_rate": 5.38538462855642e-06,
+      "loss": 0.6978,
+      "step": 4332
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.7730169314845914,
+      "learning_rate": 5.38098706417056e-06,
+      "loss": 0.807,
+      "step": 4333
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.851402696586342,
+      "learning_rate": 5.376590635008531e-06,
+      "loss": 0.8454,
+      "step": 4334
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6112912494257206,
+      "learning_rate": 5.3721953421508585e-06,
+      "loss": 0.7354,
+      "step": 4335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.68768207215782,
+      "learning_rate": 5.367801186677779e-06,
+      "loss": 0.8525,
+      "step": 4336
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.348257730715597,
+      "learning_rate": 5.363408169669253e-06,
+      "loss": 0.7565,
+      "step": 4337
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5730011308658454,
+      "learning_rate": 5.359016292204962e-06,
+      "loss": 0.7148,
+      "step": 4338
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.820490788623341,
+      "learning_rate": 5.354625555364301e-06,
+      "loss": 0.7555,
+      "step": 4339
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.407342488224494,
+      "learning_rate": 5.3502359602263975e-06,
+      "loss": 0.7254,
+      "step": 4340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.756876501448233,
+      "learning_rate": 5.345847507870087e-06,
+      "loss": 0.7881,
+      "step": 4341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.394704047096753,
+      "learning_rate": 5.341460199373925e-06,
+      "loss": 0.7396,
+      "step": 4342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.638217727897471,
+      "learning_rate": 5.337074035816197e-06,
+      "loss": 0.9129,
+      "step": 4343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5261876067230253,
+      "learning_rate": 5.332689018274892e-06,
+      "loss": 0.7824,
+      "step": 4344
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.95205594094719,
+      "learning_rate": 5.3283051478277244e-06,
+      "loss": 0.8592,
+      "step": 4345
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.646327886755659,
+      "learning_rate": 5.323922425552133e-06,
+      "loss": 0.7524,
+      "step": 4346
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.706809301435508,
+      "learning_rate": 5.319540852525264e-06,
+      "loss": 0.8686,
+      "step": 4347
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.881161427865816,
+      "learning_rate": 5.315160429823984e-06,
+      "loss": 0.8364,
+      "step": 4348
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6895405962270083,
+      "learning_rate": 5.3107811585248806e-06,
+      "loss": 0.7875,
+      "step": 4349
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.628565569216531,
+      "learning_rate": 5.306403039704249e-06,
+      "loss": 0.7816,
+      "step": 4350
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6822366932950317,
+      "learning_rate": 5.302026074438118e-06,
+      "loss": 0.7017,
+      "step": 4351
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.39036593803038,
+      "learning_rate": 5.297650263802218e-06,
+      "loss": 0.7353,
+      "step": 4352
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8448153356191828,
+      "learning_rate": 5.293275608871997e-06,
+      "loss": 0.8072,
+      "step": 4353
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.664536900824033,
+      "learning_rate": 5.28890211072263e-06,
+      "loss": 0.8026,
+      "step": 4354
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.513815124466424,
+      "learning_rate": 5.284529770428997e-06,
+      "loss": 0.7549,
+      "step": 4355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6869729516841416,
+      "learning_rate": 5.280158589065691e-06,
+      "loss": 0.8434,
+      "step": 4356
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.658999602038804,
+      "learning_rate": 5.275788567707036e-06,
+      "loss": 0.9011,
+      "step": 4357
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.625186278881105,
+      "learning_rate": 5.271419707427056e-06,
+      "loss": 0.8522,
+      "step": 4358
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8677496677154295,
+      "learning_rate": 5.267052009299494e-06,
+      "loss": 0.827,
+      "step": 4359
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.046371336945512,
+      "learning_rate": 5.2626854743978065e-06,
+      "loss": 0.7203,
+      "step": 4360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.631751934773863,
+      "learning_rate": 5.258320103795162e-06,
+      "loss": 0.8268,
+      "step": 4361
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.581228394407631,
+      "learning_rate": 5.253955898564456e-06,
+      "loss": 0.8345,
+      "step": 4362
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8536806404973762,
+      "learning_rate": 5.249592859778281e-06,
+      "loss": 0.9339,
+      "step": 4363
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.381748068008239,
+      "learning_rate": 5.245230988508947e-06,
+      "loss": 0.8614,
+      "step": 4364
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6552696050449454,
+      "learning_rate": 5.240870285828489e-06,
+      "loss": 0.7775,
+      "step": 4365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5855320318893997,
+      "learning_rate": 5.236510752808639e-06,
+      "loss": 0.6907,
+      "step": 4366
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.844571555995099,
+      "learning_rate": 5.232152390520845e-06,
+      "loss": 0.8022,
+      "step": 4367
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6640972709004926,
+      "learning_rate": 5.227795200036279e-06,
+      "loss": 0.7807,
+      "step": 4368
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5939197694825897,
+      "learning_rate": 5.223439182425809e-06,
+      "loss": 0.7405,
+      "step": 4369
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4464154437815178,
+      "learning_rate": 5.219084338760025e-06,
+      "loss": 0.8124,
+      "step": 4370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6917914716514706,
+      "learning_rate": 5.214730670109227e-06,
+      "loss": 0.8367,
+      "step": 4371
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3822181150948696,
+      "learning_rate": 5.210378177543416e-06,
+      "loss": 0.7319,
+      "step": 4372
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6997257949647064,
+      "learning_rate": 5.206026862132324e-06,
+      "loss": 0.7314,
+      "step": 4373
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.603477960959868,
+      "learning_rate": 5.201676724945377e-06,
+      "loss": 0.8084,
+      "step": 4374
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6164521221019186,
+      "learning_rate": 5.197327767051713e-06,
+      "loss": 0.7139,
+      "step": 4375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.661036351334042,
+      "learning_rate": 5.192979989520193e-06,
+      "loss": 0.7833,
+      "step": 4376
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.609577223160548,
+      "learning_rate": 5.188633393419371e-06,
+      "loss": 0.8232,
+      "step": 4377
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.481068145446453,
+      "learning_rate": 5.184287979817519e-06,
+      "loss": 0.7904,
+      "step": 4378
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.989034486269615,
+      "learning_rate": 5.179943749782623e-06,
+      "loss": 0.803,
+      "step": 4379
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7008982367335648,
+      "learning_rate": 5.175600704382371e-06,
+      "loss": 0.804,
+      "step": 4380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5881479536178684,
+      "learning_rate": 5.171258844684155e-06,
+      "loss": 0.8173,
+      "step": 4381
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.813293310899786,
+      "learning_rate": 5.166918171755097e-06,
+      "loss": 0.7653,
+      "step": 4382
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.844039680058116,
+      "learning_rate": 5.1625786866619955e-06,
+      "loss": 0.8475,
+      "step": 4383
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2181719756764,
+      "learning_rate": 5.158240390471385e-06,
+      "loss": 0.7642,
+      "step": 4384
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.606973473588288,
+      "learning_rate": 5.153903284249495e-06,
+      "loss": 0.712,
+      "step": 4385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2000365788149314,
+      "learning_rate": 5.149567369062261e-06,
+      "loss": 0.686,
+      "step": 4386
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9467566513323096,
+      "learning_rate": 5.145232645975336e-06,
+      "loss": 0.8953,
+      "step": 4387
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7557886814227865,
+      "learning_rate": 5.140899116054068e-06,
+      "loss": 0.8869,
+      "step": 4388
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.674056263695096,
+      "learning_rate": 5.136566780363515e-06,
+      "loss": 0.7777,
+      "step": 4389
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.617838450703711,
+      "learning_rate": 5.1322356399684525e-06,
+      "loss": 0.8102,
+      "step": 4390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8643287708523495,
+      "learning_rate": 5.127905695933343e-06,
+      "loss": 0.7876,
+      "step": 4391
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.567823294869154,
+      "learning_rate": 5.123576949322375e-06,
+      "loss": 0.9056,
+      "step": 4392
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.687122412429552,
+      "learning_rate": 5.119249401199428e-06,
+      "loss": 0.7666,
+      "step": 4393
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5083630162065336,
+      "learning_rate": 5.114923052628092e-06,
+      "loss": 0.7502,
+      "step": 4394
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.464663395166503,
+      "learning_rate": 5.110597904671664e-06,
+      "loss": 0.745,
+      "step": 4395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8353050509079107,
+      "learning_rate": 5.106273958393142e-06,
+      "loss": 0.7786,
+      "step": 4396
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.872102247273557,
+      "learning_rate": 5.101951214855226e-06,
+      "loss": 0.8282,
+      "step": 4397
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.730776359515011,
+      "learning_rate": 5.097629675120336e-06,
+      "loss": 0.8991,
+      "step": 4398
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7406506612238153,
+      "learning_rate": 5.093309340250578e-06,
+      "loss": 0.8758,
+      "step": 4399
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.773308560374021,
+      "learning_rate": 5.0889902113077695e-06,
+      "loss": 0.8074,
+      "step": 4400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.845091034005279,
+      "learning_rate": 5.084672289353435e-06,
+      "loss": 0.8554,
+      "step": 4401
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7187441930311462,
+      "learning_rate": 5.080355575448792e-06,
+      "loss": 0.7727,
+      "step": 4402
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.5786635844494543,
+      "learning_rate": 5.076040070654778e-06,
+      "loss": 0.8989,
+      "step": 4403
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5742272068838417,
+      "learning_rate": 5.071725776032015e-06,
+      "loss": 0.7542,
+      "step": 4404
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.389808986839853,
+      "learning_rate": 5.067412692640839e-06,
+      "loss": 0.7614,
+      "step": 4405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.754195864365841,
+      "learning_rate": 5.063100821541281e-06,
+      "loss": 0.8037,
+      "step": 4406
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.867365347124313,
+      "learning_rate": 5.058790163793083e-06,
+      "loss": 0.744,
+      "step": 4407
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.629332336481846,
+      "learning_rate": 5.054480720455677e-06,
+      "loss": 0.7491,
+      "step": 4408
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.603839244071808,
+      "learning_rate": 5.05017249258821e-06,
+      "loss": 0.9164,
+      "step": 4409
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.702141953179618,
+      "learning_rate": 5.045865481249523e-06,
+      "loss": 0.784,
+      "step": 4410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5178397816644646,
+      "learning_rate": 5.041559687498152e-06,
+      "loss": 0.8736,
+      "step": 4411
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5196806497253896,
+      "learning_rate": 5.03725511239235e-06,
+      "loss": 0.8599,
+      "step": 4412
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4533128431782645,
+      "learning_rate": 5.03295175699005e-06,
+      "loss": 0.8331,
+      "step": 4413
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5491905479689083,
+      "learning_rate": 5.0286496223489075e-06,
+      "loss": 0.8176,
+      "step": 4414
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.667131326142279,
+      "learning_rate": 5.0243487095262615e-06,
+      "loss": 0.8368,
+      "step": 4415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.507835981837654,
+      "learning_rate": 5.020049019579154e-06,
+      "loss": 0.8017,
+      "step": 4416
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7554033825056803,
+      "learning_rate": 5.015750553564331e-06,
+      "loss": 0.8758,
+      "step": 4417
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5901696392138494,
+      "learning_rate": 5.011453312538233e-06,
+      "loss": 0.7932,
+      "step": 4418
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6740393141937977,
+      "learning_rate": 5.007157297556997e-06,
+      "loss": 0.725,
+      "step": 4419
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.805224200929592,
+      "learning_rate": 5.002862509676471e-06,
+      "loss": 0.8622,
+      "step": 4420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.675242937946728,
+      "learning_rate": 4.998568949952192e-06,
+      "loss": 0.8288,
+      "step": 4421
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6625074354824894,
+      "learning_rate": 4.99427661943939e-06,
+      "loss": 0.8326,
+      "step": 4422
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4338348325708035,
+      "learning_rate": 4.989985519193008e-06,
+      "loss": 0.7297,
+      "step": 4423
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.762201107569858,
+      "learning_rate": 4.98569565026767e-06,
+      "loss": 0.8451,
+      "step": 4424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6072736360156683,
+      "learning_rate": 4.981407013717714e-06,
+      "loss": 0.7758,
+      "step": 4425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.758225364934191,
+      "learning_rate": 4.977119610597163e-06,
+      "loss": 0.8361,
+      "step": 4426
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.542021886033243,
+      "learning_rate": 4.972833441959739e-06,
+      "loss": 0.7494,
+      "step": 4427
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8372559940748414,
+      "learning_rate": 4.968548508858863e-06,
+      "loss": 0.7811,
+      "step": 4428
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.851293790673043,
+      "learning_rate": 4.964264812347651e-06,
+      "loss": 0.6002,
+      "step": 4429
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5694384446828313,
+      "learning_rate": 4.959982353478911e-06,
+      "loss": 0.7817,
+      "step": 4430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8666717398217996,
+      "learning_rate": 4.955701133305162e-06,
+      "loss": 0.8643,
+      "step": 4431
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.521520650821122,
+      "learning_rate": 4.9514211528786e-06,
+      "loss": 0.8003,
+      "step": 4432
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.361515154255389,
+      "learning_rate": 4.9471424132511224e-06,
+      "loss": 0.7376,
+      "step": 4433
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.510640568551166,
+      "learning_rate": 4.942864915474331e-06,
+      "loss": 0.827,
+      "step": 4434
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.53168907265518,
+      "learning_rate": 4.9385886605995075e-06,
+      "loss": 0.7962,
+      "step": 4435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.392625479649822,
+      "learning_rate": 4.934313649677641e-06,
+      "loss": 0.6972,
+      "step": 4436
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.537304145927643,
+      "learning_rate": 4.9300398837594076e-06,
+      "loss": 0.771,
+      "step": 4437
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.610679280539628,
+      "learning_rate": 4.925767363895179e-06,
+      "loss": 0.7721,
+      "step": 4438
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9022231835808507,
+      "learning_rate": 4.9214960911350185e-06,
+      "loss": 0.753,
+      "step": 4439
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7496681277602684,
+      "learning_rate": 4.917226066528689e-06,
+      "loss": 0.7837,
+      "step": 4440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6893999122809604,
+      "learning_rate": 4.912957291125635e-06,
+      "loss": 0.7454,
+      "step": 4441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.740717636172206,
+      "learning_rate": 4.908689765975012e-06,
+      "loss": 0.807,
+      "step": 4442
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4983590928601926,
+      "learning_rate": 4.904423492125653e-06,
+      "loss": 0.7373,
+      "step": 4443
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6187129268123734,
+      "learning_rate": 4.900158470626085e-06,
+      "loss": 0.7673,
+      "step": 4444
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6988355035420315,
+      "learning_rate": 4.895894702524538e-06,
+      "loss": 0.7332,
+      "step": 4445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.802231068626159,
+      "learning_rate": 4.891632188868921e-06,
+      "loss": 0.7753,
+      "step": 4446
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4975003336574,
+      "learning_rate": 4.887370930706845e-06,
+      "loss": 0.7797,
+      "step": 4447
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9680313023184275,
+      "learning_rate": 4.8831109290856046e-06,
+      "loss": 0.8899,
+      "step": 4448
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.727326968393698,
+      "learning_rate": 4.8788521850521904e-06,
+      "loss": 0.826,
+      "step": 4449
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7302688744437136,
+      "learning_rate": 4.874594699653281e-06,
+      "loss": 0.8682,
+      "step": 4450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.492593665785906,
+      "learning_rate": 4.870338473935246e-06,
+      "loss": 0.7884,
+      "step": 4451
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5666077831783607,
+      "learning_rate": 4.866083508944145e-06,
+      "loss": 0.7734,
+      "step": 4452
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9646458483096367,
+      "learning_rate": 4.8618298057257355e-06,
+      "loss": 0.6841,
+      "step": 4453
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.566984184996757,
+      "learning_rate": 4.857577365325452e-06,
+      "loss": 0.7161,
+      "step": 4454
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.576733376569184,
+      "learning_rate": 4.853326188788425e-06,
+      "loss": 0.8825,
+      "step": 4455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8115505630557847,
+      "learning_rate": 4.849076277159481e-06,
+      "loss": 0.9294,
+      "step": 4456
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.944850469212431,
+      "learning_rate": 4.844827631483121e-06,
+      "loss": 0.7999,
+      "step": 4457
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.708885597821722,
+      "learning_rate": 4.840580252803552e-06,
+      "loss": 0.7353,
+      "step": 4458
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.944819532038206,
+      "learning_rate": 4.836334142164654e-06,
+      "loss": 0.8249,
+      "step": 4459
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.739948538835541,
+      "learning_rate": 4.832089300610003e-06,
+      "loss": 0.8552,
+      "step": 4460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7541977968082025,
+      "learning_rate": 4.8278457291828625e-06,
+      "loss": 0.7478,
+      "step": 4461
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.765934044481929,
+      "learning_rate": 4.823603428926185e-06,
+      "loss": 0.8427,
+      "step": 4462
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.62684539168531,
+      "learning_rate": 4.819362400882602e-06,
+      "loss": 0.7443,
+      "step": 4463
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.666412253891732,
+      "learning_rate": 4.815122646094448e-06,
+      "loss": 0.8608,
+      "step": 4464
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.58003459090976,
+      "learning_rate": 4.8108841656037295e-06,
+      "loss": 0.829,
+      "step": 4465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.540279478810456,
+      "learning_rate": 4.806646960452151e-06,
+      "loss": 0.7526,
+      "step": 4466
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5625038092096126,
+      "learning_rate": 4.802411031681099e-06,
+      "loss": 0.8165,
+      "step": 4467
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9700586139058855,
+      "learning_rate": 4.798176380331638e-06,
+      "loss": 0.7515,
+      "step": 4468
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8773827775527314,
+      "learning_rate": 4.793943007444536e-06,
+      "loss": 0.6777,
+      "step": 4469
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6582906590904454,
+      "learning_rate": 4.789710914060234e-06,
+      "loss": 0.8083,
+      "step": 4470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.528804350071189,
+      "learning_rate": 4.785480101218861e-06,
+      "loss": 0.7438,
+      "step": 4471
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.542769406287971,
+      "learning_rate": 4.781250569960233e-06,
+      "loss": 0.8081,
+      "step": 4472
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.693399220911003,
+      "learning_rate": 4.777022321323849e-06,
+      "loss": 0.9128,
+      "step": 4473
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6851606809802746,
+      "learning_rate": 4.77279535634889e-06,
+      "loss": 0.8111,
+      "step": 4474
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5435598261018204,
+      "learning_rate": 4.768569676074235e-06,
+      "loss": 0.764,
+      "step": 4475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.584009644323771,
+      "learning_rate": 4.764345281538428e-06,
+      "loss": 0.7162,
+      "step": 4476
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.62466241921889,
+      "learning_rate": 4.760122173779715e-06,
+      "loss": 0.8222,
+      "step": 4477
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.533256258740026,
+      "learning_rate": 4.755900353836015e-06,
+      "loss": 0.7704,
+      "step": 4478
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.861712170497619,
+      "learning_rate": 4.751679822744928e-06,
+      "loss": 0.8504,
+      "step": 4479
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7018656507970547,
+      "learning_rate": 4.747460581543749e-06,
+      "loss": 0.7879,
+      "step": 4480
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7388923120779887,
+      "learning_rate": 4.743242631269445e-06,
+      "loss": 0.7889,
+      "step": 4481
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.855257683308518,
+      "learning_rate": 4.739025972958673e-06,
+      "loss": 0.8472,
+      "step": 4482
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.845673836028014,
+      "learning_rate": 4.734810607647766e-06,
+      "loss": 0.7232,
+      "step": 4483
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.825542516978436,
+      "learning_rate": 4.730596536372745e-06,
+      "loss": 0.7726,
+      "step": 4484
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3857235505175716,
+      "learning_rate": 4.726383760169304e-06,
+      "loss": 0.7034,
+      "step": 4485
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.530643930906867,
+      "learning_rate": 4.722172280072835e-06,
+      "loss": 0.7931,
+      "step": 4486
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.657113264744651,
+      "learning_rate": 4.717962097118394e-06,
+      "loss": 0.8575,
+      "step": 4487
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.713785602161177,
+      "learning_rate": 4.713753212340732e-06,
+      "loss": 0.8137,
+      "step": 4488
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5858389823910195,
+      "learning_rate": 4.709545626774273e-06,
+      "loss": 0.8343,
+      "step": 4489
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.556127744178826,
+      "learning_rate": 4.705339341453119e-06,
+      "loss": 0.7033,
+      "step": 4490
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.370308093136086,
+      "learning_rate": 4.701134357411065e-06,
+      "loss": 0.7872,
+      "step": 4491
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.740060254795805,
+      "learning_rate": 4.696930675681571e-06,
+      "loss": 0.7007,
+      "step": 4492
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.632838666024672,
+      "learning_rate": 4.692728297297785e-06,
+      "loss": 0.8182,
+      "step": 4493
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.703412372670818,
+      "learning_rate": 4.6885272232925426e-06,
+      "loss": 0.8383,
+      "step": 4494
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.549367641767443,
+      "learning_rate": 4.68432745469834e-06,
+      "loss": 0.8128,
+      "step": 4495
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5904592992356466,
+      "learning_rate": 4.68012899254736e-06,
+      "loss": 0.7876,
+      "step": 4496
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8401619973861996,
+      "learning_rate": 4.675931837871477e-06,
+      "loss": 0.7499,
+      "step": 4497
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0015771959374127,
+      "learning_rate": 4.671735991702225e-06,
+      "loss": 0.9088,
+      "step": 4498
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5500184314648617,
+      "learning_rate": 4.667541455070834e-06,
+      "loss": 0.7474,
+      "step": 4499
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6213439037770105,
+      "learning_rate": 4.663348229008199e-06,
+      "loss": 0.7026,
+      "step": 4500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6287716051634447,
+      "learning_rate": 4.659156314544893e-06,
+      "loss": 0.8177,
+      "step": 4501
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.459424953526767,
+      "learning_rate": 4.654965712711178e-06,
+      "loss": 0.7863,
+      "step": 4502
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.587294325712288,
+      "learning_rate": 4.6507764245369855e-06,
+      "loss": 0.8057,
+      "step": 4503
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.396969034639195,
+      "learning_rate": 4.646588451051919e-06,
+      "loss": 0.7544,
+      "step": 4504
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6668852980000266,
+      "learning_rate": 4.642401793285271e-06,
+      "loss": 0.8272,
+      "step": 4505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5722160264858815,
+      "learning_rate": 4.6382164522660055e-06,
+      "loss": 0.7214,
+      "step": 4506
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8202500748321375,
+      "learning_rate": 4.634032429022758e-06,
+      "loss": 0.8385,
+      "step": 4507
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.265580443189982,
+      "learning_rate": 4.629849724583846e-06,
+      "loss": 0.9186,
+      "step": 4508
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7453839692216855,
+      "learning_rate": 4.625668339977255e-06,
+      "loss": 0.7932,
+      "step": 4509
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.625301799758122,
+      "learning_rate": 4.621488276230662e-06,
+      "loss": 0.8853,
+      "step": 4510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4718008492512764,
+      "learning_rate": 4.617309534371404e-06,
+      "loss": 0.7119,
+      "step": 4511
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.641255755127557,
+      "learning_rate": 4.613132115426496e-06,
+      "loss": 0.7424,
+      "step": 4512
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5223496054941363,
+      "learning_rate": 4.608956020422638e-06,
+      "loss": 0.7684,
+      "step": 4513
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.561888272311924,
+      "learning_rate": 4.604781250386191e-06,
+      "loss": 0.8153,
+      "step": 4514
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4367407854667174,
+      "learning_rate": 4.600607806343196e-06,
+      "loss": 0.682,
+      "step": 4515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7086669415237523,
+      "learning_rate": 4.596435689319374e-06,
+      "loss": 0.8494,
+      "step": 4516
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.840924918015386,
+      "learning_rate": 4.59226490034011e-06,
+      "loss": 0.8094,
+      "step": 4517
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3463652209327797,
+      "learning_rate": 4.588095440430469e-06,
+      "loss": 0.7552,
+      "step": 4518
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.608067751840558,
+      "learning_rate": 4.583927310615185e-06,
+      "loss": 0.8524,
+      "step": 4519
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5728811216388427,
+      "learning_rate": 4.579760511918666e-06,
+      "loss": 0.7234,
+      "step": 4520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.688885009168033,
+      "learning_rate": 4.575595045365e-06,
+      "loss": 0.8409,
+      "step": 4521
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5940511926032985,
+      "learning_rate": 4.5714309119779385e-06,
+      "loss": 0.7412,
+      "step": 4522
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6332325498634135,
+      "learning_rate": 4.567268112780906e-06,
+      "loss": 0.8335,
+      "step": 4523
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7237828697539728,
+      "learning_rate": 4.563106648797008e-06,
+      "loss": 0.8518,
+      "step": 4524
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5714001553129493,
+      "learning_rate": 4.558946521049011e-06,
+      "loss": 0.8423,
+      "step": 4525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6647391578755824,
+      "learning_rate": 4.554787730559357e-06,
+      "loss": 0.8729,
+      "step": 4526
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.669061939275215,
+      "learning_rate": 4.550630278350165e-06,
+      "loss": 0.9408,
+      "step": 4527
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.508578076573883,
+      "learning_rate": 4.546474165443219e-06,
+      "loss": 0.8723,
+      "step": 4528
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.687921052688467,
+      "learning_rate": 4.542319392859972e-06,
+      "loss": 0.7999,
+      "step": 4529
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.492568968275793,
+      "learning_rate": 4.538165961621552e-06,
+      "loss": 0.7753,
+      "step": 4530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.526572412400771,
+      "learning_rate": 4.5340138727487505e-06,
+      "loss": 0.8778,
+      "step": 4531
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.872178179667303,
+      "learning_rate": 4.529863127262045e-06,
+      "loss": 0.7821,
+      "step": 4532
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6176834783726637,
+      "learning_rate": 4.525713726181567e-06,
+      "loss": 0.7875,
+      "step": 4533
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7386225860310667,
+      "learning_rate": 4.521565670527119e-06,
+      "loss": 0.896,
+      "step": 4534
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.5227923641482493,
+      "learning_rate": 4.517418961318185e-06,
+      "loss": 0.8407,
+      "step": 4535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.91481708532559,
+      "learning_rate": 4.513273599573906e-06,
+      "loss": 0.853,
+      "step": 4536
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4922437195264537,
+      "learning_rate": 4.50912958631309e-06,
+      "loss": 0.7683,
+      "step": 4537
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.526874141534352,
+      "learning_rate": 4.504986922554229e-06,
+      "loss": 0.7781,
+      "step": 4538
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.4559635537854567,
+      "learning_rate": 4.500845609315468e-06,
+      "loss": 0.8497,
+      "step": 4539
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.524651553011375,
+      "learning_rate": 4.496705647614628e-06,
+      "loss": 0.6886,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8576812336465296,
+      "learning_rate": 4.492567038469194e-06,
+      "loss": 0.7955,
+      "step": 4541
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7588167195369295,
+      "learning_rate": 4.488429782896315e-06,
+      "loss": 0.7579,
+      "step": 4542
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5805819596962962,
+      "learning_rate": 4.484293881912823e-06,
+      "loss": 0.8285,
+      "step": 4543
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5669736328115,
+      "learning_rate": 4.4801593365352e-06,
+      "loss": 0.8195,
+      "step": 4544
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.645156566188588,
+      "learning_rate": 4.4760261477796e-06,
+      "loss": 0.782,
+      "step": 4545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8599751400315565,
+      "learning_rate": 4.47189431666185e-06,
+      "loss": 0.7474,
+      "step": 4546
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.618919615293604,
+      "learning_rate": 4.4677638441974344e-06,
+      "loss": 0.7622,
+      "step": 4547
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5745283079506494,
+      "learning_rate": 4.463634731401506e-06,
+      "loss": 0.7358,
+      "step": 4548
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.768223479539422,
+      "learning_rate": 4.459506979288891e-06,
+      "loss": 0.8084,
+      "step": 4549
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.66595138445493,
+      "learning_rate": 4.455380588874072e-06,
+      "loss": 0.859,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5302475002778815,
+      "learning_rate": 4.451255561171199e-06,
+      "loss": 0.8463,
+      "step": 4551
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.828214551033672,
+      "learning_rate": 4.447131897194089e-06,
+      "loss": 0.8653,
+      "step": 4552
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.577556922731538,
+      "learning_rate": 4.443009597956219e-06,
+      "loss": 0.8231,
+      "step": 4553
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.616616110701205,
+      "learning_rate": 4.4388886644707415e-06,
+      "loss": 0.8405,
+      "step": 4554
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6355056781615405,
+      "learning_rate": 4.434769097750463e-06,
+      "loss": 0.9045,
+      "step": 4555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.464200613183843,
+      "learning_rate": 4.4306508988078545e-06,
+      "loss": 0.7632,
+      "step": 4556
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5987579746298515,
+      "learning_rate": 4.42653406865506e-06,
+      "loss": 0.7914,
+      "step": 4557
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.396848709039805,
+      "learning_rate": 4.422418608303878e-06,
+      "loss": 0.8756,
+      "step": 4558
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.564281271524151,
+      "learning_rate": 4.418304518765768e-06,
+      "loss": 0.766,
+      "step": 4559
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.893019951497073,
+      "learning_rate": 4.414191801051868e-06,
+      "loss": 0.9191,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.643477472378628,
+      "learning_rate": 4.410080456172963e-06,
+      "loss": 0.7728,
+      "step": 4561
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.191959381783351,
+      "learning_rate": 4.4059704851395066e-06,
+      "loss": 0.7745,
+      "step": 4562
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7803203795034017,
+      "learning_rate": 4.401861888961614e-06,
+      "loss": 0.8424,
+      "step": 4563
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6668926762762557,
+      "learning_rate": 4.39775466864906e-06,
+      "loss": 0.8434,
+      "step": 4564
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2259726953810026,
+      "learning_rate": 4.393648825211292e-06,
+      "loss": 0.8012,
+      "step": 4565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.485281211765713,
+      "learning_rate": 4.389544359657406e-06,
+      "loss": 0.7123,
+      "step": 4566
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7361707651748524,
+      "learning_rate": 4.385441272996163e-06,
+      "loss": 0.8163,
+      "step": 4567
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6342215617089617,
+      "learning_rate": 4.381339566235991e-06,
+      "loss": 0.7742,
+      "step": 4568
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4953478324533265,
+      "learning_rate": 4.3772392403849725e-06,
+      "loss": 0.8489,
+      "step": 4569
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8254277012049007,
+      "learning_rate": 4.373140296450849e-06,
+      "loss": 0.7949,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.447166580679297,
+      "learning_rate": 4.369042735441034e-06,
+      "loss": 0.6744,
+      "step": 4571
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5148934014001267,
+      "learning_rate": 4.364946558362587e-06,
+      "loss": 0.7639,
+      "step": 4572
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.631815990263399,
+      "learning_rate": 4.360851766222236e-06,
+      "loss": 0.7963,
+      "step": 4573
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7870802311280447,
+      "learning_rate": 4.356758360026364e-06,
+      "loss": 0.7633,
+      "step": 4574
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6983783957529743,
+      "learning_rate": 4.352666340781014e-06,
+      "loss": 0.7501,
+      "step": 4575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7473802269485996,
+      "learning_rate": 4.348575709491895e-06,
+      "loss": 0.7177,
+      "step": 4576
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7337053133225835,
+      "learning_rate": 4.344486467164367e-06,
+      "loss": 0.8403,
+      "step": 4577
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5316406942155516,
+      "learning_rate": 4.340398614803446e-06,
+      "loss": 0.7514,
+      "step": 4578
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.616088844833526,
+      "learning_rate": 4.336312153413821e-06,
+      "loss": 0.735,
+      "step": 4579
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5590860827417563,
+      "learning_rate": 4.332227083999824e-06,
+      "loss": 0.7324,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.823463575405153,
+      "learning_rate": 4.328143407565446e-06,
+      "loss": 0.8872,
+      "step": 4581
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5845287530717598,
+      "learning_rate": 4.3240611251143504e-06,
+      "loss": 0.8051,
+      "step": 4582
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8261552773288603,
+      "learning_rate": 4.319980237649842e-06,
+      "loss": 0.8085,
+      "step": 4583
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.281298269414887,
+      "learning_rate": 4.315900746174891e-06,
+      "loss": 0.7678,
+      "step": 4584
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7115424664869425,
+      "learning_rate": 4.311822651692117e-06,
+      "loss": 0.8394,
+      "step": 4585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7292516435578658,
+      "learning_rate": 4.307745955203802e-06,
+      "loss": 0.7851,
+      "step": 4586
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.867908725524442,
+      "learning_rate": 4.30367065771189e-06,
+      "loss": 0.9014,
+      "step": 4587
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.539392293579164,
+      "learning_rate": 4.299596760217971e-06,
+      "loss": 0.9133,
+      "step": 4588
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.595803782835126,
+      "learning_rate": 4.295524263723289e-06,
+      "loss": 0.8473,
+      "step": 4589
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.051305067876277,
+      "learning_rate": 4.291453169228761e-06,
+      "loss": 0.953,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.556206565324167,
+      "learning_rate": 4.287383477734941e-06,
+      "loss": 0.767,
+      "step": 4591
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4954943318160723,
+      "learning_rate": 4.283315190242043e-06,
+      "loss": 0.8481,
+      "step": 4592
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.58254843562837,
+      "learning_rate": 4.279248307749942e-06,
+      "loss": 0.7255,
+      "step": 4593
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.438739075744588,
+      "learning_rate": 4.275182831258165e-06,
+      "loss": 0.7969,
+      "step": 4594
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7256701612456022,
+      "learning_rate": 4.2711187617658874e-06,
+      "loss": 0.7933,
+      "step": 4595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.501946269627395,
+      "learning_rate": 4.2670561002719465e-06,
+      "loss": 0.7289,
+      "step": 4596
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6861185679974433,
+      "learning_rate": 4.262994847774826e-06,
+      "loss": 0.7463,
+      "step": 4597
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6513030959735073,
+      "learning_rate": 4.258935005272677e-06,
+      "loss": 0.86,
+      "step": 4598
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5170003857043337,
+      "learning_rate": 4.254876573763287e-06,
+      "loss": 0.8417,
+      "step": 4599
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5795342270973713,
+      "learning_rate": 4.250819554244104e-06,
+      "loss": 0.7703,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2948253468316304,
+      "learning_rate": 4.2467639477122365e-06,
+      "loss": 0.7226,
+      "step": 4601
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.426464752391824,
+      "learning_rate": 4.242709755164436e-06,
+      "loss": 0.7803,
+      "step": 4602
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.542434876466958,
+      "learning_rate": 4.238656977597104e-06,
+      "loss": 0.7254,
+      "step": 4603
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6932781263047523,
+      "learning_rate": 4.234605616006309e-06,
+      "loss": 0.8503,
+      "step": 4604
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.596101290817688,
+      "learning_rate": 4.230555671387752e-06,
+      "loss": 0.7327,
+      "step": 4605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.517592945930184,
+      "learning_rate": 4.2265071447368075e-06,
+      "loss": 0.7414,
+      "step": 4606
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6225552220576747,
+      "learning_rate": 4.222460037048481e-06,
+      "loss": 0.7787,
+      "step": 4607
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.682718897322715,
+      "learning_rate": 4.218414349317435e-06,
+      "loss": 0.8763,
+      "step": 4608
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.539578696281544,
+      "learning_rate": 4.214370082537996e-06,
+      "loss": 0.7671,
+      "step": 4609
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4552266766915336,
+      "learning_rate": 4.210327237704127e-06,
+      "loss": 0.7058,
+      "step": 4610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.450334006455967,
+      "learning_rate": 4.206285815809442e-06,
+      "loss": 0.7708,
+      "step": 4611
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5675380169728568,
+      "learning_rate": 4.202245817847216e-06,
+      "loss": 0.7293,
+      "step": 4612
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8543763293592983,
+      "learning_rate": 4.198207244810359e-06,
+      "loss": 0.7996,
+      "step": 4613
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5847582955676844,
+      "learning_rate": 4.194170097691449e-06,
+      "loss": 0.8363,
+      "step": 4614
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.607985959941781,
+      "learning_rate": 4.190134377482697e-06,
+      "loss": 0.8314,
+      "step": 4615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5181490067330885,
+      "learning_rate": 4.186100085175969e-06,
+      "loss": 0.7393,
+      "step": 4616
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7186874436591073,
+      "learning_rate": 4.182067221762787e-06,
+      "loss": 0.7471,
+      "step": 4617
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.631842047505717,
+      "learning_rate": 4.17803578823431e-06,
+      "loss": 0.8025,
+      "step": 4618
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.681078894285222,
+      "learning_rate": 4.174005785581355e-06,
+      "loss": 0.8648,
+      "step": 4619
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5036096911340064,
+      "learning_rate": 4.169977214794381e-06,
+      "loss": 0.7002,
+      "step": 4620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4533717993242927,
+      "learning_rate": 4.165950076863498e-06,
+      "loss": 0.7863,
+      "step": 4621
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5486747992946803,
+      "learning_rate": 4.161924372778461e-06,
+      "loss": 0.8162,
+      "step": 4622
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8385199190995167,
+      "learning_rate": 4.157900103528681e-06,
+      "loss": 0.826,
+      "step": 4623
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.652735195152087,
+      "learning_rate": 4.153877270103205e-06,
+      "loss": 0.8017,
+      "step": 4624
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8035117018749363,
+      "learning_rate": 4.149855873490739e-06,
+      "loss": 0.7838,
+      "step": 4625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8177744943240155,
+      "learning_rate": 4.145835914679624e-06,
+      "loss": 0.7602,
+      "step": 4626
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.702208093817233,
+      "learning_rate": 4.141817394657854e-06,
+      "loss": 0.7792,
+      "step": 4627
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4106613142451736,
+      "learning_rate": 4.137800314413072e-06,
+      "loss": 0.7998,
+      "step": 4628
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4824405757189627,
+      "learning_rate": 4.1337846749325615e-06,
+      "loss": 0.7364,
+      "step": 4629
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.422022193044217,
+      "learning_rate": 4.129770477203253e-06,
+      "loss": 0.7391,
+      "step": 4630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.391710828345458,
+      "learning_rate": 4.125757722211725e-06,
+      "loss": 0.751,
+      "step": 4631
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.650588884023421,
+      "learning_rate": 4.1217464109442e-06,
+      "loss": 0.7595,
+      "step": 4632
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5578125265610416,
+      "learning_rate": 4.11773654438654e-06,
+      "loss": 0.8018,
+      "step": 4633
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.728508450144264,
+      "learning_rate": 4.113728123524269e-06,
+      "loss": 0.7575,
+      "step": 4634
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8877626647567696,
+      "learning_rate": 4.1097211493425335e-06,
+      "loss": 0.8282,
+      "step": 4635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6971660585681816,
+      "learning_rate": 4.105715622826144e-06,
+      "loss": 0.8925,
+      "step": 4636
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6891685662456877,
+      "learning_rate": 4.101711544959544e-06,
+      "loss": 0.8454,
+      "step": 4637
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6052488039847974,
+      "learning_rate": 4.097708916726818e-06,
+      "loss": 0.7644,
+      "step": 4638
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.500823823477732,
+      "learning_rate": 4.093707739111709e-06,
+      "loss": 0.7763,
+      "step": 4639
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1481662271475703,
+      "learning_rate": 4.0897080130975885e-06,
+      "loss": 0.7999,
+      "step": 4640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3819445548136193,
+      "learning_rate": 4.0857097396674785e-06,
+      "loss": 0.6348,
+      "step": 4641
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0016153917782913,
+      "learning_rate": 4.081712919804042e-06,
+      "loss": 0.8293,
+      "step": 4642
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.361916696935901,
+      "learning_rate": 4.077717554489585e-06,
+      "loss": 0.8334,
+      "step": 4643
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0327023999462233,
+      "learning_rate": 4.073723644706052e-06,
+      "loss": 0.8209,
+      "step": 4644
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7807531192439456,
+      "learning_rate": 4.069731191435043e-06,
+      "loss": 0.83,
+      "step": 4645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6526815820931104,
+      "learning_rate": 4.065740195657784e-06,
+      "loss": 0.7878,
+      "step": 4646
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.441045340728361,
+      "learning_rate": 4.061750658355154e-06,
+      "loss": 0.7812,
+      "step": 4647
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6057736597483467,
+      "learning_rate": 4.057762580507669e-06,
+      "loss": 0.8392,
+      "step": 4648
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5718849496433167,
+      "learning_rate": 4.053775963095482e-06,
+      "loss": 0.7951,
+      "step": 4649
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9414300092669694,
+      "learning_rate": 4.049790807098398e-06,
+      "loss": 0.7309,
+      "step": 4650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6655740741989957,
+      "learning_rate": 4.045807113495854e-06,
+      "loss": 0.7683,
+      "step": 4651
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.620533608701822,
+      "learning_rate": 4.04182488326693e-06,
+      "loss": 0.7355,
+      "step": 4652
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6881651036970546,
+      "learning_rate": 4.037844117390346e-06,
+      "loss": 0.7604,
+      "step": 4653
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1868902215429236,
+      "learning_rate": 4.033864816844464e-06,
+      "loss": 0.8145,
+      "step": 4654
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.596476042767454,
+      "learning_rate": 4.029886982607278e-06,
+      "loss": 0.7789,
+      "step": 4655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.479689038232687,
+      "learning_rate": 4.025910615656439e-06,
+      "loss": 0.7502,
+      "step": 4656
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.330613762856095,
+      "learning_rate": 4.021935716969217e-06,
+      "loss": 0.9225,
+      "step": 4657
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4228114155369878,
+      "learning_rate": 4.017962287522538e-06,
+      "loss": 0.7204,
+      "step": 4658
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4409170638890645,
+      "learning_rate": 4.013990328292956e-06,
+      "loss": 0.7502,
+      "step": 4659
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.702036399975697,
+      "learning_rate": 4.010019840256665e-06,
+      "loss": 0.7252,
+      "step": 4660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7340055852492298,
+      "learning_rate": 4.006050824389504e-06,
+      "loss": 0.7979,
+      "step": 4661
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6003021686125054,
+      "learning_rate": 4.002083281666944e-06,
+      "loss": 0.8506,
+      "step": 4662
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.7024305304131317,
+      "learning_rate": 3.998117213064096e-06,
+      "loss": 0.775,
+      "step": 4663
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3818260809227723,
+      "learning_rate": 3.994152619555708e-06,
+      "loss": 0.7644,
+      "step": 4664
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.4210541149434954,
+      "learning_rate": 3.990189502116165e-06,
+      "loss": 0.6938,
+      "step": 4665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8001810887021468,
+      "learning_rate": 3.986227861719489e-06,
+      "loss": 0.7897,
+      "step": 4666
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.638390741878354,
+      "learning_rate": 3.982267699339344e-06,
+      "loss": 0.8287,
+      "step": 4667
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5404188437109227,
+      "learning_rate": 3.978309015949022e-06,
+      "loss": 0.7848,
+      "step": 4668
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.725734600082859,
+      "learning_rate": 3.974351812521462e-06,
+      "loss": 0.8366,
+      "step": 4669
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6029752762499974,
+      "learning_rate": 3.970396090029231e-06,
+      "loss": 0.8097,
+      "step": 4670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5952620908186947,
+      "learning_rate": 3.9664418494445325e-06,
+      "loss": 0.8636,
+      "step": 4671
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7985463975433755,
+      "learning_rate": 3.9624890917392125e-06,
+      "loss": 0.877,
+      "step": 4672
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.290152520574593,
+      "learning_rate": 3.958537817884744e-06,
+      "loss": 0.8129,
+      "step": 4673
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6227573346481288,
+      "learning_rate": 3.9545880288522435e-06,
+      "loss": 0.78,
+      "step": 4674
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.245561388143318,
+      "learning_rate": 3.950639725612453e-06,
+      "loss": 0.827,
+      "step": 4675
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6497248469105936,
+      "learning_rate": 3.9466929091357585e-06,
+      "loss": 0.8241,
+      "step": 4676
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7854720270745683,
+      "learning_rate": 3.942747580392172e-06,
+      "loss": 0.76,
+      "step": 4677
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.453180020069248,
+      "learning_rate": 3.938803740351351e-06,
+      "loss": 0.8078,
+      "step": 4678
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.537980143611081,
+      "learning_rate": 3.934861389982576e-06,
+      "loss": 0.779,
+      "step": 4679
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.012175652906596,
+      "learning_rate": 3.93092053025477e-06,
+      "loss": 0.7289,
+      "step": 4680
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5317863879725326,
+      "learning_rate": 3.926981162136485e-06,
+      "loss": 0.703,
+      "step": 4681
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8787146878385266,
+      "learning_rate": 3.923043286595902e-06,
+      "loss": 0.7585,
+      "step": 4682
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7757103864171007,
+      "learning_rate": 3.919106904600849e-06,
+      "loss": 0.8586,
+      "step": 4683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.675438143895143,
+      "learning_rate": 3.915172017118773e-06,
+      "loss": 0.7801,
+      "step": 4684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.523180775382848,
+      "learning_rate": 3.91123862511676e-06,
+      "loss": 0.6941,
+      "step": 4685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5315978159852053,
+      "learning_rate": 3.907306729561528e-06,
+      "loss": 0.8646,
+      "step": 4686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.519907492637914,
+      "learning_rate": 3.903376331419421e-06,
+      "loss": 0.6856,
+      "step": 4687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5523043949939592,
+      "learning_rate": 3.89944743165643e-06,
+      "loss": 0.77,
+      "step": 4688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5257770082553246,
+      "learning_rate": 3.895520031238163e-06,
+      "loss": 0.7387,
+      "step": 4689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7803624676951646,
+      "learning_rate": 3.891594131129863e-06,
+      "loss": 0.8689,
+      "step": 4690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.652024181046161,
+      "learning_rate": 3.8876697322964115e-06,
+      "loss": 0.8326,
+      "step": 4691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.517253856703637,
+      "learning_rate": 3.883746835702314e-06,
+      "loss": 0.6858,
+      "step": 4692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.472269564693469,
+      "learning_rate": 3.879825442311704e-06,
+      "loss": 0.7663,
+      "step": 4693
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6853471906503454,
+      "learning_rate": 3.875905553088357e-06,
+      "loss": 0.7294,
+      "step": 4694
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.796561653655558,
+      "learning_rate": 3.871987168995668e-06,
+      "loss": 0.8129,
+      "step": 4695
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6928289478681053,
+      "learning_rate": 3.868070290996666e-06,
+      "loss": 0.8477,
+      "step": 4696
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.84870512570839,
+      "learning_rate": 3.86415492005401e-06,
+      "loss": 0.8197,
+      "step": 4697
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.739562097828992,
+      "learning_rate": 3.860241057129987e-06,
+      "loss": 0.7791,
+      "step": 4698
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.897960792207024,
+      "learning_rate": 3.856328703186517e-06,
+      "loss": 0.8865,
+      "step": 4699
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3354076946120386,
+      "learning_rate": 3.852417859185148e-06,
+      "loss": 0.6997,
+      "step": 4700
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4224047577876915,
+      "learning_rate": 3.848508526087049e-06,
+      "loss": 0.7737,
+      "step": 4701
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7918517638326215,
+      "learning_rate": 3.844600704853033e-06,
+      "loss": 0.7162,
+      "step": 4702
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8138739928062635,
+      "learning_rate": 3.840694396443529e-06,
+      "loss": 0.8866,
+      "step": 4703
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.456079097072732,
+      "learning_rate": 3.836789601818596e-06,
+      "loss": 0.8912,
+      "step": 4704
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6774767199549454,
+      "learning_rate": 3.832886321937927e-06,
+      "loss": 0.8849,
+      "step": 4705
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5867290705841226,
+      "learning_rate": 3.828984557760838e-06,
+      "loss": 0.7677,
+      "step": 4706
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.547855076935108,
+      "learning_rate": 3.825084310246271e-06,
+      "loss": 0.7806,
+      "step": 4707
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.373974068705608,
+      "learning_rate": 3.8211855803528e-06,
+      "loss": 0.7564,
+      "step": 4708
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5217411177994697,
+      "learning_rate": 3.817288369038617e-06,
+      "loss": 0.8766,
+      "step": 4709
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5738411725657246,
+      "learning_rate": 3.813392677261557e-06,
+      "loss": 0.8039,
+      "step": 4710
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7923442043533715,
+      "learning_rate": 3.8094985059790656e-06,
+      "loss": 0.7569,
+      "step": 4711
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4916064610001065,
+      "learning_rate": 3.8056058561482203e-06,
+      "loss": 0.8459,
+      "step": 4712
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6814032775087977,
+      "learning_rate": 3.8017147287257294e-06,
+      "loss": 0.7843,
+      "step": 4713
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7524991249628963,
+      "learning_rate": 3.7978251246679223e-06,
+      "loss": 0.7788,
+      "step": 4714
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2629708574261396,
+      "learning_rate": 3.7939370449307477e-06,
+      "loss": 0.8294,
+      "step": 4715
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4710880212646478,
+      "learning_rate": 3.7900504904697964e-06,
+      "loss": 0.8517,
+      "step": 4716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.693293620784226,
+      "learning_rate": 3.78616546224027e-06,
+      "loss": 0.8109,
+      "step": 4717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.02761216391464,
+      "learning_rate": 3.782281961197e-06,
+      "loss": 0.8265,
+      "step": 4718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.646484955340347,
+      "learning_rate": 3.77839998829444e-06,
+      "loss": 0.7394,
+      "step": 4719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4578067727484534,
+      "learning_rate": 3.774519544486669e-06,
+      "loss": 0.8447,
+      "step": 4720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6699138109227407,
+      "learning_rate": 3.7706406307273978e-06,
+      "loss": 0.8107,
+      "step": 4721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.630772578360136,
+      "learning_rate": 3.76676324796995e-06,
+      "loss": 0.7524,
+      "step": 4722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7318104679393693,
+      "learning_rate": 3.762887397167274e-06,
+      "loss": 0.7799,
+      "step": 4723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.762270216428474,
+      "learning_rate": 3.759013079271955e-06,
+      "loss": 0.8176,
+      "step": 4724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4553305532030305,
+      "learning_rate": 3.7551402952361837e-06,
+      "loss": 0.728,
+      "step": 4725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.6450089954589697,
+      "learning_rate": 3.751269046011782e-06,
+      "loss": 0.8179,
+      "step": 4726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.343328022644456,
+      "learning_rate": 3.7473993325502e-06,
+      "loss": 0.789,
+      "step": 4727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9286072628824007,
+      "learning_rate": 3.7435311558025013e-06,
+      "loss": 0.9128,
+      "step": 4728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.879903474320907,
+      "learning_rate": 3.739664516719371e-06,
+      "loss": 0.7496,
+      "step": 4729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4443020986868316,
+      "learning_rate": 3.735799416251129e-06,
+      "loss": 0.7422,
+      "step": 4730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7236021790499807,
+      "learning_rate": 3.7319358553477036e-06,
+      "loss": 0.6851,
+      "step": 4731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.419018248039533,
+      "learning_rate": 3.728073834958652e-06,
+      "loss": 0.7411,
+      "step": 4732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.024391110209093,
+      "learning_rate": 3.7242133560331474e-06,
+      "loss": 0.7883,
+      "step": 4733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9023134680031646,
+      "learning_rate": 3.720354419519986e-06,
+      "loss": 0.768,
+      "step": 4734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.576821852078959,
+      "learning_rate": 3.7164970263675927e-06,
+      "loss": 0.65,
+      "step": 4735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.59943955832557,
+      "learning_rate": 3.7126411775240034e-06,
+      "loss": 0.8005,
+      "step": 4736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.733894188548558,
+      "learning_rate": 3.7087868739368727e-06,
+      "loss": 0.7918,
+      "step": 4737
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5152087590772974,
+      "learning_rate": 3.704934116553488e-06,
+      "loss": 0.7243,
+      "step": 4738
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.270981881776336,
+      "learning_rate": 3.7010829063207464e-06,
+      "loss": 0.6764,
+      "step": 4739
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.092183390138597,
+      "learning_rate": 3.6972332441851633e-06,
+      "loss": 0.8128,
+      "step": 4740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6356895097666913,
+      "learning_rate": 3.6933851310928835e-06,
+      "loss": 0.7741,
+      "step": 4741
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5035855263563644,
+      "learning_rate": 3.6895385679896634e-06,
+      "loss": 0.7324,
+      "step": 4742
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.709479031414277,
+      "learning_rate": 3.6856935558208805e-06,
+      "loss": 0.7627,
+      "step": 4743
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.536976789978986,
+      "learning_rate": 3.6818500955315295e-06,
+      "loss": 0.811,
+      "step": 4744
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.613504187688428,
+      "learning_rate": 3.678008188066222e-06,
+      "loss": 0.794,
+      "step": 4745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6157031627687535,
+      "learning_rate": 3.6741678343691987e-06,
+      "loss": 0.8401,
+      "step": 4746
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.667752066641231,
+      "learning_rate": 3.670329035384308e-06,
+      "loss": 0.7574,
+      "step": 4747
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.542434226381904,
+      "learning_rate": 3.6664917920550138e-06,
+      "loss": 0.6986,
+      "step": 4748
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6856041087077585,
+      "learning_rate": 3.6626561053244102e-06,
+      "loss": 0.9379,
+      "step": 4749
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4464421333190356,
+      "learning_rate": 3.6588219761351997e-06,
+      "loss": 0.765,
+      "step": 4750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4903426439491416,
+      "learning_rate": 3.6549894054296987e-06,
+      "loss": 0.751,
+      "step": 4751
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7515001060158353,
+      "learning_rate": 3.651158394149852e-06,
+      "loss": 0.7518,
+      "step": 4752
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0337445733466644,
+      "learning_rate": 3.6473289432372127e-06,
+      "loss": 0.7475,
+      "step": 4753
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.395247071177547,
+      "learning_rate": 3.643501053632952e-06,
+      "loss": 0.6996,
+      "step": 4754
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.489066127142281,
+      "learning_rate": 3.6396747262778565e-06,
+      "loss": 0.7531,
+      "step": 4755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.840828936358385,
+      "learning_rate": 3.635849962112329e-06,
+      "loss": 0.7807,
+      "step": 4756
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4415939387484573,
+      "learning_rate": 3.6320267620763948e-06,
+      "loss": 0.8277,
+      "step": 4757
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7229664144214722,
+      "learning_rate": 3.628205127109685e-06,
+      "loss": 0.8241,
+      "step": 4758
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5781982638832446,
+      "learning_rate": 3.6243850581514497e-06,
+      "loss": 0.825,
+      "step": 4759
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5764834094870714,
+      "learning_rate": 3.620566556140558e-06,
+      "loss": 0.8209,
+      "step": 4760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.68885570646689,
+      "learning_rate": 3.616749622015486e-06,
+      "loss": 0.6946,
+      "step": 4761
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4970988849394127,
+      "learning_rate": 3.6129342567143354e-06,
+      "loss": 0.7678,
+      "step": 4762
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6849650774427007,
+      "learning_rate": 3.609120461174813e-06,
+      "loss": 0.7518,
+      "step": 4763
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6180254178226208,
+      "learning_rate": 3.6053082363342416e-06,
+      "loss": 0.7673,
+      "step": 4764
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7240023181640893,
+      "learning_rate": 3.601497583129561e-06,
+      "loss": 0.8813,
+      "step": 4765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4256243336055787,
+      "learning_rate": 3.5976885024973196e-06,
+      "loss": 0.7399,
+      "step": 4766
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5106095364830523,
+      "learning_rate": 3.593880995373683e-06,
+      "loss": 0.6945,
+      "step": 4767
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.789038453137751,
+      "learning_rate": 3.5900750626944346e-06,
+      "loss": 0.8602,
+      "step": 4768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7024293451721766,
+      "learning_rate": 3.586270705394962e-06,
+      "loss": 0.8042,
+      "step": 4769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5976861428370888,
+      "learning_rate": 3.5824679244102677e-06,
+      "loss": 0.7232,
+      "step": 4770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.71288602620368,
+      "learning_rate": 3.5786667206749736e-06,
+      "loss": 0.8084,
+      "step": 4771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6366829492883608,
+      "learning_rate": 3.5748670951233043e-06,
+      "loss": 0.7183,
+      "step": 4772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.804276556323464,
+      "learning_rate": 3.5710690486891066e-06,
+      "loss": 0.8149,
+      "step": 4773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6738106485757993,
+      "learning_rate": 3.5672725823058297e-06,
+      "loss": 0.8348,
+      "step": 4774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6069776651382854,
+      "learning_rate": 3.56347769690654e-06,
+      "loss": 0.713,
+      "step": 4775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6979836985747507,
+      "learning_rate": 3.5596843934239133e-06,
+      "loss": 0.7818,
+      "step": 4776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7598334527062054,
+      "learning_rate": 3.5558926727902366e-06,
+      "loss": 0.8516,
+      "step": 4777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3131812110524224,
+      "learning_rate": 3.5521025359374074e-06,
+      "loss": 0.8712,
+      "step": 4778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4949648400791204,
+      "learning_rate": 3.548313983796938e-06,
+      "loss": 0.7782,
+      "step": 4779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5630884401307656,
+      "learning_rate": 3.544527017299949e-06,
+      "loss": 0.8788,
+      "step": 4780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.911939959801433,
+      "learning_rate": 3.5407416373771643e-06,
+      "loss": 0.8761,
+      "step": 4781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.742301645711473,
+      "learning_rate": 3.5369578449589325e-06,
+      "loss": 0.7713,
+      "step": 4782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.418614212609283,
+      "learning_rate": 3.533175640975196e-06,
+      "loss": 0.7763,
+      "step": 4783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0920357376682004,
+      "learning_rate": 3.529395026355521e-06,
+      "loss": 0.8437,
+      "step": 4784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.458293186069364,
+      "learning_rate": 3.525616002029073e-06,
+      "loss": 0.7363,
+      "step": 4785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.8596590025593063,
+      "learning_rate": 3.5218385689246326e-06,
+      "loss": 0.8724,
+      "step": 4786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9193255631234507,
+      "learning_rate": 3.5180627279705835e-06,
+      "loss": 0.7714,
+      "step": 4787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.79476377638554,
+      "learning_rate": 3.514288480094924e-06,
+      "loss": 0.8012,
+      "step": 4788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6235771755809973,
+      "learning_rate": 3.5105158262252537e-06,
+      "loss": 0.8237,
+      "step": 4789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5787053452761564,
+      "learning_rate": 3.506744767288792e-06,
+      "loss": 0.7667,
+      "step": 4790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6403708336838543,
+      "learning_rate": 3.502975304212357e-06,
+      "loss": 0.7564,
+      "step": 4791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.637237157385171,
+      "learning_rate": 3.499207437922373e-06,
+      "loss": 0.73,
+      "step": 4792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5619777184569874,
+      "learning_rate": 3.495441169344882e-06,
+      "loss": 0.8118,
+      "step": 4793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.688433925814969,
+      "learning_rate": 3.491676499405522e-06,
+      "loss": 0.7487,
+      "step": 4794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.603657981617799,
+      "learning_rate": 3.487913429029548e-06,
+      "loss": 0.8036,
+      "step": 4795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.876068212142599,
+      "learning_rate": 3.484151959141815e-06,
+      "loss": 0.8381,
+      "step": 4796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7161938139739186,
+      "learning_rate": 3.4803920906667865e-06,
+      "loss": 0.7448,
+      "step": 4797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4956209199321697,
+      "learning_rate": 3.4766338245285335e-06,
+      "loss": 0.6899,
+      "step": 4798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.591375363966042,
+      "learning_rate": 3.4728771616507317e-06,
+      "loss": 0.7904,
+      "step": 4799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7401520195665023,
+      "learning_rate": 3.46912210295666e-06,
+      "loss": 0.7921,
+      "step": 4800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4448228135941834,
+      "learning_rate": 3.4653686493692128e-06,
+      "loss": 0.7103,
+      "step": 4801
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.962985582453866,
+      "learning_rate": 3.461616801810882e-06,
+      "loss": 0.7812,
+      "step": 4802
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5877756856635648,
+      "learning_rate": 3.457866561203761e-06,
+      "loss": 0.7902,
+      "step": 4803
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.879005707932505,
+      "learning_rate": 3.4541179284695624e-06,
+      "loss": 0.8495,
+      "step": 4804
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.589230441682058,
+      "learning_rate": 3.4503709045295874e-06,
+      "loss": 0.7214,
+      "step": 4805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.599459866748018,
+      "learning_rate": 3.4466254903047558e-06,
+      "loss": 0.7829,
+      "step": 4806
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.790768839452864,
+      "learning_rate": 3.4428816867155835e-06,
+      "loss": 0.7689,
+      "step": 4807
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1540500764058215,
+      "learning_rate": 3.43913949468219e-06,
+      "loss": 0.8027,
+      "step": 4808
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.816907689724522,
+      "learning_rate": 3.4353989151243027e-06,
+      "loss": 0.7883,
+      "step": 4809
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.72704029894962,
+      "learning_rate": 3.431659948961251e-06,
+      "loss": 0.8028,
+      "step": 4810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.583137093090374,
+      "learning_rate": 3.4279225971119654e-06,
+      "loss": 0.7438,
+      "step": 4811
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4804854730280765,
+      "learning_rate": 3.424186860494987e-06,
+      "loss": 0.7215,
+      "step": 4812
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.658327666872961,
+      "learning_rate": 3.4204527400284537e-06,
+      "loss": 0.7776,
+      "step": 4813
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.693389347062505,
+      "learning_rate": 3.4167202366301023e-06,
+      "loss": 0.703,
+      "step": 4814
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.566126480384414,
+      "learning_rate": 3.4129893512172853e-06,
+      "loss": 0.8688,
+      "step": 4815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.585160035850379,
+      "learning_rate": 3.4092600847069423e-06,
+      "loss": 0.6803,
+      "step": 4816
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6303955696871135,
+      "learning_rate": 3.4055324380156307e-06,
+      "loss": 0.8045,
+      "step": 4817
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4160626378845333,
+      "learning_rate": 3.401806412059496e-06,
+      "loss": 0.6905,
+      "step": 4818
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.441459359441314,
+      "learning_rate": 3.398082007754292e-06,
+      "loss": 0.7945,
+      "step": 4819
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.625386501191365,
+      "learning_rate": 3.3943592260153734e-06,
+      "loss": 0.7157,
+      "step": 4820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.440194805471606,
+      "learning_rate": 3.3906380677576946e-06,
+      "loss": 0.8603,
+      "step": 4821
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.694702386247108,
+      "learning_rate": 3.386918533895809e-06,
+      "loss": 0.7045,
+      "step": 4822
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1686746071454057,
+      "learning_rate": 3.38320062534388e-06,
+      "loss": 0.7979,
+      "step": 4823
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4610000489643724,
+      "learning_rate": 3.379484343015662e-06,
+      "loss": 0.7281,
+      "step": 4824
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.880245536591487,
+      "learning_rate": 3.3757696878245105e-06,
+      "loss": 0.7813,
+      "step": 4825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5397694675753377,
+      "learning_rate": 3.3720566606833893e-06,
+      "loss": 0.7543,
+      "step": 4826
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5502516158533943,
+      "learning_rate": 3.3683452625048508e-06,
+      "loss": 0.7224,
+      "step": 4827
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6288489855836685,
+      "learning_rate": 3.3646354942010574e-06,
+      "loss": 0.8899,
+      "step": 4828
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.724002193740201,
+      "learning_rate": 3.360927356683763e-06,
+      "loss": 0.7548,
+      "step": 4829
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6394485502907252,
+      "learning_rate": 3.3572208508643254e-06,
+      "loss": 0.8039,
+      "step": 4830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7206984933518865,
+      "learning_rate": 3.3535159776536998e-06,
+      "loss": 0.8256,
+      "step": 4831
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5872363477593403,
+      "learning_rate": 3.349812737962438e-06,
+      "loss": 0.7648,
+      "step": 4832
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.772472674946281,
+      "learning_rate": 3.34611113270069e-06,
+      "loss": 0.7739,
+      "step": 4833
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5276411112383186,
+      "learning_rate": 3.3424111627782153e-06,
+      "loss": 0.8324,
+      "step": 4834
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.738238132481366,
+      "learning_rate": 3.3387128291043567e-06,
+      "loss": 0.8597,
+      "step": 4835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6733033129669286,
+      "learning_rate": 3.335016132588058e-06,
+      "loss": 0.8107,
+      "step": 4836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3693822759565277,
+      "learning_rate": 3.331321074137872e-06,
+      "loss": 0.7716,
+      "step": 4837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4285933704356033,
+      "learning_rate": 3.3276276546619324e-06,
+      "loss": 0.7562,
+      "step": 4838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.652023893212776,
+      "learning_rate": 3.323935875067984e-06,
+      "loss": 0.7635,
+      "step": 4839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4578290945488543,
+      "learning_rate": 3.320245736263361e-06,
+      "loss": 0.6576,
+      "step": 4840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7942404086760533,
+      "learning_rate": 3.3165572391549915e-06,
+      "loss": 0.838,
+      "step": 4841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8569696135870015,
+      "learning_rate": 3.3128703846494115e-06,
+      "loss": 0.725,
+      "step": 4842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7900629597578375,
+      "learning_rate": 3.3091851736527434e-06,
+      "loss": 0.7868,
+      "step": 4843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8046779017462407,
+      "learning_rate": 3.3055016070707103e-06,
+      "loss": 0.7461,
+      "step": 4844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8927421334982313,
+      "learning_rate": 3.301819685808626e-06,
+      "loss": 0.7093,
+      "step": 4845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.858699181190039,
+      "learning_rate": 3.2981394107714025e-06,
+      "loss": 0.8473,
+      "step": 4846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3621659698589026,
+      "learning_rate": 3.2944607828635554e-06,
+      "loss": 0.7192,
+      "step": 4847
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.3742739290029724,
+      "learning_rate": 3.2907838029891835e-06,
+      "loss": 0.7438,
+      "step": 4848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.606599052695424,
+      "learning_rate": 3.287108472051982e-06,
+      "loss": 0.709,
+      "step": 4849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6729760499150745,
+      "learning_rate": 3.2834347909552524e-06,
+      "loss": 0.794,
+      "step": 4850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.90045193616181,
+      "learning_rate": 3.2797627606018766e-06,
+      "loss": 0.7958,
+      "step": 4851
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6004243439972416,
+      "learning_rate": 3.2760923818943356e-06,
+      "loss": 0.703,
+      "step": 4852
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.754123503758659,
+      "learning_rate": 3.272423655734711e-06,
+      "loss": 0.7644,
+      "step": 4853
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.442792399877945,
+      "learning_rate": 3.2687565830246704e-06,
+      "loss": 0.8134,
+      "step": 4854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7555431485595028,
+      "learning_rate": 3.265091164665477e-06,
+      "loss": 0.7632,
+      "step": 4855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4544350325423854,
+      "learning_rate": 3.261427401557987e-06,
+      "loss": 0.7764,
+      "step": 4856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9536442126066604,
+      "learning_rate": 3.2577652946026483e-06,
+      "loss": 0.8453,
+      "step": 4857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9074987910434906,
+      "learning_rate": 3.2541048446995104e-06,
+      "loss": 0.7189,
+      "step": 4858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.387376823817222,
+      "learning_rate": 3.2504460527482062e-06,
+      "loss": 0.8473,
+      "step": 4859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.630468030984959,
+      "learning_rate": 3.246788919647961e-06,
+      "loss": 0.7018,
+      "step": 4860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.673732710500758,
+      "learning_rate": 3.2431334462976007e-06,
+      "loss": 0.7881,
+      "step": 4861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4794806478398836,
+      "learning_rate": 3.239479633595536e-06,
+      "loss": 0.6952,
+      "step": 4862
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7444790121034806,
+      "learning_rate": 3.2358274824397685e-06,
+      "loss": 0.7908,
+      "step": 4863
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7718138842390156,
+      "learning_rate": 3.232176993727901e-06,
+      "loss": 0.7955,
+      "step": 4864
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.496774754790833,
+      "learning_rate": 3.2285281683571178e-06,
+      "loss": 0.7177,
+      "step": 4865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5301033397244064,
+      "learning_rate": 3.2248810072241974e-06,
+      "loss": 0.8109,
+      "step": 4866
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.926696010840226,
+      "learning_rate": 3.22123551122551e-06,
+      "loss": 0.7954,
+      "step": 4867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8720728648076888,
+      "learning_rate": 3.2175916812570128e-06,
+      "loss": 0.8188,
+      "step": 4868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7452597000369656,
+      "learning_rate": 3.2139495182142656e-06,
+      "loss": 0.7985,
+      "step": 4869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.622907404479844,
+      "learning_rate": 3.2103090229924028e-06,
+      "loss": 0.8043,
+      "step": 4870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4188680852947604,
+      "learning_rate": 3.206670196486156e-06,
+      "loss": 0.6735,
+      "step": 4871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4322841023452044,
+      "learning_rate": 3.203033039589851e-06,
+      "loss": 0.6663,
+      "step": 4872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.508207235025389,
+      "learning_rate": 3.1993975531973986e-06,
+      "loss": 0.7658,
+      "step": 4873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7555177905513784,
+      "learning_rate": 3.1957637382022934e-06,
+      "loss": 0.7667,
+      "step": 4874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5579291429948676,
+      "learning_rate": 3.1921315954976317e-06,
+      "loss": 0.695,
+      "step": 4875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5855212329014883,
+      "learning_rate": 3.1885011259760913e-06,
+      "loss": 0.8272,
+      "step": 4876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6520622459619694,
+      "learning_rate": 3.1848723305299377e-06,
+      "loss": 0.7435,
+      "step": 4877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6051923472077285,
+      "learning_rate": 3.1812452100510283e-06,
+      "loss": 0.769,
+      "step": 4878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7947194109042437,
+      "learning_rate": 3.177619765430804e-06,
+      "loss": 0.8326,
+      "step": 4879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.802329035693689,
+      "learning_rate": 3.1739959975603028e-06,
+      "loss": 0.832,
+      "step": 4880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7431713512017355,
+      "learning_rate": 3.1703739073301443e-06,
+      "loss": 0.8379,
+      "step": 4881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.653079167074889,
+      "learning_rate": 3.166753495630531e-06,
+      "loss": 0.8242,
+      "step": 4882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.801604761107011,
+      "learning_rate": 3.1631347633512665e-06,
+      "loss": 0.7701,
+      "step": 4883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4733006429141358,
+      "learning_rate": 3.1595177113817298e-06,
+      "loss": 0.7112,
+      "step": 4884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1772595038733957,
+      "learning_rate": 3.1559023406108892e-06,
+      "loss": 0.7305,
+      "step": 4885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.225116065554904,
+      "learning_rate": 3.1522886519273053e-06,
+      "loss": 0.8426,
+      "step": 4886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0385149713556774,
+      "learning_rate": 3.14867664621912e-06,
+      "loss": 0.7618,
+      "step": 4887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.662748289735663,
+      "learning_rate": 3.145066324374062e-06,
+      "loss": 0.7236,
+      "step": 4888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0617462390536136,
+      "learning_rate": 3.141457687279448e-06,
+      "loss": 0.6848,
+      "step": 4889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.49603952714078,
+      "learning_rate": 3.1378507358221765e-06,
+      "loss": 0.7738,
+      "step": 4890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.996649925386801,
+      "learning_rate": 3.1342454708887404e-06,
+      "loss": 0.8548,
+      "step": 4891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.936472847178242,
+      "learning_rate": 3.1306418933652105e-06,
+      "loss": 0.7815,
+      "step": 4892
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7740070586751817,
+      "learning_rate": 3.127040004137242e-06,
+      "loss": 0.8209,
+      "step": 4893
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6643704083806368,
+      "learning_rate": 3.1234398040900836e-06,
+      "loss": 0.7166,
+      "step": 4894
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4987987550750006,
+      "learning_rate": 3.119841294108562e-06,
+      "loss": 0.6859,
+      "step": 4895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.551914873986552,
+      "learning_rate": 3.116244475077086e-06,
+      "loss": 0.7637,
+      "step": 4896
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.574786283194233,
+      "learning_rate": 3.112649347879658e-06,
+      "loss": 0.7759,
+      "step": 4897
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5550201444355434,
+      "learning_rate": 3.1090559133998576e-06,
+      "loss": 0.8019,
+      "step": 4898
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.54132320674916,
+      "learning_rate": 3.10546417252085e-06,
+      "loss": 0.8308,
+      "step": 4899
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.981833030314064,
+      "learning_rate": 3.1018741261253836e-06,
+      "loss": 0.8591,
+      "step": 4900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7911752802943677,
+      "learning_rate": 3.0982857750957895e-06,
+      "loss": 0.8839,
+      "step": 4901
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7415236410740995,
+      "learning_rate": 3.0946991203139896e-06,
+      "loss": 0.7958,
+      "step": 4902
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4625353307618836,
+      "learning_rate": 3.0911141626614792e-06,
+      "loss": 0.6676,
+      "step": 4903
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5561719544460404,
+      "learning_rate": 3.087530903019337e-06,
+      "loss": 0.8083,
+      "step": 4904
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8144420696732704,
+      "learning_rate": 3.0839493422682344e-06,
+      "loss": 0.7585,
+      "step": 4905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9425670291569945,
+      "learning_rate": 3.0803694812884167e-06,
+      "loss": 0.7633,
+      "step": 4906
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.743831645079609,
+      "learning_rate": 3.0767913209597076e-06,
+      "loss": 0.8813,
+      "step": 4907
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7116331823136397,
+      "learning_rate": 3.0732148621615266e-06,
+      "loss": 0.7728,
+      "step": 4908
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.474561448824525,
+      "learning_rate": 3.069640105772864e-06,
+      "loss": 0.7739,
+      "step": 4909
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4641070171397588,
+      "learning_rate": 3.066067052672295e-06,
+      "loss": 0.8748,
+      "step": 4910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5126521048603783,
+      "learning_rate": 3.0624957037379733e-06,
+      "loss": 0.8056,
+      "step": 4911
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.646426069683561,
+      "learning_rate": 3.0589260598476354e-06,
+      "loss": 0.7136,
+      "step": 4912
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5569589171109284,
+      "learning_rate": 3.0553581218786053e-06,
+      "loss": 0.8465,
+      "step": 4913
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.383925636701246,
+      "learning_rate": 3.0517918907077805e-06,
+      "loss": 0.6981,
+      "step": 4914
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6317329134965637,
+      "learning_rate": 3.0482273672116347e-06,
+      "loss": 0.8363,
+      "step": 4915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.599309419546993,
+      "learning_rate": 3.0446645522662356e-06,
+      "loss": 0.8145,
+      "step": 4916
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.298720627729177,
+      "learning_rate": 3.041103446747219e-06,
+      "loss": 0.7698,
+      "step": 4917
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.792177265740668,
+      "learning_rate": 3.037544051529804e-06,
+      "loss": 0.8006,
+      "step": 4918
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.869378994570629,
+      "learning_rate": 3.0339863674887938e-06,
+      "loss": 0.6888,
+      "step": 4919
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9143939203489646,
+      "learning_rate": 3.0304303954985658e-06,
+      "loss": 0.8653,
+      "step": 4920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.68410731957122,
+      "learning_rate": 3.026876136433078e-06,
+      "loss": 0.8332,
+      "step": 4921
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.544048861240414,
+      "learning_rate": 3.0233235911658665e-06,
+      "loss": 0.7344,
+      "step": 4922
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.441477009221491,
+      "learning_rate": 3.0197727605700457e-06,
+      "loss": 0.8102,
+      "step": 4923
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6620877456991425,
+      "learning_rate": 3.016223645518315e-06,
+      "loss": 0.7534,
+      "step": 4924
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8841099980051874,
+      "learning_rate": 3.012676246882945e-06,
+      "loss": 0.8603,
+      "step": 4925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6406222549052223,
+      "learning_rate": 3.009130565535784e-06,
+      "loss": 0.7859,
+      "step": 4926
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.779876928459848,
+      "learning_rate": 3.005586602348266e-06,
+      "loss": 0.8433,
+      "step": 4927
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1990330226074764,
+      "learning_rate": 3.002044358191396e-06,
+      "loss": 0.822,
+      "step": 4928
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.653865192531658,
+      "learning_rate": 2.998503833935754e-06,
+      "loss": 0.8725,
+      "step": 4929
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7786270658771475,
+      "learning_rate": 2.9949650304515098e-06,
+      "loss": 0.8057,
+      "step": 4930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.693592035006558,
+      "learning_rate": 2.9914279486083963e-06,
+      "loss": 0.8249,
+      "step": 4931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6704855809519694,
+      "learning_rate": 2.9878925892757316e-06,
+      "loss": 0.868,
+      "step": 4932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.268918586765798,
+      "learning_rate": 2.9843589533224047e-06,
+      "loss": 0.8498,
+      "step": 4933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.651243223030544,
+      "learning_rate": 2.980827041616884e-06,
+      "loss": 0.796,
+      "step": 4934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.469076464190495,
+      "learning_rate": 2.97729685502722e-06,
+      "loss": 0.6783,
+      "step": 4935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6319493654941577,
+      "learning_rate": 2.9737683944210293e-06,
+      "loss": 0.8822,
+      "step": 4936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.586829052141713,
+      "learning_rate": 2.970241660665506e-06,
+      "loss": 0.7955,
+      "step": 4937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6870831720388932,
+      "learning_rate": 2.9667166546274286e-06,
+      "loss": 0.9148,
+      "step": 4938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.754533272622671,
+      "learning_rate": 2.9631933771731414e-06,
+      "loss": 0.8005,
+      "step": 4939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5926012760865684,
+      "learning_rate": 2.9596718291685655e-06,
+      "loss": 0.9103,
+      "step": 4940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.62728719562524,
+      "learning_rate": 2.9561520114792032e-06,
+      "loss": 0.7476,
+      "step": 4941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.62973819085915,
+      "learning_rate": 2.952633924970121e-06,
+      "loss": 0.8022,
+      "step": 4942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8037242977558896,
+      "learning_rate": 2.949117570505975e-06,
+      "loss": 0.8509,
+      "step": 4943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.660119092568261,
+      "learning_rate": 2.9456029489509773e-06,
+      "loss": 0.8049,
+      "step": 4944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7454690787999785,
+      "learning_rate": 2.942090061168925e-06,
+      "loss": 0.8604,
+      "step": 4945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7323098990553603,
+      "learning_rate": 2.938578908023192e-06,
+      "loss": 0.8264,
+      "step": 4946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7730572623711573,
+      "learning_rate": 2.9350694903767185e-06,
+      "loss": 0.7949,
+      "step": 4947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.961591688622995,
+      "learning_rate": 2.9315618090920173e-06,
+      "loss": 0.6742,
+      "step": 4948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.650539798329177,
+      "learning_rate": 2.9280558650311842e-06,
+      "loss": 0.7325,
+      "step": 4949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.712772938021309,
+      "learning_rate": 2.9245516590558796e-06,
+      "loss": 0.8108,
+      "step": 4950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.613491635699674,
+      "learning_rate": 2.9210491920273365e-06,
+      "loss": 0.7417,
+      "step": 4951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.572729005128316,
+      "learning_rate": 2.917548464806368e-06,
+      "loss": 0.8002,
+      "step": 4952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4466386915599383,
+      "learning_rate": 2.914049478253349e-06,
+      "loss": 0.766,
+      "step": 4953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.615461090522433,
+      "learning_rate": 2.910552233228239e-06,
+      "loss": 0.8398,
+      "step": 4954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7812802922894964,
+      "learning_rate": 2.907056730590558e-06,
+      "loss": 0.8144,
+      "step": 4955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6422650497261895,
+      "learning_rate": 2.903562971199405e-06,
+      "loss": 0.7422,
+      "step": 4956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6107459832204496,
+      "learning_rate": 2.9000709559134456e-06,
+      "loss": 0.7618,
+      "step": 4957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.537852760775243,
+      "learning_rate": 2.8965806855909207e-06,
+      "loss": 0.8506,
+      "step": 4958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.247518579755631,
+      "learning_rate": 2.8930921610896366e-06,
+      "loss": 0.8605,
+      "step": 4959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.649138417092869,
+      "learning_rate": 2.8896053832669822e-06,
+      "loss": 0.7862,
+      "step": 4960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.812294876476703,
+      "learning_rate": 2.8861203529799052e-06,
+      "loss": 0.8908,
+      "step": 4961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4486042948844307,
+      "learning_rate": 2.8826370710849274e-06,
+      "loss": 0.7694,
+      "step": 4962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.767682856019066,
+      "learning_rate": 2.8791555384381466e-06,
+      "loss": 0.8236,
+      "step": 4963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.853722056482501,
+      "learning_rate": 2.8756757558952186e-06,
+      "loss": 0.7457,
+      "step": 4964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6353513595010014,
+      "learning_rate": 2.8721977243113854e-06,
+      "loss": 0.7987,
+      "step": 4965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9905809231839355,
+      "learning_rate": 2.8687214445414434e-06,
+      "loss": 0.7522,
+      "step": 4966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.418550038242517,
+      "learning_rate": 2.8652469174397667e-06,
+      "loss": 0.7241,
+      "step": 4967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4665763541728247,
+      "learning_rate": 2.8617741438602964e-06,
+      "loss": 0.736,
+      "step": 4968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5783211822392196,
+      "learning_rate": 2.8583031246565417e-06,
+      "loss": 0.8074,
+      "step": 4969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7056606174723274,
+      "learning_rate": 2.8548338606815805e-06,
+      "loss": 0.7845,
+      "step": 4970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.6595235853664403,
+      "learning_rate": 2.8513663527880653e-06,
+      "loss": 0.771,
+      "step": 4971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.66613898311809,
+      "learning_rate": 2.8479006018282096e-06,
+      "loss": 0.6243,
+      "step": 4972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.692923874491351,
+      "learning_rate": 2.8444366086537943e-06,
+      "loss": 0.8159,
+      "step": 4973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5376497677560717,
+      "learning_rate": 2.840974374116179e-06,
+      "loss": 0.711,
+      "step": 4974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.863142652970658,
+      "learning_rate": 2.8375138990662766e-06,
+      "loss": 0.8243,
+      "step": 4975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.488143347726353,
+      "learning_rate": 2.8340551843545817e-06,
+      "loss": 0.7433,
+      "step": 4976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8879241474099566,
+      "learning_rate": 2.8305982308311453e-06,
+      "loss": 0.8774,
+      "step": 4977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5762550206887824,
+      "learning_rate": 2.827143039345591e-06,
+      "loss": 0.7337,
+      "step": 4978
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.595931442952882,
+      "learning_rate": 2.823689610747108e-06,
+      "loss": 0.8634,
+      "step": 4979
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3186033910243693,
+      "learning_rate": 2.820237945884451e-06,
+      "loss": 0.7204,
+      "step": 4980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.614217693941196,
+      "learning_rate": 2.8167880456059394e-06,
+      "loss": 0.7471,
+      "step": 4981
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.4858416594026043,
+      "learning_rate": 2.8133399107594683e-06,
+      "loss": 0.7809,
+      "step": 4982
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.937018679007705,
+      "learning_rate": 2.80989354219249e-06,
+      "loss": 0.7293,
+      "step": 4983
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.670884778953924,
+      "learning_rate": 2.8064489407520225e-06,
+      "loss": 0.764,
+      "step": 4984
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5856640266782973,
+      "learning_rate": 2.803006107284657e-06,
+      "loss": 0.8138,
+      "step": 4985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.225279699315726,
+      "learning_rate": 2.7995650426365405e-06,
+      "loss": 0.7173,
+      "step": 4986
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.611351350122502,
+      "learning_rate": 2.7961257476533954e-06,
+      "loss": 0.7806,
+      "step": 4987
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.671165795967523,
+      "learning_rate": 2.792688223180502e-06,
+      "loss": 0.8442,
+      "step": 4988
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.221238185390751,
+      "learning_rate": 2.7892524700627053e-06,
+      "loss": 0.8327,
+      "step": 4989
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7434210137367905,
+      "learning_rate": 2.7858184891444197e-06,
+      "loss": 0.7641,
+      "step": 4990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.642891673485018,
+      "learning_rate": 2.7823862812696203e-06,
+      "loss": 0.7117,
+      "step": 4991
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.713233880054568,
+      "learning_rate": 2.7789558472818435e-06,
+      "loss": 0.7222,
+      "step": 4992
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7191943051310097,
+      "learning_rate": 2.775527188024201e-06,
+      "loss": 0.8237,
+      "step": 4993
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9476597995929237,
+      "learning_rate": 2.772100304339355e-06,
+      "loss": 0.8084,
+      "step": 4994
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9476214832791774,
+      "learning_rate": 2.7686751970695427e-06,
+      "loss": 0.7891,
+      "step": 4995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0800081298559814,
+      "learning_rate": 2.7652518670565577e-06,
+      "loss": 0.8109,
+      "step": 4996
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.455886649514224,
+      "learning_rate": 2.7618303151417534e-06,
+      "loss": 0.6759,
+      "step": 4997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.53600811818245,
+      "learning_rate": 2.758410542166059e-06,
+      "loss": 0.8115,
+      "step": 4998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.819024745363265,
+      "learning_rate": 2.754992548969956e-06,
+      "loss": 0.8393,
+      "step": 4999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6405712089148254,
+      "learning_rate": 2.75157633639349e-06,
+      "loss": 0.7871,
+      "step": 5000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.689984644938119,
+      "learning_rate": 2.748161905276271e-06,
+      "loss": 0.7925,
+      "step": 5001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5912458575941604,
+      "learning_rate": 2.7447492564574708e-06,
+      "loss": 0.8126,
+      "step": 5002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.922850248609202,
+      "learning_rate": 2.7413383907758183e-06,
+      "loss": 0.8764,
+      "step": 5003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9523305387496404,
+      "learning_rate": 2.737929309069616e-06,
+      "loss": 0.7544,
+      "step": 5004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7388750693266837,
+      "learning_rate": 2.7345220121767136e-06,
+      "loss": 0.7313,
+      "step": 5005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5705254543989833,
+      "learning_rate": 2.7311165009345362e-06,
+      "loss": 0.6219,
+      "step": 5006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6715192568830886,
+      "learning_rate": 2.7277127761800592e-06,
+      "loss": 0.8486,
+      "step": 5007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.596380510041663,
+      "learning_rate": 2.7243108387498207e-06,
+      "loss": 0.7227,
+      "step": 5008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.586844529728933,
+      "learning_rate": 2.7209106894799253e-06,
+      "loss": 0.8468,
+      "step": 5009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.099774070134371,
+      "learning_rate": 2.7175123292060335e-06,
+      "loss": 0.8458,
+      "step": 5010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.568013495598628,
+      "learning_rate": 2.714115758763366e-06,
+      "loss": 0.7795,
+      "step": 5011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5389843195512642,
+      "learning_rate": 2.710720978986705e-06,
+      "loss": 0.8222,
+      "step": 5012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4481461619079146,
+      "learning_rate": 2.7073279907103913e-06,
+      "loss": 0.6732,
+      "step": 5013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.572469747307365,
+      "learning_rate": 2.703936794768325e-06,
+      "loss": 0.7647,
+      "step": 5014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.510351138783795,
+      "learning_rate": 2.7005473919939706e-06,
+      "loss": 0.7419,
+      "step": 5015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.535945112086226,
+      "learning_rate": 2.6971597832203434e-06,
+      "loss": 0.7646,
+      "step": 5016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7867749494925027,
+      "learning_rate": 2.693773969280029e-06,
+      "loss": 0.6355,
+      "step": 5017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.788432665774386,
+      "learning_rate": 2.6903899510051624e-06,
+      "loss": 0.913,
+      "step": 5018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4875282017699964,
+      "learning_rate": 2.6870077292274366e-06,
+      "loss": 0.6947,
+      "step": 5019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.75155649877325,
+      "learning_rate": 2.6836273047781137e-06,
+      "loss": 0.7729,
+      "step": 5020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7143198356277995,
+      "learning_rate": 2.6802486784880044e-06,
+      "loss": 0.8094,
+      "step": 5021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6107155069240435,
+      "learning_rate": 2.676871851187479e-06,
+      "loss": 0.8053,
+      "step": 5022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9969143394969984,
+      "learning_rate": 2.6734968237064686e-06,
+      "loss": 0.806,
+      "step": 5023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.752371634653211,
+      "learning_rate": 2.6701235968744587e-06,
+      "loss": 0.6662,
+      "step": 5024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.552224215391848,
+      "learning_rate": 2.6667521715204914e-06,
+      "loss": 0.7168,
+      "step": 5025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.620940328135656,
+      "learning_rate": 2.6633825484731746e-06,
+      "loss": 0.8322,
+      "step": 5026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.580152864684537,
+      "learning_rate": 2.6600147285606625e-06,
+      "loss": 0.8152,
+      "step": 5027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3639360061856247,
+      "learning_rate": 2.6566487126106745e-06,
+      "loss": 0.6962,
+      "step": 5028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.406455472298094,
+      "learning_rate": 2.6532845014504814e-06,
+      "loss": 0.7394,
+      "step": 5029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6169693952537822,
+      "learning_rate": 2.6499220959069085e-06,
+      "loss": 0.7742,
+      "step": 5030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4927676361698934,
+      "learning_rate": 2.6465614968063456e-06,
+      "loss": 0.781,
+      "step": 5031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.62965426461676,
+      "learning_rate": 2.6432027049747333e-06,
+      "loss": 0.8843,
+      "step": 5032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7037687891142763,
+      "learning_rate": 2.639845721237566e-06,
+      "loss": 0.8107,
+      "step": 5033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.643649919678502,
+      "learning_rate": 2.6364905464198987e-06,
+      "loss": 0.7893,
+      "step": 5034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.691892689329616,
+      "learning_rate": 2.6331371813463356e-06,
+      "loss": 0.8265,
+      "step": 5035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6975452736113454,
+      "learning_rate": 2.6297856268410406e-06,
+      "loss": 0.81,
+      "step": 5036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4158822610707325,
+      "learning_rate": 2.626435883727735e-06,
+      "loss": 0.6869,
+      "step": 5037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4939839844165905,
+      "learning_rate": 2.623087952829688e-06,
+      "loss": 0.8316,
+      "step": 5038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.776427207607364,
+      "learning_rate": 2.619741834969731e-06,
+      "loss": 0.7877,
+      "step": 5039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3779010383973707,
+      "learning_rate": 2.616397530970244e-06,
+      "loss": 0.6899,
+      "step": 5040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6083802441754464,
+      "learning_rate": 2.6130550416531597e-06,
+      "loss": 0.7778,
+      "step": 5041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.56450784477195,
+      "learning_rate": 2.609714367839975e-06,
+      "loss": 0.7963,
+      "step": 5042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5459301905849654,
+      "learning_rate": 2.606375510351731e-06,
+      "loss": 0.7842,
+      "step": 5043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5217148421897546,
+      "learning_rate": 2.6030384700090238e-06,
+      "loss": 0.7401,
+      "step": 5044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.45823283836036,
+      "learning_rate": 2.599703247632005e-06,
+      "loss": 0.7311,
+      "step": 5045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9579669428110598,
+      "learning_rate": 2.596369844040378e-06,
+      "loss": 0.7747,
+      "step": 5046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5750011906269923,
+      "learning_rate": 2.5930382600533998e-06,
+      "loss": 0.7482,
+      "step": 5047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7888629756635206,
+      "learning_rate": 2.5897084964898835e-06,
+      "loss": 0.7846,
+      "step": 5048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5721294503502157,
+      "learning_rate": 2.586380554168185e-06,
+      "loss": 0.6925,
+      "step": 5049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7618620787962493,
+      "learning_rate": 2.583054433906228e-06,
+      "loss": 0.7054,
+      "step": 5050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6276942856050938,
+      "learning_rate": 2.5797301365214742e-06,
+      "loss": 0.7645,
+      "step": 5051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.616392198716273,
+      "learning_rate": 2.576407662830942e-06,
+      "loss": 0.7362,
+      "step": 5052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.711256026542573,
+      "learning_rate": 2.5730870136512055e-06,
+      "loss": 0.7994,
+      "step": 5053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6991497732690743,
+      "learning_rate": 2.5697681897983862e-06,
+      "loss": 0.8072,
+      "step": 5054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4775791838568586,
+      "learning_rate": 2.566451192088156e-06,
+      "loss": 0.7789,
+      "step": 5055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.388324935782896,
+      "learning_rate": 2.5631360213357425e-06,
+      "loss": 0.8261,
+      "step": 5056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.514421673855606,
+      "learning_rate": 2.5598226783559198e-06,
+      "loss": 0.7355,
+      "step": 5057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.536006040702953,
+      "learning_rate": 2.5565111639630125e-06,
+      "loss": 0.6696,
+      "step": 5058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7285987787744275,
+      "learning_rate": 2.5532014789709027e-06,
+      "loss": 0.8219,
+      "step": 5059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.476472443285063,
+      "learning_rate": 2.549893624193014e-06,
+      "loss": 0.7924,
+      "step": 5060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.662355526288266,
+      "learning_rate": 2.5465876004423285e-06,
+      "loss": 0.7893,
+      "step": 5061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.507656262345981,
+      "learning_rate": 2.543283408531373e-06,
+      "loss": 0.7732,
+      "step": 5062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8257941998664964,
+      "learning_rate": 2.53998104927222e-06,
+      "loss": 0.7729,
+      "step": 5063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7337037395372334,
+      "learning_rate": 2.5366805234765047e-06,
+      "loss": 0.7656,
+      "step": 5064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5500710978463883,
+      "learning_rate": 2.5333818319554002e-06,
+      "loss": 0.6465,
+      "step": 5065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7708922146707167,
+      "learning_rate": 2.530084975519629e-06,
+      "loss": 0.8558,
+      "step": 5066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7554935025214697,
+      "learning_rate": 2.5267899549794728e-06,
+      "loss": 0.754,
+      "step": 5067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4565508163642993,
+      "learning_rate": 2.523496771144751e-06,
+      "loss": 0.8833,
+      "step": 5068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5094279415765417,
+      "learning_rate": 2.5202054248248362e-06,
+      "loss": 0.7924,
+      "step": 5069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.172596205277227,
+      "learning_rate": 2.5169159168286503e-06,
+      "loss": 0.7781,
+      "step": 5070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.780052244901856,
+      "learning_rate": 2.513628247964658e-06,
+      "loss": 0.8596,
+      "step": 5071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6389858730839517,
+      "learning_rate": 2.510342419040881e-06,
+      "loss": 0.7318,
+      "step": 5072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0811186388492193,
+      "learning_rate": 2.5070584308648828e-06,
+      "loss": 0.8587,
+      "step": 5073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.53080728953482,
+      "learning_rate": 2.503776284243772e-06,
+      "loss": 0.7536,
+      "step": 5074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7781051409831026,
+      "learning_rate": 2.5004959799842133e-06,
+      "loss": 0.8435,
+      "step": 5075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7029578113303256,
+      "learning_rate": 2.497217518892412e-06,
+      "loss": 0.7928,
+      "step": 5076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7198966214625666,
+      "learning_rate": 2.493940901774118e-06,
+      "loss": 0.7304,
+      "step": 5077
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.482025583931032,
+      "learning_rate": 2.490666129434638e-06,
+      "loss": 0.828,
+      "step": 5078
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.46772119696312,
+      "learning_rate": 2.4873932026788162e-06,
+      "loss": 0.7316,
+      "step": 5079
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5961425268325375,
+      "learning_rate": 2.4841221223110467e-06,
+      "loss": 0.8223,
+      "step": 5080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5679783685419544,
+      "learning_rate": 2.4808528891352677e-06,
+      "loss": 0.7975,
+      "step": 5081
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7506846436308914,
+      "learning_rate": 2.4775855039549647e-06,
+      "loss": 0.7494,
+      "step": 5082
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5964912817122605,
+      "learning_rate": 2.4743199675731722e-06,
+      "loss": 0.7492,
+      "step": 5083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8934879578568884,
+      "learning_rate": 2.4710562807924664e-06,
+      "loss": 0.7708,
+      "step": 5084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6202697209531114,
+      "learning_rate": 2.4677944444149683e-06,
+      "loss": 0.7256,
+      "step": 5085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.626432453067587,
+      "learning_rate": 2.464534459242348e-06,
+      "loss": 0.8259,
+      "step": 5086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.521069808194475,
+      "learning_rate": 2.4612763260758187e-06,
+      "loss": 0.696,
+      "step": 5087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.64835206059012,
+      "learning_rate": 2.458020045716134e-06,
+      "loss": 0.7756,
+      "step": 5088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3416846206287163,
+      "learning_rate": 2.4547656189636014e-06,
+      "loss": 0.7253,
+      "step": 5089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.942055576016125,
+      "learning_rate": 2.451513046618067e-06,
+      "loss": 0.7859,
+      "step": 5090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8389802865849383,
+      "learning_rate": 2.44826232947892e-06,
+      "loss": 0.8852,
+      "step": 5091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.511118745358438,
+      "learning_rate": 2.4450134683450957e-06,
+      "loss": 0.729,
+      "step": 5092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3110347546272845,
+      "learning_rate": 2.4417664640150695e-06,
+      "loss": 0.7976,
+      "step": 5093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.440522966930613,
+      "learning_rate": 2.4385213172868716e-06,
+      "loss": 0.7789,
+      "step": 5094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6106319712037207,
+      "learning_rate": 2.4352780289580647e-06,
+      "loss": 0.8299,
+      "step": 5095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5840976790034924,
+      "learning_rate": 2.4320365998257543e-06,
+      "loss": 0.7839,
+      "step": 5096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.68223277349005,
+      "learning_rate": 2.4287970306865994e-06,
+      "loss": 0.7534,
+      "step": 5097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4092525951805706,
+      "learning_rate": 2.4255593223367923e-06,
+      "loss": 0.7492,
+      "step": 5098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6820676394906435,
+      "learning_rate": 2.4223234755720672e-06,
+      "loss": 0.7327,
+      "step": 5099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.567354858149662,
+      "learning_rate": 2.4190894911877105e-06,
+      "loss": 0.7733,
+      "step": 5100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.647653908124238,
+      "learning_rate": 2.4158573699785427e-06,
+      "loss": 0.7113,
+      "step": 5101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6788491642529397,
+      "learning_rate": 2.412627112738928e-06,
+      "loss": 0.7993,
+      "step": 5102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9280387750202412,
+      "learning_rate": 2.4093987202627735e-06,
+      "loss": 0.8553,
+      "step": 5103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.581053853352887,
+      "learning_rate": 2.4061721933435246e-06,
+      "loss": 0.8247,
+      "step": 5104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.580506365930907,
+      "learning_rate": 2.4029475327741758e-06,
+      "loss": 0.8352,
+      "step": 5105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.711539323105799,
+      "learning_rate": 2.3997247393472557e-06,
+      "loss": 0.7786,
+      "step": 5106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6544888378906997,
+      "learning_rate": 2.3965038138548346e-06,
+      "loss": 0.8774,
+      "step": 5107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.781610147223123,
+      "learning_rate": 2.3932847570885307e-06,
+      "loss": 0.7689,
+      "step": 5108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.838519577882593,
+      "learning_rate": 2.390067569839496e-06,
+      "loss": 0.7351,
+      "step": 5109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.605314160377416,
+      "learning_rate": 2.3868522528984207e-06,
+      "loss": 0.7761,
+      "step": 5110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7188592167400154,
+      "learning_rate": 2.383638807055545e-06,
+      "loss": 0.8472,
+      "step": 5111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5482203378844366,
+      "learning_rate": 2.380427233100643e-06,
+      "loss": 0.7605,
+      "step": 5112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.704449992334734,
+      "learning_rate": 2.3772175318230272e-06,
+      "loss": 0.8089,
+      "step": 5113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8402849358160167,
+      "learning_rate": 2.374009704011554e-06,
+      "loss": 0.6916,
+      "step": 5114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.864028803777469,
+      "learning_rate": 2.3708037504546124e-06,
+      "loss": 0.799,
+      "step": 5115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.501326567246752,
+      "learning_rate": 2.367599671940144e-06,
+      "loss": 0.6988,
+      "step": 5116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2004108035764625,
+      "learning_rate": 2.364397469255617e-06,
+      "loss": 0.8027,
+      "step": 5117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6490640456468046,
+      "learning_rate": 2.3611971431880407e-06,
+      "loss": 0.806,
+      "step": 5118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5612708686624686,
+      "learning_rate": 2.357998694523972e-06,
+      "loss": 0.6694,
+      "step": 5119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.0401653824639516,
+      "learning_rate": 2.3548021240494955e-06,
+      "loss": 0.7947,
+      "step": 5120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.8307884578484686,
+      "learning_rate": 2.3516074325502368e-06,
+      "loss": 0.8022,
+      "step": 5121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.72977536780408,
+      "learning_rate": 2.348414620811367e-06,
+      "loss": 0.7399,
+      "step": 5122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4748006126897417,
+      "learning_rate": 2.345223689617586e-06,
+      "loss": 0.6904,
+      "step": 5123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4064619793053934,
+      "learning_rate": 2.342034639753137e-06,
+      "loss": 0.7289,
+      "step": 5124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9327774960950537,
+      "learning_rate": 2.3388474720017963e-06,
+      "loss": 0.8078,
+      "step": 5125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4273391870497125,
+      "learning_rate": 2.33566218714688e-06,
+      "loss": 0.8186,
+      "step": 5126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.339466964529913,
+      "learning_rate": 2.3324787859712462e-06,
+      "loss": 0.8604,
+      "step": 5127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7521085637109732,
+      "learning_rate": 2.3292972692572833e-06,
+      "loss": 0.7874,
+      "step": 5128
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7279640075919307,
+      "learning_rate": 2.3261176377869165e-06,
+      "loss": 0.7593,
+      "step": 5129
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.598691379497955,
+      "learning_rate": 2.3229398923416136e-06,
+      "loss": 0.7745,
+      "step": 5130
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.519874664868604,
+      "learning_rate": 2.319764033702375e-06,
+      "loss": 0.8348,
+      "step": 5131
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.601307397164149,
+      "learning_rate": 2.316590062649734e-06,
+      "loss": 0.7321,
+      "step": 5132
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4969779020558853,
+      "learning_rate": 2.313417979963768e-06,
+      "loss": 0.728,
+      "step": 5133
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.650113640353963,
+      "learning_rate": 2.310247786424086e-06,
+      "loss": 0.7826,
+      "step": 5134
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5214125575947963,
+      "learning_rate": 2.3070794828098285e-06,
+      "loss": 0.7338,
+      "step": 5135
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9428606849207837,
+      "learning_rate": 2.3039130698996802e-06,
+      "loss": 0.8268,
+      "step": 5136
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6185692051050182,
+      "learning_rate": 2.3007485484718505e-06,
+      "loss": 0.6381,
+      "step": 5137
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6834657323433992,
+      "learning_rate": 2.297585919304097e-06,
+      "loss": 0.7596,
+      "step": 5138
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.781398509030959,
+      "learning_rate": 2.294425183173703e-06,
+      "loss": 0.7946,
+      "step": 5139
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.561953880258535,
+      "learning_rate": 2.2912663408574843e-06,
+      "loss": 0.787,
+      "step": 5140
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.618970065482725,
+      "learning_rate": 2.288109393131802e-06,
+      "loss": 0.7402,
+      "step": 5141
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.428394606083279,
+      "learning_rate": 2.2849543407725403e-06,
+      "loss": 0.6512,
+      "step": 5142
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8219441478736917,
+      "learning_rate": 2.2818011845551293e-06,
+      "loss": 0.8186,
+      "step": 5143
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5187641567693984,
+      "learning_rate": 2.27864992525452e-06,
+      "loss": 0.7891,
+      "step": 5144
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.59796954673457,
+      "learning_rate": 2.2755005636452067e-06,
+      "loss": 0.7758,
+      "step": 5145
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.652727950464101,
+      "learning_rate": 2.2723531005012133e-06,
+      "loss": 0.8253,
+      "step": 5146
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7847114123727943,
+      "learning_rate": 2.2692075365960974e-06,
+      "loss": 0.8096,
+      "step": 5147
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.692857244476947,
+      "learning_rate": 2.2660638727029484e-06,
+      "loss": 0.8111,
+      "step": 5148
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.879953323799714,
+      "learning_rate": 2.2629221095943952e-06,
+      "loss": 0.7733,
+      "step": 5149
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.710345885860965,
+      "learning_rate": 2.2597822480425934e-06,
+      "loss": 0.7381,
+      "step": 5150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.532688579438755,
+      "learning_rate": 2.2566442888192306e-06,
+      "loss": 0.8801,
+      "step": 5151
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6739871284912713,
+      "learning_rate": 2.2535082326955325e-06,
+      "loss": 0.7638,
+      "step": 5152
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6604332292354353,
+      "learning_rate": 2.2503740804422504e-06,
+      "loss": 0.7703,
+      "step": 5153
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1129370913861996,
+      "learning_rate": 2.247241832829675e-06,
+      "loss": 0.81,
+      "step": 5154
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.777618927551345,
+      "learning_rate": 2.244111490627623e-06,
+      "loss": 0.8235,
+      "step": 5155
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6995224648073917,
+      "learning_rate": 2.2409830546054456e-06,
+      "loss": 0.6747,
+      "step": 5156
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.670008968578099,
+      "learning_rate": 2.2378565255320226e-06,
+      "loss": 0.8097,
+      "step": 5157
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7697821647266103,
+      "learning_rate": 2.2347319041757675e-06,
+      "loss": 0.8285,
+      "step": 5158
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.749625518855146,
+      "learning_rate": 2.2316091913046235e-06,
+      "loss": 0.713,
+      "step": 5159
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.594875868036443,
+      "learning_rate": 2.2284883876860707e-06,
+      "loss": 0.7735,
+      "step": 5160
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.527292720733303,
+      "learning_rate": 2.2253694940871106e-06,
+      "loss": 0.7583,
+      "step": 5161
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.417019669741224,
+      "learning_rate": 2.22225251127428e-06,
+      "loss": 0.7943,
+      "step": 5162
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.642528698366732,
+      "learning_rate": 2.219137440013649e-06,
+      "loss": 0.8459,
+      "step": 5163
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5632337289614746,
+      "learning_rate": 2.2160242810708098e-06,
+      "loss": 0.8122,
+      "step": 5164
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5843935751636216,
+      "learning_rate": 2.212913035210895e-06,
+      "loss": 0.761,
+      "step": 5165
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6664617878227626,
+      "learning_rate": 2.2098037031985598e-06,
+      "loss": 0.6748,
+      "step": 5166
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5021842064630033,
+      "learning_rate": 2.20669628579799e-06,
+      "loss": 0.7371,
+      "step": 5167
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6612014852259103,
+      "learning_rate": 2.203590783772902e-06,
+      "loss": 0.7361,
+      "step": 5168
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.562022582459715,
+      "learning_rate": 2.2004871978865407e-06,
+      "loss": 0.7548,
+      "step": 5169
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8162698580900045,
+      "learning_rate": 2.1973855289016767e-06,
+      "loss": 0.8065,
+      "step": 5170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.667831941143919,
+      "learning_rate": 2.1942857775806215e-06,
+      "loss": 0.827,
+      "step": 5171
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4825779833500237,
+      "learning_rate": 2.1911879446852016e-06,
+      "loss": 0.7071,
+      "step": 5172
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.874618733898814,
+      "learning_rate": 2.1880920309767763e-06,
+      "loss": 0.723,
+      "step": 5173
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6996800389012137,
+      "learning_rate": 2.1849980372162393e-06,
+      "loss": 0.7636,
+      "step": 5174
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.922952745301586,
+      "learning_rate": 2.1819059641640015e-06,
+      "loss": 0.7544,
+      "step": 5175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4509868820621143,
+      "learning_rate": 2.178815812580015e-06,
+      "loss": 0.8044,
+      "step": 5176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7097921123798425,
+      "learning_rate": 2.175727583223748e-06,
+      "loss": 0.7816,
+      "step": 5177
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7726244992765974,
+      "learning_rate": 2.1726412768541984e-06,
+      "loss": 0.8708,
+      "step": 5178
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.646471864402857,
+      "learning_rate": 2.1695568942298984e-06,
+      "loss": 0.7968,
+      "step": 5179
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7006362811759965,
+      "learning_rate": 2.1664744361089042e-06,
+      "loss": 0.7764,
+      "step": 5180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4489036380967026,
+      "learning_rate": 2.1633939032487883e-06,
+      "loss": 0.7578,
+      "step": 5181
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8565073056265224,
+      "learning_rate": 2.160315296406669e-06,
+      "loss": 0.7507,
+      "step": 5182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.033558611119768,
+      "learning_rate": 2.1572386163391767e-06,
+      "loss": 0.8208,
+      "step": 5183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5760594105590156,
+      "learning_rate": 2.154163863802471e-06,
+      "loss": 0.7048,
+      "step": 5184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.572431198217354,
+      "learning_rate": 2.1510910395522454e-06,
+      "loss": 0.7405,
+      "step": 5185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.68710261224045,
+      "learning_rate": 2.1480201443437097e-06,
+      "loss": 0.7985,
+      "step": 5186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.7139550604337064,
+      "learning_rate": 2.144951178931608e-06,
+      "loss": 0.7096,
+      "step": 5187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.853358959872608,
+      "learning_rate": 2.1418841440702032e-06,
+      "loss": 0.8113,
+      "step": 5188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6227026638339734,
+      "learning_rate": 2.1388190405132835e-06,
+      "loss": 0.7304,
+      "step": 5189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.755493062706885,
+      "learning_rate": 2.135755869014171e-06,
+      "loss": 0.8311,
+      "step": 5190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.007649818068204,
+      "learning_rate": 2.1326946303257055e-06,
+      "loss": 0.7688,
+      "step": 5191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3892465626955914,
+      "learning_rate": 2.1296353252002535e-06,
+      "loss": 0.8222,
+      "step": 5192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8379742097250498,
+      "learning_rate": 2.126577954389706e-06,
+      "loss": 0.6913,
+      "step": 5193
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.603497319834362,
+      "learning_rate": 2.123522518645478e-06,
+      "loss": 0.8435,
+      "step": 5194
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.653831764050139,
+      "learning_rate": 2.1204690187185083e-06,
+      "loss": 0.8059,
+      "step": 5195
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.826818041431396,
+      "learning_rate": 2.1174174553592662e-06,
+      "loss": 0.8316,
+      "step": 5196
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7569910967979547,
+      "learning_rate": 2.114367829317737e-06,
+      "loss": 0.7594,
+      "step": 5197
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6456699660607916,
+      "learning_rate": 2.111320141343437e-06,
+      "loss": 0.7005,
+      "step": 5198
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6894829901775865,
+      "learning_rate": 2.1082743921853986e-06,
+      "loss": 0.8221,
+      "step": 5199
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4653459998874787,
+      "learning_rate": 2.1052305825921814e-06,
+      "loss": 0.7755,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3475070217631435,
+      "learning_rate": 2.1021887133118724e-06,
+      "loss": 0.7037,
+      "step": 5201
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5835424412431998,
+      "learning_rate": 2.0991487850920744e-06,
+      "loss": 0.8099,
+      "step": 5202
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.529540829600717,
+      "learning_rate": 2.0961107986799177e-06,
+      "loss": 0.7818,
+      "step": 5203
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5303647554019335,
+      "learning_rate": 2.0930747548220544e-06,
+      "loss": 0.8147,
+      "step": 5204
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.680295338295072,
+      "learning_rate": 2.0900406542646578e-06,
+      "loss": 0.7119,
+      "step": 5205
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4917329323772948,
+      "learning_rate": 2.0870084977534234e-06,
+      "loss": 0.709,
+      "step": 5206
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.680142357862478,
+      "learning_rate": 2.0839782860335744e-06,
+      "loss": 0.7774,
+      "step": 5207
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7202509491980726,
+      "learning_rate": 2.0809500198498465e-06,
+      "loss": 0.7406,
+      "step": 5208
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.538821119730595,
+      "learning_rate": 2.077923699946508e-06,
+      "loss": 0.6594,
+      "step": 5209
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.696771041993024,
+      "learning_rate": 2.0748993270673413e-06,
+      "loss": 0.6496,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9231310393079006,
+      "learning_rate": 2.0718769019556497e-06,
+      "loss": 0.7194,
+      "step": 5211
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4163140505938148,
+      "learning_rate": 2.0688564253542665e-06,
+      "loss": 0.7744,
+      "step": 5212
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4284110144498947,
+      "learning_rate": 2.065837898005536e-06,
+      "loss": 0.7628,
+      "step": 5213
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.564184491525239,
+      "learning_rate": 2.0628213206513283e-06,
+      "loss": 0.7292,
+      "step": 5214
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5657572703609413,
+      "learning_rate": 2.059806694033033e-06,
+      "loss": 0.7039,
+      "step": 5215
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.61890696078291,
+      "learning_rate": 2.056794018891559e-06,
+      "loss": 0.7752,
+      "step": 5216
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.841991160663711,
+      "learning_rate": 2.053783295967342e-06,
+      "loss": 0.8468,
+      "step": 5217
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.632280940282483,
+      "learning_rate": 2.050774526000331e-06,
+      "loss": 0.7935,
+      "step": 5218
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7062111362509427,
+      "learning_rate": 2.0477677097299944e-06,
+      "loss": 0.8167,
+      "step": 5219
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6616929561759517,
+      "learning_rate": 2.0447628478953285e-06,
+      "loss": 0.8678,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6295559456205364,
+      "learning_rate": 2.041759941234842e-06,
+      "loss": 0.7249,
+      "step": 5221
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.854155984098713,
+      "learning_rate": 2.0387589904865624e-06,
+      "loss": 0.8524,
+      "step": 5222
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.915882148566005,
+      "learning_rate": 2.035759996388044e-06,
+      "loss": 0.8829,
+      "step": 5223
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5580168750836223,
+      "learning_rate": 2.0327629596763522e-06,
+      "loss": 0.7832,
+      "step": 5224
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.667148793752107,
+      "learning_rate": 2.029767881088076e-06,
+      "loss": 0.7752,
+      "step": 5225
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.485601095948484,
+      "learning_rate": 2.0267747613593216e-06,
+      "loss": 0.7068,
+      "step": 5226
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3539907143857755,
+      "learning_rate": 2.0237836012257094e-06,
+      "loss": 0.851,
+      "step": 5227
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.207716113325043,
+      "learning_rate": 2.02079440142239e-06,
+      "loss": 0.7128,
+      "step": 5228
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6476668136581853,
+      "learning_rate": 2.0178071626840222e-06,
+      "loss": 0.7198,
+      "step": 5229
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.524164666724541,
+      "learning_rate": 2.014821885744782e-06,
+      "loss": 0.8026,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.433542763429434,
+      "learning_rate": 2.0118385713383717e-06,
+      "loss": 0.6877,
+      "step": 5231
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7388708012854384,
+      "learning_rate": 2.008857220198004e-06,
+      "loss": 0.8663,
+      "step": 5232
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.546278045195248,
+      "learning_rate": 2.0058778330564087e-06,
+      "loss": 0.7761,
+      "step": 5233
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5100886555285506,
+      "learning_rate": 2.002900410645842e-06,
+      "loss": 0.745,
+      "step": 5234
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.832589902344868,
+      "learning_rate": 1.999924953698067e-06,
+      "loss": 0.7755,
+      "step": 5235
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7081162850341634,
+      "learning_rate": 1.9969514629443676e-06,
+      "loss": 0.815,
+      "step": 5236
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.612861915456119,
+      "learning_rate": 1.9939799391155447e-06,
+      "loss": 0.7633,
+      "step": 5237
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.968076668129109,
+      "learning_rate": 1.9910103829419136e-06,
+      "loss": 0.7578,
+      "step": 5238
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7880471818672863,
+      "learning_rate": 1.9880427951533123e-06,
+      "loss": 0.7828,
+      "step": 5239
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5007751028079,
+      "learning_rate": 1.9850771764790866e-06,
+      "loss": 0.7722,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7926472560728772,
+      "learning_rate": 1.982113527648103e-06,
+      "loss": 0.7983,
+      "step": 5241
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.688729300603576,
+      "learning_rate": 1.9791518493887464e-06,
+      "loss": 0.7109,
+      "step": 5242
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.526455965144725,
+      "learning_rate": 1.976192142428912e-06,
+      "loss": 0.7415,
+      "step": 5243
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.361831127463669,
+      "learning_rate": 1.9732344074960106e-06,
+      "loss": 0.6652,
+      "step": 5244
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8803546372423576,
+      "learning_rate": 1.9702786453169753e-06,
+      "loss": 0.7467,
+      "step": 5245
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.606351929515565,
+      "learning_rate": 1.967324856618247e-06,
+      "loss": 0.7258,
+      "step": 5246
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5139030510363156,
+      "learning_rate": 1.9643730421257836e-06,
+      "loss": 0.714,
+      "step": 5247
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.561178712202148,
+      "learning_rate": 1.961423202565059e-06,
+      "loss": 0.6892,
+      "step": 5248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6508261197310112,
+      "learning_rate": 1.958475338661059e-06,
+      "loss": 0.6926,
+      "step": 5249
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.1385624324100405,
+      "learning_rate": 1.9555294511382895e-06,
+      "loss": 0.8052,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5151851416799134,
+      "learning_rate": 1.952585540720765e-06,
+      "loss": 0.7427,
+      "step": 5251
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7096528444167545,
+      "learning_rate": 1.949643608132015e-06,
+      "loss": 0.854,
+      "step": 5252
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6073748284771137,
+      "learning_rate": 1.946703654095087e-06,
+      "loss": 0.7681,
+      "step": 5253
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.647365142933058,
+      "learning_rate": 1.943765679332539e-06,
+      "loss": 0.7355,
+      "step": 5254
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.554651512314461,
+      "learning_rate": 1.9408296845664374e-06,
+      "loss": 0.7792,
+      "step": 5255
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7749366497128714,
+      "learning_rate": 1.937895670518374e-06,
+      "loss": 0.7521,
+      "step": 5256
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.6113040123040587,
+      "learning_rate": 1.934963637909445e-06,
+      "loss": 0.7682,
+      "step": 5257
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.68125144170482,
+      "learning_rate": 1.9320335874602615e-06,
+      "loss": 0.6771,
+      "step": 5258
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.79670004343362,
+      "learning_rate": 1.929105519890948e-06,
+      "loss": 0.7837,
+      "step": 5259
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.858531857582643,
+      "learning_rate": 1.9261794359211385e-06,
+      "loss": 0.797,
+      "step": 5260
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.601601436421406,
+      "learning_rate": 1.9232553362699867e-06,
+      "loss": 0.8745,
+      "step": 5261
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7155178948988015,
+      "learning_rate": 1.920333221656152e-06,
+      "loss": 0.7517,
+      "step": 5262
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.71136245899951,
+      "learning_rate": 1.9174130927978073e-06,
+      "loss": 0.8722,
+      "step": 5263
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6266359171360882,
+      "learning_rate": 1.914494950412642e-06,
+      "loss": 0.769,
+      "step": 5264
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.676658527106433,
+      "learning_rate": 1.9115787952178513e-06,
+      "loss": 0.8158,
+      "step": 5265
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7060334105432813,
+      "learning_rate": 1.9086646279301414e-06,
+      "loss": 0.7251,
+      "step": 5266
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6210489060040496,
+      "learning_rate": 1.9057524492657386e-06,
+      "loss": 0.7473,
+      "step": 5267
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.662061165888899,
+      "learning_rate": 1.902842259940373e-06,
+      "loss": 0.88,
+      "step": 5268
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5082720864633803,
+      "learning_rate": 1.899934060669285e-06,
+      "loss": 0.6801,
+      "step": 5269
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0120117717607733,
+      "learning_rate": 1.8970278521672314e-06,
+      "loss": 0.7691,
+      "step": 5270
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4061591232928166,
+      "learning_rate": 1.8941236351484727e-06,
+      "loss": 0.632,
+      "step": 5271
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5620260182509225,
+      "learning_rate": 1.891221410326789e-06,
+      "loss": 0.826,
+      "step": 5272
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6932349967864324,
+      "learning_rate": 1.888321178415463e-06,
+      "loss": 0.7042,
+      "step": 5273
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5244052706215725,
+      "learning_rate": 1.8854229401272883e-06,
+      "loss": 0.766,
+      "step": 5274
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7826439118860393,
+      "learning_rate": 1.8825266961745759e-06,
+      "loss": 0.7974,
+      "step": 5275
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.813590114997112,
+      "learning_rate": 1.8796324472691386e-06,
+      "loss": 0.7621,
+      "step": 5276
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4679542676635147,
+      "learning_rate": 1.8767401941222996e-06,
+      "loss": 0.7581,
+      "step": 5277
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6626177268115923,
+      "learning_rate": 1.873849937444897e-06,
+      "loss": 0.6859,
+      "step": 5278
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.514550595473596,
+      "learning_rate": 1.870961677947274e-06,
+      "loss": 0.7612,
+      "step": 5279
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.429325244420133,
+      "learning_rate": 1.8680754163392821e-06,
+      "loss": 0.8087,
+      "step": 5280
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.78224370282897,
+      "learning_rate": 1.8651911533302858e-06,
+      "loss": 0.7911,
+      "step": 5281
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6394580071077534,
+      "learning_rate": 1.8623088896291508e-06,
+      "loss": 0.7007,
+      "step": 5282
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.08473638196854,
+      "learning_rate": 1.859428625944264e-06,
+      "loss": 0.8031,
+      "step": 5283
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6230923065312197,
+      "learning_rate": 1.8565503629835102e-06,
+      "loss": 0.7139,
+      "step": 5284
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.484455187154463,
+      "learning_rate": 1.8536741014542825e-06,
+      "loss": 0.7813,
+      "step": 5285
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9025005355173255,
+      "learning_rate": 1.850799842063492e-06,
+      "loss": 0.8687,
+      "step": 5286
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7855494735084583,
+      "learning_rate": 1.847927585517546e-06,
+      "loss": 0.8703,
+      "step": 5287
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7774861500730554,
+      "learning_rate": 1.845057332522364e-06,
+      "loss": 0.7098,
+      "step": 5288
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5571608259583045,
+      "learning_rate": 1.8421890837833789e-06,
+      "loss": 0.7756,
+      "step": 5289
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6508095224438284,
+      "learning_rate": 1.8393228400055197e-06,
+      "loss": 0.7547,
+      "step": 5290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4331685506681624,
+      "learning_rate": 1.836458601893234e-06,
+      "loss": 0.6196,
+      "step": 5291
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8057215466078,
+      "learning_rate": 1.833596370150469e-06,
+      "loss": 0.8488,
+      "step": 5292
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8768433031219436,
+      "learning_rate": 1.8307361454806815e-06,
+      "loss": 0.8106,
+      "step": 5293
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6425605653461894,
+      "learning_rate": 1.8278779285868332e-06,
+      "loss": 0.8511,
+      "step": 5294
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2196702056179816,
+      "learning_rate": 1.8250217201713938e-06,
+      "loss": 0.7947,
+      "step": 5295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6762413209747815,
+      "learning_rate": 1.8221675209363376e-06,
+      "loss": 0.7019,
+      "step": 5296
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.218514063148928,
+      "learning_rate": 1.8193153315831514e-06,
+      "loss": 0.7803,
+      "step": 5297
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7580589992515314,
+      "learning_rate": 1.8164651528128208e-06,
+      "loss": 0.7302,
+      "step": 5298
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.738998097998686,
+      "learning_rate": 1.8136169853258379e-06,
+      "loss": 0.7869,
+      "step": 5299
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5623291055732937,
+      "learning_rate": 1.810770829822206e-06,
+      "loss": 0.776,
+      "step": 5300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.544255059308854,
+      "learning_rate": 1.8079266870014266e-06,
+      "loss": 0.8067,
+      "step": 5301
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6241190831600134,
+      "learning_rate": 1.8050845575625142e-06,
+      "loss": 0.6629,
+      "step": 5302
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.8966106388304853,
+      "learning_rate": 1.802244442203983e-06,
+      "loss": 0.8091,
+      "step": 5303
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.536447184938207,
+      "learning_rate": 1.7994063416238528e-06,
+      "loss": 0.7255,
+      "step": 5304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.628368067949291,
+      "learning_rate": 1.7965702565196496e-06,
+      "loss": 0.7316,
+      "step": 5305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7213475845824426,
+      "learning_rate": 1.793736187588404e-06,
+      "loss": 0.8211,
+      "step": 5306
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5743627911226468,
+      "learning_rate": 1.7909041355266465e-06,
+      "loss": 0.7483,
+      "step": 5307
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6514697681954487,
+      "learning_rate": 1.7880741010304236e-06,
+      "loss": 0.8514,
+      "step": 5308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6117066555165516,
+      "learning_rate": 1.7852460847952745e-06,
+      "loss": 0.7341,
+      "step": 5309
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7937497971549448,
+      "learning_rate": 1.7824200875162435e-06,
+      "loss": 0.8227,
+      "step": 5310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6601227382781683,
+      "learning_rate": 1.7795961098878867e-06,
+      "loss": 0.6883,
+      "step": 5311
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.555270181944585,
+      "learning_rate": 1.7767741526042537e-06,
+      "loss": 0.7181,
+      "step": 5312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.505665661355078,
+      "learning_rate": 1.7739542163589074e-06,
+      "loss": 0.7687,
+      "step": 5313
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.898396487815443,
+      "learning_rate": 1.7711363018449068e-06,
+      "loss": 0.8078,
+      "step": 5314
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3112582709990637,
+      "learning_rate": 1.7683204097548157e-06,
+      "loss": 0.8103,
+      "step": 5315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6724151435389834,
+      "learning_rate": 1.7655065407807025e-06,
+      "loss": 0.7417,
+      "step": 5316
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6001799710561384,
+      "learning_rate": 1.7626946956141355e-06,
+      "loss": 0.8023,
+      "step": 5317
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5608651624482253,
+      "learning_rate": 1.759884874946187e-06,
+      "loss": 0.7915,
+      "step": 5318
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4450320366094944,
+      "learning_rate": 1.7570770794674352e-06,
+      "loss": 0.8027,
+      "step": 5319
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7623894648539227,
+      "learning_rate": 1.7542713098679564e-06,
+      "loss": 0.7392,
+      "step": 5320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7112253465641376,
+      "learning_rate": 1.7514675668373272e-06,
+      "loss": 0.7352,
+      "step": 5321
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.462287508288741,
+      "learning_rate": 1.7486658510646337e-06,
+      "loss": 0.8257,
+      "step": 5322
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3514294846209958,
+      "learning_rate": 1.7458661632384532e-06,
+      "loss": 0.7096,
+      "step": 5323
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6405501271234275,
+      "learning_rate": 1.7430685040468775e-06,
+      "loss": 0.755,
+      "step": 5324
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8005781186999488,
+      "learning_rate": 1.7402728741774887e-06,
+      "loss": 0.681,
+      "step": 5325
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7722066542218786,
+      "learning_rate": 1.737479274317375e-06,
+      "loss": 0.7788,
+      "step": 5326
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5659134830313954,
+      "learning_rate": 1.7346877051531241e-06,
+      "loss": 0.7842,
+      "step": 5327
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.438173108512466,
+      "learning_rate": 1.731898167370827e-06,
+      "loss": 0.8067,
+      "step": 5328
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5327512937131007,
+      "learning_rate": 1.7291106616560693e-06,
+      "loss": 0.7078,
+      "step": 5329
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.836323440661082,
+      "learning_rate": 1.726325188693948e-06,
+      "loss": 0.7786,
+      "step": 5330
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.800206264294481,
+      "learning_rate": 1.7235417491690509e-06,
+      "loss": 0.8309,
+      "step": 5331
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5362830414927586,
+      "learning_rate": 1.7207603437654674e-06,
+      "loss": 0.6841,
+      "step": 5332
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.426639483725904,
+      "learning_rate": 1.7179809731667952e-06,
+      "loss": 0.7774,
+      "step": 5333
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.562833126398352,
+      "learning_rate": 1.7152036380561176e-06,
+      "loss": 0.7171,
+      "step": 5334
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7560900086185023,
+      "learning_rate": 1.7124283391160335e-06,
+      "loss": 0.7865,
+      "step": 5335
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6461379808218592,
+      "learning_rate": 1.7096550770286302e-06,
+      "loss": 0.781,
+      "step": 5336
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.56642023701555,
+      "learning_rate": 1.7068838524754961e-06,
+      "loss": 0.7718,
+      "step": 5337
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5198940553270335,
+      "learning_rate": 1.7041146661377229e-06,
+      "loss": 0.8147,
+      "step": 5338
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2246821208943555,
+      "learning_rate": 1.7013475186958983e-06,
+      "loss": 0.7905,
+      "step": 5339
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.488770000107356,
+      "learning_rate": 1.6985824108301063e-06,
+      "loss": 0.7767,
+      "step": 5340
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.579072452946949,
+      "learning_rate": 1.6958193432199377e-06,
+      "loss": 0.714,
+      "step": 5341
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.636564346699858,
+      "learning_rate": 1.6930583165444759e-06,
+      "loss": 0.8383,
+      "step": 5342
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.763894488128038,
+      "learning_rate": 1.6902993314823003e-06,
+      "loss": 0.8249,
+      "step": 5343
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.841520959324338,
+      "learning_rate": 1.6875423887114973e-06,
+      "loss": 0.7667,
+      "step": 5344
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.721729548909902,
+      "learning_rate": 1.6847874889096404e-06,
+      "loss": 0.8535,
+      "step": 5345
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.353445242020036,
+      "learning_rate": 1.682034632753813e-06,
+      "loss": 0.7144,
+      "step": 5346
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.86263924929456,
+      "learning_rate": 1.6792838209205865e-06,
+      "loss": 0.7944,
+      "step": 5347
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6271226482000403,
+      "learning_rate": 1.6765350540860336e-06,
+      "loss": 0.8035,
+      "step": 5348
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7871625156641815,
+      "learning_rate": 1.6737883329257242e-06,
+      "loss": 0.7978,
+      "step": 5349
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7585265680058004,
+      "learning_rate": 1.6710436581147248e-06,
+      "loss": 0.7069,
+      "step": 5350
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.973963568363898,
+      "learning_rate": 1.6683010303275982e-06,
+      "loss": 0.7856,
+      "step": 5351
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.640924319007717,
+      "learning_rate": 1.665560450238408e-06,
+      "loss": 0.7277,
+      "step": 5352
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6543214807904603,
+      "learning_rate": 1.6628219185207127e-06,
+      "loss": 0.7371,
+      "step": 5353
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5983459295205256,
+      "learning_rate": 1.6600854358475615e-06,
+      "loss": 0.6777,
+      "step": 5354
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5050712357400977,
+      "learning_rate": 1.6573510028915118e-06,
+      "loss": 0.7303,
+      "step": 5355
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.581829671919241,
+      "learning_rate": 1.654618620324604e-06,
+      "loss": 0.8259,
+      "step": 5356
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.757472544258648,
+      "learning_rate": 1.6518882888183874e-06,
+      "loss": 0.743,
+      "step": 5357
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6302986483984796,
+      "learning_rate": 1.6491600090438976e-06,
+      "loss": 0.7763,
+      "step": 5358
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.392032692849635,
+      "learning_rate": 1.646433781671669e-06,
+      "loss": 0.752,
+      "step": 5359
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4315483130409152,
+      "learning_rate": 1.643709607371733e-06,
+      "loss": 0.7202,
+      "step": 5360
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6533577527278807,
+      "learning_rate": 1.6409874868136132e-06,
+      "loss": 0.7539,
+      "step": 5361
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7936297349799672,
+      "learning_rate": 1.6382674206663308e-06,
+      "loss": 0.867,
+      "step": 5362
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.486355890016533,
+      "learning_rate": 1.6355494095984049e-06,
+      "loss": 0.6968,
+      "step": 5363
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.992793892900273,
+      "learning_rate": 1.6328334542778423e-06,
+      "loss": 0.8133,
+      "step": 5364
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0974063076184413,
+      "learning_rate": 1.6301195553721528e-06,
+      "loss": 0.7223,
+      "step": 5365
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5535242999454044,
+      "learning_rate": 1.6274077135483336e-06,
+      "loss": 0.8412,
+      "step": 5366
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0316245210841513,
+      "learning_rate": 1.6246979294728793e-06,
+      "loss": 0.7508,
+      "step": 5367
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7099534979043196,
+      "learning_rate": 1.6219902038117807e-06,
+      "loss": 0.7069,
+      "step": 5368
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0971035301260312,
+      "learning_rate": 1.619284537230521e-06,
+      "loss": 0.7463,
+      "step": 5369
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4461091608804377,
+      "learning_rate": 1.6165809303940761e-06,
+      "loss": 0.8191,
+      "step": 5370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8631386451165426,
+      "learning_rate": 1.6138793839669165e-06,
+      "loss": 0.6966,
+      "step": 5371
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.970022929368473,
+      "learning_rate": 1.6111798986130067e-06,
+      "loss": 0.7699,
+      "step": 5372
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.752622230132209,
+      "learning_rate": 1.6084824749958017e-06,
+      "loss": 0.7213,
+      "step": 5373
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.689384664158231,
+      "learning_rate": 1.6057871137782578e-06,
+      "loss": 0.7925,
+      "step": 5374
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.3902986847975325,
+      "learning_rate": 1.603093815622815e-06,
+      "loss": 0.7179,
+      "step": 5375
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6156407562968806,
+      "learning_rate": 1.6004025811914147e-06,
+      "loss": 0.7825,
+      "step": 5376
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8870697987925333,
+      "learning_rate": 1.5977134111454839e-06,
+      "loss": 0.7597,
+      "step": 5377
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5329585787430418,
+      "learning_rate": 1.5950263061459437e-06,
+      "loss": 0.8816,
+      "step": 5378
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.706232341302241,
+      "learning_rate": 1.5923412668532135e-06,
+      "loss": 0.7472,
+      "step": 5379
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.538461844289609,
+      "learning_rate": 1.5896582939271976e-06,
+      "loss": 0.7767,
+      "step": 5380
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6899633155459206,
+      "learning_rate": 1.5869773880272964e-06,
+      "loss": 0.7833,
+      "step": 5381
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6689412056515254,
+      "learning_rate": 1.584298549812402e-06,
+      "loss": 0.7906,
+      "step": 5382
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.616921439622051,
+      "learning_rate": 1.5816217799408962e-06,
+      "loss": 0.7398,
+      "step": 5383
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5860191831094523,
+      "learning_rate": 1.578947079070652e-06,
+      "loss": 0.785,
+      "step": 5384
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7409743051749684,
+      "learning_rate": 1.576274447859041e-06,
+      "loss": 0.7139,
+      "step": 5385
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.5230935232058465,
+      "learning_rate": 1.5736038869629168e-06,
+      "loss": 0.7318,
+      "step": 5386
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8231121808138373,
+      "learning_rate": 1.5709353970386322e-06,
+      "loss": 0.885,
+      "step": 5387
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.539737814038569,
+      "learning_rate": 1.5682689787420246e-06,
+      "loss": 0.7765,
+      "step": 5388
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.623946125219109,
+      "learning_rate": 1.5656046327284225e-06,
+      "loss": 0.7303,
+      "step": 5389
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6049099707804535,
+      "learning_rate": 1.5629423596526528e-06,
+      "loss": 0.8136,
+      "step": 5390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.759094162043502,
+      "learning_rate": 1.5602821601690254e-06,
+      "loss": 0.7784,
+      "step": 5391
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8019148720081737,
+      "learning_rate": 1.557624034931342e-06,
+      "loss": 0.7514,
+      "step": 5392
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.559663404527581,
+      "learning_rate": 1.5549679845928956e-06,
+      "loss": 0.8213,
+      "step": 5393
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6417069273809766,
+      "learning_rate": 1.5523140098064692e-06,
+      "loss": 0.8017,
+      "step": 5394
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8132581020986205,
+      "learning_rate": 1.5496621112243327e-06,
+      "loss": 0.842,
+      "step": 5395
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.539662982926937,
+      "learning_rate": 1.547012289498252e-06,
+      "loss": 0.8017,
+      "step": 5396
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2025744239552973,
+      "learning_rate": 1.5443645452794754e-06,
+      "loss": 0.7549,
+      "step": 5397
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7559697768379205,
+      "learning_rate": 1.5417188792187488e-06,
+      "loss": 0.803,
+      "step": 5398
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.603866972428103,
+      "learning_rate": 1.539075291966301e-06,
+      "loss": 0.9145,
+      "step": 5399
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.600906822519969,
+      "learning_rate": 1.536433784171848e-06,
+      "loss": 0.8172,
+      "step": 5400
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7699414581326964,
+      "learning_rate": 1.5337943564846035e-06,
+      "loss": 0.716,
+      "step": 5401
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4063352904234256,
+      "learning_rate": 1.5311570095532636e-06,
+      "loss": 0.8076,
+      "step": 5402
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5692469504876687,
+      "learning_rate": 1.5285217440260092e-06,
+      "loss": 0.7613,
+      "step": 5403
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.572772925076905,
+      "learning_rate": 1.5258885605505226e-06,
+      "loss": 0.7561,
+      "step": 5404
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.513167661863724,
+      "learning_rate": 1.5232574597739635e-06,
+      "loss": 0.7256,
+      "step": 5405
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.566421922523046,
+      "learning_rate": 1.5206284423429773e-06,
+      "loss": 0.7812,
+      "step": 5406
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3353428043208844,
+      "learning_rate": 1.5180015089037093e-06,
+      "loss": 0.8176,
+      "step": 5407
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5207035562280757,
+      "learning_rate": 1.5153766601017816e-06,
+      "loss": 0.8464,
+      "step": 5408
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5939553659111687,
+      "learning_rate": 1.512753896582313e-06,
+      "loss": 0.7761,
+      "step": 5409
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4308392420491383,
+      "learning_rate": 1.5101332189899032e-06,
+      "loss": 0.7191,
+      "step": 5410
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.752464101088354,
+      "learning_rate": 1.507514627968638e-06,
+      "loss": 0.7498,
+      "step": 5411
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.573045601314764,
+      "learning_rate": 1.5048981241620996e-06,
+      "loss": 0.7168,
+      "step": 5412
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.374466306394915,
+      "learning_rate": 1.5022837082133479e-06,
+      "loss": 0.8401,
+      "step": 5413
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.908769349069825,
+      "learning_rate": 1.499671380764931e-06,
+      "loss": 0.7582,
+      "step": 5414
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5939245885102267,
+      "learning_rate": 1.49706114245889e-06,
+      "loss": 0.8097,
+      "step": 5415
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.456523353585106,
+      "learning_rate": 1.494452993936747e-06,
+      "loss": 0.7621,
+      "step": 5416
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.479603609725367,
+      "learning_rate": 1.4918469358395104e-06,
+      "loss": 0.8219,
+      "step": 5417
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7942623118175147,
+      "learning_rate": 1.4892429688076771e-06,
+      "loss": 0.7696,
+      "step": 5418
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.46734362607315,
+      "learning_rate": 1.486641093481227e-06,
+      "loss": 0.6862,
+      "step": 5419
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.955292126554179,
+      "learning_rate": 1.4840413104996322e-06,
+      "loss": 0.7807,
+      "step": 5420
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.656829684311352,
+      "learning_rate": 1.4814436205018435e-06,
+      "loss": 0.6895,
+      "step": 5421
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.379219039208738,
+      "learning_rate": 1.4788480241262992e-06,
+      "loss": 0.7099,
+      "step": 5422
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.513725790883196,
+      "learning_rate": 1.4762545220109292e-06,
+      "loss": 0.6907,
+      "step": 5423
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.895226797679898,
+      "learning_rate": 1.473663114793139e-06,
+      "loss": 0.7467,
+      "step": 5424
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5764193868923346,
+      "learning_rate": 1.4710738031098226e-06,
+      "loss": 0.6766,
+      "step": 5425
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5882262502356967,
+      "learning_rate": 1.4684865875973663e-06,
+      "loss": 0.7805,
+      "step": 5426
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5245263727879115,
+      "learning_rate": 1.4659014688916306e-06,
+      "loss": 0.7652,
+      "step": 5427
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4390552809494097,
+      "learning_rate": 1.463318447627966e-06,
+      "loss": 0.6893,
+      "step": 5428
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4830323556842315,
+      "learning_rate": 1.460737524441207e-06,
+      "loss": 0.8335,
+      "step": 5429
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.580296423991282,
+      "learning_rate": 1.4581586999656706e-06,
+      "loss": 0.7148,
+      "step": 5430
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.531672934729575,
+      "learning_rate": 1.455581974835162e-06,
+      "loss": 0.8116,
+      "step": 5431
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.097529937171779,
+      "learning_rate": 1.4530073496829667e-06,
+      "loss": 0.7705,
+      "step": 5432
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.462077058668238,
+      "learning_rate": 1.4504348251418532e-06,
+      "loss": 0.7132,
+      "step": 5433
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.510920740948573,
+      "learning_rate": 1.4478644018440813e-06,
+      "loss": 0.7715,
+      "step": 5434
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7183210859443965,
+      "learning_rate": 1.4452960804213844e-06,
+      "loss": 0.7191,
+      "step": 5435
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4965139774054697,
+      "learning_rate": 1.4427298615049834e-06,
+      "loss": 0.7392,
+      "step": 5436
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4493008215582424,
+      "learning_rate": 1.440165745725588e-06,
+      "loss": 0.6751,
+      "step": 5437
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8053401617693123,
+      "learning_rate": 1.4376037337133818e-06,
+      "loss": 0.8177,
+      "step": 5438
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6707495812946984,
+      "learning_rate": 1.435043826098037e-06,
+      "loss": 0.718,
+      "step": 5439
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.766260855397848,
+      "learning_rate": 1.4324860235087069e-06,
+      "loss": 0.6813,
+      "step": 5440
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7959908546395873,
+      "learning_rate": 1.4299303265740238e-06,
+      "loss": 0.7618,
+      "step": 5441
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5790556035434795,
+      "learning_rate": 1.4273767359221125e-06,
+      "loss": 0.7761,
+      "step": 5442
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.667404676892974,
+      "learning_rate": 1.4248252521805716e-06,
+      "loss": 0.7206,
+      "step": 5443
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9916119664187026,
+      "learning_rate": 1.422275875976482e-06,
+      "loss": 0.6933,
+      "step": 5444
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9269245266415003,
+      "learning_rate": 1.4197286079364125e-06,
+      "loss": 0.7127,
+      "step": 5445
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7534623467804686,
+      "learning_rate": 1.4171834486864089e-06,
+      "loss": 0.7684,
+      "step": 5446
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8876537105947166,
+      "learning_rate": 1.4146403988519963e-06,
+      "loss": 0.7838,
+      "step": 5447
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8182121582454123,
+      "learning_rate": 1.4120994590581916e-06,
+      "loss": 0.7686,
+      "step": 5448
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.082920784480041,
+      "learning_rate": 1.4095606299294827e-06,
+      "loss": 0.8127,
+      "step": 5449
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8232693322075004,
+      "learning_rate": 1.4070239120898433e-06,
+      "loss": 0.7653,
+      "step": 5450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1856411989759135,
+      "learning_rate": 1.4044893061627263e-06,
+      "loss": 0.8263,
+      "step": 5451
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.676050937118328,
+      "learning_rate": 1.4019568127710659e-06,
+      "loss": 0.6561,
+      "step": 5452
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6926069696599217,
+      "learning_rate": 1.399426432537283e-06,
+      "loss": 0.7164,
+      "step": 5453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.633397695011211,
+      "learning_rate": 1.3968981660832693e-06,
+      "loss": 0.7313,
+      "step": 5454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7503718548301737,
+      "learning_rate": 1.3943720140304018e-06,
+      "loss": 0.8687,
+      "step": 5455
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7667013768472226,
+      "learning_rate": 1.3918479769995418e-06,
+      "loss": 0.759,
+      "step": 5456
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7654801691226893,
+      "learning_rate": 1.3893260556110243e-06,
+      "loss": 0.7636,
+      "step": 5457
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5775808923066905,
+      "learning_rate": 1.3868062504846646e-06,
+      "loss": 0.7662,
+      "step": 5458
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6147303801081803,
+      "learning_rate": 1.384288562239765e-06,
+      "loss": 0.7888,
+      "step": 5459
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.555246259881729,
+      "learning_rate": 1.3817729914950995e-06,
+      "loss": 0.7107,
+      "step": 5460
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6749101810783924,
+      "learning_rate": 1.3792595388689267e-06,
+      "loss": 0.8295,
+      "step": 5461
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9072465544160893,
+      "learning_rate": 1.3767482049789804e-06,
+      "loss": 0.7472,
+      "step": 5462
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7431607296729603,
+      "learning_rate": 1.3742389904424747e-06,
+      "loss": 0.7544,
+      "step": 5463
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4523455376569108,
+      "learning_rate": 1.3717318958761094e-06,
+      "loss": 0.7079,
+      "step": 5464
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6762219222918024,
+      "learning_rate": 1.3692269218960553e-06,
+      "loss": 0.8438,
+      "step": 5465
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5159210232208418,
+      "learning_rate": 1.3667240691179618e-06,
+      "loss": 0.8016,
+      "step": 5466
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7143958324723036,
+      "learning_rate": 1.3642233381569657e-06,
+      "loss": 0.7513,
+      "step": 5467
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.30670046856752,
+      "learning_rate": 1.3617247296276737e-06,
+      "loss": 0.805,
+      "step": 5468
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.711816364111367,
+      "learning_rate": 1.3592282441441707e-06,
+      "loss": 0.6686,
+      "step": 5469
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8587264851624865,
+      "learning_rate": 1.3567338823200293e-06,
+      "loss": 0.7559,
+      "step": 5470
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9278007847640493,
+      "learning_rate": 1.3542416447682893e-06,
+      "loss": 0.7923,
+      "step": 5471
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.784064503676124,
+      "learning_rate": 1.351751532101473e-06,
+      "loss": 0.8089,
+      "step": 5472
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.469197716587361,
+      "learning_rate": 1.3492635449315817e-06,
+      "loss": 0.7145,
+      "step": 5473
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5894330340506713,
+      "learning_rate": 1.3467776838700896e-06,
+      "loss": 0.7227,
+      "step": 5474
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.471371802458738,
+      "learning_rate": 1.344293949527956e-06,
+      "loss": 0.678,
+      "step": 5475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.798773140145835,
+      "learning_rate": 1.3418123425156115e-06,
+      "loss": 0.7428,
+      "step": 5476
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5164154651123667,
+      "learning_rate": 1.3393328634429636e-06,
+      "loss": 0.7172,
+      "step": 5477
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.917342816664016,
+      "learning_rate": 1.3368555129194016e-06,
+      "loss": 0.8129,
+      "step": 5478
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.638060194697016,
+      "learning_rate": 1.3343802915537885e-06,
+      "loss": 0.8024,
+      "step": 5479
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7588261673114265,
+      "learning_rate": 1.3319071999544607e-06,
+      "loss": 0.8135,
+      "step": 5480
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.551590169459782,
+      "learning_rate": 1.3294362387292391e-06,
+      "loss": 0.7706,
+      "step": 5481
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.944194600995512,
+      "learning_rate": 1.3269674084854156e-06,
+      "loss": 0.8076,
+      "step": 5482
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6657774464142485,
+      "learning_rate": 1.324500709829759e-06,
+      "loss": 0.8025,
+      "step": 5483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0273054989836354,
+      "learning_rate": 1.3220361433685137e-06,
+      "loss": 0.8681,
+      "step": 5484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2634353936725327,
+      "learning_rate": 1.319573709707399e-06,
+      "loss": 0.8016,
+      "step": 5485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.773258396954878,
+      "learning_rate": 1.317113409451618e-06,
+      "loss": 0.7462,
+      "step": 5486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4504422966436388,
+      "learning_rate": 1.31465524320584e-06,
+      "loss": 0.7048,
+      "step": 5487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6226489048346395,
+      "learning_rate": 1.3121992115742122e-06,
+      "loss": 0.7727,
+      "step": 5488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.499913869641089,
+      "learning_rate": 1.3097453151603602e-06,
+      "loss": 0.8013,
+      "step": 5489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7080732867459343,
+      "learning_rate": 1.3072935545673836e-06,
+      "loss": 0.8594,
+      "step": 5490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6214148178831653,
+      "learning_rate": 1.3048439303978534e-06,
+      "loss": 0.8599,
+      "step": 5491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8228550531261165,
+      "learning_rate": 1.3023964432538216e-06,
+      "loss": 0.7535,
+      "step": 5492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6231774742032004,
+      "learning_rate": 1.2999510937368109e-06,
+      "loss": 0.742,
+      "step": 5493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7395097723549884,
+      "learning_rate": 1.2975078824478181e-06,
+      "loss": 0.765,
+      "step": 5494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5325901699463103,
+      "learning_rate": 1.2950668099873186e-06,
+      "loss": 0.8974,
+      "step": 5495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7576079244145997,
+      "learning_rate": 1.2926278769552558e-06,
+      "loss": 0.8176,
+      "step": 5496
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6983001920413723,
+      "learning_rate": 1.2901910839510557e-06,
+      "loss": 0.7643,
+      "step": 5497
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7065562167614767,
+      "learning_rate": 1.2877564315736114e-06,
+      "loss": 0.8274,
+      "step": 5498
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.592440020847409,
+      "learning_rate": 1.2853239204212908e-06,
+      "loss": 0.7146,
+      "step": 5499
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6082475382596924,
+      "learning_rate": 1.2828935510919393e-06,
+      "loss": 0.6917,
+      "step": 5500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7716965757094347,
+      "learning_rate": 1.2804653241828724e-06,
+      "loss": 0.7834,
+      "step": 5501
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4278847914452375,
+      "learning_rate": 1.2780392402908793e-06,
+      "loss": 0.7978,
+      "step": 5502
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6464039784162,
+      "learning_rate": 1.2756153000122252e-06,
+      "loss": 0.7255,
+      "step": 5503
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.638944099478603,
+      "learning_rate": 1.2731935039426469e-06,
+      "loss": 0.744,
+      "step": 5504
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.65867797896158,
+      "learning_rate": 1.2707738526773528e-06,
+      "loss": 0.7639,
+      "step": 5505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.81615400279156,
+      "learning_rate": 1.268356346811025e-06,
+      "loss": 0.9008,
+      "step": 5506
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6510825781269673,
+      "learning_rate": 1.2659409869378159e-06,
+      "loss": 0.818,
+      "step": 5507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.866416891237285,
+      "learning_rate": 1.2635277736513596e-06,
+      "loss": 0.804,
+      "step": 5508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0494698466774426,
+      "learning_rate": 1.2611167075447527e-06,
+      "loss": 0.7794,
+      "step": 5509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7386950159586627,
+      "learning_rate": 1.258707789210566e-06,
+      "loss": 0.7816,
+      "step": 5510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.785828567783617,
+      "learning_rate": 1.2563010192408487e-06,
+      "loss": 0.7821,
+      "step": 5511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5718903666323296,
+      "learning_rate": 1.2538963982271135e-06,
+      "loss": 0.7565,
+      "step": 5512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.124178037199288,
+      "learning_rate": 1.2514939267603489e-06,
+      "loss": 0.823,
+      "step": 5513
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7898308110383474,
+      "learning_rate": 1.2490936054310176e-06,
+      "loss": 0.7378,
+      "step": 5514
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4373883835923507,
+      "learning_rate": 1.2466954348290473e-06,
+      "loss": 0.7807,
+      "step": 5515
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.466570363766371,
+      "learning_rate": 1.244299415543846e-06,
+      "loss": 0.7284,
+      "step": 5516
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5189153820994323,
+      "learning_rate": 1.2419055481642873e-06,
+      "loss": 0.7953,
+      "step": 5517
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6885918862881444,
+      "learning_rate": 1.2395138332787105e-06,
+      "loss": 0.7116,
+      "step": 5518
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5023705356369255,
+      "learning_rate": 1.2371242714749388e-06,
+      "loss": 0.7292,
+      "step": 5519
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5056392809495147,
+      "learning_rate": 1.2347368633402578e-06,
+      "loss": 0.7093,
+      "step": 5520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6732155135175915,
+      "learning_rate": 1.2323516094614218e-06,
+      "loss": 0.8043,
+      "step": 5521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5131745570401867,
+      "learning_rate": 1.229968510424665e-06,
+      "loss": 0.6928,
+      "step": 5522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3139445552542974,
+      "learning_rate": 1.2275875668156812e-06,
+      "loss": 0.7818,
+      "step": 5523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5637986429478112,
+      "learning_rate": 1.2252087792196432e-06,
+      "loss": 0.624,
+      "step": 5524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.720954635132861,
+      "learning_rate": 1.2228321482211903e-06,
+      "loss": 0.7613,
+      "step": 5525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.421612728710161,
+      "learning_rate": 1.2204576744044284e-06,
+      "loss": 0.802,
+      "step": 5526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.727338927023116,
+      "learning_rate": 1.2180853583529394e-06,
+      "loss": 0.826,
+      "step": 5527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9617344064018982,
+      "learning_rate": 1.215715200649773e-06,
+      "loss": 0.7746,
+      "step": 5528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6211203203398132,
+      "learning_rate": 1.2133472018774439e-06,
+      "loss": 0.7474,
+      "step": 5529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.060772179332325,
+      "learning_rate": 1.2109813626179434e-06,
+      "loss": 0.7757,
+      "step": 5530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4640680019112717,
+      "learning_rate": 1.2086176834527252e-06,
+      "loss": 0.7388,
+      "step": 5531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4614625902107177,
+      "learning_rate": 1.2062561649627158e-06,
+      "loss": 0.7391,
+      "step": 5532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6876423721050724,
+      "learning_rate": 1.203896807728313e-06,
+      "loss": 0.773,
+      "step": 5533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7163326211812153,
+      "learning_rate": 1.2015396123293766e-06,
+      "loss": 0.8139,
+      "step": 5534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.71148568189279,
+      "learning_rate": 1.1991845793452438e-06,
+      "loss": 0.7613,
+      "step": 5535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.751931038453203,
+      "learning_rate": 1.1968317093547133e-06,
+      "loss": 0.8681,
+      "step": 5536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6658751770376754,
+      "learning_rate": 1.1944810029360532e-06,
+      "loss": 0.7588,
+      "step": 5537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.668427297252355,
+      "learning_rate": 1.1921324606670037e-06,
+      "loss": 0.7865,
+      "step": 5538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.89947410474797,
+      "learning_rate": 1.1897860831247686e-06,
+      "loss": 0.7112,
+      "step": 5539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5563106453481006,
+      "learning_rate": 1.1874418708860237e-06,
+      "loss": 0.7953,
+      "step": 5540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.702224895601683,
+      "learning_rate": 1.1850998245269096e-06,
+      "loss": 0.8173,
+      "step": 5541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7219772816758465,
+      "learning_rate": 1.1827599446230354e-06,
+      "loss": 0.7415,
+      "step": 5542
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.811823473183657,
+      "learning_rate": 1.180422231749475e-06,
+      "loss": 0.7129,
+      "step": 5543
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8051088616260356,
+      "learning_rate": 1.1780866864807795e-06,
+      "loss": 0.7613,
+      "step": 5544
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.516067073289398,
+      "learning_rate": 1.1757533093909535e-06,
+      "loss": 0.768,
+      "step": 5545
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5930686110231003,
+      "learning_rate": 1.1734221010534807e-06,
+      "loss": 0.7489,
+      "step": 5546
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.893175999553963,
+      "learning_rate": 1.1710930620413053e-06,
+      "loss": 0.7489,
+      "step": 5547
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8849620602373243,
+      "learning_rate": 1.1687661929268367e-06,
+      "loss": 0.7634,
+      "step": 5548
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4784255720944017,
+      "learning_rate": 1.166441494281959e-06,
+      "loss": 0.7371,
+      "step": 5549
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.771587298983736,
+      "learning_rate": 1.1641189666780151e-06,
+      "loss": 0.7193,
+      "step": 5550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8362983127978527,
+      "learning_rate": 1.161798610685818e-06,
+      "loss": 0.8037,
+      "step": 5551
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.583812039759552,
+      "learning_rate": 1.1594804268756455e-06,
+      "loss": 0.6902,
+      "step": 5552
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7400986861722365,
+      "learning_rate": 1.1571644158172435e-06,
+      "loss": 0.7665,
+      "step": 5553
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8029871469307004,
+      "learning_rate": 1.154850578079818e-06,
+      "loss": 0.7821,
+      "step": 5554
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5630396676092575,
+      "learning_rate": 1.152538914232052e-06,
+      "loss": 0.6432,
+      "step": 5555
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.754520435755249,
+      "learning_rate": 1.150229424842082e-06,
+      "loss": 0.7511,
+      "step": 5556
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.668987688707433,
+      "learning_rate": 1.1479221104775195e-06,
+      "loss": 0.7309,
+      "step": 5557
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5638068256997326,
+      "learning_rate": 1.1456169717054378e-06,
+      "loss": 0.6905,
+      "step": 5558
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.690726492060148,
+      "learning_rate": 1.143314009092371e-06,
+      "loss": 0.7855,
+      "step": 5559
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4735568575923454,
+      "learning_rate": 1.141013223204328e-06,
+      "loss": 0.6855,
+      "step": 5560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5690760259180982,
+      "learning_rate": 1.138714614606775e-06,
+      "loss": 0.7458,
+      "step": 5561
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4077556748136186,
+      "learning_rate": 1.136418183864646e-06,
+      "loss": 0.7496,
+      "step": 5562
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.468467344759974,
+      "learning_rate": 1.134123931542339e-06,
+      "loss": 0.7601,
+      "step": 5563
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7634747107670807,
+      "learning_rate": 1.1318318582037168e-06,
+      "loss": 0.7783,
+      "step": 5564
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.559302157414664,
+      "learning_rate": 1.129541964412104e-06,
+      "loss": 0.756,
+      "step": 5565
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.699526488995371,
+      "learning_rate": 1.1272542507302985e-06,
+      "loss": 0.7129,
+      "step": 5566
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.974066739013522,
+      "learning_rate": 1.1249687177205493e-06,
+      "loss": 0.8527,
+      "step": 5567
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9764748728047463,
+      "learning_rate": 1.1226853659445824e-06,
+      "loss": 0.7279,
+      "step": 5568
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.857166506430615,
+      "learning_rate": 1.1204041959635791e-06,
+      "loss": 0.7442,
+      "step": 5569
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.896244832002908,
+      "learning_rate": 1.1181252083381844e-06,
+      "loss": 0.7165,
+      "step": 5570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3254418710580094,
+      "learning_rate": 1.1158484036285134e-06,
+      "loss": 0.7253,
+      "step": 5571
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4974170064587518,
+      "learning_rate": 1.1135737823941405e-06,
+      "loss": 0.7949,
+      "step": 5572
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.777471113005515,
+      "learning_rate": 1.1113013451941024e-06,
+      "loss": 0.7766,
+      "step": 5573
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.921270848027808,
+      "learning_rate": 1.1090310925869009e-06,
+      "loss": 0.7718,
+      "step": 5574
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.701574184000981,
+      "learning_rate": 1.1067630251304996e-06,
+      "loss": 0.8679,
+      "step": 5575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.787188494578665,
+      "learning_rate": 1.104497143382325e-06,
+      "loss": 0.7404,
+      "step": 5576
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.551177086497416,
+      "learning_rate": 1.1022334478992702e-06,
+      "loss": 0.7234,
+      "step": 5577
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.232648327249752,
+      "learning_rate": 1.099971939237685e-06,
+      "loss": 0.759,
+      "step": 5578
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5073434658427947,
+      "learning_rate": 1.0977126179533892e-06,
+      "loss": 0.7839,
+      "step": 5579
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.541437023443651,
+      "learning_rate": 1.0954554846016575e-06,
+      "loss": 0.7859,
+      "step": 5580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5804987815371825,
+      "learning_rate": 1.0932005397372282e-06,
+      "loss": 0.7573,
+      "step": 5581
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.838112090513552,
+      "learning_rate": 1.090947783914308e-06,
+      "loss": 0.7422,
+      "step": 5582
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.659588045163355,
+      "learning_rate": 1.0886972176865585e-06,
+      "loss": 0.7169,
+      "step": 5583
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6839338857706836,
+      "learning_rate": 1.0864488416071061e-06,
+      "loss": 0.731,
+      "step": 5584
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3940028914051665,
+      "learning_rate": 1.084202656228538e-06,
+      "loss": 0.6844,
+      "step": 5585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7620213694437195,
+      "learning_rate": 1.0819586621029043e-06,
+      "loss": 0.7331,
+      "step": 5586
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.774375832645594,
+      "learning_rate": 1.0797168597817143e-06,
+      "loss": 0.6923,
+      "step": 5587
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1324940080961343,
+      "learning_rate": 1.0774772498159424e-06,
+      "loss": 0.7889,
+      "step": 5588
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.439717577235111,
+      "learning_rate": 1.0752398327560199e-06,
+      "loss": 0.6448,
+      "step": 5589
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6372831755194044,
+      "learning_rate": 1.0730046091518442e-06,
+      "loss": 0.7257,
+      "step": 5590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.809764341873967,
+      "learning_rate": 1.0707715795527685e-06,
+      "loss": 0.7913,
+      "step": 5591
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.530267851629339,
+      "learning_rate": 1.0685407445076067e-06,
+      "loss": 0.6289,
+      "step": 5592
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7837106686303845,
+      "learning_rate": 1.0663121045646397e-06,
+      "loss": 0.7535,
+      "step": 5593
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8086295330354774,
+      "learning_rate": 1.0640856602716021e-06,
+      "loss": 0.7656,
+      "step": 5594
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4661618844981947,
+      "learning_rate": 1.0618614121756932e-06,
+      "loss": 0.7703,
+      "step": 5595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8436514864476696,
+      "learning_rate": 1.059639360823569e-06,
+      "loss": 0.8581,
+      "step": 5596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.331913971552762,
+      "learning_rate": 1.057419506761347e-06,
+      "loss": 0.815,
+      "step": 5597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9949193504323883,
+      "learning_rate": 1.0552018505346074e-06,
+      "loss": 0.7474,
+      "step": 5598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6932560263632497,
+      "learning_rate": 1.0529863926883865e-06,
+      "loss": 0.8571,
+      "step": 5599
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.72659374994844,
+      "learning_rate": 1.05077313376718e-06,
+      "loss": 0.6961,
+      "step": 5600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.386739434798129,
+      "learning_rate": 1.0485620743149494e-06,
+      "loss": 0.7768,
+      "step": 5601
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7055649797455272,
+      "learning_rate": 1.0463532148751076e-06,
+      "loss": 0.7969,
+      "step": 5602
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.753211825418324,
+      "learning_rate": 1.0441465559905295e-06,
+      "loss": 0.7847,
+      "step": 5603
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.629614530438831,
+      "learning_rate": 1.0419420982035545e-06,
+      "loss": 0.7199,
+      "step": 5604
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.634367194061076,
+      "learning_rate": 1.0397398420559724e-06,
+      "loss": 0.7798,
+      "step": 5605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7751002146810633,
+      "learning_rate": 1.037539788089037e-06,
+      "loss": 0.763,
+      "step": 5606
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6799197719112042,
+      "learning_rate": 1.0353419368434614e-06,
+      "loss": 0.7823,
+      "step": 5607
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7524900554080403,
+      "learning_rate": 1.0331462888594112e-06,
+      "loss": 0.8375,
+      "step": 5608
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.917619249390206,
+      "learning_rate": 1.0309528446765206e-06,
+      "loss": 0.7743,
+      "step": 5609
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8815542439790574,
+      "learning_rate": 1.0287616048338743e-06,
+      "loss": 0.8678,
+      "step": 5610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8823740915476574,
+      "learning_rate": 1.026572569870017e-06,
+      "loss": 0.7844,
+      "step": 5611
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7691903823345037,
+      "learning_rate": 1.024385740322954e-06,
+      "loss": 0.6978,
+      "step": 5612
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.748898140798564,
+      "learning_rate": 1.022201116730145e-06,
+      "loss": 0.8205,
+      "step": 5613
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.502627608195643,
+      "learning_rate": 1.0200186996285077e-06,
+      "loss": 0.7322,
+      "step": 5614
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7651574060507675,
+      "learning_rate": 1.0178384895544235e-06,
+      "loss": 0.7698,
+      "step": 5615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6948552555209138,
+      "learning_rate": 1.0156604870437247e-06,
+      "loss": 0.8678,
+      "step": 5616
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4570240141927884,
+      "learning_rate": 1.0134846926317022e-06,
+      "loss": 0.7441,
+      "step": 5617
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5496315754472625,
+      "learning_rate": 1.0113111068531068e-06,
+      "loss": 0.7771,
+      "step": 5618
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4844705154268096,
+      "learning_rate": 1.0091397302421412e-06,
+      "loss": 0.6951,
+      "step": 5619
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9074753927062607,
+      "learning_rate": 1.0069705633324745e-06,
+      "loss": 0.8395,
+      "step": 5620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.870455798197107,
+      "learning_rate": 1.004803606657223e-06,
+      "loss": 0.7325,
+      "step": 5621
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.792599450580263,
+      "learning_rate": 1.0026388607489646e-06,
+      "loss": 0.7905,
+      "step": 5622
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0586931850734733,
+      "learning_rate": 1.0004763261397355e-06,
+      "loss": 0.8382,
+      "step": 5623
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.816909936823562,
+      "learning_rate": 9.98316003361025e-07,
+      "loss": 0.8835,
+      "step": 5624
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.657891923324879,
+      "learning_rate": 9.961578929437764e-07,
+      "loss": 0.7731,
+      "step": 5625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7264248273390295,
+      "learning_rate": 9.940019954183977e-07,
+      "loss": 0.8202,
+      "step": 5626
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6457426310497985,
+      "learning_rate": 9.918483113147447e-07,
+      "loss": 0.7751,
+      "step": 5627
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0766698726447506,
+      "learning_rate": 9.896968411621332e-07,
+      "loss": 0.8389,
+      "step": 5628
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.844765688371593,
+      "learning_rate": 9.87547585489338e-07,
+      "loss": 0.721,
+      "step": 5629
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4495238218117894,
+      "learning_rate": 9.854005448245796e-07,
+      "loss": 0.6981,
+      "step": 5630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.43118087190644,
+      "learning_rate": 9.832557196955438e-07,
+      "loss": 0.754,
+      "step": 5631
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7459618919864357,
+      "learning_rate": 9.811131106293691e-07,
+      "loss": 0.7378,
+      "step": 5632
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6027613647952124,
+      "learning_rate": 9.789727181526453e-07,
+      "loss": 0.767,
+      "step": 5633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.827180784194871,
+      "learning_rate": 9.76834542791425e-07,
+      "loss": 0.826,
+      "step": 5634
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2869195172489576,
+      "learning_rate": 9.746985850712099e-07,
+      "loss": 0.7602,
+      "step": 5635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.518787706950005,
+      "learning_rate": 9.725648455169568e-07,
+      "loss": 0.7141,
+      "step": 5636
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.909684582981514,
+      "learning_rate": 9.704333246530828e-07,
+      "loss": 0.7257,
+      "step": 5637
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0101097572734803,
+      "learning_rate": 9.683040230034536e-07,
+      "loss": 0.7605,
+      "step": 5638
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8883948525240966,
+      "learning_rate": 9.661769410913913e-07,
+      "loss": 0.8731,
+      "step": 5639
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.799359360157228,
+      "learning_rate": 9.640520794396746e-07,
+      "loss": 0.7386,
+      "step": 5640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.515485834092157,
+      "learning_rate": 9.619294385705336e-07,
+      "loss": 0.7184,
+      "step": 5641
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3554919634425526,
+      "learning_rate": 9.598090190056553e-07,
+      "loss": 0.6601,
+      "step": 5642
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5777835963991556,
+      "learning_rate": 9.576908212661784e-07,
+      "loss": 0.7231,
+      "step": 5643
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.544230852123378,
+      "learning_rate": 9.555748458726944e-07,
+      "loss": 0.7102,
+      "step": 5644
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4889271852604997,
+      "learning_rate": 9.534610933452548e-07,
+      "loss": 0.7839,
+      "step": 5645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.011943460171467,
+      "learning_rate": 9.513495642033599e-07,
+      "loss": 0.755,
+      "step": 5646
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5838351272639097,
+      "learning_rate": 9.4924025896596e-07,
+      "loss": 0.8452,
+      "step": 5647
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.513812262444184,
+      "learning_rate": 9.471331781514681e-07,
+      "loss": 0.779,
+      "step": 5648
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6039552895971103,
+      "learning_rate": 9.450283222777445e-07,
+      "loss": 0.7665,
+      "step": 5649
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5535137404506805,
+      "learning_rate": 9.429256918621011e-07,
+      "loss": 0.7957,
+      "step": 5650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.814064035545963,
+      "learning_rate": 9.408252874213097e-07,
+      "loss": 0.7728,
+      "step": 5651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.437786969150219,
+      "learning_rate": 9.387271094715877e-07,
+      "loss": 0.735,
+      "step": 5652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7107592789968256,
+      "learning_rate": 9.366311585286103e-07,
+      "loss": 0.7294,
+      "step": 5653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6797509758975924,
+      "learning_rate": 9.345374351075009e-07,
+      "loss": 0.8449,
+      "step": 5654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.702157913427348,
+      "learning_rate": 9.324459397228391e-07,
+      "loss": 0.7923,
+      "step": 5655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.667032279356962,
+      "learning_rate": 9.303566728886571e-07,
+      "loss": 0.7211,
+      "step": 5656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.483166462186134,
+      "learning_rate": 9.282696351184383e-07,
+      "loss": 0.6771,
+      "step": 5657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.632341617741301,
+      "learning_rate": 9.26184826925114e-07,
+      "loss": 0.7805,
+      "step": 5658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.803600915223199,
+      "learning_rate": 9.241022488210772e-07,
+      "loss": 0.7631,
+      "step": 5659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.668583864035917,
+      "learning_rate": 9.220219013181642e-07,
+      "loss": 0.6902,
+      "step": 5660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9654201592454927,
+      "learning_rate": 9.199437849276649e-07,
+      "loss": 0.7293,
+      "step": 5661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4537593166619436,
+      "learning_rate": 9.178679001603252e-07,
+      "loss": 0.6717,
+      "step": 5662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7924386936278873,
+      "learning_rate": 9.15794247526337e-07,
+      "loss": 0.8249,
+      "step": 5663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5150201116445263,
+      "learning_rate": 9.137228275353471e-07,
+      "loss": 0.715,
+      "step": 5664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.782277092977421,
+      "learning_rate": 9.116536406964527e-07,
+      "loss": 0.8094,
+      "step": 5665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6310827622808963,
+      "learning_rate": 9.095866875181991e-07,
+      "loss": 0.6985,
+      "step": 5666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7013028558305043,
+      "learning_rate": 9.07521968508589e-07,
+      "loss": 0.7311,
+      "step": 5667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.504527699009366,
+      "learning_rate": 9.054594841750707e-07,
+      "loss": 0.6948,
+      "step": 5668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9194448959018002,
+      "learning_rate": 9.033992350245435e-07,
+      "loss": 0.8246,
+      "step": 5669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.580593402067048,
+      "learning_rate": 9.013412215633633e-07,
+      "loss": 0.7519,
+      "step": 5670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.645194130581552,
+      "learning_rate": 8.992854442973264e-07,
+      "loss": 0.7213,
+      "step": 5671
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4402723740062746,
+      "learning_rate": 8.972319037316901e-07,
+      "loss": 0.6859,
+      "step": 5672
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6777006267138024,
+      "learning_rate": 8.95180600371156e-07,
+      "loss": 0.7776,
+      "step": 5673
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6668439690078825,
+      "learning_rate": 8.931315347198754e-07,
+      "loss": 0.7065,
+      "step": 5674
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.558904510622878,
+      "learning_rate": 8.910847072814521e-07,
+      "loss": 0.8571,
+      "step": 5675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8898235434369774,
+      "learning_rate": 8.890401185589393e-07,
+      "loss": 0.7485,
+      "step": 5676
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6045615465844207,
+      "learning_rate": 8.86997769054836e-07,
+      "loss": 0.74,
+      "step": 5677
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5371702561275455,
+      "learning_rate": 8.849576592710996e-07,
+      "loss": 0.7178,
+      "step": 5678
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9241583684169945,
+      "learning_rate": 8.82919789709129e-07,
+      "loss": 0.7445,
+      "step": 5679
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5870081381342502,
+      "learning_rate": 8.808841608697749e-07,
+      "loss": 0.8143,
+      "step": 5680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6290340495413433,
+      "learning_rate": 8.788507732533413e-07,
+      "loss": 0.8404,
+      "step": 5681
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.592317293842339,
+      "learning_rate": 8.768196273595719e-07,
+      "loss": 0.7048,
+      "step": 5682
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6255917103261073,
+      "learning_rate": 8.747907236876718e-07,
+      "loss": 0.8538,
+      "step": 5683
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8929535869985923,
+      "learning_rate": 8.727640627362854e-07,
+      "loss": 0.7605,
+      "step": 5684
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5282572880948555,
+      "learning_rate": 8.707396450035099e-07,
+      "loss": 0.6847,
+      "step": 5685
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.571181113520221,
+      "learning_rate": 8.687174709868895e-07,
+      "loss": 0.821,
+      "step": 5686
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.640049537712735,
+      "learning_rate": 8.666975411834188e-07,
+      "loss": 0.7872,
+      "step": 5687
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.15738494350599,
+      "learning_rate": 8.646798560895376e-07,
+      "loss": 0.8605,
+      "step": 5688
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.660662027497053,
+      "learning_rate": 8.626644162011399e-07,
+      "loss": 0.7281,
+      "step": 5689
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6983042556708496,
+      "learning_rate": 8.606512220135621e-07,
+      "loss": 0.7855,
+      "step": 5690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4483957728925616,
+      "learning_rate": 8.586402740215893e-07,
+      "loss": 0.7887,
+      "step": 5691
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.778073690955736,
+      "learning_rate": 8.566315727194607e-07,
+      "loss": 0.8097,
+      "step": 5692
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7771629099321484,
+      "learning_rate": 8.546251186008536e-07,
+      "loss": 0.701,
+      "step": 5693
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.498328459699538,
+      "learning_rate": 8.526209121589024e-07,
+      "loss": 0.7905,
+      "step": 5694
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6098896557905844,
+      "learning_rate": 8.506189538861831e-07,
+      "loss": 0.8243,
+      "step": 5695
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.677295059627817,
+      "learning_rate": 8.486192442747221e-07,
+      "loss": 0.7308,
+      "step": 5696
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5776146690712345,
+      "learning_rate": 8.466217838159896e-07,
+      "loss": 0.8068,
+      "step": 5697
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.571734748384213,
+      "learning_rate": 8.446265730009074e-07,
+      "loss": 0.6534,
+      "step": 5698
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7287346704837434,
+      "learning_rate": 8.426336123198386e-07,
+      "loss": 0.7079,
+      "step": 5699
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5595723305614015,
+      "learning_rate": 8.406429022626028e-07,
+      "loss": 0.7684,
+      "step": 5700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.519081924016723,
+      "learning_rate": 8.386544433184573e-07,
+      "loss": 0.7304,
+      "step": 5701
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6865616162166055,
+      "learning_rate": 8.36668235976108e-07,
+      "loss": 0.6706,
+      "step": 5702
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6769547229461383,
+      "learning_rate": 8.346842807237132e-07,
+      "loss": 0.799,
+      "step": 5703
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.525744181272662,
+      "learning_rate": 8.327025780488696e-07,
+      "loss": 0.6805,
+      "step": 5704
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4046255837962276,
+      "learning_rate": 8.307231284386264e-07,
+      "loss": 0.7225,
+      "step": 5705
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.515854941221048,
+      "learning_rate": 8.287459323794777e-07,
+      "loss": 0.661,
+      "step": 5706
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4848884090304417,
+      "learning_rate": 8.267709903573606e-07,
+      "loss": 0.7989,
+      "step": 5707
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8341237442734837,
+      "learning_rate": 8.247983028576612e-07,
+      "loss": 0.8496,
+      "step": 5708
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7306413767137143,
+      "learning_rate": 8.228278703652115e-07,
+      "loss": 0.743,
+      "step": 5709
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7006097693456637,
+      "learning_rate": 8.208596933642854e-07,
+      "loss": 0.7519,
+      "step": 5710
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.667161566314598,
+      "learning_rate": 8.188937723386104e-07,
+      "loss": 0.7382,
+      "step": 5711
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4897380903135375,
+      "learning_rate": 8.16930107771352e-07,
+      "loss": 0.6433,
+      "step": 5712
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6686472482951227,
+      "learning_rate": 8.149687001451223e-07,
+      "loss": 0.7862,
+      "step": 5713
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6432076739400863,
+      "learning_rate": 8.130095499419843e-07,
+      "loss": 0.768,
+      "step": 5714
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.623041787253728,
+      "learning_rate": 8.110526576434386e-07,
+      "loss": 0.8018,
+      "step": 5715
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.361320310110989,
+      "learning_rate": 8.090980237304369e-07,
+      "loss": 0.693,
+      "step": 5716
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.53061217832916,
+      "learning_rate": 8.071456486833729e-07,
+      "loss": 0.7252,
+      "step": 5717
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7445462678102412,
+      "learning_rate": 8.051955329820849e-07,
+      "loss": 0.8299,
+      "step": 5718
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.678869645619878,
+      "learning_rate": 8.032476771058572e-07,
+      "loss": 0.779,
+      "step": 5719
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.597783234585979,
+      "learning_rate": 8.013020815334182e-07,
+      "loss": 0.7771,
+      "step": 5720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5709764296981987,
+      "learning_rate": 7.993587467429387e-07,
+      "loss": 0.6913,
+      "step": 5721
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8878615808643717,
+      "learning_rate": 7.974176732120386e-07,
+      "loss": 0.8386,
+      "step": 5722
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4739802793848904,
+      "learning_rate": 7.954788614177789e-07,
+      "loss": 0.7136,
+      "step": 5723
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5254962861324204,
+      "learning_rate": 7.935423118366625e-07,
+      "loss": 0.7467,
+      "step": 5724
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.59825821177977,
+      "learning_rate": 7.916080249446434e-07,
+      "loss": 0.749,
+      "step": 5725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6911630767002617,
+      "learning_rate": 7.896760012171101e-07,
+      "loss": 0.8131,
+      "step": 5726
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7269735173965386,
+      "learning_rate": 7.87746241128905e-07,
+      "loss": 0.8272,
+      "step": 5727
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.655547149168662,
+      "learning_rate": 7.858187451543064e-07,
+      "loss": 0.7408,
+      "step": 5728
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7887679312481746,
+      "learning_rate": 7.838935137670378e-07,
+      "loss": 0.8403,
+      "step": 5729
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8238335497241875,
+      "learning_rate": 7.819705474402695e-07,
+      "loss": 0.8438,
+      "step": 5730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7864697800450675,
+      "learning_rate": 7.800498466466099e-07,
+      "loss": 0.7503,
+      "step": 5731
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9423775361485385,
+      "learning_rate": 7.781314118581141e-07,
+      "loss": 0.6889,
+      "step": 5732
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6177300541736197,
+      "learning_rate": 7.762152435462821e-07,
+      "loss": 0.7418,
+      "step": 5733
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5682224575904224,
+      "learning_rate": 7.743013421820522e-07,
+      "loss": 0.7338,
+      "step": 5734
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.730093806909977,
+      "learning_rate": 7.723897082358067e-07,
+      "loss": 0.7525,
+      "step": 5735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.773570026247756,
+      "learning_rate": 7.704803421773743e-07,
+      "loss": 0.8018,
+      "step": 5736
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8167438145337145,
+      "learning_rate": 7.685732444760197e-07,
+      "loss": 0.749,
+      "step": 5737
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8460419034294295,
+      "learning_rate": 7.666684156004589e-07,
+      "loss": 0.7815,
+      "step": 5738
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.581997146323194,
+      "learning_rate": 7.647658560188431e-07,
+      "loss": 0.6945,
+      "step": 5739
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6421514833054456,
+      "learning_rate": 7.628655661987661e-07,
+      "loss": 0.8141,
+      "step": 5740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.500537719209546,
+      "learning_rate": 7.609675466072719e-07,
+      "loss": 0.7023,
+      "step": 5741
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4936945926955283,
+      "learning_rate": 7.590717977108342e-07,
+      "loss": 0.7819,
+      "step": 5742
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.59145708475709,
+      "learning_rate": 7.571783199753746e-07,
+      "loss": 0.7877,
+      "step": 5743
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8791756410563267,
+      "learning_rate": 7.552871138662621e-07,
+      "loss": 0.8033,
+      "step": 5744
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.410861373680256,
+      "learning_rate": 7.53398179848297e-07,
+      "loss": 0.6554,
+      "step": 5745
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6788493966132716,
+      "learning_rate": 7.515115183857302e-07,
+      "loss": 0.8059,
+      "step": 5746
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.509424351655514,
+      "learning_rate": 7.496271299422498e-07,
+      "loss": 0.6685,
+      "step": 5747
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.487752174672153,
+      "learning_rate": 7.477450149809818e-07,
+      "loss": 0.6712,
+      "step": 5748
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8476622128192064,
+      "learning_rate": 7.458651739645017e-07,
+      "loss": 0.8459,
+      "step": 5749
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.559185180461762,
+      "learning_rate": 7.439876073548192e-07,
+      "loss": 0.6805,
+      "step": 5750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.459377863705062,
+      "learning_rate": 7.421123156133869e-07,
+      "loss": 0.7042,
+      "step": 5751
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4854631564887613,
+      "learning_rate": 7.402392992011032e-07,
+      "loss": 0.7098,
+      "step": 5752
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.772999848604892,
+      "learning_rate": 7.383685585782985e-07,
+      "loss": 0.8132,
+      "step": 5753
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.296150380444605,
+      "learning_rate": 7.365000942047506e-07,
+      "loss": 0.8222,
+      "step": 5754
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.083231204527326,
+      "learning_rate": 7.346339065396746e-07,
+      "loss": 0.8012,
+      "step": 5755
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4491012331318753,
+      "learning_rate": 7.327699960417256e-07,
+      "loss": 0.6899,
+      "step": 5756
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7249084898062574,
+      "learning_rate": 7.30908363169005e-07,
+      "loss": 0.6925,
+      "step": 5757
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6967437072752967,
+      "learning_rate": 7.290490083790458e-07,
+      "loss": 0.7866,
+      "step": 5758
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.556881939811135,
+      "learning_rate": 7.271919321288268e-07,
+      "loss": 0.7139,
+      "step": 5759
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.815172538512507,
+      "learning_rate": 7.253371348747662e-07,
+      "loss": 0.8048,
+      "step": 5760
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.638440578512759,
+      "learning_rate": 7.234846170727194e-07,
+      "loss": 0.7051,
+      "step": 5761
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6966172220810654,
+      "learning_rate": 7.216343791779834e-07,
+      "loss": 0.7231,
+      "step": 5762
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.5534587002684463,
+      "learning_rate": 7.197864216452965e-07,
+      "loss": 0.7466,
+      "step": 5763
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7849363578109165,
+      "learning_rate": 7.179407449288344e-07,
+      "loss": 0.7409,
+      "step": 5764
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.544837759440299,
+      "learning_rate": 7.160973494822121e-07,
+      "loss": 0.6465,
+      "step": 5765
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7231069748329695,
+      "learning_rate": 7.142562357584836e-07,
+      "loss": 0.7866,
+      "step": 5766
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.546310544712134,
+      "learning_rate": 7.124174042101428e-07,
+      "loss": 0.7346,
+      "step": 5767
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0119781270597197,
+      "learning_rate": 7.105808552891258e-07,
+      "loss": 0.8738,
+      "step": 5768
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.588257545636272,
+      "learning_rate": 7.087465894468037e-07,
+      "loss": 0.7377,
+      "step": 5769
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6439528897738858,
+      "learning_rate": 7.069146071339839e-07,
+      "loss": 0.8051,
+      "step": 5770
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7572845518844895,
+      "learning_rate": 7.050849088009216e-07,
+      "loss": 0.6755,
+      "step": 5771
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5912309063614627,
+      "learning_rate": 7.032574948973037e-07,
+      "loss": 0.7551,
+      "step": 5772
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.710962518663575,
+      "learning_rate": 7.014323658722544e-07,
+      "loss": 0.7734,
+      "step": 5773
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6644565985000748,
+      "learning_rate": 6.996095221743426e-07,
+      "loss": 0.7861,
+      "step": 5774
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7841204191393847,
+      "learning_rate": 6.977889642515711e-07,
+      "loss": 0.8666,
+      "step": 5775
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4809075380037533,
+      "learning_rate": 6.959706925513832e-07,
+      "loss": 0.6904,
+      "step": 5776
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6378329404590213,
+      "learning_rate": 6.941547075206567e-07,
+      "loss": 0.7582,
+      "step": 5777
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4058948024512397,
+      "learning_rate": 6.923410096057093e-07,
+      "loss": 0.7166,
+      "step": 5778
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7027364478522724,
+      "learning_rate": 6.905295992522998e-07,
+      "loss": 0.6934,
+      "step": 5779
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8777912482750647,
+      "learning_rate": 6.887204769056221e-07,
+      "loss": 0.8297,
+      "step": 5780
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0010290591989044,
+      "learning_rate": 6.86913643010304e-07,
+      "loss": 0.8016,
+      "step": 5781
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8797253529185736,
+      "learning_rate": 6.851090980104191e-07,
+      "loss": 0.7898,
+      "step": 5782
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8338261112987286,
+      "learning_rate": 6.833068423494727e-07,
+      "loss": 0.8343,
+      "step": 5783
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6323910473659247,
+      "learning_rate": 6.815068764704047e-07,
+      "loss": 0.8172,
+      "step": 5784
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.596841816940055,
+      "learning_rate": 6.797092008156026e-07,
+      "loss": 0.7403,
+      "step": 5785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.738457357878896,
+      "learning_rate": 6.779138158268806e-07,
+      "loss": 0.6369,
+      "step": 5786
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.915690322939979,
+      "learning_rate": 6.761207219454957e-07,
+      "loss": 0.8547,
+      "step": 5787
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.537974599329801,
+      "learning_rate": 6.743299196121389e-07,
+      "loss": 0.7582,
+      "step": 5788
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.707714310242144,
+      "learning_rate": 6.725414092669391e-07,
+      "loss": 0.7383,
+      "step": 5789
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5190834121324293,
+      "learning_rate": 6.707551913494626e-07,
+      "loss": 0.7306,
+      "step": 5790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8560257151264334,
+      "learning_rate": 6.689712662987124e-07,
+      "loss": 0.7703,
+      "step": 5791
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6256493925464923,
+      "learning_rate": 6.671896345531248e-07,
+      "loss": 0.7432,
+      "step": 5792
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4196845841221815,
+      "learning_rate": 6.654102965505782e-07,
+      "loss": 0.6261,
+      "step": 5793
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6043213927859536,
+      "learning_rate": 6.636332527283817e-07,
+      "loss": 0.7592,
+      "step": 5794
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3460123501037726,
+      "learning_rate": 6.618585035232828e-07,
+      "loss": 0.6926,
+      "step": 5795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5889018853522097,
+      "learning_rate": 6.600860493714667e-07,
+      "loss": 0.7117,
+      "step": 5796
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.515917583771837,
+      "learning_rate": 6.583158907085518e-07,
+      "loss": 0.6722,
+      "step": 5797
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7197011832215168,
+      "learning_rate": 6.565480279695946e-07,
+      "loss": 0.7158,
+      "step": 5798
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.229504657309361,
+      "learning_rate": 6.547824615890841e-07,
+      "loss": 0.7511,
+      "step": 5799
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5513386297995084,
+      "learning_rate": 6.530191920009465e-07,
+      "loss": 0.7745,
+      "step": 5800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.616216804463169,
+      "learning_rate": 6.512582196385475e-07,
+      "loss": 0.7391,
+      "step": 5801
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9565292094630533,
+      "learning_rate": 6.49499544934683e-07,
+      "loss": 0.7664,
+      "step": 5802
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.626493868178572,
+      "learning_rate": 6.477431683215841e-07,
+      "loss": 0.7836,
+      "step": 5803
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5789068803703685,
+      "learning_rate": 6.459890902309218e-07,
+      "loss": 0.6654,
+      "step": 5804
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8827734647635666,
+      "learning_rate": 6.442373110937994e-07,
+      "loss": 0.6883,
+      "step": 5805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7036197267519055,
+      "learning_rate": 6.424878313407501e-07,
+      "loss": 0.6766,
+      "step": 5806
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4623104650551375,
+      "learning_rate": 6.407406514017534e-07,
+      "loss": 0.6671,
+      "step": 5807
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.441650887017995,
+      "learning_rate": 6.389957717062145e-07,
+      "loss": 0.5803,
+      "step": 5808
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.72535880083023,
+      "learning_rate": 6.37253192682975e-07,
+      "loss": 0.7387,
+      "step": 5809
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.531639961170924,
+      "learning_rate": 6.355129147603134e-07,
+      "loss": 0.7407,
+      "step": 5810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.1542566998202872,
+      "learning_rate": 6.337749383659386e-07,
+      "loss": 0.8718,
+      "step": 5811
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5198596314597026,
+      "learning_rate": 6.32039263926999e-07,
+      "loss": 0.694,
+      "step": 5812
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5182991329606703,
+      "learning_rate": 6.303058918700744e-07,
+      "loss": 0.7802,
+      "step": 5813
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.603245498148873,
+      "learning_rate": 6.285748226211774e-07,
+      "loss": 0.7041,
+      "step": 5814
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7378847104911825,
+      "learning_rate": 6.268460566057599e-07,
+      "loss": 0.7567,
+      "step": 5815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.539181928795849,
+      "learning_rate": 6.251195942487009e-07,
+      "loss": 0.7489,
+      "step": 5816
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6082265092629866,
+      "learning_rate": 6.233954359743155e-07,
+      "loss": 0.8378,
+      "step": 5817
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.479348939551202,
+      "learning_rate": 6.216735822063569e-07,
+      "loss": 0.7271,
+      "step": 5818
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6870083550884587,
+      "learning_rate": 6.199540333680065e-07,
+      "loss": 0.7784,
+      "step": 5819
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7374816629494965,
+      "learning_rate": 6.18236789881882e-07,
+      "loss": 0.796,
+      "step": 5820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.599437310217699,
+      "learning_rate": 6.165218521700333e-07,
+      "loss": 0.8368,
+      "step": 5821
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6962671130641436,
+      "learning_rate": 6.148092206539425e-07,
+      "loss": 0.6851,
+      "step": 5822
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.546788351836191,
+      "learning_rate": 6.130988957545281e-07,
+      "loss": 0.7044,
+      "step": 5823
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4294266150794495,
+      "learning_rate": 6.113908778921407e-07,
+      "loss": 0.7229,
+      "step": 5824
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5925856274482286,
+      "learning_rate": 6.0968516748656e-07,
+      "loss": 0.7794,
+      "step": 5825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8770495316561764,
+      "learning_rate": 6.079817649570052e-07,
+      "loss": 0.7082,
+      "step": 5826
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.799481628204912,
+      "learning_rate": 6.062806707221236e-07,
+      "loss": 0.769,
+      "step": 5827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7666018489574338,
+      "learning_rate": 6.045818851999952e-07,
+      "loss": 0.7609,
+      "step": 5828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.753622059951933,
+      "learning_rate": 6.028854088081359e-07,
+      "loss": 0.7736,
+      "step": 5829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5700093823746983,
+      "learning_rate": 6.011912419634924e-07,
+      "loss": 0.6826,
+      "step": 5830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8027455356023525,
+      "learning_rate": 5.994993850824415e-07,
+      "loss": 0.8633,
+      "step": 5831
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0498463017469133,
+      "learning_rate": 5.97809838580794e-07,
+      "loss": 0.7948,
+      "step": 5832
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.787764168001535,
+      "learning_rate": 5.961226028737932e-07,
+      "loss": 0.7494,
+      "step": 5833
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0044287571788915,
+      "learning_rate": 5.944376783761164e-07,
+      "loss": 0.7201,
+      "step": 5834
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6799421290682717,
+      "learning_rate": 5.927550655018699e-07,
+      "loss": 0.7481,
+      "step": 5835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7264929525021024,
+      "learning_rate": 5.910747646645898e-07,
+      "loss": 0.8224,
+      "step": 5836
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.545838928841274,
+      "learning_rate": 5.893967762772512e-07,
+      "loss": 0.7329,
+      "step": 5837
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6500564552598376,
+      "learning_rate": 5.877211007522555e-07,
+      "loss": 0.7454,
+      "step": 5838
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.65799984455843,
+      "learning_rate": 5.86047738501433e-07,
+      "loss": 0.7239,
+      "step": 5839
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6193677259589876,
+      "learning_rate": 5.843766899360547e-07,
+      "loss": 0.8434,
+      "step": 5840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3599943054612464,
+      "learning_rate": 5.827079554668147e-07,
+      "loss": 0.7064,
+      "step": 5841
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5828320727422587,
+      "learning_rate": 5.810415355038413e-07,
+      "loss": 0.7733,
+      "step": 5842
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.954381968066485,
+      "learning_rate": 5.793774304566946e-07,
+      "loss": 0.8062,
+      "step": 5843
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5084137036333116,
+      "learning_rate": 5.777156407343621e-07,
+      "loss": 0.8645,
+      "step": 5844
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6051884694422123,
+      "learning_rate": 5.76056166745268e-07,
+      "loss": 0.8177,
+      "step": 5845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6013475332208693,
+      "learning_rate": 5.74399008897265e-07,
+      "loss": 0.7638,
+      "step": 5846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5731747532030598,
+      "learning_rate": 5.72744167597632e-07,
+      "loss": 0.7762,
+      "step": 5847
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6559945677786647,
+      "learning_rate": 5.710916432530877e-07,
+      "loss": 0.7291,
+      "step": 5848
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.575724733370775,
+      "learning_rate": 5.694414362697742e-07,
+      "loss": 0.8015,
+      "step": 5849
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6141726668382512,
+      "learning_rate": 5.677935470532636e-07,
+      "loss": 0.7739,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.869045509165927,
+      "learning_rate": 5.661479760085642e-07,
+      "loss": 0.8365,
+      "step": 5851
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.579470374575014,
+      "learning_rate": 5.645047235401091e-07,
+      "loss": 0.6617,
+      "step": 5852
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.665435683114741,
+      "learning_rate": 5.628637900517652e-07,
+      "loss": 0.7943,
+      "step": 5853
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.737771133863642,
+      "learning_rate": 5.612251759468301e-07,
+      "loss": 0.7013,
+      "step": 5854
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6553279117918236,
+      "learning_rate": 5.595888816280226e-07,
+      "loss": 0.8019,
+      "step": 5855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.523369735721981,
+      "learning_rate": 5.579549074975032e-07,
+      "loss": 0.6509,
+      "step": 5856
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.0010206491405205,
+      "learning_rate": 5.563232539568553e-07,
+      "loss": 0.8583,
+      "step": 5857
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5704308172659847,
+      "learning_rate": 5.546939214070923e-07,
+      "loss": 0.7319,
+      "step": 5858
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9387864037827605,
+      "learning_rate": 5.530669102486619e-07,
+      "loss": 0.8983,
+      "step": 5859
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.934935762272051,
+      "learning_rate": 5.514422208814352e-07,
+      "loss": 0.6828,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.334253343305994,
+      "learning_rate": 5.49819853704715e-07,
+      "loss": 0.7281,
+      "step": 5861
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8362699742929722,
+      "learning_rate": 5.481998091172358e-07,
+      "loss": 0.7761,
+      "step": 5862
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5752370173710646,
+      "learning_rate": 5.465820875171557e-07,
+      "loss": 0.7658,
+      "step": 5863
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7342653154702,
+      "learning_rate": 5.4496668930207e-07,
+      "loss": 0.818,
+      "step": 5864
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.625860994123965,
+      "learning_rate": 5.433536148689944e-07,
+      "loss": 0.7583,
+      "step": 5865
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.774371794317327,
+      "learning_rate": 5.417428646143797e-07,
+      "loss": 0.773,
+      "step": 5866
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7174080154169933,
+      "learning_rate": 5.401344389341013e-07,
+      "loss": 0.7942,
+      "step": 5867
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.928065420008455,
+      "learning_rate": 5.385283382234674e-07,
+      "loss": 0.7364,
+      "step": 5868
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.071811201597634,
+      "learning_rate": 5.369245628772079e-07,
+      "loss": 0.72,
+      "step": 5869
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.608659449286244,
+      "learning_rate": 5.35323113289491e-07,
+      "loss": 0.6901,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.684645889963231,
+      "learning_rate": 5.337239898539071e-07,
+      "loss": 0.7628,
+      "step": 5871
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6352813074432464,
+      "learning_rate": 5.321271929634719e-07,
+      "loss": 0.8173,
+      "step": 5872
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.590851206610053,
+      "learning_rate": 5.305327230106383e-07,
+      "loss": 0.6468,
+      "step": 5873
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.752933732434142,
+      "learning_rate": 5.289405803872782e-07,
+      "loss": 0.7054,
+      "step": 5874
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2451802406056287,
+      "learning_rate": 5.273507654846999e-07,
+      "loss": 0.7577,
+      "step": 5875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.853764991107126,
+      "learning_rate": 5.257632786936328e-07,
+      "loss": 0.7259,
+      "step": 5876
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4914983750520885,
+      "learning_rate": 5.241781204042362e-07,
+      "loss": 0.6737,
+      "step": 5877
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.05209228111768,
+      "learning_rate": 5.225952910060994e-07,
+      "loss": 0.7027,
+      "step": 5878
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.621867794165036,
+      "learning_rate": 5.210147908882357e-07,
+      "loss": 0.7497,
+      "step": 5879
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.922865468524496,
+      "learning_rate": 5.194366204390867e-07,
+      "loss": 0.8475,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5874957343095515,
+      "learning_rate": 5.178607800465252e-07,
+      "loss": 0.6427,
+      "step": 5881
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.825060543684903,
+      "learning_rate": 5.162872700978483e-07,
+      "loss": 0.762,
+      "step": 5882
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9238862189206114,
+      "learning_rate": 5.147160909797777e-07,
+      "loss": 0.8199,
+      "step": 5883
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1311383501295964,
+      "learning_rate": 5.13147243078469e-07,
+      "loss": 0.8092,
+      "step": 5884
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8516286192122204,
+      "learning_rate": 5.11580726779497e-07,
+      "loss": 0.7055,
+      "step": 5885
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.597901538018006,
+      "learning_rate": 5.100165424678715e-07,
+      "loss": 0.7527,
+      "step": 5886
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8139780948798947,
+      "learning_rate": 5.08454690528023e-07,
+      "loss": 0.7478,
+      "step": 5887
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.434325694543393,
+      "learning_rate": 5.06895171343812e-07,
+      "loss": 0.7304,
+      "step": 5888
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4674361312600506,
+      "learning_rate": 5.05337985298523e-07,
+      "loss": 0.7271,
+      "step": 5889
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.642364749860154,
+      "learning_rate": 5.037831327748699e-07,
+      "loss": 0.7311,
+      "step": 5890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.688287039097314,
+      "learning_rate": 5.022306141549893e-07,
+      "loss": 0.7586,
+      "step": 5891
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4741994629105313,
+      "learning_rate": 5.006804298204515e-07,
+      "loss": 0.7872,
+      "step": 5892
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.854934089901339,
+      "learning_rate": 4.991325801522429e-07,
+      "loss": 0.6969,
+      "step": 5893
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.41087337300248,
+      "learning_rate": 4.975870655307868e-07,
+      "loss": 0.7949,
+      "step": 5894
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.953241328716546,
+      "learning_rate": 4.96043886335924e-07,
+      "loss": 0.7946,
+      "step": 5895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8049058507014983,
+      "learning_rate": 4.945030429469244e-07,
+      "loss": 0.7912,
+      "step": 5896
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.595414400806504,
+      "learning_rate": 4.929645357424862e-07,
+      "loss": 0.7498,
+      "step": 5897
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.745635922213741,
+      "learning_rate": 4.914283651007312e-07,
+      "loss": 0.7651,
+      "step": 5898
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.982896718226123,
+      "learning_rate": 4.898945313992054e-07,
+      "loss": 0.7171,
+      "step": 5899
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6219179763701024,
+      "learning_rate": 4.883630350148827e-07,
+      "loss": 0.737,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.602707647768999,
+      "learning_rate": 4.868338763241631e-07,
+      "loss": 0.7237,
+      "step": 5901
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6003569589861706,
+      "learning_rate": 4.853070557028672e-07,
+      "loss": 0.7409,
+      "step": 5902
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5458129925277726,
+      "learning_rate": 4.837825735262503e-07,
+      "loss": 0.7459,
+      "step": 5903
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.758220418720208,
+      "learning_rate": 4.822604301689826e-07,
+      "loss": 0.8263,
+      "step": 5904
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.79996513102502,
+      "learning_rate": 4.807406260051672e-07,
+      "loss": 0.848,
+      "step": 5905
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7388566308633995,
+      "learning_rate": 4.792231614083287e-07,
+      "loss": 0.7703,
+      "step": 5906
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.704544888648876,
+      "learning_rate": 4.777080367514153e-07,
+      "loss": 0.8292,
+      "step": 5907
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7338348905418997,
+      "learning_rate": 4.7619525240680475e-07,
+      "loss": 0.7844,
+      "step": 5908
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.545428147432849,
+      "learning_rate": 4.746848087462963e-07,
+      "loss": 0.7442,
+      "step": 5909
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.751373583389988,
+      "learning_rate": 4.731767061411141e-07,
+      "loss": 0.7968,
+      "step": 5910
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.632331818799675,
+      "learning_rate": 4.716709449619084e-07,
+      "loss": 0.8126,
+      "step": 5911
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.397912522997727,
+      "learning_rate": 4.701675255787519e-07,
+      "loss": 0.7333,
+      "step": 5912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8191668230435645,
+      "learning_rate": 4.686664483611425e-07,
+      "loss": 0.6911,
+      "step": 5913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5749939501751276,
+      "learning_rate": 4.6716771367800507e-07,
+      "loss": 0.6782,
+      "step": 5914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6443502932254215,
+      "learning_rate": 4.656713218976838e-07,
+      "loss": 0.6271,
+      "step": 5915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0452845496714342,
+      "learning_rate": 4.641772733879535e-07,
+      "loss": 0.8101,
+      "step": 5916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.292735176357161,
+      "learning_rate": 4.626855685160059e-07,
+      "loss": 0.8173,
+      "step": 5917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5951150490455013,
+      "learning_rate": 4.611962076484611e-07,
+      "loss": 0.7186,
+      "step": 5918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0111442088519453,
+      "learning_rate": 4.5970919115136406e-07,
+      "loss": 0.7697,
+      "step": 5919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5973671680062185,
+      "learning_rate": 4.582245193901802e-07,
+      "loss": 0.7933,
+      "step": 5920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.475572384849162,
+      "learning_rate": 4.567421927297999e-07,
+      "loss": 0.6779,
+      "step": 5921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.4497504442552986,
+      "learning_rate": 4.5526221153453845e-07,
+      "loss": 0.7951,
+      "step": 5922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2379051155154412,
+      "learning_rate": 4.5378457616813255e-07,
+      "loss": 0.8106,
+      "step": 5923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0168271076344393,
+      "learning_rate": 4.52309286993744e-07,
+      "loss": 0.8453,
+      "step": 5924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5571350055712943,
+      "learning_rate": 4.508363443739583e-07,
+      "loss": 0.7459,
+      "step": 5925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.422313233053986,
+      "learning_rate": 4.493657486707814e-07,
+      "loss": 0.8195,
+      "step": 5926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6614041261222745,
+      "learning_rate": 4.478975002456465e-07,
+      "loss": 0.8475,
+      "step": 5927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5459306588993655,
+      "learning_rate": 4.4643159945940816e-07,
+      "loss": 0.7793,
+      "step": 5928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.638010072274686,
+      "learning_rate": 4.449680466723416e-07,
+      "loss": 0.7412,
+      "step": 5929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6374186218730866,
+      "learning_rate": 4.435068422441491e-07,
+      "loss": 0.7863,
+      "step": 5930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0473504441514527,
+      "learning_rate": 4.4204798653395334e-07,
+      "loss": 0.724,
+      "step": 5931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.880335924375563,
+      "learning_rate": 4.405914799002997e-07,
+      "loss": 0.8654,
+      "step": 5932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5988830192409593,
+      "learning_rate": 4.391373227011564e-07,
+      "loss": 0.8395,
+      "step": 5933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8005912873580217,
+      "learning_rate": 4.376855152939152e-07,
+      "loss": 0.8489,
+      "step": 5934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5327179535187696,
+      "learning_rate": 4.362360580353875e-07,
+      "loss": 0.7765,
+      "step": 5935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.110555066156692,
+      "learning_rate": 4.347889512818115e-07,
+      "loss": 0.7865,
+      "step": 5936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.425327650857889,
+      "learning_rate": 4.33344195388844e-07,
+      "loss": 0.7973,
+      "step": 5937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.484576397445717,
+      "learning_rate": 4.319017907115686e-07,
+      "loss": 0.6719,
+      "step": 5938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7141696380110596,
+      "learning_rate": 4.3046173760448507e-07,
+      "loss": 0.7373,
+      "step": 5939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6212725554013523,
+      "learning_rate": 4.2902403642151704e-07,
+      "loss": 0.6919,
+      "step": 5940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.622753613256637,
+      "learning_rate": 4.27588687516014e-07,
+      "loss": 0.7129,
+      "step": 5941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7304264084363306,
+      "learning_rate": 4.2615569124074385e-07,
+      "loss": 0.7973,
+      "step": 5942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6718605017803423,
+      "learning_rate": 4.2472504794789593e-07,
+      "loss": 0.764,
+      "step": 5943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0322436728986784,
+      "learning_rate": 4.232967579890823e-07,
+      "loss": 0.8199,
+      "step": 5944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7072193959741755,
+      "learning_rate": 4.2187082171533665e-07,
+      "loss": 0.8041,
+      "step": 5945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8810435881483913,
+      "learning_rate": 4.204472394771142e-07,
+      "loss": 0.7923,
+      "step": 5946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8209460055158027,
+      "learning_rate": 4.190260116242917e-07,
+      "loss": 0.5896,
+      "step": 5947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.612444919294644,
+      "learning_rate": 4.176071385061664e-07,
+      "loss": 0.7208,
+      "step": 5948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5928435503810485,
+      "learning_rate": 4.1619062047145943e-07,
+      "loss": 0.735,
+      "step": 5949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7103087781610578,
+      "learning_rate": 4.1477645786831e-07,
+      "loss": 0.7629,
+      "step": 5950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.55167539583218,
+      "learning_rate": 4.1336465104427793e-07,
+      "loss": 0.6997,
+      "step": 5951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.608092717551468,
+      "learning_rate": 4.1195520034634896e-07,
+      "loss": 0.7727,
+      "step": 5952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.578863079954882,
+      "learning_rate": 4.105481061209249e-07,
+      "loss": 0.7717,
+      "step": 5953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6419007520560576,
+      "learning_rate": 4.091433687138291e-07,
+      "loss": 0.7056,
+      "step": 5954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.674006217875925,
+      "learning_rate": 4.0774098847030875e-07,
+      "loss": 0.8579,
+      "step": 5955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6653185857072184,
+      "learning_rate": 4.063409657350281e-07,
+      "loss": 0.7532,
+      "step": 5956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.695677170700721,
+      "learning_rate": 4.0494330085207314e-07,
+      "loss": 0.7022,
+      "step": 5957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7217815025546916,
+      "learning_rate": 4.0354799416495227e-07,
+      "loss": 0.7414,
+      "step": 5958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6230826008731913,
+      "learning_rate": 4.0215504601659017e-07,
+      "loss": 0.7257,
+      "step": 5959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.201265139540046,
+      "learning_rate": 4.007644567493374e-07,
+      "loss": 0.8023,
+      "step": 5960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.534269235714635,
+      "learning_rate": 3.993762267049606e-07,
+      "loss": 0.6844,
+      "step": 5961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6808601105724374,
+      "learning_rate": 3.9799035622464674e-07,
+      "loss": 0.6956,
+      "step": 5962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.581383954042853,
+      "learning_rate": 3.9660684564900574e-07,
+      "loss": 0.7696,
+      "step": 5963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5849069154868634,
+      "learning_rate": 3.9522569531806556e-07,
+      "loss": 0.7603,
+      "step": 5964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.768738229693881,
+      "learning_rate": 3.9384690557127125e-07,
+      "loss": 0.6432,
+      "step": 5965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8027221859258016,
+      "learning_rate": 3.9247047674749625e-07,
+      "loss": 0.7137,
+      "step": 5966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6574074923851816,
+      "learning_rate": 3.9109640918502333e-07,
+      "loss": 0.7667,
+      "step": 5967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7951563356356353,
+      "learning_rate": 3.897247032215601e-07,
+      "loss": 0.8096,
+      "step": 5968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.5183737997202704,
+      "learning_rate": 3.883553591942346e-07,
+      "loss": 0.812,
+      "step": 5969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8079909667473477,
+      "learning_rate": 3.8698837743959325e-07,
+      "loss": 0.7284,
+      "step": 5970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.68043983615785,
+      "learning_rate": 3.8562375829360286e-07,
+      "loss": 0.6751,
+      "step": 5971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7758583003898907,
+      "learning_rate": 3.8426150209164624e-07,
+      "loss": 0.7877,
+      "step": 5972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7937395372284466,
+      "learning_rate": 3.8290160916852894e-07,
+      "loss": 0.8075,
+      "step": 5973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.688516227702249,
+      "learning_rate": 3.8154407985847466e-07,
+      "loss": 0.7854,
+      "step": 5974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.628765653663377,
+      "learning_rate": 3.8018891449512654e-07,
+      "loss": 0.6539,
+      "step": 5975
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7655961176352757,
+      "learning_rate": 3.788361134115448e-07,
+      "loss": 0.7628,
+      "step": 5976
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6225969425848907,
+      "learning_rate": 3.774856769402113e-07,
+      "loss": 0.7296,
+      "step": 5977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.691373260121956,
+      "learning_rate": 3.7613760541302504e-07,
+      "loss": 0.7735,
+      "step": 5978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5287047223034285,
+      "learning_rate": 3.747918991613031e-07,
+      "loss": 0.7892,
+      "step": 5979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.56744426619113,
+      "learning_rate": 3.734485585157843e-07,
+      "loss": 0.7774,
+      "step": 5980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6214997737584773,
+      "learning_rate": 3.7210758380662125e-07,
+      "loss": 0.8006,
+      "step": 5981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8131875526793033,
+      "learning_rate": 3.707689753633914e-07,
+      "loss": 0.7807,
+      "step": 5982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8500936267310752,
+      "learning_rate": 3.6943273351508604e-07,
+      "loss": 0.7625,
+      "step": 5983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6465562873541058,
+      "learning_rate": 3.680988585901124e-07,
+      "loss": 0.7722,
+      "step": 5984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5574726673727084,
+      "learning_rate": 3.66767350916305e-07,
+      "loss": 0.8831,
+      "step": 5985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.782848100485891,
+      "learning_rate": 3.654382108209087e-07,
+      "loss": 0.73,
+      "step": 5986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6344063959839676,
+      "learning_rate": 3.6411143863058773e-07,
+      "loss": 0.7311,
+      "step": 5987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6342752111204453,
+      "learning_rate": 3.6278703467142684e-07,
+      "loss": 0.7964,
+      "step": 5988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6008339547301254,
+      "learning_rate": 3.6146499926892786e-07,
+      "loss": 0.7708,
+      "step": 5989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7388841845665794,
+      "learning_rate": 3.6014533274800867e-07,
+      "loss": 0.7886,
+      "step": 5990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4069664083059545,
+      "learning_rate": 3.588280354330065e-07,
+      "loss": 0.6621,
+      "step": 5991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8839315931631804,
+      "learning_rate": 3.5751310764767567e-07,
+      "loss": 0.7736,
+      "step": 5992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.733277195734112,
+      "learning_rate": 3.56200549715191e-07,
+      "loss": 0.8026,
+      "step": 5993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.540055857905672,
+      "learning_rate": 3.5489036195814007e-07,
+      "loss": 0.6752,
+      "step": 5994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8417129277346667,
+      "learning_rate": 3.535825446985297e-07,
+      "loss": 0.7915,
+      "step": 5995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5360667168506597,
+      "learning_rate": 3.522770982577872e-07,
+      "loss": 0.7955,
+      "step": 5996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6673994000706913,
+      "learning_rate": 3.5097402295675373e-07,
+      "loss": 0.7334,
+      "step": 5997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.472030136770235,
+      "learning_rate": 3.496733191156876e-07,
+      "loss": 0.7367,
+      "step": 5998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6281808987607693,
+      "learning_rate": 3.483749870542663e-07,
+      "loss": 0.811,
+      "step": 5999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.084336288946746,
+      "learning_rate": 3.470790270915836e-07,
+      "loss": 0.791,
+      "step": 6000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.69540151361276,
+      "learning_rate": 3.4578543954615017e-07,
+      "loss": 0.8372,
+      "step": 6001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6477593650807663,
+      "learning_rate": 3.444942247358918e-07,
+      "loss": 0.7617,
+      "step": 6002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6109865773332137,
+      "learning_rate": 3.4320538297815454e-07,
+      "loss": 0.8041,
+      "step": 6003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.564118447432275,
+      "learning_rate": 3.419189145896995e-07,
+      "loss": 0.6351,
+      "step": 6004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7347820160761946,
+      "learning_rate": 3.4063481988670375e-07,
+      "loss": 0.85,
+      "step": 6005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6586665554457767,
+      "learning_rate": 3.3935309918476155e-07,
+      "loss": 0.7468,
+      "step": 6006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6617987820098747,
+      "learning_rate": 3.3807375279888644e-07,
+      "loss": 0.7237,
+      "step": 6007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6600573016886737,
+      "learning_rate": 3.3679678104350353e-07,
+      "loss": 0.7518,
+      "step": 6008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5286339031019405,
+      "learning_rate": 3.355221842324552e-07,
+      "loss": 0.7932,
+      "step": 6009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.454421060746707,
+      "learning_rate": 3.342499626790052e-07,
+      "loss": 0.8156,
+      "step": 6010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7733675155642494,
+      "learning_rate": 3.329801166958291e-07,
+      "loss": 0.8301,
+      "step": 6011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1470563228210944,
+      "learning_rate": 3.317126465950171e-07,
+      "loss": 0.7118,
+      "step": 6012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7036527420458616,
+      "learning_rate": 3.3044755268808013e-07,
+      "loss": 0.7611,
+      "step": 6013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.697254921913109,
+      "learning_rate": 3.291848352859406e-07,
+      "loss": 0.7057,
+      "step": 6014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5999225248707623,
+      "learning_rate": 3.279244946989424e-07,
+      "loss": 0.7057,
+      "step": 6015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8870135590746515,
+      "learning_rate": 3.266665312368389e-07,
+      "loss": 0.6807,
+      "step": 6016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7602796195978123,
+      "learning_rate": 3.2541094520880166e-07,
+      "loss": 0.8041,
+      "step": 6017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4683812993866825,
+      "learning_rate": 3.241577369234228e-07,
+      "loss": 0.8085,
+      "step": 6018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4856380251395604,
+      "learning_rate": 3.229069066887014e-07,
+      "loss": 0.7617,
+      "step": 6019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4488210107592954,
+      "learning_rate": 3.2165845481205826e-07,
+      "loss": 0.7211,
+      "step": 6020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4634891340354304,
+      "learning_rate": 3.2041238160032793e-07,
+      "loss": 0.6962,
+      "step": 6021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7593153098002845,
+      "learning_rate": 3.191686873597599e-07,
+      "loss": 0.7758,
+      "step": 6022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.38830097389013,
+      "learning_rate": 3.1792737239601965e-07,
+      "loss": 0.6635,
+      "step": 6023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7072742284457427,
+      "learning_rate": 3.166884370141876e-07,
+      "loss": 0.7502,
+      "step": 6024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.690475824319762,
+      "learning_rate": 3.1545188151875795e-07,
+      "loss": 0.7552,
+      "step": 6025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.846891691572372,
+      "learning_rate": 3.142177062136431e-07,
+      "loss": 0.707,
+      "step": 6026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0359662518289827,
+      "learning_rate": 3.1298591140216827e-07,
+      "loss": 0.7993,
+      "step": 6027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.746866925967777,
+      "learning_rate": 3.117564973870735e-07,
+      "loss": 0.875,
+      "step": 6028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5842931925443122,
+      "learning_rate": 3.105294644705148e-07,
+      "loss": 0.6753,
+      "step": 6029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7795678238783204,
+      "learning_rate": 3.093048129540632e-07,
+      "loss": 0.8573,
+      "step": 6030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6587604849304656,
+      "learning_rate": 3.0808254313870133e-07,
+      "loss": 0.7591,
+      "step": 6031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6921282826190547,
+      "learning_rate": 3.068626553248311e-07,
+      "loss": 0.8027,
+      "step": 6032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.650607136409856,
+      "learning_rate": 3.0564514981226703e-07,
+      "loss": 0.7448,
+      "step": 6033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5959315484385255,
+      "learning_rate": 3.0443002690023537e-07,
+      "loss": 0.6463,
+      "step": 6034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4771076355337986,
+      "learning_rate": 3.032172868873817e-07,
+      "loss": 0.7658,
+      "step": 6035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.560831138005353,
+      "learning_rate": 3.0200693007176097e-07,
+      "loss": 0.7482,
+      "step": 6036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6691916248453276,
+      "learning_rate": 3.0079895675084734e-07,
+      "loss": 0.7337,
+      "step": 6037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4902436658315352,
+      "learning_rate": 2.995933672215257e-07,
+      "loss": 0.7613,
+      "step": 6038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4597485662812892,
+      "learning_rate": 2.983901617800955e-07,
+      "loss": 0.6547,
+      "step": 6039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.889902140838847,
+      "learning_rate": 2.971893407222737e-07,
+      "loss": 0.8037,
+      "step": 6040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8326105413501033,
+      "learning_rate": 2.9599090434318523e-07,
+      "loss": 0.7842,
+      "step": 6041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6873547730949623,
+      "learning_rate": 2.947948529373723e-07,
+      "loss": 0.791,
+      "step": 6042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.812710976846079,
+      "learning_rate": 2.9360118679879315e-07,
+      "loss": 0.7425,
+      "step": 6043
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.720941580967908,
+      "learning_rate": 2.9240990622081634e-07,
+      "loss": 0.7981,
+      "step": 6044
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1581410361795386,
+      "learning_rate": 2.912210114962244e-07,
+      "loss": 0.7823,
+      "step": 6045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.656038443141141,
+      "learning_rate": 2.900345029172158e-07,
+      "loss": 0.6951,
+      "step": 6046
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7266040236326963,
+      "learning_rate": 2.888503807753984e-07,
+      "loss": 0.7339,
+      "step": 6047
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7799392585020386,
+      "learning_rate": 2.8766864536179937e-07,
+      "loss": 0.7675,
+      "step": 6048
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.737208483644531,
+      "learning_rate": 2.8648929696685535e-07,
+      "loss": 0.8165,
+      "step": 6049
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.842487845149167,
+      "learning_rate": 2.853123358804144e-07,
+      "loss": 0.7047,
+      "step": 6050
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7143236985630113,
+      "learning_rate": 2.8413776239174404e-07,
+      "loss": 0.8682,
+      "step": 6051
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.556683398994706,
+      "learning_rate": 2.82965576789519e-07,
+      "loss": 0.6815,
+      "step": 6052
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9004282615196924,
+      "learning_rate": 2.81795779361832e-07,
+      "loss": 0.6865,
+      "step": 6053
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5260827904674867,
+      "learning_rate": 2.806283703961854e-07,
+      "loss": 0.8114,
+      "step": 6054
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.911421961080375,
+      "learning_rate": 2.794633501794952e-07,
+      "loss": 0.7076,
+      "step": 6055
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5387998313691216,
+      "learning_rate": 2.783007189980902e-07,
+      "loss": 0.7606,
+      "step": 6056
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.45474959202561,
+      "learning_rate": 2.77140477137714e-07,
+      "loss": 0.7467,
+      "step": 6057
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.966599236143366,
+      "learning_rate": 2.759826248835196e-07,
+      "loss": 0.7403,
+      "step": 6058
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4783423814613297,
+      "learning_rate": 2.74827162520076e-07,
+      "loss": 0.6726,
+      "step": 6059
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.638212372845662,
+      "learning_rate": 2.7367409033136395e-07,
+      "loss": 0.7919,
+      "step": 6060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.565673526069858,
+      "learning_rate": 2.725234086007744e-07,
+      "loss": 0.8529,
+      "step": 6061
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.495147047946822,
+      "learning_rate": 2.713751176111146e-07,
+      "loss": 0.6873,
+      "step": 6062
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6994013441494182,
+      "learning_rate": 2.7022921764459977e-07,
+      "loss": 0.6993,
+      "step": 6063
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8122848045588733,
+      "learning_rate": 2.6908570898286355e-07,
+      "loss": 0.7469,
+      "step": 6064
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.593524412271433,
+      "learning_rate": 2.679445919069457e-07,
+      "loss": 0.7646,
+      "step": 6065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6157359363362693,
+      "learning_rate": 2.668058666973017e-07,
+      "loss": 0.6645,
+      "step": 6066
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.685971547296702,
+      "learning_rate": 2.6566953363379777e-07,
+      "loss": 0.7256,
+      "step": 6067
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.505877258203053,
+      "learning_rate": 2.6453559299571276e-07,
+      "loss": 0.8303,
+      "step": 6068
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.510661056965399,
+      "learning_rate": 2.63404045061737e-07,
+      "loss": 0.7109,
+      "step": 6069
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.650980374114757,
+      "learning_rate": 2.622748901099759e-07,
+      "loss": 0.6944,
+      "step": 6070
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.705362907402246,
+      "learning_rate": 2.611481284179407e-07,
+      "loss": 0.7836,
+      "step": 6071
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7305925750666393,
+      "learning_rate": 2.6002376026255883e-07,
+      "loss": 0.7874,
+      "step": 6072
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6614969532801633,
+      "learning_rate": 2.5890178592016925e-07,
+      "loss": 0.8704,
+      "step": 6073
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4483769604343886,
+      "learning_rate": 2.5778220566652025e-07,
+      "loss": 0.686,
+      "step": 6074
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4934663469292087,
+      "learning_rate": 2.5666501977677614e-07,
+      "loss": 0.721,
+      "step": 6075
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.459218609307165,
+      "learning_rate": 2.5555022852550736e-07,
+      "loss": 0.6552,
+      "step": 6076
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8604168333650875,
+      "learning_rate": 2.5443783218669804e-07,
+      "loss": 0.6455,
+      "step": 6077
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5098450736455997,
+      "learning_rate": 2.5332783103374725e-07,
+      "loss": 0.6351,
+      "step": 6078
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.841494576735548,
+      "learning_rate": 2.52220225339459e-07,
+      "loss": 0.7896,
+      "step": 6079
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.692905809508813,
+      "learning_rate": 2.511150153760522e-07,
+      "loss": 0.7215,
+      "step": 6080
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.727313980169833,
+      "learning_rate": 2.5001220141515736e-07,
+      "loss": 0.7061,
+      "step": 6081
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5469765476370787,
+      "learning_rate": 2.489117837278143e-07,
+      "loss": 0.8545,
+      "step": 6082
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.559331431246218,
+      "learning_rate": 2.4781376258447564e-07,
+      "loss": 0.754,
+      "step": 6083
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.722703605969057,
+      "learning_rate": 2.4671813825500324e-07,
+      "loss": 0.6871,
+      "step": 6084
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.596003743401093,
+      "learning_rate": 2.456249110086717e-07,
+      "loss": 0.6845,
+      "step": 6085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.083109926834688,
+      "learning_rate": 2.4453408111416497e-07,
+      "loss": 0.7946,
+      "step": 6086
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7663008724269544,
+      "learning_rate": 2.4344564883957976e-07,
+      "loss": 0.7694,
+      "step": 6087
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8746332918626205,
+      "learning_rate": 2.4235961445241987e-07,
+      "loss": 0.7514,
+      "step": 6088
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6013942389612517,
+      "learning_rate": 2.41275978219605e-07,
+      "loss": 0.6923,
+      "step": 6089
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.912273621633104,
+      "learning_rate": 2.4019474040746004e-07,
+      "loss": 0.8051,
+      "step": 6090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7464518435612963,
+      "learning_rate": 2.391159012817246e-07,
+      "loss": 0.7467,
+      "step": 6091
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7813044398351425,
+      "learning_rate": 2.3803946110754649e-07,
+      "loss": 0.7753,
+      "step": 6092
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.518752845522808,
+      "learning_rate": 2.369654201494842e-07,
+      "loss": 0.7387,
+      "step": 6093
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.725181935975217,
+      "learning_rate": 2.3589377867150543e-07,
+      "loss": 0.8623,
+      "step": 6094
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1544796740668186,
+      "learning_rate": 2.3482453693699282e-07,
+      "loss": 0.7932,
+      "step": 6095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4788389266335646,
+      "learning_rate": 2.3375769520873393e-07,
+      "loss": 0.7703,
+      "step": 6096
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.842756149992874,
+      "learning_rate": 2.3269325374892903e-07,
+      "loss": 0.6308,
+      "step": 6097
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5921408232583545,
+      "learning_rate": 2.3163121281918888e-07,
+      "loss": 0.7522,
+      "step": 6098
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.633161071717052,
+      "learning_rate": 2.3057157268053133e-07,
+      "loss": 0.6943,
+      "step": 6099
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.552057063916039,
+      "learning_rate": 2.2951433359338805e-07,
+      "loss": 0.7337,
+      "step": 6100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.520467762762636,
+      "learning_rate": 2.2845949581759785e-07,
+      "loss": 0.7381,
+      "step": 6101
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.5453403536853623,
+      "learning_rate": 2.2740705961241006e-07,
+      "loss": 0.7681,
+      "step": 6102
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0029521200394957,
+      "learning_rate": 2.2635702523648552e-07,
+      "loss": 0.6868,
+      "step": 6103
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8302795535117573,
+      "learning_rate": 2.253093929478911e-07,
+      "loss": 0.7434,
+      "step": 6104
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4786546039372634,
+      "learning_rate": 2.2426416300410537e-07,
+      "loss": 0.7557,
+      "step": 6105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.48657198357177,
+      "learning_rate": 2.2322133566201941e-07,
+      "loss": 0.7145,
+      "step": 6106
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6497695414229825,
+      "learning_rate": 2.2218091117792718e-07,
+      "loss": 0.668,
+      "step": 6107
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6722487924038067,
+      "learning_rate": 2.2114288980753962e-07,
+      "loss": 0.7089,
+      "step": 6108
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.714821583484799,
+      "learning_rate": 2.2010727180597157e-07,
+      "loss": 0.7621,
+      "step": 6109
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8942861918137464,
+      "learning_rate": 2.1907405742774723e-07,
+      "loss": 0.7278,
+      "step": 6110
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.280632946907079,
+      "learning_rate": 2.1804324692680345e-07,
+      "loss": 0.7916,
+      "step": 6111
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5931959273453886,
+      "learning_rate": 2.1701484055648536e-07,
+      "loss": 0.7815,
+      "step": 6112
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7709121512981065,
+      "learning_rate": 2.1598883856954523e-07,
+      "loss": 0.7502,
+      "step": 6113
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9441111783976277,
+      "learning_rate": 2.1496524121814576e-07,
+      "loss": 0.8539,
+      "step": 6114
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6898849255122323,
+      "learning_rate": 2.13944048753858e-07,
+      "loss": 0.7489,
+      "step": 6115
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9925758163754987,
+      "learning_rate": 2.1292526142766223e-07,
+      "loss": 0.7994,
+      "step": 6116
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6208054097627214,
+      "learning_rate": 2.1190887948994822e-07,
+      "loss": 0.6947,
+      "step": 6117
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8860097331924224,
+      "learning_rate": 2.1089490319051497e-07,
+      "loss": 0.7517,
+      "step": 6118
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.49078276280568,
+      "learning_rate": 2.0988333277856877e-07,
+      "loss": 0.6892,
+      "step": 6119
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7127390041528097,
+      "learning_rate": 2.0887416850272514e-07,
+      "loss": 0.8339,
+      "step": 6120
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7664440908223717,
+      "learning_rate": 2.078674106110079e-07,
+      "loss": 0.788,
+      "step": 6121
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7691557943619647,
+      "learning_rate": 2.068630593508514e-07,
+      "loss": 0.8135,
+      "step": 6122
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.519939135167319,
+      "learning_rate": 2.05861114969097e-07,
+      "loss": 0.7414,
+      "step": 6123
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.875431563743744,
+      "learning_rate": 2.0486157771199223e-07,
+      "loss": 0.7245,
+      "step": 6124
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6094402261215643,
+      "learning_rate": 2.0386444782519722e-07,
+      "loss": 0.7991,
+      "step": 6125
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.859394466977294,
+      "learning_rate": 2.0286972555377704e-07,
+      "loss": 0.748,
+      "step": 6126
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9414054266730982,
+      "learning_rate": 2.0187741114220837e-07,
+      "loss": 0.7258,
+      "step": 6127
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6658460640061454,
+      "learning_rate": 2.0088750483437393e-07,
+      "loss": 0.7896,
+      "step": 6128
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.478030327430569,
+      "learning_rate": 1.9990000687356348e-07,
+      "loss": 0.7599,
+      "step": 6129
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.772471185791872,
+      "learning_rate": 1.9891491750247738e-07,
+      "loss": 0.8028,
+      "step": 6130
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6702866352287553,
+      "learning_rate": 1.9793223696322306e-07,
+      "loss": 0.6865,
+      "step": 6131
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.640238395629171,
+      "learning_rate": 1.969519654973151e-07,
+      "loss": 0.6392,
+      "step": 6132
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5352033004938774,
+      "learning_rate": 1.9597410334567746e-07,
+      "loss": 0.7221,
+      "step": 6133
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7111068335551245,
+      "learning_rate": 1.9499865074864122e-07,
+      "loss": 0.7778,
+      "step": 6134
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7186845902814403,
+      "learning_rate": 1.9402560794594572e-07,
+      "loss": 0.7479,
+      "step": 6135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6369963760573523,
+      "learning_rate": 1.9305497517673633e-07,
+      "loss": 0.7727,
+      "step": 6136
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.550199010640346,
+      "learning_rate": 1.9208675267956666e-07,
+      "loss": 0.736,
+      "step": 6137
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9158190657987237,
+      "learning_rate": 1.9112094069240084e-07,
+      "loss": 0.7635,
+      "step": 6138
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.894596711016241,
+      "learning_rate": 1.9015753945260784e-07,
+      "loss": 0.8388,
+      "step": 6139
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6370639749811517,
+      "learning_rate": 1.8919654919696383e-07,
+      "loss": 0.7775,
+      "step": 6140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9681643766004524,
+      "learning_rate": 1.8823797016165435e-07,
+      "loss": 0.7151,
+      "step": 6141
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7649931358599034,
+      "learning_rate": 1.872818025822709e-07,
+      "loss": 0.7196,
+      "step": 6142
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6534712725876854,
+      "learning_rate": 1.863280466938111e-07,
+      "loss": 0.7542,
+      "step": 6143
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.868580015988256,
+      "learning_rate": 1.8537670273068298e-07,
+      "loss": 0.7451,
+      "step": 6144
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6025947830881266,
+      "learning_rate": 1.844277709267006e-07,
+      "loss": 0.7171,
+      "step": 6145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0658603765662704,
+      "learning_rate": 1.834812515150841e-07,
+      "loss": 0.7431,
+      "step": 6146
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.603293519933533,
+      "learning_rate": 1.8253714472846184e-07,
+      "loss": 0.7527,
+      "step": 6147
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.71605864580305,
+      "learning_rate": 1.81595450798866e-07,
+      "loss": 0.7696,
+      "step": 6148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.537355350931297,
+      "learning_rate": 1.806561699577436e-07,
+      "loss": 0.6957,
+      "step": 6149
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0873686115208168,
+      "learning_rate": 1.7971930243593893e-07,
+      "loss": 0.7041,
+      "step": 6150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8757638527449303,
+      "learning_rate": 1.7878484846371001e-07,
+      "loss": 0.7434,
+      "step": 6151
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6085872075147596,
+      "learning_rate": 1.7785280827071982e-07,
+      "loss": 0.6898,
+      "step": 6152
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5746517753489577,
+      "learning_rate": 1.769231820860362e-07,
+      "loss": 0.74,
+      "step": 6153
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.667581760238676,
+      "learning_rate": 1.759959701381353e-07,
+      "loss": 0.6863,
+      "step": 6154
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.799957147402608,
+      "learning_rate": 1.7507117265490148e-07,
+      "loss": 0.6544,
+      "step": 6155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6982037422852576,
+      "learning_rate": 1.7414878986362294e-07,
+      "loss": 0.6762,
+      "step": 6156
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.976137628722369,
+      "learning_rate": 1.732288219909961e-07,
+      "loss": 0.8104,
+      "step": 6157
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.559301938412268,
+      "learning_rate": 1.7231126926312235e-07,
+      "loss": 0.6893,
+      "step": 6158
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9686201652518025,
+      "learning_rate": 1.7139613190550906e-07,
+      "loss": 0.7745,
+      "step": 6159
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.741683295872832,
+      "learning_rate": 1.7048341014307523e-07,
+      "loss": 0.6785,
+      "step": 6160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8906234050957478,
+      "learning_rate": 1.6957310420013918e-07,
+      "loss": 0.734,
+      "step": 6161
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6247075942288927,
+      "learning_rate": 1.686652143004286e-07,
+      "loss": 0.6899,
+      "step": 6162
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.558633377476938,
+      "learning_rate": 1.6775974066707833e-07,
+      "loss": 0.7669,
+      "step": 6163
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6371825253937002,
+      "learning_rate": 1.668566835226293e-07,
+      "loss": 0.6652,
+      "step": 6164
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7199806217099205,
+      "learning_rate": 1.6595604308902613e-07,
+      "loss": 0.8627,
+      "step": 6165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6341507876042325,
+      "learning_rate": 1.650578195876218e-07,
+      "loss": 0.7276,
+      "step": 6166
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1415513717540815,
+      "learning_rate": 1.6416201323917413e-07,
+      "loss": 0.7839,
+      "step": 6167
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4806645277468076,
+      "learning_rate": 1.6326862426384705e-07,
+      "loss": 0.7079,
+      "step": 6168
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.567681957131878,
+      "learning_rate": 1.6237765288121044e-07,
+      "loss": 0.7302,
+      "step": 6169
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.749034703344554,
+      "learning_rate": 1.6148909931024026e-07,
+      "loss": 0.7658,
+      "step": 6170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7358323500883133,
+      "learning_rate": 1.606029637693185e-07,
+      "loss": 0.7315,
+      "step": 6171
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.783968035090616,
+      "learning_rate": 1.5971924647623204e-07,
+      "loss": 0.8276,
+      "step": 6172
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8293213016993257,
+      "learning_rate": 1.5883794764817272e-07,
+      "loss": 0.7387,
+      "step": 6173
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8949693147568425,
+      "learning_rate": 1.5795906750174062e-07,
+      "loss": 0.814,
+      "step": 6174
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2324609639018966,
+      "learning_rate": 1.570826062529407e-07,
+      "loss": 0.7689,
+      "step": 6175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5822059542555635,
+      "learning_rate": 1.5620856411717954e-07,
+      "loss": 0.6804,
+      "step": 6176
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.777882167249254,
+      "learning_rate": 1.5533694130927424e-07,
+      "loss": 0.7848,
+      "step": 6177
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1167534290212546,
+      "learning_rate": 1.544677380434445e-07,
+      "loss": 0.6931,
+      "step": 6178
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8203501178896424,
+      "learning_rate": 1.5360095453331724e-07,
+      "loss": 0.6658,
+      "step": 6179
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5251340746824225,
+      "learning_rate": 1.5273659099192317e-07,
+      "loss": 0.6895,
+      "step": 6180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.599439628395259,
+      "learning_rate": 1.5187464763169678e-07,
+      "loss": 0.8254,
+      "step": 6181
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.804159345742806,
+      "learning_rate": 1.5101512466448197e-07,
+      "loss": 0.7402,
+      "step": 6182
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6898781919312302,
+      "learning_rate": 1.501580223015242e-07,
+      "loss": 0.7552,
+      "step": 6183
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7033042374066762,
+      "learning_rate": 1.4930334075347497e-07,
+      "loss": 0.7328,
+      "step": 6184
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.473450686295006,
+      "learning_rate": 1.4845108023039178e-07,
+      "loss": 0.8645,
+      "step": 6185
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.556613318569326,
+      "learning_rate": 1.47601240941736e-07,
+      "loss": 0.7423,
+      "step": 6186
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6293302467512274,
+      "learning_rate": 1.4675382309637386e-07,
+      "loss": 0.6899,
+      "step": 6187
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9116476239785314,
+      "learning_rate": 1.4590882690257768e-07,
+      "loss": 0.7177,
+      "step": 6188
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6550641452752664,
+      "learning_rate": 1.4506625256802355e-07,
+      "loss": 0.6495,
+      "step": 6189
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6021279534255637,
+      "learning_rate": 1.4422610029979244e-07,
+      "loss": 0.7871,
+      "step": 6190
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.76386657740609,
+      "learning_rate": 1.4338837030437147e-07,
+      "loss": 0.7694,
+      "step": 6191
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7179569750741943,
+      "learning_rate": 1.4255306278764813e-07,
+      "loss": 0.7451,
+      "step": 6192
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7081378794105344,
+      "learning_rate": 1.4172017795492044e-07,
+      "loss": 0.8077,
+      "step": 6193
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.263696185374978,
+      "learning_rate": 1.4088971601088796e-07,
+      "loss": 0.8088,
+      "step": 6194
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7987236616266005,
+      "learning_rate": 1.400616771596519e-07,
+      "loss": 0.7513,
+      "step": 6195
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7879734412104082,
+      "learning_rate": 1.3923606160472504e-07,
+      "loss": 0.7186,
+      "step": 6196
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.675492212746545,
+      "learning_rate": 1.3841286954901834e-07,
+      "loss": 0.7624,
+      "step": 6197
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5732871399578583,
+      "learning_rate": 1.3759210119485e-07,
+      "loss": 0.7256,
+      "step": 6198
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.719745063329388,
+      "learning_rate": 1.3677375674394088e-07,
+      "loss": 0.8057,
+      "step": 6199
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8279350287112672,
+      "learning_rate": 1.359578363974179e-07,
+      "loss": 0.7716,
+      "step": 6200
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6119360982876247,
+      "learning_rate": 1.351443403558117e-07,
+      "loss": 0.7593,
+      "step": 6201
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.548879171886614,
+      "learning_rate": 1.3433326881905683e-07,
+      "loss": 0.6883,
+      "step": 6202
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.768212090009197,
+      "learning_rate": 1.3352462198649163e-07,
+      "loss": 0.8027,
+      "step": 6203
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6002595913786246,
+      "learning_rate": 1.327184000568582e-07,
+      "loss": 0.7569,
+      "step": 6204
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8178858099550923,
+      "learning_rate": 1.3191460322830364e-07,
+      "loss": 0.7394,
+      "step": 6205
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.758556782006995,
+      "learning_rate": 1.3111323169837875e-07,
+      "loss": 0.855,
+      "step": 6206
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8694004245810385,
+      "learning_rate": 1.3031428566403824e-07,
+      "loss": 0.7254,
+      "step": 6207
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.875488729219647,
+      "learning_rate": 1.2951776532164062e-07,
+      "loss": 0.7762,
+      "step": 6208
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.542501426417321,
+      "learning_rate": 1.2872367086694704e-07,
+      "loss": 0.6791,
+      "step": 6209
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.637201824519699,
+      "learning_rate": 1.2793200249512693e-07,
+      "loss": 0.7678,
+      "step": 6210
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6419157596436613,
+      "learning_rate": 1.271427604007458e-07,
+      "loss": 0.7553,
+      "step": 6211
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.639956132902984,
+      "learning_rate": 1.263559447777818e-07,
+      "loss": 0.7385,
+      "step": 6212
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.936008032052921,
+      "learning_rate": 1.2557155581960913e-07,
+      "loss": 0.745,
+      "step": 6213
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.565648666042382,
+      "learning_rate": 1.247895937190091e-07,
+      "loss": 0.7251,
+      "step": 6214
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6059344542245975,
+      "learning_rate": 1.2401005866816586e-07,
+      "loss": 0.7442,
+      "step": 6215
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5199423385190025,
+      "learning_rate": 1.2323295085866827e-07,
+      "loss": 0.7081,
+      "step": 6216
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.766694260160205,
+      "learning_rate": 1.2245827048150694e-07,
+      "loss": 0.8202,
+      "step": 6217
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7075914640665246,
+      "learning_rate": 1.2168601772707622e-07,
+      "loss": 0.7708,
+      "step": 6218
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.623914655169994,
+      "learning_rate": 1.209161927851754e-07,
+      "loss": 0.836,
+      "step": 6219
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5301146209908945,
+      "learning_rate": 1.2014879584500427e-07,
+      "loss": 0.6935,
+      "step": 6220
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.884730413869238,
+      "learning_rate": 1.193838270951686e-07,
+      "loss": 0.7502,
+      "step": 6221
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.862630433813254,
+      "learning_rate": 1.1862128672367579e-07,
+      "loss": 0.7959,
+      "step": 6222
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7660713652174667,
+      "learning_rate": 1.1786117491793702e-07,
+      "loss": 0.7949,
+      "step": 6223
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8777554137925074,
+      "learning_rate": 1.1710349186476732e-07,
+      "loss": 0.7505,
+      "step": 6224
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6031163077151893,
+      "learning_rate": 1.1634823775038218e-07,
+      "loss": 0.7089,
+      "step": 6225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2398709920206232,
+      "learning_rate": 1.155954127604042e-07,
+      "loss": 0.7993,
+      "step": 6226
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7285117964587773,
+      "learning_rate": 1.148450170798543e-07,
+      "loss": 0.7544,
+      "step": 6227
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0268938632735596,
+      "learning_rate": 1.1409705089315826e-07,
+      "loss": 0.6847,
+      "step": 6228
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6551913985471,
+      "learning_rate": 1.1335151438414905e-07,
+      "loss": 0.7259,
+      "step": 6229
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9149685796855356,
+      "learning_rate": 1.1260840773605452e-07,
+      "loss": 0.7453,
+      "step": 6230
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.519006685781597,
+      "learning_rate": 1.1186773113151083e-07,
+      "loss": 0.7421,
+      "step": 6231
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5509960965895147,
+      "learning_rate": 1.1112948475255681e-07,
+      "loss": 0.7092,
+      "step": 6232
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5465021594958617,
+      "learning_rate": 1.1039366878063062e-07,
+      "loss": 0.7868,
+      "step": 6233
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.7905289925773573,
+      "learning_rate": 1.0966028339657763e-07,
+      "loss": 0.7965,
+      "step": 6234
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.847803802086891,
+      "learning_rate": 1.0892932878064144e-07,
+      "loss": 0.8927,
+      "step": 6235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8952086928132865,
+      "learning_rate": 1.0820080511247055e-07,
+      "loss": 0.888,
+      "step": 6236
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.726833808841528,
+      "learning_rate": 1.0747471257111619e-07,
+      "loss": 0.756,
+      "step": 6237
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6147257812951286,
+      "learning_rate": 1.0675105133503116e-07,
+      "loss": 0.6977,
+      "step": 6238
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5060729867642375,
+      "learning_rate": 1.0602982158206987e-07,
+      "loss": 0.825,
+      "step": 6239
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.403041630415913,
+      "learning_rate": 1.0531102348949273e-07,
+      "loss": 0.6932,
+      "step": 6240
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5488272121277284,
+      "learning_rate": 1.0459465723395956e-07,
+      "loss": 0.7599,
+      "step": 6241
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6571821875824044,
+      "learning_rate": 1.0388072299153174e-07,
+      "loss": 0.7615,
+      "step": 6242
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.871469208438707,
+      "learning_rate": 1.0316922093767556e-07,
+      "loss": 0.7261,
+      "step": 6243
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6783501614843437,
+      "learning_rate": 1.0246015124725672e-07,
+      "loss": 0.7352,
+      "step": 6244
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.862113879224904,
+      "learning_rate": 1.0175351409454693e-07,
+      "loss": 0.7957,
+      "step": 6245
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5378602149474845,
+      "learning_rate": 1.0104930965321724e-07,
+      "loss": 0.7019,
+      "step": 6246
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9359078320937506,
+      "learning_rate": 1.0034753809634035e-07,
+      "loss": 0.6046,
+      "step": 6247
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2423297564849984,
+      "learning_rate": 9.96481995963916e-08,
+      "loss": 0.738,
+      "step": 6248
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.491630187666835,
+      "learning_rate": 9.89512943252513e-08,
+      "loss": 0.71,
+      "step": 6249
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6773863279396015,
+      "learning_rate": 9.825682245419576e-08,
+      "loss": 0.7626,
+      "step": 6250
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7660942543458504,
+      "learning_rate": 9.756478415390847e-08,
+      "loss": 0.7542,
+      "step": 6251
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.897450522419471,
+      "learning_rate": 9.687517959447446e-08,
+      "loss": 0.7878,
+      "step": 6252
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.499684952206905,
+      "learning_rate": 9.618800894537594e-08,
+      "loss": 0.7011,
+      "step": 6253
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.691932857931784,
+      "learning_rate": 9.550327237550339e-08,
+      "loss": 0.792,
+      "step": 6254
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4931560935692882,
+      "learning_rate": 9.482097005314328e-08,
+      "loss": 0.8742,
+      "step": 6255
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.820129725912161,
+      "learning_rate": 9.414110214598815e-08,
+      "loss": 0.6588,
+      "step": 6256
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7829327595354405,
+      "learning_rate": 9.3463668821131e-08,
+      "loss": 0.732,
+      "step": 6257
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8675835655307327,
+      "learning_rate": 9.278867024506421e-08,
+      "loss": 0.7629,
+      "step": 6258
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.736723949619082,
+      "learning_rate": 9.211610658368397e-08,
+      "loss": 0.7268,
+      "step": 6259
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5946668669950257,
+      "learning_rate": 9.144597800228693e-08,
+      "loss": 0.7606,
+      "step": 6260
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7121437423369064,
+      "learning_rate": 9.077828466557359e-08,
+      "loss": 0.8796,
+      "step": 6261
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.449222452836164,
+      "learning_rate": 9.011302673764266e-08,
+      "loss": 0.7207,
+      "step": 6262
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5211066598013563,
+      "learning_rate": 8.94502043819967e-08,
+      "loss": 0.7476,
+      "step": 6263
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7837470093262406,
+      "learning_rate": 8.878981776153872e-08,
+      "loss": 0.7562,
+      "step": 6264
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6387684488289946,
+      "learning_rate": 8.813186703857334e-08,
+      "loss": 0.73,
+      "step": 6265
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.514520139552492,
+      "learning_rate": 8.747635237480679e-08,
+      "loss": 0.7326,
+      "step": 6266
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.642189124321444,
+      "learning_rate": 8.68232739313446e-08,
+      "loss": 0.7199,
+      "step": 6267
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.658159002696627,
+      "learning_rate": 8.617263186869728e-08,
+      "loss": 0.8156,
+      "step": 6268
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5569705911088567,
+      "learning_rate": 8.552442634677361e-08,
+      "loss": 0.7859,
+      "step": 6269
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.507594117809746,
+      "learning_rate": 8.487865752488277e-08,
+      "loss": 0.6458,
+      "step": 6270
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7535773207528407,
+      "learning_rate": 8.423532556173896e-08,
+      "loss": 0.8386,
+      "step": 6271
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4238202953270656,
+      "learning_rate": 8.359443061545348e-08,
+      "loss": 0.6733,
+      "step": 6272
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6580601292273744,
+      "learning_rate": 8.29559728435414e-08,
+      "loss": 0.7644,
+      "step": 6273
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5288605289633383,
+      "learning_rate": 8.231995240291613e-08,
+      "loss": 0.7586,
+      "step": 6274
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7429246054717367,
+      "learning_rate": 8.168636944989595e-08,
+      "loss": 0.8357,
+      "step": 6275
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.836135023842705,
+      "learning_rate": 8.105522414019629e-08,
+      "loss": 0.7551,
+      "step": 6276
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.601433221659957,
+      "learning_rate": 8.04265166289353e-08,
+      "loss": 0.7847,
+      "step": 6277
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8694538043296087,
+      "learning_rate": 7.980024707063161e-08,
+      "loss": 0.7775,
+      "step": 6278
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6236106508897428,
+      "learning_rate": 7.91764156192043e-08,
+      "loss": 0.7561,
+      "step": 6279
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7156377840729413,
+      "learning_rate": 7.85550224279752e-08,
+      "loss": 0.7812,
+      "step": 6280
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8693681860985936,
+      "learning_rate": 7.793606764966321e-08,
+      "loss": 0.7878,
+      "step": 6281
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.859687806470752,
+      "learning_rate": 7.731955143639225e-08,
+      "loss": 0.7898,
+      "step": 6282
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.753093997291165,
+      "learning_rate": 7.67054739396822e-08,
+      "loss": 0.7418,
+      "step": 6283
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.697051038845374,
+      "learning_rate": 7.609383531045788e-08,
+      "loss": 0.7606,
+      "step": 6284
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6081068715682965,
+      "learning_rate": 7.548463569904241e-08,
+      "loss": 0.7459,
+      "step": 6285
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.493989378746843,
+      "learning_rate": 7.487787525516044e-08,
+      "loss": 0.8218,
+      "step": 6286
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1768599720968322,
+      "learning_rate": 7.427355412793602e-08,
+      "loss": 0.7762,
+      "step": 6287
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8093572987998923,
+      "learning_rate": 7.367167246589479e-08,
+      "loss": 0.7584,
+      "step": 6288
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5319574525103574,
+      "learning_rate": 7.307223041696177e-08,
+      "loss": 0.7272,
+      "step": 6289
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7085439960965894,
+      "learning_rate": 7.247522812846353e-08,
+      "loss": 0.7297,
+      "step": 6290
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.656018882388865,
+      "learning_rate": 7.188066574712604e-08,
+      "loss": 0.743,
+      "step": 6291
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.854981479330724,
+      "learning_rate": 7.128854341907688e-08,
+      "loss": 0.8181,
+      "step": 6292
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7236793623907425,
+      "learning_rate": 7.069886128984294e-08,
+      "loss": 0.7819,
+      "step": 6293
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.592995981693292,
+      "learning_rate": 7.011161950435053e-08,
+      "loss": 0.7192,
+      "step": 6294
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.73348258434267,
+      "learning_rate": 6.952681820692753e-08,
+      "loss": 0.7506,
+      "step": 6295
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.794527286958416,
+      "learning_rate": 6.89444575413023e-08,
+      "loss": 0.7659,
+      "step": 6296
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6722789816838124,
+      "learning_rate": 6.836453765060258e-08,
+      "loss": 0.7245,
+      "step": 6297
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.6796626725806423,
+      "learning_rate": 6.778705867735657e-08,
+      "loss": 0.7804,
+      "step": 6298
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.795808761513518,
+      "learning_rate": 6.721202076349187e-08,
+      "loss": 0.7924,
+      "step": 6299
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.578487552690196,
+      "learning_rate": 6.663942405033763e-08,
+      "loss": 0.7203,
+      "step": 6300
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.912508034176297,
+      "learning_rate": 6.606926867862129e-08,
+      "loss": 0.62,
+      "step": 6301
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5523100748647063,
+      "learning_rate": 6.550155478847075e-08,
+      "loss": 0.7968,
+      "step": 6302
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8976628589496327,
+      "learning_rate": 6.493628251941442e-08,
+      "loss": 0.7717,
+      "step": 6303
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4732197073116837,
+      "learning_rate": 6.437345201038115e-08,
+      "loss": 0.7558,
+      "step": 6304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6265852771064067,
+      "learning_rate": 6.381306339969806e-08,
+      "loss": 0.6971,
+      "step": 6305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.573298694817957,
+      "learning_rate": 6.325511682509277e-08,
+      "loss": 0.7839,
+      "step": 6306
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0305915808088173,
+      "learning_rate": 6.269961242369338e-08,
+      "loss": 0.8317,
+      "step": 6307
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.924612836068098,
+      "learning_rate": 6.214655033202732e-08,
+      "loss": 0.7161,
+      "step": 6308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6344706369779183,
+      "learning_rate": 6.159593068602255e-08,
+      "loss": 0.6954,
+      "step": 6309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4740253678071262,
+      "learning_rate": 6.104775362100301e-08,
+      "loss": 0.6835,
+      "step": 6310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.807355565461176,
+      "learning_rate": 6.050201927169875e-08,
+      "loss": 0.699,
+      "step": 6311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.689978770639496,
+      "learning_rate": 5.995872777223466e-08,
+      "loss": 0.8269,
+      "step": 6312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.728285657498286,
+      "learning_rate": 5.941787925613507e-08,
+      "loss": 0.748,
+      "step": 6313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6673535571875577,
+      "learning_rate": 5.8879473856328084e-08,
+      "loss": 0.8318,
+      "step": 6314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.858044430958844,
+      "learning_rate": 5.8343511705136746e-08,
+      "loss": 0.755,
+      "step": 6315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6411821646605738,
+      "learning_rate": 5.780999293428569e-08,
+      "loss": 0.7643,
+      "step": 6316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8070989561640527,
+      "learning_rate": 5.727891767490001e-08,
+      "loss": 0.8535,
+      "step": 6317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5051496078033573,
+      "learning_rate": 5.675028605750199e-08,
+      "loss": 0.6572,
+      "step": 6318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.635854501865161,
+      "learning_rate": 5.622409821201436e-08,
+      "loss": 0.7762,
+      "step": 6319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9390171766065847,
+      "learning_rate": 5.5700354267760326e-08,
+      "loss": 0.7604,
+      "step": 6320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9946880040924264,
+      "learning_rate": 5.517905435345916e-08,
+      "loss": 0.7079,
+      "step": 6321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.561259049882208,
+      "learning_rate": 5.4660198597235035e-08,
+      "loss": 0.795,
+      "step": 6322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6513264646111168,
+      "learning_rate": 5.414378712660706e-08,
+      "loss": 0.6741,
+      "step": 6323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.741628132065017,
+      "learning_rate": 5.36298200684926e-08,
+      "loss": 0.8026,
+      "step": 6324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7135972268576647,
+      "learning_rate": 5.3118297549212826e-08,
+      "loss": 0.7927,
+      "step": 6325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.848539842298522,
+      "learning_rate": 5.260921969448496e-08,
+      "loss": 0.8676,
+      "step": 6326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.789041773355471,
+      "learning_rate": 5.210258662942669e-08,
+      "loss": 0.815,
+      "step": 6327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5460210484795236,
+      "learning_rate": 5.159839847855175e-08,
+      "loss": 0.7364,
+      "step": 6328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7098943104095854,
+      "learning_rate": 5.109665536577768e-08,
+      "loss": 0.6325,
+      "step": 6329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.855953216059735,
+      "learning_rate": 5.059735741441807e-08,
+      "loss": 0.6845,
+      "step": 6330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7627575881511905,
+      "learning_rate": 5.0100504747186974e-08,
+      "loss": 0.751,
+      "step": 6331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6643373656366363,
+      "learning_rate": 4.9606097486195604e-08,
+      "loss": 0.7861,
+      "step": 6332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.840794267648225,
+      "learning_rate": 4.911413575295787e-08,
+      "loss": 0.7249,
+      "step": 6333
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.635273402121381,
+      "learning_rate": 4.8624619668381504e-08,
+      "loss": 0.6318,
+      "step": 6334
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7977870462313095,
+      "learning_rate": 4.813754935277581e-08,
+      "loss": 0.8487,
+      "step": 6335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6463144854142757,
+      "learning_rate": 4.765292492585172e-08,
+      "loss": 0.6637,
+      "step": 6336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.065956639630013,
+      "learning_rate": 4.717074650671394e-08,
+      "loss": 0.8322,
+      "step": 6337
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7850962673987762,
+      "learning_rate": 4.6691014213868794e-08,
+      "loss": 0.7225,
+      "step": 6338
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9330373089277932,
+      "learning_rate": 4.621372816522196e-08,
+      "loss": 0.8766,
+      "step": 6339
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.653237142876649,
+      "learning_rate": 4.573888847807517e-08,
+      "loss": 0.7426,
+      "step": 6340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4358443741832705,
+      "learning_rate": 4.5266495269132846e-08,
+      "loss": 0.7164,
+      "step": 6341
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.392735653487316,
+      "learning_rate": 4.479654865449545e-08,
+      "loss": 0.6767,
+      "step": 6342
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.972324375045306,
+      "learning_rate": 4.432904874966171e-08,
+      "loss": 0.8116,
+      "step": 6343
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.380025455047499,
+      "learning_rate": 4.386399566953081e-08,
+      "loss": 0.6901,
+      "step": 6344
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.771678928999211,
+      "learning_rate": 4.340138952839912e-08,
+      "loss": 0.7367,
+      "step": 6345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.557862003826193,
+      "learning_rate": 4.294123043996235e-08,
+      "loss": 0.777,
+      "step": 6346
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.840869855641693,
+      "learning_rate": 4.248351851731558e-08,
+      "loss": 0.7093,
+      "step": 6347
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7641169505849392,
+      "learning_rate": 4.2028253872949954e-08,
+      "loss": 0.813,
+      "step": 6348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7348826985320174,
+      "learning_rate": 4.157543661875929e-08,
+      "loss": 0.7661,
+      "step": 6349
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7141646574375673,
+      "learning_rate": 4.112506686603013e-08,
+      "loss": 0.7166,
+      "step": 6350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.433956966142259,
+      "learning_rate": 4.067714472545281e-08,
+      "loss": 0.6179,
+      "step": 6351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5873762825931,
+      "learning_rate": 4.023167030711484e-08,
+      "loss": 0.7205,
+      "step": 6352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5709124602943914,
+      "learning_rate": 3.978864372049973e-08,
+      "loss": 0.7412,
+      "step": 6353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.894106131249421,
+      "learning_rate": 3.934806507449041e-08,
+      "loss": 0.7344,
+      "step": 6354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.649700374213911,
+      "learning_rate": 3.8909934477370234e-08,
+      "loss": 0.7375,
+      "step": 6355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7352121319525016,
+      "learning_rate": 3.847425203681865e-08,
+      "loss": 0.6973,
+      "step": 6356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.546676278554904,
+      "learning_rate": 3.804101785991443e-08,
+      "loss": 0.8082,
+      "step": 6357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.731368752423272,
+      "learning_rate": 3.7610232053135745e-08,
+      "loss": 0.7807,
+      "step": 6358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.600089510254957,
+      "learning_rate": 3.7181894722355674e-08,
+      "loss": 0.7551,
+      "step": 6359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.710146340660481,
+      "learning_rate": 3.675600597284779e-08,
+      "loss": 0.7711,
+      "step": 6360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.728594166988212,
+      "learning_rate": 3.6332565909283915e-08,
+      "loss": 0.7805,
+      "step": 6361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1455008400453077,
+      "learning_rate": 3.591157463573303e-08,
+      "loss": 0.7706,
+      "step": 6362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7493337696123508,
+      "learning_rate": 3.5493032255665694e-08,
+      "loss": 0.7219,
+      "step": 6363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.582480537155428,
+      "learning_rate": 3.507693887194408e-08,
+      "loss": 0.7119,
+      "step": 6364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9089662705341968,
+      "learning_rate": 3.4663294586835264e-08,
+      "loss": 0.7601,
+      "step": 6365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.579269195629373,
+      "learning_rate": 3.425209950200015e-08,
+      "loss": 0.719,
+      "step": 6366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9961981717863577,
+      "learning_rate": 3.3843353718499004e-08,
+      "loss": 0.7513,
+      "step": 6367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7274870617059004,
+      "learning_rate": 3.343705733679037e-08,
+      "loss": 0.7911,
+      "step": 6368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6028781172920112,
+      "learning_rate": 3.3033210456729915e-08,
+      "loss": 0.8072,
+      "step": 6369
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5206585417321614,
+      "learning_rate": 3.263181317757269e-08,
+      "loss": 0.7499,
+      "step": 6370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.89486420226331,
+      "learning_rate": 3.2232865597969786e-08,
+      "loss": 0.763,
+      "step": 6371
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4263650727508166,
+      "learning_rate": 3.183636781597277e-08,
+      "loss": 0.7045,
+      "step": 6372
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.688965635402593,
+      "learning_rate": 3.144231992902813e-08,
+      "loss": 0.8026,
+      "step": 6373
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9726884382447123,
+      "learning_rate": 3.105072203398285e-08,
+      "loss": 0.8126,
+      "step": 6374
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5089298805790174,
+      "learning_rate": 3.0661574227081046e-08,
+      "loss": 0.82,
+      "step": 6375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6440762681304477,
+      "learning_rate": 3.0274876603962885e-08,
+      "loss": 0.7883,
+      "step": 6376
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.514536365196918,
+      "learning_rate": 2.989062925966901e-08,
+      "loss": 0.7623,
+      "step": 6377
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4245283656582233,
+      "learning_rate": 2.95088322886361e-08,
+      "loss": 0.7147,
+      "step": 6378
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.896910981296976,
+      "learning_rate": 2.9129485784699096e-08,
+      "loss": 0.8048,
+      "step": 6379
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.694298363324352,
+      "learning_rate": 2.8752589841092306e-08,
+      "loss": 0.7603,
+      "step": 6380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7538631724452465,
+      "learning_rate": 2.8378144550443852e-08,
+      "loss": 0.8414,
+      "step": 6381
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.607625116690004,
+      "learning_rate": 2.8006150004782352e-08,
+      "loss": 0.7615,
+      "step": 6382
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7452716377783775,
+      "learning_rate": 2.7636606295534664e-08,
+      "loss": 0.7256,
+      "step": 6383
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6681791298014645,
+      "learning_rate": 2.7269513513523692e-08,
+      "loss": 0.7814,
+      "step": 6384
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.785266623894804,
+      "learning_rate": 2.6904871748970607e-08,
+      "loss": 0.6774,
+      "step": 6385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.548359718667786,
+      "learning_rate": 2.6542681091493715e-08,
+      "loss": 0.6805,
+      "step": 6386
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8727987933271812,
+      "learning_rate": 2.6182941630109594e-08,
+      "loss": 0.7931,
+      "step": 6387
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.954983354484794,
+      "learning_rate": 2.5825653453233067e-08,
+      "loss": 0.84,
+      "step": 6388
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4906140254521496,
+      "learning_rate": 2.547081664867501e-08,
+      "loss": 0.6294,
+      "step": 6389
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8461489930565773,
+      "learning_rate": 2.511843130364233e-08,
+      "loss": 0.8024,
+      "step": 6390
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.539588939846344,
+      "learning_rate": 2.4768497504744637e-08,
+      "loss": 0.7963,
+      "step": 6391
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.551724315283847,
+      "learning_rate": 2.4421015337984244e-08,
+      "loss": 0.7367,
+      "step": 6392
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.792982810540755,
+      "learning_rate": 2.4075984888762838e-08,
+      "loss": 0.7481,
+      "step": 6393
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6340251459926645,
+      "learning_rate": 2.3733406241878145e-08,
+      "loss": 0.7786,
+      "step": 6394
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.795483266771458,
+      "learning_rate": 2.3393279481527255e-08,
+      "loss": 0.7786,
+      "step": 6395
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7073715073593534,
+      "learning_rate": 2.305560469130552e-08,
+      "loss": 0.7669,
+      "step": 6396
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5520780876011058,
+      "learning_rate": 2.2720381954201008e-08,
+      "loss": 0.6954,
+      "step": 6397
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6048219054047697,
+      "learning_rate": 2.2387611352603365e-08,
+      "loss": 0.7811,
+      "step": 6398
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6189944256832876,
+      "learning_rate": 2.2057292968298284e-08,
+      "loss": 0.7038,
+      "step": 6399
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.615322166675953,
+      "learning_rate": 2.1729426882468596e-08,
+      "loss": 0.7856,
+      "step": 6400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.950085694256778,
+      "learning_rate": 2.1404013175694292e-08,
+      "loss": 0.7141,
+      "step": 6401
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7197599076092187,
+      "learning_rate": 2.1081051927953623e-08,
+      "loss": 0.726,
+      "step": 6402
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9021177208654376,
+      "learning_rate": 2.0760543218621976e-08,
+      "loss": 0.7622,
+      "step": 6403
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5346673680250844,
+      "learning_rate": 2.044248712646968e-08,
+      "loss": 0.6864,
+      "step": 6404
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7369348943349383,
+      "learning_rate": 2.0126883729667534e-08,
+      "loss": 0.7364,
+      "step": 6405
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.534941329247269,
+      "learning_rate": 1.9813733105780163e-08,
+      "loss": 0.741,
+      "step": 6406
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9269037015104162,
+      "learning_rate": 1.950303533177378e-08,
+      "loss": 0.7941,
+      "step": 6407
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6677188601438586,
+      "learning_rate": 1.91947904840073e-08,
+      "loss": 0.7969,
+      "step": 6408
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5997357128251033,
+      "learning_rate": 1.88889986382379e-08,
+      "loss": 0.7151,
+      "step": 6409
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7504245622344503,
+      "learning_rate": 1.858565986962324e-08,
+      "loss": 0.7679,
+      "step": 6410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.626561601579103,
+      "learning_rate": 1.8284774252713688e-08,
+      "loss": 0.7941,
+      "step": 6411
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.771230928418999,
+      "learning_rate": 1.7986341861458976e-08,
+      "loss": 0.7155,
+      "step": 6412
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.696959350794206,
+      "learning_rate": 1.7690362769205993e-08,
+      "loss": 0.747,
+      "step": 6413
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.13649995719672,
+      "learning_rate": 1.7396837048696547e-08,
+      "loss": 0.8031,
+      "step": 6414
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6578860908672857,
+      "learning_rate": 1.710576477207293e-08,
+      "loss": 0.803,
+      "step": 6415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.844554639142442,
+      "learning_rate": 1.6817146010871255e-08,
+      "loss": 0.7004,
+      "step": 6416
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5950797820771916,
+      "learning_rate": 1.653098083602478e-08,
+      "loss": 0.7316,
+      "step": 6417
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.939752922140131,
+      "learning_rate": 1.6247269317868353e-08,
+      "loss": 0.7377,
+      "step": 6418
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.654338826560342,
+      "learning_rate": 1.596601152612731e-08,
+      "loss": 0.8092,
+      "step": 6419
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.55037502648767,
+      "learning_rate": 1.5687207529927472e-08,
+      "loss": 0.6677,
+      "step": 6420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4745890974533937,
+      "learning_rate": 1.54108573977918e-08,
+      "loss": 0.6845,
+      "step": 6421
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.352244373978386,
+      "learning_rate": 1.51369611976393e-08,
+      "loss": 0.7089,
+      "step": 6422
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.67104638009709,
+      "learning_rate": 1.4865518996786122e-08,
+      "loss": 0.8284,
+      "step": 6423
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.9781450975243584,
+      "learning_rate": 1.4596530861944458e-08,
+      "loss": 0.7792,
+      "step": 6424
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5664080284926336,
+      "learning_rate": 1.432999685922365e-08,
+      "loss": 0.8157,
+      "step": 6425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7769335025657664,
+      "learning_rate": 1.4065917054132405e-08,
+      "loss": 0.8154,
+      "step": 6426
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.649007252116273,
+      "learning_rate": 1.3804291511572144e-08,
+      "loss": 0.8174,
+      "step": 6427
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.926878045065561,
+      "learning_rate": 1.3545120295843651e-08,
+      "loss": 0.7972,
+      "step": 6428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8927224247284036,
+      "learning_rate": 1.3288403470643751e-08,
+      "loss": 0.7388,
+      "step": 6429
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7844682550817415,
+      "learning_rate": 1.3034141099066422e-08,
+      "loss": 0.7739,
+      "step": 6430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6199308848828426,
+      "learning_rate": 1.2782333243601675e-08,
+      "loss": 0.8188,
+      "step": 6431
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5090276146240362,
+      "learning_rate": 1.2532979966138892e-08,
+      "loss": 0.7832,
+      "step": 6432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7804406982385927,
+      "learning_rate": 1.2286081327959055e-08,
+      "loss": 0.6882,
+      "step": 6433
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2823754032742363,
+      "learning_rate": 1.2041637389745842e-08,
+      "loss": 0.8219,
+      "step": 6434
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.592159073695726,
+      "learning_rate": 1.1799648211574533e-08,
+      "loss": 0.8091,
+      "step": 6435
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.667384164685074,
+      "learning_rate": 1.1560113852919774e-08,
+      "loss": 0.7555,
+      "step": 6436
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6250053411416956,
+      "learning_rate": 1.1323034372653364e-08,
+      "loss": 0.7027,
+      "step": 6437
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.558165207717662,
+      "learning_rate": 1.1088409829042023e-08,
+      "loss": 0.6945,
+      "step": 6438
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.554335993292053,
+      "learning_rate": 1.0856240279750741e-08,
+      "loss": 0.7743,
+      "step": 6439
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6211979677469794,
+      "learning_rate": 1.0626525781838316e-08,
+      "loss": 0.6903,
+      "step": 6440
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5971448988828993,
+      "learning_rate": 1.0399266391764029e-08,
+      "loss": 0.7464,
+      "step": 6441
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5474481779950495,
+      "learning_rate": 1.0174462165380983e-08,
+      "loss": 0.8297,
+      "step": 6442
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4791215754553084,
+      "learning_rate": 9.952113157940534e-09,
+      "loss": 0.7697,
+      "step": 6443
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.479645255730251,
+      "learning_rate": 9.732219424087863e-09,
+      "loss": 0.7865,
+      "step": 6444
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.620899042191569,
+      "learning_rate": 9.514781017869734e-09,
+      "loss": 0.6784,
+      "step": 6445
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7710163417045157,
+      "learning_rate": 9.299797992724514e-09,
+      "loss": 0.7089,
+      "step": 6446
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6119013634252335,
+      "learning_rate": 9.087270401488823e-09,
+      "loss": 0.7556,
+      "step": 6447
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5940185225330477,
+      "learning_rate": 8.877198296396438e-09,
+      "loss": 0.6927,
+      "step": 6448
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5758034981359446,
+      "learning_rate": 8.66958172907717e-09,
+      "loss": 0.705,
+      "step": 6449
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4943602171675057,
+      "learning_rate": 8.464420750556868e-09,
+      "loss": 0.8198,
+      "step": 6450
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5946902014802995,
+      "learning_rate": 8.26171541125964e-09,
+      "loss": 0.7201,
+      "step": 6451
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5390879531895627,
+      "learning_rate": 8.061465761003417e-09,
+      "loss": 0.8203,
+      "step": 6452
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.982939394034132,
+      "learning_rate": 7.863671849004384e-09,
+      "loss": 0.8149,
+      "step": 6453
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6976834349346293,
+      "learning_rate": 7.668333723874765e-09,
+      "loss": 0.8409,
+      "step": 6454
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7794087904559786,
+      "learning_rate": 7.475451433623936e-09,
+      "loss": 0.7434,
+      "step": 6455
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7798006349430144,
+      "learning_rate": 7.2850250256562e-09,
+      "loss": 0.8585,
+      "step": 6456
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.749781684197471,
+      "learning_rate": 7.097054546773008e-09,
+      "loss": 0.7369,
+      "step": 6457
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.754818961661538,
+      "learning_rate": 6.911540043171849e-09,
+      "loss": 0.7442,
+      "step": 6458
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4299632918254526,
+      "learning_rate": 6.728481560448474e-09,
+      "loss": 0.5999,
+      "step": 6459
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.691427186417734,
+      "learning_rate": 6.54787914359134e-09,
+      "loss": 0.7845,
+      "step": 6460
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.47678081900891,
+      "learning_rate": 6.369732836989384e-09,
+      "loss": 0.7412,
+      "step": 6461
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7436779125657464,
+      "learning_rate": 6.194042684425361e-09,
+      "loss": 0.7747,
+      "step": 6462
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8750941731143174,
+      "learning_rate": 6.0208087290780645e-09,
+      "loss": 0.7888,
+      "step": 6463
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.770040007515961,
+      "learning_rate": 5.850031013524549e-09,
+      "loss": 0.6909,
+      "step": 6464
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6684774674425613,
+      "learning_rate": 5.681709579737904e-09,
+      "loss": 0.7019,
+      "step": 6465
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0274878532422287,
+      "learning_rate": 5.515844469085041e-09,
+      "loss": 0.7444,
+      "step": 6466
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5402296932243495,
+      "learning_rate": 5.352435722332238e-09,
+      "loss": 0.6953,
+      "step": 6467
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6199879604104117,
+      "learning_rate": 5.191483379639595e-09,
+      "loss": 0.7152,
+      "step": 6468
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5531274187025073,
+      "learning_rate": 5.0329874805654656e-09,
+      "loss": 0.7096,
+      "step": 6469
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.634831711496866,
+      "learning_rate": 4.876948064064246e-09,
+      "loss": 0.7461,
+      "step": 6470
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3984961627115484,
+      "learning_rate": 4.723365168485261e-09,
+      "loss": 0.7232,
+      "step": 6471
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4114883364205943,
+      "learning_rate": 4.572238831574982e-09,
+      "loss": 0.717,
+      "step": 6472
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5541673170994525,
+      "learning_rate": 4.4235690904759206e-09,
+      "loss": 0.8181,
+      "step": 6473
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7352721716539095,
+      "learning_rate": 4.277355981727738e-09,
+      "loss": 0.6996,
+      "step": 6474
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7112077492868942,
+      "learning_rate": 4.133599541265021e-09,
+      "loss": 0.7444,
+      "step": 6475
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8709744528710486,
+      "learning_rate": 3.992299804418398e-09,
+      "loss": 0.7478,
+      "step": 6476
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.830825498249224,
+      "learning_rate": 3.853456805915645e-09,
+      "loss": 0.8736,
+      "step": 6477
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7189102796459337,
+      "learning_rate": 3.7170705798816875e-09,
+      "loss": 0.717,
+      "step": 6478
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2342398789984927,
+      "learning_rate": 3.583141159834158e-09,
+      "loss": 0.801,
+      "step": 6479
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.80624204172367,
+      "learning_rate": 3.4516685786922798e-09,
+      "loss": 0.8927,
+      "step": 6480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8881355207422845,
+      "learning_rate": 3.322652868764653e-09,
+      "loss": 0.7925,
+      "step": 6481
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.833888014886496,
+      "learning_rate": 3.196094061762578e-09,
+      "loss": 0.8012,
+      "step": 6482
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6885761878683825,
+      "learning_rate": 3.071992188790063e-09,
+      "loss": 0.8459,
+      "step": 6483
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6920429486726296,
+      "learning_rate": 2.9503472803471546e-09,
+      "loss": 0.7366,
+      "step": 6484
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.793627620980447,
+      "learning_rate": 2.831159366331049e-09,
+      "loss": 0.7325,
+      "step": 6485
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4166025338417283,
+      "learning_rate": 2.7144284760349804e-09,
+      "loss": 0.6311,
+      "step": 6486
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.788670358621249,
+      "learning_rate": 2.600154638148222e-09,
+      "loss": 0.7814,
+      "step": 6487
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7429804605417925,
+      "learning_rate": 2.488337880754976e-09,
+      "loss": 0.8176,
+      "step": 6488
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.699912986291421,
+      "learning_rate": 2.378978231338813e-09,
+      "loss": 0.7569,
+      "step": 6489
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6965672451161726,
+      "learning_rate": 2.272075716774902e-09,
+      "loss": 0.711,
+      "step": 6490
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.637709653615361,
+      "learning_rate": 2.167630363338891e-09,
+      "loss": 0.786,
+      "step": 6491
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7864743894646486,
+      "learning_rate": 2.065642196699136e-09,
+      "loss": 0.7065,
+      "step": 6492
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5309378266446063,
+      "learning_rate": 1.966111241922253e-09,
+      "loss": 0.7645,
+      "step": 6493
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9183202809954496,
+      "learning_rate": 1.8690375234697854e-09,
+      "loss": 0.7266,
+      "step": 6494
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8177372518135932,
+      "learning_rate": 1.7744210651993167e-09,
+      "loss": 0.7258,
+      "step": 6495
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.797322429322555,
+      "learning_rate": 1.6822618903655773e-09,
+      "loss": 0.7366,
+      "step": 6496
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6958226240614174,
+      "learning_rate": 1.592560021618228e-09,
+      "loss": 0.7854,
+      "step": 6497
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5733350687801146,
+      "learning_rate": 1.5053154810040772e-09,
+      "loss": 0.6753,
+      "step": 6498
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8514235606033753,
+      "learning_rate": 1.4205282899659722e-09,
+      "loss": 0.7067,
+      "step": 6499
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0375905676960664,
+      "learning_rate": 1.3381984693405793e-09,
+      "loss": 0.7599,
+      "step": 6500
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.483747166388334,
+      "learning_rate": 1.2583260393628226e-09,
+      "loss": 0.7857,
+      "step": 6501
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8775232736931855,
+      "learning_rate": 1.1809110196636663e-09,
+      "loss": 0.8218,
+      "step": 6502
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0171976204757547,
+      "learning_rate": 1.1059534292690022e-09,
+      "loss": 0.7009,
+      "step": 6503
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5790803766513988,
+      "learning_rate": 1.0334532866007608e-09,
+      "loss": 0.7804,
+      "step": 6504
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8344092373568186,
+      "learning_rate": 9.634106094791317e-10,
+      "loss": 0.7493,
+      "step": 6505
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5103505676785516,
+      "learning_rate": 8.958254151170131e-10,
+      "loss": 0.7996,
+      "step": 6506
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6642999642212963,
+      "learning_rate": 8.30697720126672e-10,
+      "loss": 0.6592,
+      "step": 6507
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.861760072908654,
+      "learning_rate": 7.680275405130833e-10,
+      "loss": 0.749,
+      "step": 6508
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5797148250536144,
+      "learning_rate": 7.078148916783711e-10,
+      "loss": 0.7418,
+      "step": 6509
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.572541953632293,
+      "learning_rate": 6.500597884229188e-10,
+      "loss": 0.7554,
+      "step": 6510
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.605496462403277,
+      "learning_rate": 5.947622449409274e-10,
+      "loss": 0.8051,
+      "step": 6511
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.316767161854669,
+      "learning_rate": 5.419222748226372e-10,
+      "loss": 0.7009,
+      "step": 6512
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.547132182034208,
+      "learning_rate": 4.915398910532166e-10,
+      "loss": 0.8116,
+      "step": 6513
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6269179574986836,
+      "learning_rate": 4.436151060183136e-10,
+      "loss": 0.7224,
+      "step": 6514
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8338788225773315,
+      "learning_rate": 3.9814793149295373e-10,
+      "loss": 0.7944,
+      "step": 6515
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8067035466603394,
+      "learning_rate": 3.5513837865486236e-10,
+      "loss": 0.8032,
+      "step": 6516
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.550073722813949,
+      "learning_rate": 3.145864580722524e-10,
+      "loss": 0.7207,
+      "step": 6517
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.636265469611505,
+      "learning_rate": 2.764921797138165e-10,
+      "loss": 0.7546,
+      "step": 6518
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.616310209319721,
+      "learning_rate": 2.4085555293984484e-10,
+      "loss": 0.8064,
+      "step": 6519
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.817876730596737,
+      "learning_rate": 2.0767658650999723e-10,
+      "loss": 0.7837,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4123576010128955,
+      "learning_rate": 1.7695528857886169e-10,
+      "loss": 0.7163,
+      "step": 6521
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.801149751350197,
+      "learning_rate": 1.4869166669595482e-10,
+      "loss": 0.791,
+      "step": 6522
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.669239979861841,
+      "learning_rate": 1.2288572780905227e-10,
+      "loss": 0.6997,
+      "step": 6523
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0859647495600746,
+      "learning_rate": 9.95374782586378e-11,
+      "loss": 0.7652,
+      "step": 6524
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.8567736532925334,
+      "learning_rate": 7.86469237845644e-11,
+      "loss": 0.8433,
+      "step": 6525
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.95683604369921,
+      "learning_rate": 6.021406952161357e-11,
+      "loss": 0.8292,
+      "step": 6526
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5110923123790236,
+      "learning_rate": 4.423891999838503e-11,
+      "loss": 0.7575,
+      "step": 6527
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.3604571293366843,
+      "learning_rate": 3.072147914284784e-11,
+      "loss": 0.6375,
+      "step": 6528
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.6002150822854406,
+      "learning_rate": 1.966175027567907e-11,
+      "loss": 0.6676,
+      "step": 6529
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.781193187571927,
+      "learning_rate": 1.1059736115814901e-11,
+      "loss": 0.7323,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.670757867546929,
+      "learning_rate": 4.915438777119974e-12,
+      "loss": 0.7783,
+      "step": 6531
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.786026911255166,
+      "learning_rate": 1.2288597694976034e-12,
+      "loss": 0.6062,
+      "step": 6532
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.022792001865195,
+      "learning_rate": 0.0,
+      "loss": 0.8392,
+      "step": 6533
+    },
+    {
+      "epoch": 1.0,
+      "step": 6533,
+      "total_flos": 6195132391227392.0,
+      "train_loss": 0.8881614911151154,
+      "train_runtime": 203520.9615,
+      "train_samples_per_second": 4.109,
+      "train_steps_per_second": 0.032
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6533,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 400,
+  "total_flos": 6195132391227392.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}